Academia.eduAcademia.edu
{v} Dealing with Cognitive Penetration ChrisTucker190113OUS.indd 223 5/10/2013 11:46:46 AM ChrisTucker190113OUS.indd 224 5/10/2013 11:46:46 AM { 10 } Phenomenal Conservatism and Cognitive Penetration: he “Bad Basis” Counterexamples Matthew McGrath 1. Introduction Seemings and beliefs are related in important ways. Like beliefs, seemings are content-bearing mental states of a subject that have truth as their correctness condition: a seeming that P is correct, accurate, only if P is true. Not only this, but when it seems that P, this normally inclines one to believe that P. It is not a mystery why they should so incline one, because seemings “have the feel of truth, the feel of a state whose content reveals how things really are” (Tolhurst 1988, 298). Because of these relationships, a subject can be said to “conserve” a seeming by believing its content. A key question in recent epistemology is whether one is always prima facie justiied in conserving one’s seemings in this way. According to what Michael Huemer (2001) calls “phenomenal conservatism,” the answer is yes: PC if it seems to you that P, then you are prima facie justiied in believing that P. Huemer and others see signiicant epistemological implications in PC. For Huemer, it provides a single simple overarching foundationalist principle. For Chris Tucker, it provides a way to solve the “speckled hen” problem for foundationalism. For James Pryor, the principle of dogmatism, which restricts PC to perceptual seemings,1 provides the basis for a diagnosis of where the best skeptical arguments go wrong. Philosophers divide sharply over the plausibility of PC. On the one hand, Huemer (2001, 103–4) claims the principle is self-evident when understood 1 Pryor (2000, 519) writes: “he dogmatist about perceptual justiication says that when it perceptually seems to you if p is the case, you have a kind of justiication for believing p that does not presuppose or rest on your justiication for anything else, which could be cited in an argument—even an ampliative argument!—for p. To have this justiication for believing p, you need only have an experience that represents p as being the case.” Pryor assumes that experiences have propositional content. I will follow him in this assumption here. ChrisTucker190113OUS.indd 225 5/10/2013 11:46:46 AM 226 Dealing with Cognitive Penetration correctly. To understand the principle correctly is to appreciate that the sort of justiication involved is epistemic rather than moral or prudential justiication and that the sort of justiication at issue is justiication from the subject’s point of view. A good ordinary language gloss on the sort of justiication Huemer has in mind is justiication in the sense of what is “reasonable” for one to think, or in a more philosophical parlance, justiication in the sense of the doxastic attitude it is reasonable for a subject to have.2 We might add that understanding the principle correctly also requires understanding that “seeming” is being used for states that meet Tolhurst’s description: they feel as if they reveal how things are, not merely how things might be or which are somewhat likely. With all this in place, let us think through just what PC says. Suppose it seems to you that P and you have no defeaters (i.e., no good evidence for not-P and no good evidence that this seeming is unreliable as to whether P). Which doxastic attitude would it be reasonable for you to have toward P? Disbelieve P, without good evidence for not-P? Withhold judgment on P? It does seem to you that P, and you lack evidence for not-P and for the unreliability of the seeming with respect to P? he only reasonable attitude to take is belief. Even if we do not ind this answer clearly correct, let alone self-evident, it does have intuitive plausibility. Its seeming to one that P seems to be at least some evidence weighing in favor of one’s believing P. here are several objections that arise immediately. One concerns the liberality of PC. Any seeming? What if the seeming just overtakes one out of the blue (unbeknownst to one)? Is it then evidence? Or if it is evidence, is it good enough evidence for belief? A second worry concerns the need for higher-order reasons, reasons to think seemings are reliable, or at least the relevant particular seeming is reliable on the occasion in question. he irst objection might make one back of from PC to restricted versions of it, such as dogmatism about perceptual justiication: if it perceptually seems to you that P then you are prima facie justiied in believing P. (Other possibilities privilege memorial seemings, introspective seemings, intuitional seemings.) Surely, if the thing looks plainly like a bird, this is evidence for you that it is a bird. If we do not require too much conidence for belief, isn’t this good enough evidence for belief, in the absence of defeaters?3,4 he second objection is familiar from debates over moderate foundationalism, and it applies both to PC and its restricted versions. However, the 2 Other glosses: what one ought to think, what it is rational to think, what it makes sense to think. his answer is arguably less plausible if a great deal hinges on whether P. Perhaps the prima facie rider is unsatisied in high stakes situations. I ignore the relevance of stakes in this paper. 4 Alternatively, the irst objection might be met by retreating from “justiied in believing” to “has at least some degree of justiication to believe.” he safest reply might be to combine the two ways of responding, scaling back PC as follows: if it perceptually (and perhaps memorially, intuitionally) seems to you that P, then prima facie you have at least some degree of justiication to believe P. his is a substantial retreat. If Huemer is right in claiming that a great many of our ordinary beliefs are based on seemings, this retreat provides an opening for skepticism. he skeptic may argue that even if PC, 3 ChrisTucker190113OUS.indd 226 5/10/2013 11:46:46 AM he “Bad Basis” Counterexamples 227 objection is hardly decisive. Even if such higher-level requirements could be defended without falling into skepticism (a big “if ”), there are proposals for combining moderate foundationalism with higher-level requirements. Suppose a source S could not justify one in believing P unless one was also justiied in believing S is reliable. his supposition does not rule out S’s providing foundational justiication to believe P. For instance, it may well be a necessary condition on being justiied in believing anything that one is justiied in believing that one exists.5 his doesn’t show that justiication to believe things other than “I exist” cannot be foundational. Similarly, it may be a necessary condition on being justiied through source S that one is justiied in believing S is reliable. his would not show that justiication through S cannot be foundational. More threatening to PC than these two objections are apparent counterexamples involving cognitive penetration of perceptual seemings. Peter Markie gives an example: Mental processes that are incapable of producing prima facie justiied beliefs can nonetheless determine how things seem to us. Suppose that we are prospecting for gold. You have learned to identify a gold nugget on sight but I have no such knowledge. As the water washes out of my pan, we both look at a pebble, which is in fact a gold nugget. My desire to discover gold makes it seem to me as if the pebble is gold; your learned identiication skills make it seem that way to you. According to (PC), the belief that it is gold has prima facie justiication for both of us. Yet, certainly, my wishful thinking should not gain my perceptual belief the same positive epistemic status of defeasible justiication as your learned identiication skills. (Markie 2006, 356–57) Susanna Siegel (2012) cites examples involving cognitive penetration of visual experience. If Jill unjustiiably thinks Jack is angry, she might expect him to look angry, and this might lead her to “see” him as angry, to his looking angry to her. If it does, surely her belief does not become justiied if she re-bases it on his looking angry. Or think of the preformationist looking through the microscope who “sees” tiny human embryos in an ordinary cell, due to his desire that his theory be conirmed. Intuitively, if he forms the belief that there is a tiny embryo in the cell on the basis of such an experience, his belief is not justiied. If we understand Siegel’s cases to involve visual seemings—its visually seeming that Jack is angry, that there is an embryo in the cell—then her cases are potential counterexamples to PC as well as its restriction to perceptual seemings. so modiied, is true, these ordinary beliefs are not justiied because the degree of justiication seemings provide is not enough for outright belief. See Huemer (2001, 39–41). 5 Cf. Silins 2008, McGrath forthcoming. ChrisTucker190113OUS.indd 227 5/10/2013 11:46:46 AM 228 Dealing with Cognitive Penetration We are thus faced with a puzzle. PC is plausible, especially if it is restricted to perceptual seemings, but so are the apparent counterexamples involving cognitively penetrated perceptual seemings—the “bad basis” counterexamples, as I’ll call them. Anyone, whether a proponent of PC or not, ought to want some account of how and why both the principle and the apparent counterexamples to it should be plausible. his chapter explores the options for proponents of PC (or its restriction).6 I will consider three options: (1) the defeater approach, which takes the purported counterexamples to be cases in which the prima facie proviso is unsatisied because the subject has a defeater; (2) the “distinguish the epistemic statuses” approach, which takes the examples to refute principles resembling PC but which concern some epistemic status other than the “favored” status of reasonableness to believe; and (3) the receptivity approach, which modiies PC by restricting it to “receptive” seemings, where a seeming’s receptivity is not understood in terms of the kind of seeming it is (e.g., perceptual, memorial) but rather in terms it is lacking a certain sort of basis, what I will call a “quasi-inferential” one. I will argue that the irst two approaches, despite enabling conservatives to show that PC withstands many “bad basis” cases, do not save them from all. I agree with Markie and Siegel that PC stands refuted, even when restricted to perceptual seemings. However, I hope to show that its prospects brighten when it is restricted to receptive seemings. Our discussion bears on a considerable number of views defended in epistemology. As Siegel (2012) notes, many epistemological theories seem vulnerable to “bad basis” counterexamples. Many philosophers have argued that how things are presented in perceptual experience prima facie justiies the subject in believing things are that way. Among these views are leading forms of moderate foundationalism, including not only Pryor’s (2000) dogmatism but Pollock’s (1975) direct realism and Robert Audi’s (1993) fallibilistic foundationalism. On other views (e.g., White 2006, Wright 2004), experiences by themselves do not provide immediate justiication but can justify only together with independently justiied beliefs or assumptions about their general reliability (these background beliefs or assumptions might be taken to be a priori justiied or justiied because of an “unearned” warrant). he counterexamples discussed here threaten these views as well. We may imagine that Markie’s wishful gold-digger has not only a gold-experience but also whatever White or Wright require in the way of justiied background beliefs or assumptions. If so, and if this doesn’t afect the intuition that his belief isn’t justiied (or isn’t as justiied as it would be if the experience resulted from the exercise of perceptual identiication skills), then this is a problem for White and Wright. I claim that these views, too, can take advantage of the receptivity approach and improve their 6 I leave this qualiication tacit in what follows. ChrisTucker190113OUS.indd 228 5/10/2013 11:46:46 AM he “Bad Basis” Counterexamples 229 prospects in the face of the bad basis counterexamples. Finally, in an appendix, I briely discuss how proponents of doxastic conservatism might attempt to appeal to the idea of receptivity. 2. A Preliminary Worry Before considering the three approaches, I want to address a worry about the phenomenal conservative’s favored status of what it is reasonable for the subject to think. One might doubt whether the intuitive correctness of such claims— claims made in my motivation for PC—tracks any normative epistemic status as opposed to the demands of coherence.7 So, even if the motivations I gave for PC motivate something, the worry is that they do not motivate a principle with any normative “oomph.” Moreover, if throughout this paper I am going to appeal to intuitive claims about what it is reasonable to think in various situations, these appeals will be questionable as well. he general worry is about traicking in the language of “reasonableness” at all in doing epistemology. Here is a way to explain one source of the worry. A coherent set of beliefs can be out of keeping with one’s evidence. Suppose my evidence strongly points against P and against Q, but that nevertheless I believe P. hen, if I continue to believe P, coherence requires that I believe the disjunctive proposition P or Q if I entertain the question Suppose I entertain the question of whether P or Q is true. We want to say that it is reasonable for me to believe P or Q, insofar as I already have a belief that P. How could it be reasonable not to believe P or Q given that I already believe P? On the one hand, surely it isn’t reasonable for me to believe P or Q, because ater all, I have good evidence against P and evidence against Q! So, we want to say that it isn’t reasonable for me to believe P or Q, and yet also feel it is reasonable insofar as I already have a belief that P. John Broome (1999, 2002) explains such apparently conlicting intuitions as follows. If I believe P and consider the question of whether the disjunction P or Q is true, and then I fail to believe P or Q (by disbelieving it or by suspending judgment), then I am being incoherent; my cognitive states will clash. To keep the terminology clear, I will use the label “coherence requirement,” in place of Broome’s “rational requirement.” So, believing P requires me, on pain of incoherence, to believe P or Q at least if I consider the question of whether it is true that P or Q. However, although my belief in P imposes this coherence requirement to believe P or Q, this does not make believing P or Q epistemically appropriate for me in a normative sense: my evidence against P and against Q together give me suicient epistemic reason not to believe P or Q. here is no contradiction here, according to Broome: it is possible, if one 7 Such a worry is developed in Jackson (2011). ChrisTucker190113OUS.indd 229 5/10/2013 11:46:46 AM 230 Dealing with Cognitive Penetration takes a normatively inappropriate attitude, to be such that coherence requires taking another normatively inappropriate attitude. he intuitive judgment, “I am not reasonable to believe P or Q” picks up on the epistemically normative status he intuitive judgment, “I am reasonable to believe P or Q insofar as I believe P” picks up on the demands of Broomestyle coherence requirements. Whether such requirements have normative standing is a matter of dispute: are there epistemic reasons to be coherent in this sense; is there an epistemic obligation to be coherent?8 Even if both intuitive judgments of reasonableness track something of normative signiicance, though, the sources of the normativity are diferent. In the one case, the source is the particular concrete evidence against P and against Q. In the other case, the source is some more abstract reason or obligation to be coherent.9 Let us return to the original worry about my use of “reasonable.” Broome’s distinction is clear, but in ordinary life we can use the same terms—“reasonable,” “ought,” “should”—for both what is required for coherence and for epistemic normative appropriateness (deriving from concrete garden-variety reasons rather than some abstract reason to be coherent). What reason is there to think that when I motivated PC, I wasn’t picking up on coherence requirements only rather than normative appropriateness? What reason is there to think that when I go on to consider cognitive penetration cases I won’t be simply gathering intuitions about coherence requirements? I make two points in response. First, the fact that we ordinarily use “reasonable to believe” to cover cases of epistemic normative appropriateness and cases of what’s required for coherence is by itself no good reason to abandon all such talk in doing epistemology. We need only to take caution not to confuse the two. When “reasonable” is used merely for what is required for coherence, there is an implicit relativization to other mental states which themselves can be assessed for their normative epistemic status. When we ind it natural to say that believing P “is reasonable,” we need to check to see whether we want to say the likes of “reasonable insofar as one believes P” or “reasonable given that one doubts Q.” he relativizations are to the mere holding of these 8 See Kolodny (2005) on the question of the normative force of Broome-style coherence requirements. 9 Similarly, if I continue to intend to achieve some end, against all the many powerful reasons not to, then if I know M is a necessary means to that end, practical coherence demands me to intend to take means M (Broome 1999, 2002; Dancy 2000). (his is a bit rough. We may need to add the belief that unless I intend to take M I will not do M.) Here again, a judgment of reasonableness picks up on the following fact: I hold other states which are such that, to maintain consistency, if I add to them a “practical” attitude toward M, it must be intention. Such facts can hold even if it is normatively inappropriate to intend M, because I have good reasons not to intend the end in the irst place. Which attitude it is normatively appropriate in a practical sense to have is one thing. Which attitude I need to have to maintain coherence with another attitude I already have is quite another thing. Intuitive judgments of practical reasonableness can pick up on either. ChrisTucker190113OUS.indd 230 5/10/2013 11:46:46 AM he “Bad Basis” Counterexamples 231 psychological states, not to the epistemically appropriate/inappropriate holding of them. Coherence requirements are blind to the normative statuses of the states involved. Consider, again, the above motivation for PC. he objector agrees that we do ind it natural to say that one is “reasonable” to believe P when it seems to her that P and she lack defeaters. So, we have to ask if it feels necessary to add a relativization to other mental states assessable for their normative epistemic statuses. Assuming seemings are not assessable in this way (can there be an epistemic reason for it to seem to one that P?), there is no such relativization.10 One might say, “She is reasonable to believe P given that it seems to her that P,” but here the “given that” only functions to indicate the various features of the person’s situation which might help to make it reasonable for her to believe P in the normatively loaded sense. Second, coherence requirements are presumably not brute facts. hey need explanation: why is it that combining such and such attitudes is incoherent, if it is? In some cases, the explanation proceeds solely in terms of the nature of the attitudes and the logical relations between their contents. Perhaps we can explain the coherence requirement not to believe ~P if one believes P by pointing to the fact that P and ~P are contradictories and that belief aims at truth. However, other coherence requirements cannot be explained in this fashion. Doubting whether a testiier is sincere rationally requires me not to believe what the person says merely on his word (so that if I do believe the testiier solely on his word while believing he is insincere my mental states clash). It is far from clear how this requirement is explainable solely in terms of logical relations between the contents involved and the natures of the states of doubt and belief. A paradigm way to explain coherence requirements is by referring to a corresponding normative principle. (We use “reasonable” hereater only for the normatively loaded status.) If I am reasonable to doubt that a testiier is being insincere in testifying that P to me, then the testiier’s mere word that P is not enough to make me reasonable in believing P. hat’s why there is something incoherent about doubting his sincerity while accepting P based on solely on his word. In so doing, I would be guaranteeing that either my doubt or my belief was unreasonable. In fact, wherever there is a true normative epistemic principle about reasonableness, it grounds a corresponding coherence requirement. he fact that being reasonable in believing P suices for being reasonable in believing P or Q (if you entertain the question) grounds the fact that believing P requires believing P or Q (if you entertain the question) If one louts the coherence requirement, one would be guaranteeing that one falls short with respect to reasonableness either in believing P or in not believing P or Q. 10 he question whether some seemings can be normatively assessable is revisited in note 22. ChrisTucker190113OUS.indd 231 5/10/2013 11:46:46 AM 232 Dealing with Cognitive Penetration hus, where there is a coherence requirement that appears inexplicable merely in terms of logical relations between contents and the nature of belief, a plausible potential explanation for that requirement is the corresponding normative principle. Suppose our objector agrees that there is a coherence requirement to believe P when it seems to one that P and one lacks defeaters. What is its explanation or ground? It is diicult to explain it in terms of logical relations among contents together with the nature of belief and seeming. However, there is a ready explanation that appeals to the normative principle PC. he critic who concedes that the motivations I have provided for PC motivate accepting a coherence requirement concerning seeming and believing thus may very well have to appeal to PC to explain that requirement, where PC is understood as a normative epistemic principle. With this preliminary worry addressed, I turn to the three approaches for coping with the “bad basis” counterexamples. In what follows, I oten write of “justiication,” because of its familiarity and its convenient grammatical features.11 But the reader should see this terminology merely as a convenient way of talking about what a subject is reasonable to think, where “reasonable” is understood in the epistemically normative sense. 3. he Defeat Approach Appealing to defeaters handles many familiar “bad basis” cases. Consider wishful thinking. Within psychology, wishful thinking is understood as judgment inluenced by a directional goal (a goal of reaching a particular conclusion rather than an accurate conclusion on a given question), and there is some consensus that one primary way that directional goals work is by biasing the evidence the person brings to mind at the time of judgment (Kunda 1990, 1999). In particular, when we have defeasible evidence favoring a desired belief that P—say, it seems to us that P—we tend not to search memory as hard, if at all, for any counterevidence we might have. his need not always have the desired efect, of course, because sometimes clearly disconirming evidence springs to mind anyway, despite the biased memory search.12 Still, it can work, and when it does, the belief formed is intuitively not justiied. here 11 hus, we can speak of something justifying a belief and of one having justiication to believe something. All these could be re-expressed using unlovely neologisms derived from “reasonable.” 12 Research into judgment inluenced by directional goals shows there are limits to its power. Awareness of decisive counterevidence, and even the anticipation of decisive evidence one way or the other in the future, drastically reduces its efects. It is also limited in a number of other ways (e.g., it operates less efectively when the question at hand is a hard factual question rather than an open-ended one). See Kunda (1990, 1999). ChrisTucker190113OUS.indd 232 5/10/2013 11:46:46 AM he “Bad Basis” Counterexamples 233 is no mystery about why: the subject has a defeater, one that isn’t brought to mind, but a defeater all the same.13 However, there could well be “bad basis” cases in which the subject lacks defeaters, even unconscious ones. he examples from Markie and Siegel appear to fall into this category. 4. he “Distinguish the Epistemic Statuses” Approach To have a counterexample to PC, we need a case in which it seems to the subject that P, the subject lacks defeaters, but it isn’t reasonable for her to believe P. he hope behind the “distinguish the epistemic statuses” approach is to grant that in those “bad basis” cases in which the subject lacks, defeaters the subject may well lack many important positive epistemic statuses, but not the favored one of being reasonable to believe. When we take the “bad basis” examples to refute PC—the claim is—we conlate distinct epistemic statuses. he counterexamples may well refute principles like PC about those other statuses, but not PC itself. We will discuss a number of ways to pursue this approach. One strategy is to press into service the distinction between propositional and doxastic justiication. PC concerns propositional justiication, i.e., justiication to believe, but the counterexamples show, at least in the irst instance, only that the subjects lack doxastic justiication, i.e., their beliefs are not justiied. Markie claims that the belief of the gold-digger who relies on the wishful seeming surely does not have the “same positive epistemic status” as that of the gold-digger who believes on the basis of an appearance grounded in the exercise of perceptual learning. Siegel, too, makes claims about the epistemic 13 Taking such a line would require the phenomenal conservative to agree that one can have defeaters that one is not consciously entertaining. his is a plausible line in any case (though see Feldman, 1988, for an opposing view). In light of the relevance of stored evidence to justiication, it is important to distinguish a number of related assessments of unreasonableness, which vary in the degree of associated intuitive irrationality. If a person believes something that isn’t reasonable to believe in light of her conscious information at the time, this is clearly more irrational than if a person believes something that it is reasonable to believe in light of conscious evidence but not in light of her total possessed evidence. To take an example from Goldman (1988, 202–3), Melanie knows the library normally opens at 7:00 am, but she has also noticed it opens at 1:00 pm on Sundays, although she rarely goes there Sundays. his Sunday she has a special reason to go to the library. She is heading of to the library shortly ater 7:00 am. She inds it closed and thinks, “How silly of me! I knew the library doesn’t open on Sundays till 1:00 pm.” Her action seems far less irrational than it would if she had not recalled the Sunday hours and still went to the library. Yet even in Goldman’s case, Melanie’s action is unreasonable: she is doing something she has evidence will not turn out well. Compare this to a case in which Melanie had never known about the Sunday hours, or to a case in which years ago Melanie knew of the Sunday hours but had “completely forgotten,” so that if you asked her beforehand whether the Sunday hours were the same as the regular hours she would say yes. here are many interesting intermediary cases in which one seems to have “forgotten” and yet is tempted to say, ater being reminded, “I knew that!” ChrisTucker190113OUS.indd 233 5/10/2013 11:46:46 AM 234 Dealing with Cognitive Penetration status of beliefs in her examples. he examples, then, might be taken to refute only a principle like PC about doxastic justiication, but not PC itself. However, as Siegel notes in her second paper (2013), doxastic and propositional justiication are standardly thought to be related. If you are propositionally justiied in believing P—if P is something you are justiied to believe—then it seems you should have a route available to a doxastically justiied belief (or at least a belief that isn’t doxastically unjustiied because of its basis in what propositionally justiied you).14 If the unhappy look on your face when I serve you lentils and cabbage for dinner provides me propositional justiication to believe you don’t like lentils with cabbage, then if I form this belief, based on my perception of your unhappy look, my belief should be justiied. he conservative’s favored notion of reasonableness to believe itself has a doxastic counterpart, viz., reasonable belief. So, if propositional and doxastic justiication are related in the way described, the conservative cannot say both that a seeming which is the result of wishful thinking provides you propositional justiication and that the belief based on this seeming is doxastically unjustiied.15 Conservatives would do better to argue that the “bad basis” cases are cases of justiication in the favored sense (both propositional and doxastic) and that intuitions of epistemic inappropriateness pick up on distinct negative epistemic statuses found in the cases. Christopher Tucker suggests that wishful seemings do provide propositional justiication, and so provide routes to doxastic justiication, but they fail to put one in a position to have knowledge. Tucker (2010, 538–42) reports a response from Markie in which Markie contrasts two brains in vats (BIVs), one whose gold-seemings rely on something like perceptual training (imagine that if the brain were embodied, the embodied subject was placed in a normal situation involving gold on Earth and he would be able to recognize gold nuggets) and one whose gold-seemings are wishful. If both rely on their seemings, neither would be in a position to know, or even to believe 14 I thank Siegel for making me see the need for the parenthetical. See her (2013) for an explanation of why further adjustments are needed to the linking thesis, insofar as propositional justiication is localized in a way the doxastic justiication is not. See also the footnote that follows. 15 John Turri (2010) proposes counterexamples to the standard assumption about the link between propositional and doxastic justiication. In one example, a juror bases his belief in the guilt of the defendant on a set of reasons not because those reasons make guilt likely but only because the tea leaves say that they do. In this case, the juror has propositional justiication owing to a certain factor but although he bases belief on that factor his belief is not doxastically justiied. Turri recommends weakening the principle so that what having propositional justiication to believe p ensures is that the subject “currently possesses at least one means of coming to believe p such that, were [she] to believe p in one of those ways, [her] belief would thereby be doxastically justiied” (2010, 320). his is plausible—the basing must be epistemically kosher. However, this minor ix to the standard assumption is not enough for the conservative. In Markie’s gold-digger case, the subject already has the seeming, based on wishful thinking. He would have to re-base it on something else, but there seems no adequate base available. Turri’s weaker link between propositional and doxastic justiication is enough to get the diiculty up and running. ChrisTucker190113OUS.indd 234 5/10/2013 11:46:46 AM he “Bad Basis” Counterexamples 235 reliably, or to have any externalist positive epistemic status, and yet the belief of the BIV who relies on wishful seemings is still epistemically worse. In reply, Tucker appeals to a diference in blameworthiness. he wishful BIV is blameworthy for his false beliefs in a way that the “expert” isn’t. Being blameworthy for getting it wrong, Tucker claims, is consistent with being justiied (reasonable) in one’s belief. Instead of blameworthiness, the conservative might appeal to epistemic statuses less clearly tied to voluntariness, e.g., epistemically vicious belief or bad cognitive functioning. My belief might have these statuses even if I did my best and so I am excused and thus not blameworthy. (Perhaps the wishful BIV gold-digger was trying his very best to be objective.) One could then recast Tucker’s suggestion as follows: in cases in which one’s evidence-assembling processes are biased by the wish to arrive at a belief that P one’s belief is viciously formed but reasonable all the same. he guiding assumption behind approaches like Tucker’s is the plausible one that the following two questions are independent: (1) did one believe reasonably in light of the evidence one had at the time? (2) Was one’s belief based on evidence that was obtained in an epistemically acceptable (non-vicious) way? In the “bad basis” cases in which the subject lacks defeaters, the hope would be that the answer to (1) is always airmative and that claims to the contrary wrongly take the negative answer to (2) to decide the answer to (1).16 So far, so good. he problem is that there are, or could be, cognitively penetrated seemings in which (a) the subject lacks defeaters and (b) the cognitive penetration doesn’t consist in the biasing of evidence-assembling processes but is more direct than this, in a way that seems to matter to reasonableness of belief in light of the evidence one had at the time.17 Consider cases of what I will call “free enrichment” of one seeming by another, due to some cognitively penetrating state. It is natural to understand Markie’s gold-digger case as follows. he pebble looks somewhat yellowish to the gold-digger and the efect of his wish to believe it is a gold nugget is that it looks to him to be a gold nugget. he sizable gap between the yellowish-pebble-seeming and the goldnugget-seeming is closed by the wish, not by any knowledge of the observable features of gold nuggets. he gold-seeming freely enriches the yellowish-pebble seeming (if you pardon the pun).18 Similarly, a natural way to understand Siegel’s anger case is to think of Jack looking at Jill in a certain way which is 16 he independence of the two questions is well explained by Feldman (2003). hus, Siegel (2012) deines the interesting sort of cognitive penetration as follows: there is such penetration if and only if there are cases in which two subjects, with the same distal stimuli and attending to the same things, have experiences with diferent contents, owing to the efects of cognitive states. However, in her second paper, Siegel (2013) seems to want to classify some etiologies involving biasing inluences on attention as irrational. See note 20 for further discussion. 18 I use “free enrichment” only when the content of the resulting seeming “goes beyond” the content of the seeming on which it is based, and goes beyond it in an epistemically problematic way, so that if the transition was between beliefs this would count as “jumping to conclusions.” 17 ChrisTucker190113OUS.indd 235 5/10/2013 11:46:46 AM 236 Dealing with Cognitive Penetration not decisively that of someone who isn’t angry, but that Jill’s belief that he is angry leads her to expect that he will look angry, which in turn gives rise to his looking clearly angry to her. Again, we have one visual seeming being freely enriched by another, due to the presence of a cognitive state, an expectation. In free enrichment, the wish makes one “jump to conclusions.” he jumping is not jumping from one belief to another but from one seeming to another.19 he problem for the “distinguish the epistemic statuses” approach is that when there is such jumping to conclusions within seemings, it starts to seem that the “concluding” seemings aren’t really evidence; they are non-evidence because they are arrived at in epistemically sub-par ways based on the real evidence, the upstream seemings. I think the conservative must admit that free enrichment cases are counterexamples. She might revise her view.20 5. he “Receptivity” Approach When thinking about what it is reasonable for a subject to believe, a certain intuitive picture comes to mind. he subject has certain basic evidence or grounds, which are “handed” to her. Based on what is handed to her, the 19 We need not assume that there is always a temporal gap whenever a person jumps to conclusions in beliefs, and similarly for jumping to conclusions in seemings. What matters in the belief case is that there is the sort of asymmetric dependence associated with inference. Something similar holds for the seemings case (but with a notion of “quasi-inference” developed in what follows). 20 One might worry that I have assumed, too quickly, that the conservative has no trouble with biased attention cases. Siegel (2013) considers psychological experiments in which white subjects perceive anger more clearly in black faces than in white faces which are equally angry and which the subjects would see to be angry if they paid the same attention to them as they do the black faces. In such a case, Siegel suggests it might look to the white subject as though only the black faces are angry. If the subject believes this, though, he wouldn’t be justiied in believing only the black faces are angry. So, doesn’t attention here afect justiication and not merely evidence? Consider possible etiologies of the subject’s experience, his perceptual seeming. Two suggest themselves. he irst is that it seems to the subject that only the black faces are angry because it seems to him that the black faces are angry and it seems to him that the white faces are not angry. he latter seeming could well amount to a “free enrichment” due to positive bias. he brief unfocused scanning of the white faces doesn’t produce any visual seeming that would be enough to support the “enriched” seeming. Here attention isn’t the problem, free enrichment is. A second possible etiology is that it seems to the subject that only the black faces are angry because it seems to him that the black faces are angry and the subject is also aware that it doesn’t seem to him that the white ones are. Here, again, the problem isn’t attention. It’s the insuiciency of the combination of the angry-black-faces-seeming and the awareness of lacking an angry-white-faces-seeming to justify the subject in believing that only the black faces are angry. We would need to add to this combination further knowledge or justiied belief in the subject about his being suitably attentive to the white faces, but it is hard to see how that belief would be justiied in the experiments in question. I would hazard that if they were justiied well enough, then the resulting belief would be justiied as well. I should add that one way cognitive penetration via free enrichment might be implemented is by the subject’s focal attention. Perhaps Markie’s gold-digger focuses, because of biased attention, on those aspects of his experience that could provide some support for thinking the stone is gold. Still, the problem is that the support is not good enough. ChrisTucker190113OUS.indd 236 5/10/2013 11:46:46 AM he “Bad Basis” Counterexamples 237 subject “makes” something of her situation—she draws conclusions about about how things are. Given this picture, when we ask whether something provides foundational justiication, we are asking about whether the basic evidence justiies it—not whether something the subject makes of that evidence would or does justify it. he standard way of illing out this picture assumes that all experiences are part of one’s basic evidence and that what one “makes” of one’s basic evidence consists only of one’s beliefs and other doxastic attitudes. As Sosa writes: Experiences are able to provide justiication that is foundational because they lie beyond justiication and unjustiication. Since they are passively received, they cannot manifest obedience to anything, including rational norms, whether epistemic or otherwise. Since unmotivated by reasons, they can serve as foundational sources, as regress-stoppers. When they help explain the rational standing of some other state or action, they do not thereby problematize their own rational standing. Being so passive, they have no such standing. (2007, 46) he possibility of freely enriched perceptual seemings challenges this standard way of illing out the basic picture. Such seemings are experiences, but they are also part of what we make of the basic evidence, rather than parts of that basic evidence itself. hus, they cannot provide foundational justiication. Let us say that a transition from a seeming that P to a seeming that Q is “quasi-inferential” just in case the transition that would result from replacing these seemings with corresponding beliefs that P and Q would count as genuine inference by the person.21 (he beliefs here might need to be thought of as utilizing demonstrative or phenomenal concepts.) I claim that quasi-inferential transitions between seemings function epistemically in the way inference by the person does: they can at best transmit the relevant epistemic property of the inputs to the outputs; they cannot generate this property for the outputs when it isn’t possessed by the inputs. In the case of inferential transitions between beliefs, the epistemically relevant property that can be transmitted is doxastic justiication. In the case of transitions between seemings, the property is justifying the subject in believing its content. Suppose I transition from a belief that P to a belief that Q through inference, solely basing the belief that Q on the belief that P. If the input belief (P) isn’t justiied, the output (Q) isn’t either. Moreover, if the inference isn’t a good one—if P doesn’t support Q suiciently for the subject—the output belief is not justiied. When all goes well, the output belief is justiied, but only 21 he strategy of characterizing epistemically signiicant features of seemings in terms of the corresponding features of beliefs I borrow from Siegel. Again, the terminology of “transition” might suggest temporal succession: irst there’s one seeming and then it produces another. What matters, though, is an inference-like dependence, rather than any temporal relation. ChrisTucker190113OUS.indd 237 5/10/2013 11:46:46 AM 238 Dealing with Cognitive Penetration derivatively, not foundationally. Similarly, suppose I transition from a seeming that P to a seeming that Q, through quasi-inference, solely basing the seeming that Q on the seeming that P. If the input seeming doesn’t justify the subject in its content (P), then the output seeming cannot justify the subject in believing its content (Q). If the quasi-inference isn’t a good one—if the content of the input— P—doesn’t suiciently support the content of the output—Q—for the subject, then the output seeming cannot justify. When all goes well, the output seeming can justify, but as with inference, only derivatively, not foundationally.22 My way of picking out “quasi-inferential” seemings-transitions is modeled on inferential belief-transitions. To have a more intrinsic characterization, we would have to ask what makes a transition between beliefs an inference by the person and then apply this to seemings-transitions. he following are plausibly necessary: (i) the input and output states (beliefs in this case) must be mental states of the person, not merely of a sub-personal system; and (ii) there must be an explanation in terms of the person’s own mental states that “rationalizes” the transition, i.e., that allows us to see the transition as the person’s treating the content of the input state as supporting the content of the output state. One sort of explanation might appeal to the person’s grasp (perhaps good, perhaps faulty) of the support the one proposition provides for the other. Another might appeal to the person’s background “information” (true or false), and where the possession of this information could amount to know-how. A third might appeal to cognitive states that make us “jump to conclusions,” including expectations, desires, moods. No doubt there are cases in which several of these factors jointly explain the transition. So far, I have limited quasi-inferences to seemings-transitions. However, some seemings might be based on beliefs in a way that resembles inference. his presumably happens regularly for non-perceptual seemings: because I believe one thing, another thing can seem to be true. So, let us allow that belief-to-seeming transitions can be quasi-inferences, too, so long as, were the inal seeming replaced with a belief, this would be genuine inference by the person. Here, too, the output seeming would provide, at best, derivative justiication, not foundational justiication.23 22 Seemings that are quasi-inferred from other seemings have some sort of normative epistemic standing in that they either count or fail to count as providing justiication depending on whether the quasi-inference was a good one. hus, one might be required in Broome’s sense to believe P when it seems to one that P and one lacks defeaters even though one “ought not” have that seeming in a normative epistemic sense. his observation holds out hope for explaining why it might seem “reasonable” in some sense for the wishful-seeming gold-digger to believe he sees a gold nugget—it seems reasonable given his gold-seeming, but that gold-seeming itself is normatively inappropriate. hus, although his belief isn’t reasonable in the normatively oomphy sense, it is rationally required by the seeming. Here I anticipate the receptivity proposal, developed below. 23 hus the idea of quasi-inference is of some help in resisting the charge, lodged against PC, that it treats even inferentially justiied beliefs as foundationally justiied. ChrisTucker190113OUS.indd 238 5/10/2013 11:46:46 AM he “Bad Basis” Counterexamples 239 Finally, the Kantian terminology: call a seeming that is not an output of a quasi-inferential transition receptive and call one that is nonreceptive. he receptivity proposal is to take receptive seemings to be sources of justiication—to confer foundational justiication—and to take nonreceptive seemings as justifying derivatively if at all.24 If this is what the receptivity proposal amounts to, what reason is there to accept it? Later, I consider a number of diferent sorts of cases to test the proposal’s predictions. he proposal makes predictions insofar as it has implications for which beliefs are justiied in various cases. Depending on whether the beliefs in the various cases are justiied or not, the prediction is either correct or not. I argue that the predictions across a range of test cases are correct, or at least plausible. In each case, as I hope will be clear, the proposal gives us not only a plausible verdict but a plausible accompanying explanation of why the case is or is not a case of justiied (i.e., reasonable) belief. However, there are certain hard cases that suggest the possibility of broadening the notion of quasi-inference further. I discuss these at the end of the section. Let’s start with cases of free enrichment due to cognitive penetration. I take it that the beliefs in these cases are not justiied. he receptivity proposal of course predicts this (it was designed to do so). Take the gold-digger case. he pebble’s seeming somewhat yellowish would not make the novice justiied in believing it is gold (even given background knowledge). So, if, due to the wish to ind gold, this seeming is freely enriched to give rise to the pebble’s seeming to be gold, the gold-seeming is nonreceptive and so incapable of providing foundational justiication. It can at best justify derivatively. But it doesn’t even justify derivatively, because the quasi-inference is not a good one: this stone is somewhat yellowish doesn’t suiciently support this stone is a gold nugget for the subject. As the subject bases his belief on this seeming, and this seeming doesn’t justify, the subject’s belief isn’t justiied. Siegel’s preformationism and anger cases are handled analogously.25 24 My receptivity proposal draws from Siegel’s (2013) work on rationally assessable etiologies of experiences. If we think of an irrational etiology as one such that the resulting seeming cannot provide justiication, I claim that an etiology is rational or irrational just in case it involves quasi-inference, rational if the quasi-inference is good and irrational if bad. 25 Widely discussed psychological experiments by Delk and Fillenbaum (1965) suggest that subjects in certain situations see clearly gray bananas as yellow due to their knowledge that bananas are or normally look yellow (or have looked yellow in the past). here are various possible accounts of why this might be. One treats this simply as a case of making false judgments about one’s experiences: it doesn’t really look yellow to us but we judge it does because of our expectation that it will because it’s a banana and they look yellow. here is no quasi-inference involved on this account. A second account understands such cases to involve free enrichment of one color experience by another, due to the knowledge that it is a banana and beliefs about bananas being or looking yellow. his knowledge would not close the gap between its looking rather grayish and its looking yellow At best the knowledge would be part of another justiication for thinking the banana is and looks yellow. A third sort of account, closer to that suggested by Hansen et al. (2006), locates the inluence of the knowledge earlier in visual processing, prior to any visual seemings-transitions. ChrisTucker190113OUS.indd 239 5/10/2013 11:46:46 AM 240 Dealing with Cognitive Penetration By allowing that quasi-inferred seemings justify derivatively, the receptivity proposal gets cases of good quasi-inference right.26 Consider a version of homas Senor’s (2005) sunset example. In the evening, when I view the beautiful sunset over the Missouri River, I believe (and know) it is a sunset based on its looking like a sunset (let’s suppose). It looks like a sunset to me in part due to my knowledge that it is evening together with more upstream seemings. Here the content of my upstream seemings does suiciently support that it is a beautiful sunset for me. It does so, because of my background knowledge that it is evening. For someone else who didn’t know if it was evening or morning, the contents of the upstream seemings would not provide good enough evidence (assuming that sunsets and sunrises look alike). he receptivity proposal relies heavily on the personal/sub-personal distinction. I included as the “intrinsic” properties of quasi-inferential seemings-transitions the following properties of inferential belief-transitions: the seemings must be seemings of the person and the transition between them must be explainable by mental states of the person. I of course admit that there are vague cases of the personal/sub-personal distinction.27 However, there are clear cases, and in these cases the receptivity proposal’s predictions seem plausible. I consider some examples. Suppose my visual system produces a perceptual state in which it seems to me that one tree is farther away than another on the basis of “cues” such as these: facts about diference in the angle of convergence between the two eyes when ixed on the center of each of the trees, facts about the relation of the trees’ retinal images to the horizon line, facts about occlusion, etc. Presumably, I do not go through this sort of calculation when I see one of the trees as farther away; rather, some part of my sub-personal visual system does. Similarly, when I hear a sound as coming from a certain direction, I am not computing information about wave interference created by the fact that the head is an obstacle; but in ongoing theories of auditory perception, our auditory perceptual systems do make such computations. he explanation for the output seeming in these cases is not given in terms of personal mental states. hese are not quasi-inferential transitions. he receptivity approach therefore predicts that the resulting seemings, because receptive, provide foundational justiication. his seems correct. Suppose a sub-personal visual or auditory system, for some reason, “jumps to conclusions” in its calculations. his by itself does not afect whether the person’s belief is reasonable. It would afect 26 hanks to Chris Tucker for discussion here. He saved me from several mistakes. If a heuristic is at work but is not easily accessible to consciousness and is not available for central processing, is the person relying on the heuristic or rather the sub-personal system? When I anchor on a prime (say, my SSN), and use it in estimations of populations of cities, say, am I relying on a belief about my SSN as a reason? Or consider practical reasons: when I prefer items on the right, am I relying on it’s on the right as my reason? 27 ChrisTucker190113OUS.indd 240 5/10/2013 11:46:46 AM he “Bad Basis” Counterexamples 241 reasonableness, though, if the seeming produced by the system depended for its justiicatory power for the person on the quality of the sub-personal inference. Consider, also, our reactions to perceptual illusions. We do not think that someone unaware of the potential for these illusions is unreasonable to trust appearances; and our assessment would not change if we learned that what is responsible for the illusion is some overgeneralization on the part of the visual system, an instance of “jumping to conclusions.” In other cases, a perceptual system might employ information that the person does possess but the employment of which by a perceptual system does not count as the person’s employing it. For instance, we all know that lighting usually comes from above. However, if you are in a room with strange lighting, in which it is unclear from which direction the light is shining, you might still see an object as convex, as bulging outward, because of shading in its lower parts.28 Even assuming there is a seeming-transition in this case, from one personal seeming state (e.g., about shading) to another (about convexity), what explains this transition is not your possession of information about the direction from which the light is shining or usually shines, but rather your visual system’s possession of this information.29 Supposing you had not read up on this feature of the perceptual system, wouldn’t it be reasonable for you to think that the object was convex, despite having reason to doubt the lighting is from above? One might worry about how the receptivity proposal fares when we consider seemings resulting from the exercise of learned perceptual skills. Suppose through Audubon society training I learn to recognize pine warblers by sight. When I see one and it seems to me to be a pine warbler, I believe it is one on this basis. Suppose I don’t do any conscious reasoning. My eyes scan the bird, no doubt focusing on certain features of the bird, and as a result it seems to me that this is a pine warbler. Is this seeming receptive or not? First, suppose there is a seeming-transition here: the bird seems to be a pine warbler as a result of its seeming to have certain other features which it is hard to articulate, its seeming to have a pine warbler gestalt. Is this a quasiinferential transition? If it is explained by appeal to the person’s knowledge of the observable features of pine warblers, the answer is yes. Suppose someone, by virtue of this knowledge, based a belief that the bird is a pine warbler on the belief that the bird had such and such gestalt. his would count as inference, I think. When these beliefs are exchanged for seemings, the result is a quasiinferential seemings-transition and so a nonreceptive output seeming. hus, on the receptivity proposal, the output seeming can at best justify derivatively. 28 See Braisby and Gellatly (2005) for discussion. It’s surprising to learn about how perception of convexity and concavity works. Compare this to “learning” that one perceives someone as a police oicer because of the person’s uniform. 29 ChrisTucker190113OUS.indd 241 5/10/2013 11:46:46 AM 242 Dealing with Cognitive Penetration It plainly does justify, because the quasi-inference is a good one: the input seeming justiies one, and one’s know-how makes it the case that the content of that input suiciently supports the content of the output seeming for one. his verdict seems plausible: the birder who believes the bird is a pine warbler based on quasiinference is justiied, and justiied in part because of her perceptual know-how.30 Suppose, instead, that there is no seemings-transition at all. What would account for the justiication of the resultant belief? I am not sure how to answer this question. Perhaps we can make sense of seemings as being based on experiences with non-conceptual content, in such a way that the transition from these experiences to the seeming counts as close enough to be “inferential.” If so, we might hope to explain the output seeming as non-receptive. But in lieu of that, if the seeming comes from nothing like inference by the person, we would have to say that the seeming is receptive.31 On the receptivity proposal, this seeming would then be classiied as providing foundational justiication. Either way, there is no danger of ruling the birder’s belief unjustiied.32 30 If we follow Turri, we might think that more is required for quasi-inference to result in a justifying seeming. Not only must the input seeming have this property, and not only must the quasiinference be a good one, the actual transition must be based appropriately on the quasi-inference. So, in a case in which there is a good quasi-inferential relationship between a seeming that P and a seeming that Q, and in which the seeming that P can justify the subject in believing that P, the seeming that Q might not justify the subject in believing that P because the person’s transition from one seeming to the other was not appropriately based on the quasi-inferential relationship. So, suppose a gold-digger has a visual seeming in which the stone does look quite yellowish, shiny, etc., and in which the quasiinferential relationship is a good one for the person, but still the gold-digger makes the transition because of wishful thinking. If we take a Turri-style view, we will say that the resulting gold-seeming does not justify the gold-digger. 31 Top-down inluences can have the result of increasing our capacity for having rich receptive seemings. Goldstone (2010) notes experimental evidence for a number of top-down inluences on relatively early perceptual processing. he categories in our conceptual repertoire, he claims, afect perceptual unitization (the development of single functional units that are triggered by complex conigurations in the stimulus) and dimensionalization (the development of distinct functional units triggered by diferent elements of a stimulus). he former makes possible quicker and more eicient processing of a scene. he latter makes possible selective attention in experts, where no such attention is possible for most of us (e.g., color scientists who can attend to saturation and brightness separately). See Goldstone (2003, 2010). 32 What if the pine-warbler-seeming is the output of a quasi-inferential transition, but the transition is made not because of the person’s learned identiicatory skills but because of a mental state masquerading as learned identiicatory “know-how”? Would the output seeming justify? To determine this, we would have to examine the diferent possibilities for these masquerading imposters. Imposters might include unreliable dispositions resulting from poor training (e.g., the disposition to see something as a cardinal bird if one sees it as yellow and small) and reliable but unlearned dispositions (e.g., those that result, say, from “last night’s neurosurgery” as Jack Lyons (2011) imagines). In either case, these are not “know-how” but at best “belief how,” if you will. Whether the resulting seeming justiies will therefore depend on whether such “belief how” makes it the case that the content of the input seeming suiciently supports the content of the output seeming for the subject. It presumably does not if the “belief how” counts as unreasonable but may if it counts as reasonable. A key question is whether doxastic conservatism can be defended. If it can, these “beliefs how” may well count as justiied. If not, they will not. I leave this matter open, for diferent phenomenal conservatives to decide diferently. See the appendix of this chapter for a brief discussion of doxastic conservatism. ChrisTucker190113OUS.indd 242 5/10/2013 11:46:46 AM he “Bad Basis” Counterexamples 243 We have considered cases in which there is quasi-inference. he receptivity proposal handles these cases well. We have also considered cases in which there is neither quasi-inference nor top-down inluences in which cognitive states afect the goings-on at the sub-personal level which lead to the output seeming. Here, too, the receptivity proposal seems to do well. What of the remaining class of cases, cases in which a person’s cognitive states inluence the sub-personal level? Some cases in this category might be cases of what Siegel calls a-rational causation. To take a fanciful case, suppose my thinking about the number 354.333 causes certain changes in sub-personal processing prior to my visual experience. his ought not afect the capacity of the experience to provide me justiication; the experience is evidence for me, despite its a-rational and epistemically unpropitious etiology. he harder question is what to think about cases in which the cognitive penetration operates in such a way that the resultant seeming appears to be action-like, i.e., appears to be related to beliefs/desires/intentions as actions are to cognitive states in Aristotelian practical inference, in practical inference in which the conclusion is an action. It seems at least conceptually possible that wishful seeming could work this way: it might seem to one that P because one intends it to seem to one that P, and one intends it to seem to one that P because one wants to believe P and knows the way to get oneself to believe that P is to make it seem to one that P. (he story about how such Aristotelian practical inference produces the seeming might refer to efects on sub-personal inferences or on the inputs to such inferences, or both.) Suppose, fancifully again, that I could, just by willing, afect sub-personal goings on in such a way as to produce in myself gold-seemings (or other seemings that would be appropriately connected to gold-seemings). Suppose I could do it in such a way as to rid myself of the memory that I did it and to ensure that I lacked other defeaters. I do it. Am I reasonable to believe the thing I see is a gold nugget? On the receptivity proposal as it stands, the answer would be yes. It would be yes because the seeming is not quasi-inferred. We could modify the receptivity proposal so that it gave an answer of no. We could say that when a seeming is the result of Aristotelian practical inference—one whose inputs are beliefs, intentions, desires and whose output is a seeming as “action”—this too is a kind of quasi-inference. It would be a kind of quasi-inference that, because of its practical nature, cannot transmit (epistemic) justiication from any beliefs involved.33 We could make such amendments. hey would answer to the feeling that there is something considerably epistemically worse when one afects one’s seemings in this way. here is also some independent theoretical plausibility to the amendments. Ater all, if a 33 hanks to Susanna Siegel for discussion here. ChrisTucker190113OUS.indd 243 5/10/2013 11:46:46 AM 244 Dealing with Cognitive Penetration seeming is the conclusion of a piece of practical reasoning, then it is “made” by the person and not part of what the person is “handed.”34 One worry about making this modiication is that we might get certain cases wrong. Compare the standard BIV case with the case of a person who makes himself a BIV by, say, taking a pill, or even by willing himself to be one, if one can imagine this. In the standard BIV case, someone else acts intentionally to produce the subject’s experiences (acts by afecting inputs to sub-personal processes). Intuitively, in the standard BIV case, the intentions and plans of the deceiver do not prevent the subject’s experiences from justifying. hose experiences count as evidence, albeit epistemically unfortunate evidence. In the self-made BIV case, much is the same except for the fact that one has brought the situation on oneself, either with the help of a pill or through sheer willing. It seems to me, however, that one is no less reasonable in one’s beliefs in the self-made BIV case—assuming all defeaters are eliminated, of course—than in the standard case. In the self-made case, as in the standard case, the subject has epistemically unfortunate evidence, and not merely unfortunate in externalistic respects but in ways that indicate epistemic vice or improper cognitive functioning. Still, it is evidence all the same. Presumably, in the self-made BIV case the subject’s previous desires/beliefs are related to the resulting seemings by Aristotelian practical inference (if not, why not?). So, while we could alter our account of quasi-inference to allow for quasi-inference via Aristotelian practical inference, there are reasons not to do so. I leave this matter open. 6. Conclusion So, what is the best option for phenomenal conservatives in the face of the “bad basis” examples? Appealing to defeaters and to the diference between the favored notion of justiication (what it is reasonable to believe) and other epistemic statuses (e.g., blameless belief, virtuous belief, good cognitive functioning) can take the conservative some of the distance. Yet these approaches do not save her from counterexamples involving free enrichment. If Markie’s gold-digger seems to see gold because his wish to ind gold leads him to “jump” to this seeming-state from his seeming to see something yellowish, then his resulting belief is not reasonable, contrary to PC. PC is refuted by such cases. However, we may revise PC to state that one is prima facie justiied in believing P when one has a receptive seeming that P. In free enrichment cases, the output seeming isn’t receptive; it isn’t part of one’s basic evidence 34 he sense of “made” intended in our explanation of the intuitive picture in our thinking about reasonable belief was intellectual: what was “made” was concluded from, and accountable to, what was “handed” to one. ChrisTucker190113OUS.indd 244 5/10/2013 11:46:46 AM he “Bad Basis” Counterexamples 245 or the grounds one fundamentally is “handed.” It therefore cannot provide foundational justiication. Nor can it provide non-foundational justiication, because it depends on something very much like bad reasoning, viz., a bad “quasi-inference.” Appendix: Receptivity and Doxastic Conservatism Doxastic conservatism (DC) holds that one is prima facie justiied in retaining the beliefs one already has. DC is subject to “bad basis” counterexamples in droves, since belief is obviously afected by all sorts of psychological states of the person. Markie’s gold-digger surely isn’t reasonable to believe that the pebble is a gold nugget just because he does (wishfully!) believe it, even if he lacks defeaters. Is there any receptivity cure for DC’s ailments as we have argued there is, or plausibly is, for PC?35 What is needed is an account of receptivity for beliefs. he following seems to me promising: a belief is receptive, roughly, if it isn’t, at the time in question, something that is the “making” of the person, but rather is simply “handed” to her. A belief is not, at the time, an instance of what one makes of one’s situation if it is not held at that time on the basis of evidence or reasons (broadly construed so as to include non-doxastic states such as seemings). hus, we might propose that a belief is receptive so long as it is not currently held on the basis of evidence or reasons. his would have the implication that perceptual beliefs, intuitive beliefs, introspective beliefs, testimonial beliefs, along with beliefs held on the basis of an inference all count as non-receptive. Receptive beliefs would be chiely limited to beliefs retained from memory. I say “chiely” because a belief that isn’t based on evidence or reasons that simply pop into one’s head would count as receptive by the account under consideration. Some will ind this objectionable. However, here, as in our defense of the receptivity proposal for PC, there is a defense: this belief is part of what is “handed” to one at the time, not of what one makes of what is handed to one, and what is handed to one can provide foundational justiication. hese ideas are only a irst sketch. Problems arise with DC that do not arise with PC. For instance, when they are not based in a quasi-inferential way on other seemings, seemings make a good candidate for being basic evidence. However, it is harder to see how a state of belief could be evidence for its content. How could one’s take on how things are be evidence for oneself at that very time that things are that way? So, to the extent that the intuitive idea of “what is handed to one” should be understood in terms of one’s basic evidence, 35 My earlier defense of doxastic conservatism (McGrath 2007) appeals to an expanded concept of defeat to handle such cases. I now think this answer is not suicient. ChrisTucker190113OUS.indd 245 5/10/2013 11:46:46 AM 246 Dealing with Cognitive Penetration it would be harder to think of any beliefs as handed to one. I leave the sortingout of such issues for a later occasion. Acknowledgments An earlier version of this chapter was presented at the 2011 University of Arkansas Epistemology Workshop. I thank audience members for helpful discussion. I am grateful to Alex Jackson, Peter Markie, Ted Poston, Tom Senor for helpful comments. I want to express particular gratitude to Chris Tucker and especially Susanna Siegel for their trenchant criticisms of the main ideas of the chapter. References Audi, Robert. 1993. he Structure of Justiication. Cambridge: Cambridge University Press. Braisby, Nick and Angus Gellatly, eds. 2005. Cognitive Psychology. Oxford: Oxford University Press. Broome, John. 2002. “Practical Reasoning.” In José Bermùdez and Alan Millar, eds., Reason and Nature: Essays in the heory of Rationality, 85–111. Oxford: Oxford University Press. _____. 1999. “Normative Requirements.” Ratio 12: 398–419. Dancy, Jonathan. 2000. Practical Reality. Oxford: Oxford University Press. Feldman, Richard. 2003. Epistemology. Upper Saddle River, NJ: Prentice-Hall. _____. 1988. “Having Evidence.” In David Austin, ed., Philosophical Analysis, 83–104. Dordrecht: Kluwer Academic. Goldman, Alvin 1988. Epistemology and Cognition. Cambridge, MA: Harvard University Press. Goldstone, Robert L. and Andrew T. Hendrickson. 2010. “Categorical Perception.” Interdisciplinary Reviews: Cognitive Science 1: 65–78. _____. 2003. “Learning to Perceive while Perceiving to Learn.” In Ruth Kimchi, Marlene Behrmann, and Carl R Olson, eds., Perceptual Organization in Vision: Behavioral and Neural Perspectives, 233–78. Mahwah, NJ: Lawrence Erlbaum. Hansen, horsten, Maria Olkkonen, Sebastian Walter, and Karl R. Gegenfurtner. 2006. “Memory Modulates Color Experience.” Nature Neuroscience 9(11): 1367–68. _____. 2001. Skepticism and the Veil of Perception. Lanham, MD: Rowman and Littleield. Jackson, Alexander. 2011. “Appearances, Rationality and Justiied Belief.” Philosophy and Phenomenological Research 82: 564–93. Kolodny, Niko. 2005. “Why Be Rational?” Mind 114: 509–63. Kunda, Ziva. 1999. Social Cognition: Making Sense of People. Cambridge, MA: MIT Press. _____. 1990. “he Case for Motivated Reasoning.” Psychological Bulletin 108(3): 480–98. Lyons, Jack. 2011. “Circularity, Reliability and the Cognitive Penetrability of Perception.” Philosophical Issues 21: 289–311. Markie, Peter. 2006. “Epistemically Appropriate Perceptual Belief.” Nous 40: 118–42. McGrath, Matthew. Forthcoming. “Dogmatism, Underminers and Skepticism.” Philosophy ChrisTucker190113OUS.indd 246 5/10/2013 11:46:46 AM he “Bad Basis” Counterexamples 247 and Phenomenological Research. _____. 2007. “Memory and Epistemic Conservatism.” Synthese 157, 1–24. Pollock, John. 1975. Knowledge and Justiication. Princeton, NJ: Princeton University Press. Pryor, James. 2004. “What’s Wrong with Moore’s Argument.” Philosophical Issues 14(1): 349–78. _____. 2000. “he Skeptic and the Dogmatist.” Nous 34: 517–49. Senor, homas. 2005. “he Epistemological Problems of Memory.” Stanford Encyclopedia of Philosophy. http://plato.stanford.edu/archives/fall2009/entries/memory-episprob/. Siegel, Susanna.2013. “he Epistemic Impact of the Etiology of Experience.” Philosophical Studies 162: 697–722. _____. 2012. “Cognitive Penetrability and Perceptual Justiication.” Nous 46: 201–22. Silins, Nicholas. 2008. “Basic Justiication and the Moorean Response to the Skeptic.” In Tamar Szabó Gendler, John Hawthorne, eds., Oxford Studies in Epistemology, Vol. 2, 108–41. Oxford: Oxford University Press. Sosa, Ernest. 2007. A Virtue Epistemology. New York: Oxford University Press. Tolhurst, William. 1998. “Seemings.” American Philosophical Quarterly 35: 293–302. Tucker, Chris. 2010. “Why Open-Minded People Should Endorse Dogmatism.” Philosophical Perspectives 24(1): 529–45. Turri, John. 2010. “On the Relationship between Propositional and Doxastic Justiication.” Philosophy and Phenomenological Research 80(1): 312–26. White, Roger. 2006. “Problems for Dogmatism.” Philosophical Studies 31: 525–57. Wright, Crispin. 2004. “Warrant for Nothing (and Foundations for Free)?” Aristotelian Society Supplement 78(1): 167–212. ChrisTucker190113OUS.indd 247 5/10/2013 11:46:46 AM