The Zombie s Cogito: Meditations on Type-Q Materialism. (Forthcoming, Philosophical Psychology)

Transcription

1 The Zombie s Cogito: Meditations on Type-Q Materialism (Forthcoming, Philosophical Psychology) ` Josh Weisberg University of Houston Department of Philosophy 513Agnes Arnold Hall Houston TX jweisberg@uh.edu

2 1. Introduction Philosophical zombies are creatures physically identical to humans lacking phenomenal consciousness, consciousness marked by qualia, the intrinsic qualities of experience. 1 Because we are phenomenally conscious, there is something it is like to be us, for us. 2 There is nothing it is like to be a philosophical zombie for the zombie. All is dark inside. The mere possibility of these creatures stalking the metaphysical landscape may be enough to refute materialism, the claim that everything, conscious experience included, is ultimately physical in nature. Much philosophical ink has been spilled trying to rule out the possibility of zombies, but like their silver-screen namesakes, zombies have proven very hard to kill. They continue to haunt the dreams of materialistically-inclined philosophers, threatening the ontological triumph of materialism. How do zombies threaten materialism? It is generally agreed that if zombies are metaphysically possible, then materialism is false. A successful materialist theory should rule out this possibility. The crux of the zombie problem is that zombies seem to be easily conceivable we have little troubling imagining zombies. And some hold that this also provides reason to believe in their possibility: the conceivability of zombies implies their possibility. With this in mind, the zombie argument against materialism runs as follows: 1. Zombies are conceivable 2. If zombies are conceivable, they are possible 3. So, zombies are possible. 4. If zombies are possible, materialism is false. 5. Therefore, materialism is false.

3 Most materialist responses to the zombie argument reject either premise one or premise two. These two strategies are indicative of what David Chalmers has labeled type-a and type-b materialism, respectively. 3 Type-A holds that zombies are inconceivable because our concept of phenomenal consciousness is fully analyzable in functional terms to conceive of creatures functionally identical to us is conceive of creatures phenomenally identical to us, full stop. Type-B holds that zombies are conceivable, but there is no reliable link between conceivability and possibility in the zombie case; it then rejects the possibility of zombies for empirical reasons. But not all materialist ways of addressing the argument fall so neatly into these categories. Philosophers sympathetic with what Chalmers has called type-q materialism ( Q for Quinean) are dubious of the distinction between conceptual and empirical truth needed to cleanly separate the type-a and type-b materialists responses to the zombie argument. 4 In this spirit, type-q materialists hold that conceivability and possibility intuitions are theory-laden the strength of our intuitions is a mark of the entrenchment of this or that theory, not an insight into the one, true meaning of a term or the essence of some feature of reality. And this has implications for how one responds to the zombie argument and how one evaluates the effectiveness of the response. In this paper, I follow the type-q strategy in attacking the zombie argument by stressing the theory-ladeness of our concepts of consciousness and qualia. 5 I argue that our first-person access to consciousness systematically misleads us into theorizing that the distinctive qualities of conscious experience qualia are nonfunctional. Our conception of zombies is driven by this misleading theoretical picture. We mistakenly posit nonfunctional qualia, and then imagine creatures lacking qualia so conceived. But

4 if first-person experience is misleading about the nature of qualia, then the way is open for a definitive third-personal characterization of qualia, one fully characterizing them in functional terms. And if this theoretical characterization is justified, then it follows that there cannot be creatures physically identical (and so functionally identical) to us lacking qualia. This response to the zombie argument may seem straightforwardly to fall into type-a or type-b materialism. 6 But it crosscuts these categories. On the one hand, it can be seen as claiming that zombies are inconceivable: once we properly analyze the concepts of phenomenal consciousness and qualia, we discover that they are fully functional concepts. But on the other hand, it can be seen as accepting the conceivability of zombies by way of a false theory while maintaining their impossibility. But type- Q suggests that these ways of parsing the theoretical space are spurious. If we accept one theory, zombies are conceivable; if we accept another, they are not. But acceptance of theory is an empirical matter driven by questions of overall theoretical utility. The conceptual and the empirical are thus inextricably intertwined; there is no privileged place from which to make secure a priori judgments about concepts and modality. This strategy I propose is not without its worries, however. It involves positing an appearance/reality distinction at the level of consciousness itself. According to a range of philosophers otherwise differing greatly in their approaches to mind and metaphysics, such a distinction makes no sense it simply gets wrong the data to be explained. The qualities of consciousness just are the appearances; there is no reality to them beyond how they appear. To take a well-worn case, if one seems to be in pain, one is in pain.

5 That s just what being in pain amounts to: things seeming a certain way. There is no space here to open up an appearance/reality distinction. I will argue that, despite prima facie intuition to the contrary, such a distinction can and should be drawn in the case of qualia. I support this claim by appeal to the very creatures threatening materialism. I contend that zombies as conceived by the antimaterialists undermine the assertion that if one seems to be in pain, one is in pain. After all, there is a good sense in which it seems to zombies that they are in pain, even though, ex hypothesi, they are not. Further, given our physical identity to zombies, it s not clear what reason we can have for holding that we can t be wrong about the nature of qualia as well. But if we can wrong about qualia, it s not clear what reason one can have to deny the appearance/reality distinction in question. And this blocks the zombie argument. Though my argument fits well with type-q materialist attitudes, if one is not sympathetic to Quinean challenges to analyticity, a priority, etc. the argument can be viewed as a version of type-a or -B materialism. I think it is no coincidence that much of the debate over zombies involves seemingly intractable disputes about what is conceivable and what is possible; I take this fact as support for Quine s great distrust for these notions. But if one thinks that such notions have theoretical utility, the basic points of my response still hold. But it is beholden on those who invoke these difficult distinctions to provide a decent explanation of what they are talking about and how we are to settle disputes about their application. The paper is structured as follows: first, I give my response to the zombie argument and I develop the reply it provokes. Then I consider the reasoning of Rene Descartes zombie twin, in order to establish the possibility of an appearance/reality

6 distinction for qualia. I then argue that we have no good reason to deny this distinction in our own case. I close by addressing a number of objections, and by fending off a potential reemergence of the zombie argument at the level of the appearances of qualia. 2. An Appearance/Reality Distinction for Qualia Materialism is the view that all things are ultimately physical in nature. 7 A prominent strategy for defending materialism is to provide a reductive explanation for the seemingly nonphysical phenomenon in question. A reductive explanation shows how the target phenomenon is nothing over and above a complex collection of physical parts. The key to a reductive explanation, in the sense intended here, is providing a functional characterization of the target phenomenon, an adequate description in terms of the causal role the thing plays in the world. If such a characterization is available, and if it really does adequately describe the target explanandum, then the door is open to providing a story about how physical elements realize the causal role in question. If that can be provided, the explanation will have shown that the target phenomenon is nothing more than a collection of physical parts playing a particular causal role. 8 For example, a mousetrap can be characterized as an object which reliably takes live mice as input and delivers dead mice as output. Anything that plays this role is a mousetrap; that is what mousetrap means. And if we can show how physical parts could play this role say, by instantiating a gluey substance on a sheet of plastic then we have reductively explained a mousetrap. There is nothing to being a mousetrap over and above having this collection of physical parts playing this causal role. In this way, we locate the mousetrap in a materialist ontology.

7 Reductive explanations seem to be available for most phenomena in the natural world, including psychological states and processes such as belief, desire, memory, and learning. 9 However, there appears to be a problem when we try to reductively explain consciousness. The stumbling block shows up at the first step of reductive explanation. It seems impossible to give a characterization of consciousness in wholly functional terms. Consciousness does not seem to be functional, at least in crucial respects. Consciousness is marked by the presence of seemingly intrinsic qualitative properties, the properties constituting what it s like for a subject in conscious experience: qualia. Qualia, it has been argued, just are not functional. Or at the very least, a functional characterization of qualia will inevitably leave out the intrinsic conscious feel of the experience. And if we cannot provide a functional characterization of qualia, we cannot proceed to locate consciousness in a materialist ontology by way of reductive explanation. This is what David Chalmers terms the hard problem of consciousness. 10 One line of defense for the materialist pursuing a reductive explanation of consciousness is to claim that, despite appearances, qualia can be functionally characterized. The antimaterialist rejection of a functional characterization is based on introspection and reflection on the results of introspection. We access qualia introspectively and then reflect on their functionality, concluding that qualia and function can come apart at least in principle. But must we accept all that introspective reflection has to say about this matter? Perhaps introspection fails to reveal the underlying functional nature of qualia and thus misleads us when we reason about function and conscious feel. So long as the materialist can produce a plausible functional account of the mechanisms of first-person access leading us astray, there is no need to accept an

8 irreducibly nonfunctional element in conscious experience. A reductive explanation of consciousness, one that effectively locates consciousness in a physical world, can then proceed. 11 The materialist is claiming, on this line of argument, that qualia are functional even though they do not seem functional. Viewed from the perspective of neuro- or cognitive science, such a claim is not particularly surprising. Why think we can simply peer inward and grasp the underlying nature of the processes and states occurring in our minds? The last hundred years of psychology, psychiatry, and neuroscience are marked by just how little is available to the introspecting subject and just how wrong we can be about what goes on when we believe, desire, feel, or sense. To think that we should be able to read-off the underlying mechanisms of mind in first-person reflection or that we can trust the clarity of inner vision on this matter is to go against the evidence of a considerable body of empirical research. Qualia, like other aspects of the mind, may not be as they appear. 12 If the materialist is granted the claim that qualia are not as they appear, she possesses a ready answer to the zombie argument. Zombies seem conceivable because we base this conception on the deliverances of first-person reflection. Because firstperson reflection seems to reveal nonfunctional qualia, we can easily conceive of a world where the functional aspects of mind and qualia so-conceived come apart. However, if first-person reflection is not a reliable guide to the nature of qualia, the way is open for an empirically-justified characterization of qualia in functional terms. That rules out the possibility of zombies, despite how things appear.

9 What, then, are we conceiving of when we seem to conceive of zombies? The claim of this paper is that we are conceiving of a world where a false theory is true. First, we take our naïve first-person reflection to be revelatory of the nature of qualia. We reason that because qualia do not seem functional, and because introspection reveals their underlying nature, qualia are not functional. Second, on this basis, we conceive of a physical-functional duplicate to our world where qualia so-characterized are absent. Given that we believe that qualia are nonfunctional, it seems to us that we can imagine zombies nothing in our conception rules them out. But the first step of our reasoning is flawed. Introspection does not reveal the nonfunctional essence of qualia; indeed, it does not reveal their essence at all. And then if our best empirical theory dictates that qualia are functional, it does not matter that they do not seem to be so in introspection. But we can still conceive of a world lacking the things that introspection seemed to reveal. However, that is not a world without qualia; qualia are found a posteriori to be identical to particular causal role properties, so if zombies possess the states playing the right roles, they possess qualia. Zombie worlds are impossible. An analogy from the history of biology may help to clarify the position. Consider an early 20 th century vitalist arguing with a biological mechanist. She might claim that she can conceive of a world mechanistically just like ours, but nonetheless lacking life. And on this basis, she might contend that such a world is possible, and thus that mechanism in biology is false. What is the best way to characterize this case, looking back after the collapse of vitalism and the establishment of mechanism? The right answer, I contend, is that the world she is imagining is impossible: given the facts of our best science, worlds that are minimal mechanistic duplicates are life duplicates. But must

10 we also say that her proposed world was literally inconceivable? That seems too strong. Rather, she reasoned that her vitalist theory was true, and therefore that life required something additional to mere mechanism. On this basis, she conceived of a world mechanistically identical to ours, but lacking an additional property. The imagined vitalism case parallels the situation with zombies. We are able to conceive of worlds based on false or incomplete beliefs; this explains the apparent conceivability of zombies. But discovering which worlds are possible is generally an a posteriori matter, not one of a priori reflection. So the worlds conceived of on this misguided basis are not possible, despite how thing seem to those in the grip of the misguided beliefs. But it might be argued that my vitalism analogy misses the mark with qualia. 13 The vitalists believed that mechanism alone failed to explain various complex functional features of biology. It s not that they thought that biological function was fully explained and that something nonfunctional was missing. This is quite different from the case of qualia, which do not seem to be functional. Thus, the analogy breaks down. I am not sure that this truly captures what the vitalists were thinking, but even if it does, we can consider a nearby view we can label commonsense vitalism. Such a view is quite common in my intro classes and it seems to have some play in various new age holistic books in the self-help and metaphysics section of bookstores. Upon hearing a mechanistic description of a creature meeting all the criteria for life, commonsense vitalists hold that something is still missing. For them, merely performing the right functions is not enough. The spark of life or spirit is missing. We can then consider the reasoning of a commonsense vitalist conceiving of a mechanistic duplicate of our world lacking the additional feature they hold necessary for life. Again, such a world is

11 impossible, given the assumed truth of mechanism, but it seems wrong to say that the commonsense vitalist is failing to conceive of a coherent world. Those of us who have been convinced by the mechanist theory may now find such a world hard to imagine, but that does not mean that we can rule out the conceivability of such worlds to others. And it does not seem incoherent to consider what the world is like if a false theory were true, and then construct a zombie world missing the false theory s posits. However, there is another crucial difference between the vitalist case and the qualia case. Qualia are appearance properties, properties constituting how things seem to us from the first-person perspective. And as such, they do not seem open to an appearance/reality distinction. In that way, they differ fundamentally from the élan vital of the commonsense vitalist. Even if that is a nonfunctional property, it is plausibly not an appearance property. It is instead a feature of the nature of life, not how life seems to us. The analogy breaks down. A wide range of philosophers hold that qualia do not admit of an appearance/reality distinction, even those who differ on the question of the truth of materialism. For example, Saul Kripke, in arguing against materialism, writes, To be in the same epistemic situation that would obtain if one had a pain is to have a pain; to be in the same epistemic situation that would obtain in the absence of pain is not to have a pain (1980, 152, emphasis in original). Further, he writes, In the case of mental phenomena, there is no appearance beyond the mental phenomenon itself (154). In the same spirit, David Lewis, a strong defender of materialism, writes To have pain and to feel pain are one and the same. For a state to be pain and for it to feel painful are likewise one and the same. A theory of what it is for a state

12 to be pain is inescapably a theory of what it is like to be in that state, of how that state feels, of the phenomenal character of that state (1983, 130). 14 How then, can I coherently claim that qualia are not as they seem? Isn t this tantamount to claiming that the essence of pain is not in how it feels, but in some other feature? Indeed, Kripke makes his claim in the context of rejecting materialist mind-body identity. 15 Kripke famously breaks the tie between a priority and possibility. He holds that for a range of terms, the meaning of the term is given by empirical research, not by an a priori graspable set of descriptions. But he notes the importance of explaining away the appearance of contingency for these a posteriori necessities. Kripke holds that we can explain away the appearance of contingency for a posteriori necessities in the following way. Science tells us that water is H2O. However, it may seem to us that we can easily conceive of a world where something else is water, XYZ, say. But in those cases, Kripke claims, we are really conceiving of a world that seems to have water it contains something superficially just like water that nonetheless does not contain water. The distinction between water s superficial appearance and its underlying reality allow us to see that there is no world where water is not H2O. But this maneuver is not available when it comes to qualia. To conceive of a world superficially just like a world where there s pain is just to conceive of a world where there s pain. Pain just is the appearance property, so if that appearance is present, pain is present. We cannot explain away the appearance of contingency, so the contingency must be real. But, claims Kripke, mind-brain identities ought to be necessary. So there is no identity at all.

13 On a 2-dimensional interpretation of modal logic, the same point is made by claiming that the primary and secondary intensions of mental state terms like pain do not come apart. Any world where the primary intension holds the epistemic intension, characterized in terms accessible to the subject of experience the secondary intension picking out pain s essence holds. There is no space on such a characterization for an appearance/reality distinction for pain. And, by parity of reasoning, the same holds true for all terms for qualia. 16 But why should we accept that to seem to be in pain just is to be in pain? Is it given a priori in our grasp of the meaning of the term pain? That seems unlikely. First, it s not clear that to know the meaning of a term is to grasp the essence of its referent. It seems there are lots of terms we can understand and use effectively without knowing a detailed story about its underlying nature. Second, there seem to be coherent cases of both unconscious pains (pains lacking the appearance of pain) and false positives of pain (states appearing to be pain, but are not). We might have a headache all day that we fail to notice at times while we re engrossed in work. Or we might suffer an injury in a dangerous, stressful situation, and only notice that we ve been in pain at a later, calmer moment. And we might mistake another feeling for pain. Hearing the whirr of the dentist s drill might make us mistake the sensation of pressure on our jaw for pain in our totally numbed teeth. Or we might feel the hot scald of water from the tap, only to realize a moment later that the water is cold (and indeed soothing). Here we seem to have cases of pain that do not seem painful and cases that seem like pain but are not. 17 The primary and secondary intensions of pain can come apart.

14 But the proper interpretation of these cases is obviously in dispute in this context. It can be denied that unconscious pains are really pains. Likewise, one can deny the reality of false positives, or claim that in such cases the subjects really are in pain, no matter what they say at later moments. We seem left with bare intuition here. And such intuitions are notoriously subject to influence by our implicit and explicit theoretical beliefs. Is there anything else we can appeal to here to settle the issue? In the next section, I ll consider zombies as a way of supporting an appearance/reality distinction for qualia. While this may seem to move from the frying pan of philosophical intuitions into the fire of outré philosophical thought experiments, there is one hopeful factor here. The enemies of materialism already agree that zombies are conceivable. Thus, the case itself should be uncontroversial. Obviously, when I attempt in later sections to use this fact against the friends of zombies things will change. But then I believe that the choice will be to either drop the zombie case altogether, or to allow for appearance/reality distinction in question. Either way is a win for materialism. 3. The Zombie s Cogito Let us consider a particular zombie and his epistemic musings. This zombie is the molecule-for-molecule duplicate of Rene Descartes. Let us call him Rene Zescartes, or RZ for short. RZ has decided to challenge all his previously held beliefs, in order to determine if any are indubitable and thus suitable to serve as the foundations of his knowledge. RZ runs through a range of increasingly skeptical scenarios, with the goal of isolating ideas so epistemically secure that they can t be doubted. Like his conscious counterpart, RZ considers a malin génie, an evil genius comparable in omnipotent

15 power to God, using all his power to deceive. Despite all this, and even though he doubts such basic eternal truths like 2+3=5, RZ arrives at an unshakeable proposition, namely that I am, I exist, is necessarily true whenever it is put forward by me or conceived in my mind (CSM II 17, A&T 25). 18 Or, as more famously expressed in the Discourse on Method, I think, therefore I am. Cogito, ergo sum. So far RZ has exactly paralleled his non-zombie counterpart RD. And the conclusion seems as secure in either case: it seems just as true that the zombie RZ s existence is guaranteed by his running through this process of reasoning as it does for RD. And given that zombies share all our functional states, and so, all of our beliefs, it s not clear how the RZ could be in error if RD is not. Returning to RZ s meditations (or better computations to remain noncommittal): Like RD, RZ now considers just what he s established. What can he know about himself at this point, assuming the evil genius is still out there using all his deceptive cunning? He reasons that, while he can doubt that he has a body or an Aristotelian soul, he cannot doubt that he is thinking thing that he is a res cogitans. And what does this amount to? Paralleling RD, RZ reasons, But what then am I? A thing that thinks. What is that? A thing that doubts, understands, affirms, denies, is willing, is unwilling, and also imagines and has sense perceptions (CSM II 19, A&T 28, emphasis added). But he hastens to add, Lastly, it is also the same I who has sensory perceptions, or is aware of bodily things as it were through the sense. For example, I am now seeing light, hearing a noise, feeling heat. But I am asleep, so all this is false. Yet I certainly seem to see, to hear, and to be warmed. This cannot be false; what is called having a

16 sensory perception is strictly just this, and in this restricted sense of the term it is simply thinking (CSM II 19, A&T 29, emphasis in original). RZ, under the continuing threat of so-called demon doubt, nonetheless concludes that he seems to see, hear, and feel. Given that RZ is paralleling RD s reasoning here they share the same beliefs if it is reasonable for Descartes to so conclude, it is likewise reasonable for his zombie twin to do so as well. RZ the zombie is as certain of the fact that he seems to perceive as he is of anything. There is no empirical evidence that could shake this belief, nor is there any further rational or nonempirical line of evidence that could do so. Indeed, it seems that it would be irrational for RZ to doubt that he seems to see, hear, and feel. 19 To do so would be to accept what Descartes holds to be beyond the pale, that one could doubt that one seems to have sense perceptions when it seems to one that one does. 20 It is important to note that from RZ s epistemic perspective, there is nothing more that could be done to shake his belief that he is not a zombie. There is no reasonable step he should have taken, no obscure possibility he missed. Just like RD, and like us, he followed the rational method of systematic doubt to its logical and epistemological conclusion. If Rene Descartes has made no error, neither has Rene Zescartes, his zombie twin. Below I will consider challenges to these claims. But take note of what must be denied here: that RZ cannot be certain that he seems to see, even though it seems to him that he seems to see. Indeed, RZ is compelled to assent to these beliefs by the natural light of reason, in Descartes terms. But if RZ can t be certain of this, why think that RD, or the rest of us can? For we are in just that epistemic situation: we find the reasoning at the beginning of the

17 second meditation to be compelling, even if we later part ways with Descartes philosophy. Surely, to doubt something like this is to have left the road of reason; it is as crazy as thinking we are pumpkins, or made of glass (CSM II 13, A&T 19). Zombies share all of our beliefs. 21 If we have no reason to doubt we seem to be having sensory perceptions, the zombie has no reason to doubt that he seems to be having sensory perceptions. But then it follows that we do not know we are not zombies. If the zombie has followed all of our reasoning and shares all of our beliefs, yet still does not have any reason to doubt that he s a non-zombie, despite the fact that he is indeed a zombie, then we cannot rule out the possibility that we are in the same epistemic situation. We might have full empirical and rational confidence that we are not zombies, despite the fact that all is dark inside and we are not in fact phenomenally conscious! But what follows from all this? Let us now consider the claim that there is no appearance/reality distinction when it comes to qualia. If there is such a distinction, then how qualia seem to us in first-person reflection may not be a fool-proof guide to their reality. What are we to make of how things seem to RZ? According to his Computations on First Philosophy, it seems to RZ that he is in pain this he cannot doubt. But ex hypothesi he is not in pain: he lacks pain qualia, there is nothing it is like for him to feel pain. One way to analyze this situation is to hold how pain appears to the zombie comes apart from the reality of pain. That is, he seems to be in pain when he is not. But if we are in the identical epistemic situation to RZ, then it follows that we cannot be sure that how pain appears to us reveals how it really is. There is noting in our epistemic situation that rules out such a possibility. Anything we believe about pain is something our zombie twins believe as well, with the same strength of conviction. It follows that the

18 appearance of pain alone is not enough to determine whether or not the appearance and reality of pain necessarily coincide for us. We could have the same appearances and be wrong about the nature (or even existence) of our pains. Further, there is no possible additional evidence that could tell one way or the other, for us or our zombie counterparts. No empirical result could do so. If there was, the zombies could avail themselves of this information to undermine their belief in phenomenal consciousness. Indeed, the very point of the zombie thought experiment is to show that no empirical evidence could make a difference: physical reality is held constant across our world and the zombie world. But there is also no line of nonempirical reasoning that could make a difference either. RZ, in parallel with Descartes, is under the threat of demon doubt. Even mathematical and geometric truths can be doubted in such a circumstance, according to the Fathers of Modern Philosophy. But neither of us can doubt that we seem to feel, even in this epistemically dire situation. And yet one of us nonetheless does not feel, despite how things seem. But which one of us is it? How could we possibly tell? The zombie s cogito shows us that we can conceive of beings that seem to be in pain but are not. But it also shows that we can t be sure we are not zombies ourselves. So nothing in how pain appears to us guarantees that we are in pain if it did, we (and the zombies) could distinguish our epistemic situations, in contradiction to the zombie story. We cannot be sure, then, that the appearance of pain in us tracks its reality. Something more is needed to close the appearance/reality gap for qualia if the antimaterialist is to resist the line of argument I ve developed here.

19 At this point it is helpful to introduce a useful concept developed by Terry Horgan, John Tienson, and George Graham (2006). They coin the term internal world skepticism for the possibility that we might be wrong about our inner mental world parallel to the way that external world skepticism makes use skeptical of our external physical world. Horgan, Tienson, and Graham (HT&G) note that internal world skepticism fails to grip us at all it just does not seem like a viable possibility, even in the weak sense in which we entertain the idea of external world skepticism. They take this non-grippiness as a datum in need of explanation. What is interesting to note here is that the very conceivability of zombies opens up the possibility of internal world skepticism. If there is no a posteriori evidence or a priori reasoning that can rule out the possibility that we are zombies, then internal world skepticism stands as an open possibility. One reaction to this fact is to reject the conceivability of zombies altogether: internal world skepticism is so implausible as to show that something has gone wrong conceptually with the very idea of zombies. 22 I have some sympathy with this move, but I do not think it is the right way to deal with internal world skepticism. Nor will the antimaterialist, of course. Both the type-q materialist and the anti-materialist have an epistemic story about why we should not be troubled by internal world skepticism they both have theories explaining its non-grippiness. The question becomes, which theory is more plausible? And it is here that type-q materialism comes to the fore. In a Quinean spirit, belief in the existence of an external world of physical objects is best justified by its theoretical utility: it best explains our past, present, and anticipated experiences. 23 But likewise, according to type-q, belief in the existence of qualia is best explained by

20 their theoretical utility. They best explain our past, present, and anticipated introspective experience, and the reliability of our inner beliefs. However, to play this theoretical role, qualia must be functional they have to be the reliable causes of our inner beliefs. So, the best explanation of the non-grippiness of internal world skepticism is the presence of reliable mechanisms tracking the presence of functionally characterized qualia. This is not something read off of our experience; it is the best theoretical explanation of our inner mental lives. In the same way that causal contact with a world of persisting physical objects is the best theoretical explanation of our experience, causal contact with a world of persisting physical qualia is the best theoretical explanation of our inner beliefs. And if qualia are best characterized functionally, it follows that the best explanation of our intuitive resistance to this fact is a failure of first-person access to reveal the functional nature of qualia. The positing of an appearance/reality distinction for qualia, therefore, is part and parcel of the best theoretical explanation of the accuracy of our inner beliefs in general if not about the functional nature of qualia, then about the fact that we re in pain, tasting chocolate, seeing red, etc. and of the non-grippiness of internal world skepticism. This appearance/reality distinction also explains the seemingly easy conceivability of zombies, and this further supports the distinction by helping to establish materialism. Materialism has shown its theoretical value throughout the rest of science; what s more, showing that the mind is nothing over-and-above a complex functional arrangement of physical parts simplifies our overall picture of the world. For the type-q materialist, theoretical considerations of this kind are decisive in

21 selecting between views. Taken together, we have good reason to hold that qualia may not be as they seem. Anti-materialists have an alternative explanation of the non-grippiness of internal world skepticism. Further, this alternative purportedly explains why we know we re not zombies, even though the zombie shares all our beliefs. Anti-materialists appeal to a special noncausal acquaintance we have with our qualia. 24 Indeed, our internal beliefs are in part constituted by qualia, leaving no room for error in our internal world beliefs. Since we are intimately acquainted with our qualia, it cannot be that we are zombies. This acquaintance must be noncausal, in order to rule it out of bounds for zombies. But nonetheless it assures that we know we re conscious and indeed makes this knowledge more secure than any other knowledge we possess. It follows on this view that qualia are as they appear, because we are intimately acquainted with these qualia and they actually make up our fundamental beliefs about their nature. But how plausible is such a theory? Since it is not given in experience itself, the view needs additional support the appearances themselves are consistent with us being zombies. But what could that support be? No empirical evidence can be given in its favor: we are dealing with a noncausal relation. And the view must posit a sui generis link unlike any known in the rest of science. Such a link might exist, but this certainly is not the simplest or most conservative theoretical explanation available. And any additional rational argument seems inconclusive. The same line of reasoning will be open to a rationalist zombie like RZ, so it cannot remove the doubt. Finally, we cannot tell simply by introspecting that our beliefs are constituted by qualia. Zombies believe the same thing, and behave just as we do as a result. Again, if the appearances alone are

22 not enough to settle the case, something else must be appealed to in order to establish the anti-materialist s claim. Given that direct evidence and argument can t settle the issue, we must appeal to more global theoretical considerations. And here the anti-materialist comes up short. The best theory materialism tells against them. For the type-q materialist, this is decisive; for those unsympathetic to type-q s Quinean leanings, the burden is to show what other moves might save the anti-materialist. But even an anti- Quinean might wonder at the pragmatic utility of doing so. I conclude that it makes theoretical sense to posit an appearance/reality distinction for qualia. That opens the way to my rejection of the zombie argument laid out in section I: qualia only seem nonfunctional because of our misleading first-person access; they are in fact functional, and so zombies are impossible. When we seem to conceive of zombies, we are instead conceiving of creatures lacking nonfunctional qualia. Zombies are in fact inconceivable, but for a posteriori reasons, as dictated by type-q materialism. However, there are a number of key places where the anti-materialist can challenge my position. I will present and counter a number of objections in the final section. 4. Objections and Replies Objection: I ve argued that zombies appear to be in pain when they are not. But this is to misdescribe the zombie situation. Things do not appear any way at all to zombies all is dark inside. So it can t be concluded that there s an appearance/reality distinction for qualia in the zombie case.

23 Reply: While there may be a sense in which things do not appear any way at all for zombies, there is still a good sense in which things appear to them. Fred Dretske (1995) notes a distinction between a doxastic sense of terms like looks, seems, appears, etc. and a phenomenal sense of those terms. 25 The doxastic sense picks out an epistemic claim, while the phenomenal sense picks out what it s like for a subject. Zombies ex hypothesi lack appearances in the phenomenal sense. But they plausibly have appearances in the doxastic sense. Doxastic appearances are tied up with rational behavior. And zombies behave just as we do in exactly the same situations. So it s plausible that they share our doxastic appearances that is, things seem the same to them as to us, in the doxastic sense. But this is the sense that matters for the argument. How things doxastically seem to us determines our basic epistemic state. Doxastic appearances fix the basic evidence we have about the world and determine our conviction about what s true and false, given how things seem to us. Since zombies share our doxastic seemings, they will hold the same things to be true or false as we do, with the same conviction. They will cite the same evidence, appeal to the same intuitions and arguments, express incredulity at the same odd claims, and so on. So long as the doxastic appearances are shared, we cannot be sure we are not zombies things seem the same to us and them in the relevant sense. It is interesting to note that phenomenal seemings are shown by the zombie scenario to be epistemically and behaviorally inert. Whether they are present or absent, we will believe the same things and behave in the same way. This means that phenomenal seemings are epiphenomenal or at least replaceable without in any way affecting the beliefs and behavior of the individual. If zombies can share our belief and

24 behavior without sharing our phenomenal appearances, those appearances are rendered oddly irrelevant. Of course, if one holds that qualia are functional (and thus zombies are impossible) this odd dissolving of the phenomenal does not occur. Qualia are functional properties of states playing roles in perception and cognition. We would not believe or behave as we do if states with those properties were absent. But to some it may seem just unbelievable that one could lack phenomenal experience and not be aware of it. Surely we d notice a difference from the inside. However, there are actual cases of bizarre phenomena which are best interpreted as showing that it s the doxastic seemings that matter here. Cases of Anton s syndrome, also known as blindness denial, show that people can demonstrably lack the conscious experience of seeing, but not know it. 26 Indeed, they believe sincerely they are seeing, even thought they are not. In such cases, the doxastic, driven by damage to the brain, trumps the phenomenal. These subjects are vision zombies, but they do not know it. In addition, cases of so-called hysterical blindness also indicate a dominance of the doxastic over the phenomenal. These subjects deny that they can see when there is good empirical evidence that they do. 27 What it s like for these subjects how it seems to them is that they are blind, even though they are seeing. If such cases occur in the actual world, it seems reasonable to assume that someone sharing our doxastic appearances will share all our beliefs about consciousness and zombies. And that will be the case even if they are zombies. Or even if we are zombies. Note that Anton s syndrome and hysterical blindness do not undermine the claim that qualia are functional. Indeed, these case support that claim. Victims of Anton s syndrome cannot navigate the world as we do. They lack the relevant qualia, and

25 therefore, they cannot function as we do. Likewise, subjects with hysterical blindness behave in ways that show they are seeing they react in ways that require vision. And this shows that they possess qualia, even though it does not seem to them that they do. But such cases are particularly hard to explain if there is no appearance/reality distinction for qualia. Objection: Because we have qualia and zombies do not, it is a mistake to say that they share all of our beliefs. We have beliefs about qualia that they lack. And it is these beliefs that underwrite our confidence that qualia are as they appear. 28 Reply: This is to conceive of beings quite different from the usual zombies present in the philosophical literature. These new zombies will lack a large subset of our beliefs. But even if that is granted, zombies must share all our beliefs that relate directly to the production of behavior. Zombies behave exactly as we do in all situations, and so if our behavior is best explained by belief and desire, so is the behavior of zombies, by parity of reasoning. What s more, there must be some explanation of how zombies can do all that we do, even though they lack a large subset of our beliefs. This effectively renders all the beliefs not shared by our zombie brethren epiphenomenal. If the unshared subset made a difference in behavior, zombies would be externally detectable. And that would give us grounds for assigning a causal role to qualia, undermining the point of the zombie scenario. And if the beliefs are epiphenomenal with respect to behavior, they are plausibly epiphenomenal with respect to the causation of introspective belief. Introspective belief is reasonably constituted by its connection with verbally reporting

26 our mental states. 29 Since zombies make the same vocal sounds that we do and, indeed, such sounds are plausibly interpreted in just the same way as our vocalizations, given our identical behaviors it can t be that the unshared subset of beliefs are connected with reporting. And because introspection has a constitutive tie to reporting, it seems that the special subset can t be involved in introspection. And even if we give up the plausible constitutive tie between introspection and reporting, we must posit a set of states in the zombie causally responsible for its reporting behavior. These states have the precise causal profile of our introspective beliefs and they constitute the introspective doxastic appearances for the zombie. The question then remains, how do we know we are not zombies? How do we know that we possess the special subset of qualia-generated beliefs as opposed to their zombie equivalent? All our behavior and our doxastic appearances will be the same either way. What s to tell one way or another? Thus, even if we grant the objection, the defense of an appearance/reality distinction still holds. One last point: it s not clear that the revised zombies proposed here are as easily conceived of as the more standard zombie sharing all our beliefs. It may be, then, that the initial conceivability intuitions driving the antimaterialist argument are undermined by this revision. In any event, it seems less of a stretch for the materialist to deny the conceivability of such creatures. Objection: Our beliefs about qualia are special, in that they are in part constituted by the very qualia they are about. Because zombies lack qualia, they do not possess these phenomenal beliefs, and it is on this basis that we know, in a way that zombies cannot, that we are not zombies. 30

27 Reply: First, even accepting that there are such phenomenal beliefs in part constituted by qualia, the zombie scenario shows that it s at least possible that there are beings lacking such beliefs who nonetheless believe they are not zombies. And they hold this belief with a conviction surviving even Descartes demon doubt. This of course raises the question of how we know we re creatures with real phenomenal beliefs as opposed to pseudo-phenomenal beliefs like those possessed by zombies. 31 So the challenge remains. What s more, there seem to be actual cases of pseudo-phenomenal beliefs in humans. Subjects suffering from Anton s syndrome believe they can see when they are blind. Subjects with hysterical blindness believe they cannot see when they have sight. This suggests that we can mistake pseudo-phenomenal beliefs for phenomenal beliefs. And if such cases are possible in us, what assurance do we have that any of our beliefs about our qualia are constituted by qualia? Nothing about the beliefs themselves reveals this special constitution. This raises the question of what distinguishes Anton s sufferers from normal subjects if we reject the constitution picture. But here, again, we can invoke the failure of normally reliable belief-forming mechanisms and the failure of systems monitoring the constancy of our beliefs to explain the syndrome. It seems that the question of constitution vs. reliable causation is not an empirical one. 32 All the evidence we might have for one view seems to equally support the other. One camp might identify a brain state as one instantiating qualia, and thus argue that phenomenal belief is constituted by qualia. But the other side will argue that qualia are instantiated by the functional processes leading to the tokening of this brain state, thus vindicating the causal-functional

28 picture. But how are we to label the brain states in the first place? This seems to require a prior commitment for or against nonfunctional qualia, a commitment made independently of the physical evidence. From the type-q perspective, this is a matter of theory selection, and for the reasons given above, it favors the causal-functional picture. Objection: The question of whether or not we re zombies just does not arise. This is something that is so obvious as to need no justification, and it is mistake to ask for such justification. Questions need to end somewhere, and this is a reasonable place to draw the line. Thus, the challenge is unreasonable and needs no answer. 33 Reply: One of the points of this paper is to stress that the conceivability of zombies makes the question of how we know we re not zombies legitimate. Indeed, the zombie scenario, by suggesting the possibility of creatures with our beliefs and behaviors that nonetheless lack consciousness is the very skeptical scenario making internal world skepticism a live option. Once the scenario is in play, it is question-begging to say it need not be answered. The materialist hasn t brought in an illegitimate question; rather, the anti-materialist, by raising the zombie issue, has made such questions pressing. Further, by invoking a mysterious non-causal acquaintance relation to explain our knowledge of qualia, the anti-materialist has offered a theory about how we know we re not zombies. The materialist has a rival explanation: that we know we re not zombies because our knowledge of qualia is reliably caused by functionally-construed qualia themselves. It is interesting to note that some anti-materialists (and even materialist antifunctionalists) argue that it s the functionalist who can t explain our knowledge of