Zombies and the Phenomenal Concept Strategy Chair: Richard Brown Presenter: Dave Beisecker, University of Nevada -Las Vegas Dave’s paper or a larger version of his video Commentator: Sam Coleman, University of Hertfordshire A larger version of Sam’s video Advertisements Share this:FacebookTwitterEmailRedditPrintPinterestTumblrLinkedInLike this:Like Loading... Related 30 Comments I very much enjoyed the paper, as it is a refreshingly new and more radical reply to the zombie argument than what we find on the market. But I also think that it is not radical enough. I would suggest that the paper be developed not as a defense of the phenomenal concept strategy, but rather as an argument that this whole business of whether zombies are conceivable or whether conceivability entails possibility is in fact humbug, and the whole debate is a pseudo-debate, grounded in cheap definitional postulations. The key passage in the paper, in my view, which I would like to see a bit developed, is this: “It should be clear that Chalmers’ contention here gets virtually all of its mileage from definitional fiat. And the proper reply on behalf of the type Z phenomenal concepts strategist should be equally apparent. Fine; so be it. If zombies are simply defined as non-human (or not having our consciousness), then I’m perfectly content to give up the term.” The zombie argument is cheap, in other words. I can’t agree more! Consider DOMBİES, defined as creatures that are physically exactly like us, but lack digestion. Replace “zombies” with “dombies” in the zombie argument. What would we say about it? We would say that either dombies are not conceivable, because digestion is defined in physical-functional terms, or that the possibility of dombies only shows that the concept DOMBİE does not have anything to do with the concept of DİGESTİON, so not only the possibility of dombies is accepted, but that we are actually dombies (DOMBİE would mean in that case something like: physically identical to us but lacking some function that we also lack to begin with, anyway). The same we can say about zombies and CONSİOUSNESS, according to Dave Beisecker, and I think it makes perfect sense. My view now is that the conceivability premise can only hold given our knowledge of our own concepts, and that such knowledge implies knowledge of our ontological commitments by using those concepts. Conceivability claims are never prior to knowledge of the component concepts. And knowledge of such concepts is never prior to knowledge of the ontological commitments of the concepts when used in an assertion. Given this, one can’t assert that zombies are conceivable so as to derive dualism unless one already knows that the concept CONSCIOUSNESS is non-empty and corresponds to something that is not physical; i.e. without begging the question against physicalism. Contray to many people (those who don’t really do history of philosophy), I suppose, I don’t think, for instance, that Descartes was a dualist because of the conceivability of disembodiment, but that he believed in the possibility of disembodiment, because he was a dualist, i.e. because he DEFINED mind as thinking substance and body as extended substance, and believed, without any modal-epistemological or modal-metaphysical humbug, that thinking is distinct from being extended. So I wonder whether the more radical view is something you sympathize with. Hi Istvan, and thanks for being the only person so far to comment on our exchange! I’ve got two points: 1. It seems that if you think Dave B’s argument is better turned against the zombie argument itself that you agree with me that it really isn’t a defence of the phenomenal concept strategy at all. That right? 2. I don’t get the dombies thing, so I wondered if you’d explain it a bit more. It seems that dombies are straightforwardly inconceivable, as Chalmers would point out, because digestion is going to be easily captured in physical/functional terms. On the other hand it does seem conceivable, absent any special pre-theoretical commitment about its nature, that physical/functional duplicates could lack consciousness. That’s the force of the zombie argument. I just don’t see how Chalmers is only getting that intuition going through some definitional trick. Fine, deny the intuition, then you’re a type a physicalist. But I’m not seeing the distinctive position here that says zombies are conceivable, but only by cheap definition. I’m still struggling to understand the key move I think. Help! Hi Sam, Regarding your first point: yes, I don’t think the paper is a defense of the phenomenal concept strategy. Regarding your second point. The key lies in what you say when you say: “It seems that dombies are straightforwardly inconceivable, as Chalmers would point out, because digestion is going to be easily captured in physical/functional terms.” So it looks like whether dombies are conceivable is posterior to whether digestion is analyzable in physical/functional terms. This means that one’s justification for thinking that zombies or dombies are conceivable comes from one knowing whether consciousness or digestion is analyzable in physical/functional terms. How can we know such facts as whether something is analyzable in terms of something else other than by (a) postulation (“digestion is defined as ….”, “consciousness is defined as …”), or (b) by whether one thing is conceivable without the other. If (a), then we have conceivability for cheap, if (b), then we have conceivability as primitive. Option (b) is, therefore, the usable one. But notice that the conceivability or not of some proposition depends on the extensions of the component concepts in each possible world where those extensions are not empty: if one knows what belongs to each concept’s extension in each possible world where that extension is empty (by extension I mean what falls under the concept, or what belongs to the extension of the predicate whose sense the concept is), then one can or cannot conceive some proposition containing those concepts. So conceivability cannot be primitive. Further, knowing what belongs to the concepts’ respective extensions can be of two kinds: (c) by postulation, (d) empirically. The first cheap again, so let us focus on (d). Take zombies again. If (d), then in order to know wether zombies are conceivabe we have to know what things belong to the extension of “is conscious” in each possible world where that extension is empty. But how do we know which at which worlds they have their extensions empty? There is just no way to know, I think. We can’t say “is conscious” has a null extension in worlds without consciousness”, because we were suppoosed to explain consciousness via the extension of “is conscious”. More importantly, we can’t say that in zombies worlds the extension of “is conscious” is empty! We can only say it if we go by the cheap way of just postulating that that world is not conscious. If we don’t wanna go cheap, we have to accept that we have no reason to think that the zombie world is different from our world in any way. Finally, why I think this not good for the physicalist either is the following. What the physicalist does by replying as above is to take the whole zombie argument, turn it into a conditional (by conditional proof) and then, assuming physicalism, derive the concşusion that we are not actually conscious (in the sense required by the zombie argument). Here it is more formally. The zombie argument says: 1. There is consciousness. (the extension of “is conscious” is actually non-empty) 2. It is conceivable that P and there is no consciousness. 3. It is possible that P and there is no consciousness. 4. Physicalism is false. The Dave B. type physicalist puts forward the following argument. 5. If there is consciousness. (the extension of “is conscious” is actually non-empty), then Physicalism is false (conditional proof) 1-4) 6. Physicalism is true. 7. Therefore, there is no consciousness (the extension of “is conscious” is actually empty) I think this is as cheap as the zombie argument, and we have a “debate” about one proposition: whether the actual extenison of “is conscious” is empty, which, I predict, is cheap again, as it depends on what you postulate consciousness to mean. This is roughly half of my “radical” view (the other half is about whether there is a real debate about whether conceivability entails possibility – you can guess my answer). Hi Sam and Istavan, Thanks for the starting the ball rolling in my absence. Sorry I didn’t made an earlier appearance. This weekend, a non-virtual reality kept me from dropping in on our virtual proceedings. I greatly appreciate all of your endeavors to help clarify the position I’m trying to articulate. It’s been slow going, but with your help, I’ll be much further along to finding its proper rhetorical pitch. First, I have a couple of remarks about Sam’s very nice commentary. I’ll send another message about Istavan’s remarks. Sam does a much better job than I of laying out Chalmers’ 2007 argument against the phenomenal concept strategy. Chalmers finds it conceivable that there be creatures that are just like me/us in all the relevant material aspects (particularly with respect to the material basis for my/our grasping and deploying phenomenal concepts), yet lacking our epistemic perspective. As Sam puts it, it is conceivable that “C without E.” From this, Chalmers concludes that it follows that having C must not really account for E after all, and so we are led to conclude that the phenomenal concept strategy must leave something out. If I have it right, Sam suggests that the problem that I ultimately have with Chalmers’ argument is that as a zombie, I must find it hard *conceiving* of a creature materially indistinguishable to me, yet wholly lacking my epistemic perspective. So it would seem that in the end I must be endorsing some sort of Type-A position. (Am I getting this right, at least in its very broad strokes?) Now the problem I actually have with Ch’s argument here is not about what he says he can or cannot conceive of. Rather, just as in the zombie argument, I take issue with the terms he uses to characterize the spaces of conceivability and possibility. In particular, how should we characterize ‘E’ above? If we start from the outset by glossing E as “having our epistemic perspective,” then that only threatens to center us illicitly on a non-zombie, angelic world. Once more, it begs the question “by definitional fiat,” for the issue to be joined is what the nature of our epistemic position is – specifically, whether or not it is one of acquaintance with a super-material reality. So suppose instead that we pick a less loaded, more topic-neutral way to unpack E, which does not presuppose what our actual epistemic position is, but rather picks it out as supermaterialists understand it to be. For instance, let’s characterize E as “having an acquaintance with a super-material realm.” The first premise of the target argument now looks like this: (1) I find it conceivable that there be creatures materially just like me, which have an acquaintance with a supermaterial realm. I have no complaints about this premise, when thusly cast. There is a slippage between our material and phenomenal concepts that allow such conceptions (and, I would argue, possibilities) to persist. That, by the way, is the reason I think I’m pursuing a Phenomenal Concepts Strategy, though one with a non-standard slant that doesn’t seek to block the move from conceivability to possibility. But now I beg you to notice that when we cast Chalmers’ conclusion in a similarly cleansed vocabulary, it turns out being that having C – the material basis for deploying and grasping phenomenal concepts – cannot account for having an acquaintance with a supermaterial realm. With THIS conclusion, I most heartily concur (and so, BTW, would Sellars, since there is nothing known by acquaintance!). The Sellars story grafted onto the end of the paper is meant to show precisely how having C need not commit us to the more alarming, question-begging way of interpreting E. If portions of this prior paragraph sound quite a bit like Sam’s own suggestions at the end of his commentary, I would have no objections. I think that when this is all hashed out, the position Sam was offering and my own are going to converge. I just have tons more work to do trying to stop folk from pegging me as Type A. Ugh! I think I’ll stop there for the moment. I hope at least some of it was intelligible to y’unz. For my part, I found Sam’s recasting of the Chalmers 2007 argument, along with his subsequent challenge, very helpful. Glad you could join us Dave. Ok, I’m trying to get your position clear; I obviously didn’t get that near it in my reply. First, on my reading of your position: since you embrace the second horn, but deny that zombies differ with regard to E, it seemed to me that you were denying the claim that such beings are even conceivable. That’s the modality operating across the whole of Chalmers’ master argument dilemma. That’s why I labelled you type A. If you were just saying that IN FACT zombies would have E, that didn’t address Chalmers’ argument here, since it is directed at concerns having to do with conceivability and explanation. Now, leaving that behind and trying to start afresh: Why do you think that Chalmers is going to accept that he’s committed to 1? As I understand it, his E is already stated in topic-neutral terms. The only thing assumed is what you grant: that whatever we are, we are conscious, and so our phenomenal beliefs are true, justified and so on. cheers Sam Hi all. I agree with Sam that it’s easy to read Dave B’s paper as type-A. In fact, if we took Dave B at his word in saying that his views are that zombies are actual, it follows immediately that he is an eliminativist about consciousness and therefore is type-A. Here it’s worth stressing that the zombie hypothesis is just “physical duplicates without phenomenal consciousness”, and not “physical duplicates without X” where X is phenomenal consciousness as some theorist (e.g. a dualist) understands it on their theory. So if I’m a zombie, I lack consciousness by definition. I suspect that Dave doesn’t really intend this eliminativist consequence, and therefore doesn’t really intend to say that zombies (as specified here) are actual. But then he needs to give some other reply to the conceivability argument (as specified here), and I’m not sure what that reply is. (Incidentally Paul Skokowski’s paper “I, Zombie” gives a similar “zombies are actual” reply, and I think falls victim to a similar reply. But Dave and Paul are in good company — in correspondence David Lewis took a similar line!) Like others, I don’t really see that the phenomenal concept strategy plays an essential role in Dave’s paper. One point of intersection is Dave’s deflationary view of our epistemic situation with regard to consciousness, which suggests that if he was to flesh out his view with an appeal to the phenomenal concept strategy, he’d then take option 2 of denying that there’s an epistemic gap between C and E. Here it’s worth noting that E needn’t build in theoretical notions such as “acquaintance”, though. In the paper I only invoke notions such as true and false beliefs, justification, and cognitive significance. Maybe Dave thinks that when E is put this way, we can’t conceive zombies without E, or maybe he thinks that even we don’t satisfy the version of E that I say we satisfy. Again, some elaboration would be useful. Thanks all, so here’s another go at elaboration: I still just can’t shake the thought that something funky is going on with the definition of zombie, and that suspect inferential liaisons built into the concept surreptitiously commit those who use the notion toward supermaterialism. And I think this infelicity also infects Ch’s conception of E – the epistemic position of non-zombies. It just isn’t that theoretically inocuous. Perhaps this is a better way to express my overall hang-up. Let’s go back to the passage I quoted in the paper, wherein David states that zombies just cannot share our epistemic position (both intuitively, and by definition). In the portion I didn’t highlight, David acknowledges that zombies might have something analogous to phenomenal consciousness, schmonsciousness, that could ground the epistemic propriety of much of their talk that’s apparently about their own conscious states. As I see it, that means that there are two notions of *onsciousness in play (and two corresponding epistemic positions), one appropriate to thoroughly material beings (as Chalmers understands his zombies to be), and the other appropriate to supermaterial ones. And the question of interest at this point is how we *ought* to think about our own concepts of phenomenal consciousness (whether or not we do so think). Whether the anti-materialist conclusion of the zombie argument follows turns on which conception of consciousness we should adopt. Now Chalmers seems to insist that our own notion of *onsciousness simply must be the one that is more appropriate for the non-materialist conclusion (or at least one that is not amenable to the materialistic conclusion). But that just isn’t so obvious to me, and I don’t know how Chalmers is going to sustain that claim without making a substantial and controversial commitment to what the world is actually like. In other words, this is where Chalmers illicitly centers our own world on an angelic one. [I might add that as an inveterate externalist, I’m going to be deaf to David’s repeated insistence that our own concept of phenomenal consciousness simply has to be the one that is appropriate to a supermaterial world.] Echoing a point Istvan made in the very first posting (No, I haven’t forgotten you, I!), I’m not interested in challenging David about the space of conceivability or possibility (everybody just loses those battles anyway!), just the terms by which we describe those conceivable/possible worlds. Are zombies conceivable? Well, yes – if you insist on speaking in a way that’s prejudicial to supermaterialism. I can make sense of what the supermaterialist saying, though in my careful moments I wouldn’t exactly put the conception in quite those terms. Are they possible? Again, yes – they’re us! Though again in my more careful moments, I wouldn’t exactly call us zombies. David is right; I’m no eliminativist. As I prefer to understand things, I’m phenomenally conscious (though the supermaterialist might prefer to call me phenomenally schmonscious –and so not conscious by their lights). Again, I can perfectly well understand what the supermaterialist is trying to say here, though I wouldn’t express it that way. The key to the type Z proposal I’m pushing, then, is to insist (pace Chalmers) on keeping two sets of books regarding the contents (or inferential liaisons) of conceivability and possibility claims: one which reflects how the supermaterialist prefers to interpret them and the other corresponding to how the (type Z) materialist thinks our concepts ought to go. Once we do that, I think the type Z materialist can perfectly well sustain the twin claims of being phenomenally conscious (by type z lights) and also zombies (by dualistic lights). I suspect that what gets most type A and B materialists in trouble is that they don’t keep two sets of books. Even in their denial of the conceivability or possibility of zombies, they take on the vocabulary of the zombie argument without seeing fully the trouble it buys for them. This does begin to sound like the kind of thing that David Lewis might have said – I’m thinking here of “Scorekeeping in a language game”), so I’d be interested if this any of this overlaps the course of the correspondence with Lewis to which Chalmers alludes (too many David’s here!). I haven’t here touched upon or pursued the phenomenal concept strategy, so I suppose I’ll have to accept the general consensus that my paper isn’t really about the PCS. Point taken – and I apologize for the distraction. At the same time, I still resist the type A characterization, at least how I understand it. I’m no behavioral or functional reductionist. My suspicion has been all along that since I’m granting conceivability and possibility to supermaterialists, I must be challenging the third premise of the zombie argument. Is that correct? So what am I? [ – wait, don’t answer that!… or at least be nice when you do.] I think that all this stuff about the “supermaterialist conception” is irrelevant to my arguments. The arguments are just using “phenomenal consciousness” the ordinary way: someone is phenomenally conscious if there’s something it’s like to be them (no dualism built in, no claims about acquaintance built in, and so on). I think that you think we’re phenomenally conscious, in the same ordinary sense in which I think we’re phenomenally conscious. Furthermore, I think that for the ordinary notion of phenomenal consciousness, it’s conceivable that there are physical duplicates of us without consciousness. And for general philosophical reasons, I think that conceivability (of the relevant sort) entails possibility. From these two premises plus the claim that we are phenomenally conscious, the falsity of materialism follows. As a materialist who accepts that we are phenomenally conscious, you’ll presumably reject one of these two premises. My guess is that you’ll reject the first premise: at least, you’ll say that in the sense of “phenomenally conscious” in which it’s true that we’re phenomenally conscious, physical duplicates without consciousness are not conceivable. If so, that makes you a type-A materialist (and makes you answerable to the objections to that view). If not, then presumably you’ll deny the second premise, and you’ll be a type-B materialist (ditto). I guess I’ll jump in here; Chalmers is probably the one that will raise the most useful objections to my way of working things out, so if he’ll humor me I’d be much obliged. I generally agree with Dave that the zombie argument is question begging. My reasoning runs something like this: The “supermaterialist” claim is that our phenomenal consciousness cannot be explained by a purely physical account. Consciousness is realized in the “something extra” that zombies allegedly lack. I’d like then to consider the conditional “If we are conscious, then we are not purely physical creatures,” or something that says roughly that, to be the supermaterialist claim. I’d like to point out that the contraposition of this claim is “If we are purely physical creatures, then we are not conscious.” It seems to me that zombies are just the embodiment of this conditional, and so are nothing more than the supermaterialist claim with a name and a little story. Essentially what I, and I think Dave, want to say is that a commitment to zombies really translates into a commitment to supermaterialism. What seems to be weird here is that somehow the possibility of zombies is supposed to rule out materialism as a viable theory of consciousness, whereas I might be appropriately open minded about the nature of consciousness and the possibility of supermaterialism as an accurate account doesn’t seem to rule out the possibility of materialism as an accurate account. I think that this is very much the point that recent attempts at undermining the zombie argument, and here I’m thinking primarily of Keith Frankish’s article The Anti-Zombie Argument, are getting at. So I think that the articulation of why (P1) Supermaterialism is conceivable as an accurate explanation of consciousness. (P2) If supermaterialism is conceivable then supermaterialism is possible. (P3) If supermaterialism is possible, then materialism is false. (C) Materialism is false. fails as an argument against materialism should capture why the zombie argument fails as well. As far as I know, you (Chalmers) have not yet replied to the Anti-Zombie argument, or at least I can’t find the reply if it’s there. I’m very much interested in what you think the response in favor of zombies ought to be. Additionally, I think that if we’re supposed to be working primarily from intuition and we’re using consciousness as nothing more than the what it’s like we find ourselves with the materialist might be entitled to an explanation of why we should accept that zombies would lack consciousness, but I suspect we might disagree here about what it means to really being open minded theoretically, so for now I’ll set that objection aside. Tim: Just a quick point. You say “What seems to be weird here is that somehow the possibility of zombies is supposed to rule out materialism as a viable theory of consciousness.” Why is this weird? If we can read ‘physicalism’ for ‘materialism’ and understand physicalism as the thesis that it is necessary that P–>Q (where P = all the physical facts and Q = all the phenomenal facts), then the possibility of zombies (there is a possible world where P and ~Q) shows that physicalism is false. Do you mean something else by ‘materialism’ perhaps? Maybe you mean ‘epistemic possibility’ when you speak of possibility? Well, if my claim is true and the idea of a zombie isn’t in any really significant way different from a supermaterialist account of consciousness, which is what I was hoping to show with the conditionals, then it’s quite strange that someone would think that the possibility of zombies would make materialism false but recognizing that its possible that supermaterialism holds in the actual world doesn’t mean materialism has to be false. The point being that if we’re appropriately open minded about the nature of consciousness, we’d recognize that materialism and supermaterialism are both possible. However, the possibility of either doesn’t rule out the other, so if the idea of a zombie is really just a different way of expressing the idea of supermaterialism, it’s weird that we’d take that to rule out materialism. Though this is somewhat tangential to the main thrust of the discussion here, I’ve always thought that the type-A materialist should take heart in the famous “mill” example from Leibniz’ Monadology. Recall that Leibniz says: “Moreover, it must be confessed that perception and that which depends upon it are inexplicable on mechanical grounds, that is to say, by means of figures and motions. And supposing there were a machine, so constructed as to think, feel, and have perception, it might be conceived as increased in size, while keeping the same proportions, so that one might go into it as into a mill. That being so, we should, on examining its interior, find only parts which work one upon another, and never anything by which to explain a perception. Thus it is in a simple substance, and not in a compound or in a machine, that perception must be sought for. Further, nothing but this (namely, perceptions and their changes) can be found in a simple substance. It is also in this alone that all the internal activities of simple substances can consist.” This passage is often cited as an early expression of anti-materialism, and of course it is. But it is not directed just at consciousness. Earlier in the Monadology, Leibniz writes: “The passing condition, which involves and represents a multiplicity in the unit [unite] or in the simple substance, is nothing but what is called Perception, which is to be distinguished from Apperception or Consciousness, as will afterwards appear. In this matter the Cartesian view is extremely defective, for it treats as non-existent those perceptions of which we are not consciously aware.” So perception isn’t consciousness (there can be unconscious perception), and yet Leibniz thinks that perceptions cannot be given a physical explanation. Now, I grant that Leibniz may be using these terms in a somewhat different way than we do, but my sense is that a lot of contemporary philosophers (even those who grant that there is a hard problem of consciousness) would say that Leibniz was simply wrong about the possibility of explaining perception in physical terms (if perception is not consciousness, as Leibniz says). If Leibniz is wrong, where did he go wrong? I don’t think he was merely saying that we do not currently understand how our bodies can produce perception; rather, he was making the much stronger argument that no material story could work. I am inclined to say that he was making some sort of conceptual error–perception just is a matter of performing certain functions. I’m still having trouble grasping the position Tim and Dave B. i. ‘materialism and supermaterialism are both possible’ If supermaterialism is possible this seems to refute physicalism: if it’s possible that a physical duplicate could be lacking consciousness, then that shows the physical doesn’t suffice for consciousness. But physicalism claims that the physical does suffice for consciousness. ii. Tim, what do you mean by ‘a commitment to zombies’ in your first post? The relevant commitment, at least to start with, is just that zombies are conceivable. Are you denying that? Or are you asserting that even in allowing their conceivability supermaterialism is assumed? If it’s the latter disjunct, I really don’t see how it works. On the other hand a commitment to the possibility of zombies DOES amount to a commitment to supermaterialism. But this is meant to follow if we grant the premise that what’s conceivable if logically possible. It’s the conclusion of the argument that supermaterialism is true. So are you disputing the conceivability/possibility premise? It sounds like you don’t dispute this. So I’m confused. Hey everyone, nice discussion going on in here! Sam, I am confused by the same thing (re: (i)). The only way I can make sense of what Tim and Dave B. are saying is by hearing it as the claim that both zombies and shombies (my term for a physical duplicate of me that has consciousness in just the way that I do) are epistemically/prima facie conceivable even though only one is metaphysically/ideally conceivable…Now, I am sympathetic to this point (as it is the point of my presentation, but I get the feeling thatthis is not what Dave B. is going for…so, yeah, I’m confused too. Does anyone know of any recent papers that discuss Putnam’s “Mind and Machines” in light of the zombie arguments? In that paper, Putnam argued that the “Mind-Body” problem would arise for any computing system capable of answering questions about its own structure. The argument seems clearly relevant to zombies, yet I never see anyone address it (pro or anti-zombie) in the recent literature. Hey Martin, have you checked out Derek’s talk? I don’t know if he was inspired by putnam or not, but he is (I think) trying to make a similar argument. Very interesting (if tangled!) discussion. Let me have a go at some remarks I take to be in the spirit of Dave B. (Hey Dave B! Are you getting the blog equivalent of cock-eyed stares?): There is something wrong with the Zombies’ epistemic situation, such that either we must attribute to them massive and pervasive error, or they will show different behavior from us. That may indeed show Zombies to be inconceivable. But it may show that Zombies are conceivable once we fix our epistemic conception of them, via Sellars. Sellars shows how true first-person reports might work for Zombies. But then it seems plausible that we are in the same epistemic situation as Zombies–there’s no acquaintance, as Sellars showed. We access our mental states just as they do. But, says Chalmers, we are by definition not in the same epistemic situation as Zombies. We know we are phenomenally conscious (in the “ordinary” sense of the term); they cannot. But what justifies this? If Sellars works equally well for us and Zombies, it must be that Chalmers is leaning on acquaintance to distinguish us from Zombies. That’s the definitional “fiat.” So we ought to say that we are in the identical epistemic situation to zombies. To the extent that we are justified in saying we are phenomenally conscious (in the ordinary sense), so are they. Now, that could mean that we are zombies. Or it could mean that zombies are inconceivable, for Sellarsian epistemic reasons. Dave B, I take it, opts for the former–the only thing consciousness could amount to on Chalmers’s argument is something knowable by acquaintance. Since nothing in this world is knowable by acquaintance, the world with consciousness must be angelic. There is no reason to think this world is angelic, so we are zombies. Postscript on phenomenal concepts: they can then be invoked to explain why we zombies are confused about this: our phenomenal concepts “come apart” from our physical concepts. That explains the intuition that we might be angels. But we should reject that intuition once we read Uncle Wilfird. (Apologies for length and for any egregious misreadings! I hope I have not made things worse!) Wow, things have been busy here since I last looked in! These remarks are primarily in response to David’s latest. I don’t think the appeal to “our ordinary notion” of phenomenal consciousness comes so easily, or gets you nearly so far. Why couldn’t such a conception prove chimerical – not only in the sense that the very existence of a unified ordinary conception may be called into question, but also insofar as it exists at all, it most likely embraces an amalgam of disparate ways we may think about consciousness? It just doesn’t follow from the fact that (virtually) all of us would intuitively agree that we are phenomenally conscious that these judgments all flow from a single,unified core conception. Our ordinary notion – if there is one – is radically underspecified (and there are good reasons for its being that way). Now I don’t mean to be dense or stubborn here (well….maybe a little stubborn). I get how the argument is supposed to run. It’s the “From these premises plus the claim we are phenomenally conscious, it follows that materialism is false.” that I always keep hesitating over. For when it is spelled out how that inference is supposed to go, I’m always struck by how our “ordinary notion” of phenomenal consciousness has suddenly been saddled with something quite extraordinary. Thus I’m inexerobly led to the dual sets of books that you think is irrelevant. So how about the following idea? As you know, I’m perfectly happy to grant that zombies – physical duplicates of us that lack phenomenal consciousness – are conceivable (and I follow your general philosophical reasons for thinking they are possible as well). That’s because I generally accept the idea if materialism is false, then a physical duplicate of me would not be phenomenally conscious, and I’m perfectly happy to conceive of the possibility that materialism is false. So the conceivability of zombies now seems held hostage to (or under the scope of) my thinking about consciousness in a non-material way. Thinking about consciousness in such a way is something that is allowed by the “ordinary notion” of consciousness, but it isn’t required. The various “reverse” or “anti”-zombie arguments now on offer similarly exploit different ways the “ordinary notion” allows us to think about phenomenal consciousness. [Hi Richard! I’m glad to see you’re keeping tabs on us.] Thus one can reach the conclusion that dualism is false from the conceivability of zoombies, but such a conception is reached only if we have already locked ourselves into thinking about consciousness “in a material way.” I think the question-beggingness of both conceivability arguments is apparent. Doing the work required to conceive of zombies, zoombies, schombies, anti-zombies or whatever has already locked us into a permissible, yet optional, way of thinking about phenomenal consciousness. It would be a mistake to conclude that materialism, dualism, or whatever simply follows from the ordinary notion. So rather than indexing the different conceptions as “as the materialist is inclined to think about consciousness” or “as the materialist thinks about consciousness” (as I was doing earlier), why not index them as “when we operate under the conception that materialism is false” and “when we operate under the conception materialism is true”? Keep it coming! This is all proving very helpful. It’s very clear to me now that the paper I should have written is quite different from the paper I actually wrote. Martin, Sam, and Richard: Yeah, you might have noticed that I generally talk about “materialism” rather than “physicalism.” And this is largely because I take the latter to be a stronger thesis, which carries implausibly strong modal commitments in its train. The kind of materialist that I am thinking about will try to escape the zombie argument by saying that all the material facts are sufficient for our (schombic) form of consciousness. I don’t think that the materialist further needs to say that the material facts must entail our form of consciousness (angels have a different form along with a different epistemic perspective). Perhaps this lines up somewhat with Sam’s earlier challenge to Chalmers’ entailment account of explanation. Maybe the point is that there isn’t a simple dichotomy between physicalism and dualism. No, Martin, I don’t mean (mere) epistemic possibility – or at least I don’t mean to mean it. Well, don’t I then run afoul Kripkean concerns? If a sensation of red simply is such and such a brain state for us, must it then be the same thing across all possible worlds? No. One upshot of the Sellars unpacking of phenomenal concepts is that they turn out to be descriptively mediated – not rigid designators (although one may always rigidify a phenomenal concept). Here again is one respect in which it isn’t entirely inappropriate to say I’m pursuing some sort of phenomenal concept strategy (that is, using peculiarities in the concepts to explain the explanatory gap), although in an admittedly non-standard fashion. Ah, thanks Dave B., for the explanation. That clears it for me. So from your point of view both are conceivable, both are possible and the only question is which is true of the world we live in. But then it is strange that you say, get how the argument is supposed to run. It’s the “From these premises plus the claim we are phenomenally conscious, it follows that materialism is false.” that I always keep hesitating over. because given the way you use the terminology here that is NOT the argument that anyone gives. The argument they give is “from these premises…it follows that physicalism is false. You seem to agree with that so you aren’t really engaging the argument that Chalmers et. al. have (since it isn’t aimed at you anyways). So what you need to be doing is not giving anargument that the zombie argument is bad. You need to be giving an argument that materialism (in your sense) is more plausible than physicalism (in your sense). In order to do that you have to give up S5 (since it can be proven that identities are necessary in S5), which is pretty drastic (or so it seems to me). Hi Josh! Very kind of you to step in. Yes, you got me right (No surprise there, given how many times I’ve subjected you to this story!). And I do indeed hope that doesn’t make anything worse for any of us…. David’s latest, however, makes me wish I wouldn’t have to lean so heavily on acquaintance. Richard, I agree that there is a whiff of not engaging the zombie argument as it is usually conceptualized (with just the two premises – as David did in his posting, but interestingly, not as you did in your own paper!). If we read the zombie argument as an argument FOR immaterialism (or at least property dualism), then I engage it. If instead we read it as an argument AGAINST an implausibly strong (ultimately type A) physicalism, then I don’t. I don’t think I have to give up S5. I’ve been suggesting that phenomenal concepts are descriptively mediated, and Identities involving descriptions don’t need to be necessary, do they? [Let me add that the “conditional analysis” account of phenomenal concepts strikes me as trying to achieve much the same effect (designating something material in a thoroughly material world and something else in a non-material one), though I think it does so in an appallingly ham-handed fashion.] Oh, sorry for being so dramatic Dave!! I forgot about the descriptional bit. But, I never thought that the zombie argument was itself a direct argument for immaterialism. Rather the argument looks like this: (1) Either dualism or physicalism is true (2) Physicalism isn’t true (zombie arg.) (3) Therefore dualism is true You seem to be question the first premise (what about materialism?) Yeah, the detour through descriptions is crucial. Since it bars me from blocking the move in the zombie argument’s second premise from conceivability to possibility, it’s what keeps me from pursuing a type B line (it also inclines me to be generally sympathetic to Istvan’s earlier contention that the fuss over conceivability and possibility is all such “humbug”). Regarding your little disjunctive syllogism: yeah, I’m trying to articulate a (modest) form of materialism short of that which embodies implausible modal commitments. But in trying to do so, I’d like to hang onto a materialism worthy of the name. I’d rather not fall back onto some spooky (ultimately supermaterial) monism which wimpily gives up trying to say anything substantial about the relationship between the mundanely material and the phenomenal…. or more accurately, between descriptions couched in material and phenomenal terms. So I think the disjunction you take me to be challenging is not between physicalism and dualism, but rather between physicalism and supermaterialism (where the latter encompasses not just dualism, but also unsatisfyingly wimpy monisms.) At least that way of putting things is a little less trivial! Now I think my story achieves just that. In effect, my account of phenomenal concepts states that in thoroughly material worlds, phenomenal states are to be identified with circumstances picked out by particular material descriptions, though those identities need not hold in supermaterial worlds. [Again, the conditional analysis of phenomenal states seems to get the same result, though my account further explains why phenomenal concepts appear to be conditional without building the conditionals themselves into the analysis.] So I suppose that this account should be thought of as some sort of token identity thesis – at least for thoroughly material worlds. All that we need in order to give an account of phenomenal consciousness is to know all the (relevant) material facts and to make one further, materialistic assumption: that that is all there is (in our world). The desired explanation of why certain phenomenal states neatly line up with certain materialistic conditions then falls neatly into place: a certain phenomenal state co-varies with a certain neurophysiological state, because that neurophysiological state is the only plausible candidate as the one that a subject has selected to indicate a certain sensible quality to them. To my mind, other forms of wimpy, anomalous monisms offer nothing in the way of such a tidy explanation of the nature of our phenomenal states. I’ll stop here and start ducking the grenades….[Insert smiley emoticon here. I can’t bear to do it myself.] These sound like only contingent identities between phenomenal and material states. That right? Sam, Contingent identities? In a nutshell, that’s about the shape of it. [Though I’m reminded of the line commonly attributed to Morgenbesser that any position that can be put in a nutshell probably belongs there.] I reckon that contingent identity is what it would have to be, at least if I want to hold onto it as an objection to the zombie argument’s often tacit third premise. And the the quasi-Sellarsian stroy about how phenomenal concepts are tied to descriptions is meant to render intelligible the possibility that a contingent identity position can be both theoretically motivated and also genuinely explanatory. As an analogy, we might acknowledge that there may be possible worlds in which, say, Richard Brown cannot be identified as the organizer of the first Consciousness Online conference, but it would obviously be unreasonable to demand that an explanation of that fact in this world must rule those possibilities out. At the risk of opening up a whole new can of worms, I sometimes think that functionalism is best viewed as a story which lines up certain higher-order concepts with descriptions that may be satisfied by realizers expressed in relatively rigid vocabulary. And so at least some functionalist unpackings of phenomenal consciousness really might be better served construing themselves not as type A positions denying the very conceivabillty of zombies, but rather as denying the third premise that I for one find so unmotivated. I still wonder if that might have been Lewis’ thoughts about the matter. I also wonder how this might be related to your own challenge to the notion of explanation Chalmers seems to have bought into in the conceivability argument, insofar as it also admits of explanation without entailment. Hi. Dave B., I thought I was following this but some of the recent comments left me with an unclear picture. Let Q = phenomenological consciousness we have Q- = whatever consciousness a mathematically/physically describable person would have Q+ = supermaterial consciousness P –> Q- Type-A physicalists would say that Q = Q-, so P and ~Q is not conceivable, even though it seems conceivable since we find it hard to understand Q. Like Dennett they typically say that our intuitive notions about phenomenology are wrong. You seem to be saying that P and ~Q is conceivable (since it is concievable that Q = Q+) But is P and ~Q in fact possible? Only if Q = Q+. You believe Q = Q-, but can’t prove it. If you could prove it, P and ~Q would not be conceivable. Technically, that would seem to make you a type-B physicalist with the position that P&~Q is conceivable, but not possible. Usually the problem is that if P and ~Q is conceivable it’s hard to see why it might not be possible. Here the reason is explicit: In conceiving of P and ~Q, we conceived that Q = Q+. But (as per Dennett) we don’t know what Q is; it could be Q-. A zombie has P and ~Q. So if Q=Q+, there could be zombies. If Q = Q-, there couldn’t. If you define a zombie+ as having P and ~Q+, then there could be such a thing as a zombie+ regardless of whether Q = Q+ or Q-. So I think you are saying we are zombies+. However, I get the feeling I’m missing any number of subtleties 🙂 Hi Dave, Ok, having got hold of the contingent identities stuff plus Sellars I think I can now make sense of it. What a strange view though! I wonder if holding that the phenomenal is only contingently material even counts as physicalism? After all, Dave C believes that our world is entirely physically realised, I think. It’s the possibility of zombies in other worlds that he thinks falsifies physicalism. On this understanding of your view you and he agree about zombie worlds, but you don’t think they falsify physicalism (materialism as you understand it). So does that mean that the only discussion between you guys is really about what counts as physicalism/materialism? I have to say I’d be on his side in this one: it seems to me that if its possible to have all the physical stuff in place that we actually have in place and yet this doesn’t suffice for phenomenal experience then physicalism is false: for it’s false then that the physical way of the world necessitates the phenomenal way of the world. Looks more like dualism to me! As Dave C would I think put it, you’re only going to have material states bound to phenomenal states in this world by natural law. Well that’s the same structure as dualism. Hi Jacques, I like your +/- notation quite a bit. I might well see fit at some point to steal (er, I mean borrow) it! So go back to my Feb. 26, 6:53pm reply to Chalmers. I think your laying out of the situation maps onto it pretty well. [One quibble: when you ask, “But is P and ~Q possible?” I’d answer in the affirmative, but not because Q=Q+, but rather because possibly Q=Q+.] Notice how your +/- device does exactly what David C doesn’t want us to do- namely proliferate and index different possible elaborations of the “ordinary notion” of phenomenal consciousness. Your Q+ is my “consciousness as supermaterialists understand it” (or as I now prefer to put it, “consciousness when we think about it supermaterialistically). Similarly, your Q- is my “consciousness when we think about it materialistically.” Chalmers seems to want to insist that the ordinary conception Q aligns with your Q+, an intuition that folks like Brown challenge in their intuitions that schombies are conceivable. Your zombies+ then are physical duplicates that lack consciousness according to a supermaterialistic conception of it. So yes, my hunch is that we are actually zombies+ (which are also Brown’s schombies), and to be sure, that is probably a little less inflammatory than the claim that we are zombies. The more careful way of putting my conclusion is that Chalmers cannot rule out the possibility that we are zombies+, which I think is all we need to defuse the zombie argument. That way of putting things seems to show that Brown’s attempt to turn the tables on dualism is very close to my own efforts to reverse conceivability considerations, although we still might harbor disagreements about the ultimate spaces of conceivability and possibility. I suspect he’s leery about treading into the very quagmire Sam is asking about. By the by, Sam, very good to hear from you again. Once more, I’m finding that you have a nice way of reducing matters to their proverbial nutshells. So is contingent identity some form of dualism? I don’t see how it could be, since it is at heart an identity thesis. However, I admit that I still owe you an account of how a position asserting contingent identities between the phenomenal and the material could ever offer a satisfyingly materialist explanation or account of phenomenal consciousness. My initial, admittedly deflationist, inclination is to say that it should be sufficient to offer a materialistically acceptable account of how thoroughly material creatures could come to engage in phenomenal (or schenomenal) talk. The rest of the “explanation” comes from the general methodological reluctance to appeal to supermaterial entities or properties. They aren’t impossible, so much as simply unnecessary. Why, other than trying to preserve a sense of mystery in the universe, should we believe in them? We can get by without ‘em. Sigh, that probably wasn’t very satisfactory. I have more work to do on this score. Dave B., thanks for the clarification. You are certainly welcome to use my notation 🙂 You say ““But is P and ~Q possible?” I’d answer in the affirmative, but not because Q=Q+, but rather because possibly Q=Q+.” So you don’t mean to declare that it is metaphysically possible (or impossible), but epistemically possible, which to me seems the same as “concievable”. I think that when you talk about contingent identities, the same thing is going on: you mean that they are epistemically contingent, not metaphysically contingent. I think the zombie argument fails as a “proof” that Q=Q+ because there is no proof that P&~Q is metaphysically possible, even if we grant that it is epistemically possible; that is the whole physicalism/dualism debate. However, to extend the utility of my notation :), there is still at least a strong prima facie case that Q > Q-, which is the “math leads only to more math, while we have qualia that are qualitatively different” argument. This qualia argument is often invoked to shore up the zombie argument, but it should really be disentangled from that. But now without the framework of a “proof” there is more of a level playing field in which the qualia argument must contend with countervailing intuitions such as the schmonciouness = consciousness intuition and the self-stultification argument against epiphenomenalism. Indeed, I think that if we just go with armchair intuitions, only interactionism seems viable. Although the current fashion among interactionists is that it has something to do with quantum wavefunction collapse, I predict that if we ever look inside some guy’s head, we’ll see a small man inside working a whole bunch of levers, and he’s really the one who’s conscious 🙂 Either that or Dennett is right that our intuitions are off base. Well, it seems to me that as long as there is a sense of “consciousness” such that in this sense, (i) I am conscious, and (ii) a physical duplicate of me without consciousness is conceivable, then that’s all that’s needed for the anti-materialist argument to go through (assuming the conceivability-possibility premise, which you’ve been granting, and setting aside your distinction between materialism and physicalism, which I don’t fully grasp). So I think that you have to deny that there is any such sense: there’s at best one sense in which “I am conscious” is false, and another (presumably a functionally analyzable sense?) in which zombies are inconceivable. I take it that something like that may be one possible upshot of the Q+/Q- discussion. That’s an interesting view, but I think it’s pretty clearly a version of type-A materialism. Comments are closed.