Blog powered by Typepad

« Google Scholar self-indulgence | Main | Advertisement: PhD at ANU »

June 07, 2005


Istvan Aranyosi

Hello Dave! A few comments regarding a possible way out for phenomenal-concepts-based materialism (PCM from now on). I would interpret PCM not as affirming that C explains E, but rather as saying that C provides a way for physicalism to accommodate E. In other words PCM is not an explanatory theory for E, but a compatibilizing account of Physicalism and E. (Physicalism: P metaphysically determines any truth, including phenomenal ones like Q. E: It is conceivable that P & non-Q.)

Turning to your argument, we should observe that supposing C stands for a phenomenal property Q, your premises contain C as mentioned, not used --as Frege would say (see "Concept and Object") C appears as object, not as concept. PCM accommodates E within Physicalism by saying that (a) the conceivability of P & ~Q is due to the conceivability of Concept P & non-C, that is to a gap at the level of concepts, at the level where we mention them, and (b) this is compatible with Physicalism. In your argument, P appear as used--as in the zombie argument, where your first premise is that it is conceivable that P & nonQ-- while C, obviously, is Q mentioned.

The fan of PCM can then argue that his point is unaffected by the second horn of your dilemma, because

(*) It is conceivable that (It is not conceivable that P & ~C) & E.

This is enough for the purpose of accommodating Physicalism and E, so the PCM does not need your stronger condition that

(**) It is not conceivable that C & non-E.

In other words C being physically explicable is compatible with C not being explicable by Concept P, which is the same Q's not being explained by P. This way the explanatory gap is compatible with physicalism. The big question is, of course, whether (*) is true...

Your second horn would work against this compatibilist version of PCM only if it was not conceivable that Concept P & non-C. But this is the denial of PCM, it is type A materialism. At the same time, in this situation, in order not to equivocate on P, your first horn would look like:

It is conceivable that Concept P & non-C,

which does not affect Physicalism, since it is at the level of concepts as mentioned, not used. All this depends on whether PCM is a compatibilist rather than an explanatory account of E.


Hi Istvan, I'm not sure I follow all this, but the premises can all be interpreted as using rather than mentioning P, C, and E: i.e. what I say "P&~C is conceivable", this is equivalent to saying "conceivably P&~C". Re the compatibilist strategy, see my discussion of this under option 1 (which also applies to option 2). The upshot there is that my argument doesn't rule out a compatibilist version of physicalism (where qualia and phenomenal concepts are physical despite there being an explanatory gap for both), but that the phenomenal concepts strategy itself can't do any work in supporting the compatibility and rebutting anti-materialist arguments. That work will have to be done elsewhere.

Damon Woolsey

Greetings, Mr. Chalmers. I love your papers by the way, you are quite an inspiration to me.

I've come up with (what I think is) a rather good argument against physicalism. I've posted it here in bare-bones form; the premises are supported in my paper, Non-Linear Supervenience and the Mind [link:]
NLS Supplement: my explicit argument against physicalism and in favor of dual-aspect theory

1. The physical domain is causally closed (observation)
2. We observe functional taxonomies in the world that are not physical taxonomies (even though they both describe the same "stuff")
3. Physics cannot explain the functional taxonomies (otherwise they would be physical taxonomies)
4. Functional taxonomies cannot explain functional taxonomies (due to vicious circularity)
5. Therefore, physicalism is false (since it cannot explain functional taxonomies)
6. Functional taxonomies must supervene on something else (they are not fundamental)
7. The explanation for functional taxonomies is in the minds of those who observe them (there is nowhere else to look for their explanation)
8. Functional taxonomies cannot explain the mind (by 4 & 7)
9. The mind cannot be fundamental (it is the most complex thing in the known universe)
10. The physical and the mental must both supervene on something else (by 5, 6, 8, & 9)
11. There is only one fundamental substance (by simplicity)
12. Dual-aspect theory is true (10 & 11)


Hi Dave, is this paper going to appear anywhere else? I think it is a substantial position paper on a very widely held view and it shouldn't be buried in an anthology. It's also quite elegant.



Thanks, Gregg. It will also appear in a collection of my papers on consciousness, to be published next year. But the Alter/Walter collection is going to be superb. It should be the definitive book on the topic, with excellent papers by Block, Dennett, Jackson, Levine, Papineau, and many others. Also Knut Nordby (the achromat color scientist) on what he thinks about the Mary scenario, and Larry Nemirow responding to the many responses to his ability hypothesis.

qm sceptic
Let P be the complete microphysical truth about the universe: a long conjunctive sentence detailing the fundamental microphysical properties of every fundamental microphysical entity, across space and time. Let Q be an arbitrary truth about phenomenal consciousness: for example, the truth that somebody is phenomenally conscious (that is, that there is something it is like to be that person), or that I am experiencing a certain shade of phenomenal blueness.

Many puzzles of consciousness start from the observation that there is an apparent epistemic gap between P and Q: a gap between knowledge of P and knowledge of Q, or between our conception of P and our conception of Q.

Given that "P" is (at best) a description of the quantum mechanical wavefunction of the universe, and as such does not include descriptions of most everyday "classical" physical objects, the fundamental epistemic gap is not between P and Q, but between P and CP - the quantum and classical universe.

I think it is almost certainly true that P -> Q won't be solved until P -> CP is solved. After all, given what we know of P, CP looks like it is a construct of Q.

qm sceptic

Ok, lets make this even simpler:

Let P be the complete microphysical truth about the universe...

Please just tell us what that is. You can't argue about a gap between P and Q until you tell us what P is.


Hi Dave, I've read the argument of the paper and would like to make a couple of comments on it.

As I see it, the type-B materialist (i.e., your opponent in your argument) is clearly most likely to take the second horn of the dilemma you present, and in your reductio of that horn, the by far most crucial and vulnerable claim is (as you say yourself) the thesis that a zombie's epistemic situation is not the same as ours. This claim seems to me very much the Achilles' heel of your argument. I think a type-B materialist can quite plausibly reject it - at least, in all relevant respects.

I say in 'all relevant respects' because you seem to understand the term 'epistemic situation' in a relatively broad sense, and through this broad understanding, at least one irrelevant respect is brought into the picture. Namely, you seem to understand 'epistemic situation' in a way that includes not only what beliefs the person or thing in question has and inhowfar they are justified, but also whether or not they are *true*. This comes to bear when you defend said claim by pointing out that the zombie's phenomenal beliefs are false, whereas ours are true, which, on your understanding, makes for a difference in epistemic situation between the zombie and us. But this difference is (if at all) an *irrelevant* aspect of our epistemic situation, since it is not among those aspects of our epistemic situation that the type-B materialist would want to explain with his thesis C. As far as I am aware of the arguments in this area, all that the physicalists want to explain is how we come to have those phenomenal beliefs, and the resulting impression of an epistemic gap. Neither the beliefs nor this impression need to be justified, and they certainly need not be true, according to the physicalist; so, he will have no need either of explaining their being true or even their being justified.

In your defense of the epistemic difference, you are arguing that Mary's "what it's like" knowledge is "substantive", and that "Zombie Mary" could not have this knowledge. Again, your argument here has the form of 'closing the exits', where the exits are what you call the deflationary and the inflationary strategy. I think it's the inflationary strategy that deserves the most attention (because it's the one that I think is actually correct). But for it to work, all that has to be shown is that Zombie Mary, too, can have beliefs of the form "I am in such-and-such now" (as opposed to the merely indexical and thus trivial "I am in this state now"). But it's hard to see how anything could possibly show that she does *not* have such beliefs. You seem to recognize this and consequently argue that a being that has such beliefs is simply not how we normally conceive a zombie to be. Yet you can do so only by calling those beliefs part of an analog to consciousness (which you call 'schmonciousness'), and argue that all you need is the conceivability of beings that have no such analog. But what if the analog is simply (though very roughly speaking) the inner physiological workings of the being in question? How is it conceivable that a physical duplicate of a human being does not exhibit the same kinds of physiological mechanisms?

You may think that this kind of reply is blocked by your remarks on "certain naturalistic theories of the mind: perhaps a functionalist theory of belief, a causal theory of mental content, and/or a reliabilist theory of knowledge", since you say that "to appeal to these theories in this context is to beg the question". But does it? I wouldn't think so, since all we really want of a belief, for theoretical purposes, is that it be the sort of thing that can be true and justified, and that it plays a certain psychological role when brought together with desires and other beliefs, etc. I don't see where phenomenal states or experiential qualities should be needed here. Of course, you say that "consideration of the Mary situation and related matters gives us good reason to believe that consciousness is relevant to matters such as mental content and epistemic status", but as far as I could see, this consideration of the Mary situation only goes to show (if at all) that Mary's belief are *truer* than Zombie Mary's, i.e., those considerations don't say anything about those aspects of the epistemic situation that are actually relevant (as I've argued above). So, it seems to me that you can only rely on a sort of 'slippery slope argument', à la: "if phenomenal states are relevant for this aspect of epistemic situations, they'll be relevant for the other aspects as well", but it goes without saying that this is an argument that the type-B materialist would immediately reject.


But it's hard to see how anything could possibly show that she does *not* have such beliefs.

How does a zombie possess any beliefs?


How does a zombie possess any beliefs?

Well, can you assume he/she doesn't? Of course, I shouldn't really presuppose that zombies do have beliefs, either, but it will do if we cast our talk about zombies' beliefs in terms of judgments, which zombies will uncontroversially be capable of having.


**Why did my latest comment get deleted??**

I'll just paraphrase what I wrote in the deleted comment. A 'belief' is a term given to a mental disposition, which zombies, by virtue of a lack of an inner mental life, don't possess. If & when I say that a robotic arm 'believes' it is picking up a ball, it's just a figure of speech where I'm mirroring my modelling of a homologous action by me, a *conscious* agent. Same for zombies, who are just physical machines who happen to look & behave like a human.


A 'belief' is a term given to a mental disposition, which zombies, by virtue of a lack of an inner mental life, don't possess.

Well, there are certainly competing opinions on what constitutes a belief and what doesn't. For example, a functionalist would say that a belief is a functional state, and nothing precludes the possibility that zombies have the same functional states as we do.

If I understand you correctly, you are arguing along the lines: to be a zombie means not to have a mind, a fortiori no mental states; beliefs are mental states; hence, a zombie cannot have any beliefs. So far so good, but I'd presuppose a slightly different definition for zombiehood, namely one according to which to be a zombie only means not to have phenomenal consciousness (that's after all how, IIRC, Dave defines the term himself), which leaves it open (at least initially) whether or not zombies can have beliefs. So, the challenge you and Dave would have to answer is how you can make a convincing point for the thesis that having beliefs requires phenomenal consciousness. Or in fact, since we do not even know beforehand whether phenomenal consciousness is just another physico-functional property: the challenge is how you can make a convincing point for the thesis that having beliefs requires something that a human being does not already have by virtue of his or her physico-functional makeup. My point, when I said above in my reply to Dave that "it's hard to see how anything could possibly show that [Zombie-Mary] does *not* have such beliefs", was that it's presumably not possible to answer this challenge.


My comment has vanished, twice. Once more...

a functionalist would say that a belief is a functional state

Agreed. But a mental functional state.

namely one according to which to be a zombie only means not to have phenomenal consciousness (that's after all how, IIRC, Dave defines the term himself), which leaves it open (at least initially) whether or not zombies can have beliefs.

'Mind' is just the totality of the objects of phenomenal consciousness. No pheno-consc, no mind. Including the (modelled) activities that interpolate between conscious states as part of the 'mind' is a relatively recent corruption of its meaning.


/No pheno-consc, no mind. Including the (modelled) activities that interpolate between conscious states as part of the 'mind' is a relatively recent corruption of its meaning./

I'm sympathetic to this view, but in the context of Dave's philosophy, we should probably stick to his distinction between the 'psychological' and the 'phenomenal' aspects of the mind that he put forward in the first chapters of The Conscious Mind. And with respect to this distinction, beliefs should probably fall on the side of the ('merely') psychological aspect (for, otherwise, what else would be left to fall there?). But, as I said, if you insist to the contrary that beliefs involve phenomenal consciousness practically by definition, I'm not terribly opposed to this way of thinking. But then, the question would (still) be how you can be sure that phenomenal consciousness does not supervene on physico-functional properties.


Actually, in The Conscious Mind I don't take for granted that beliefs are psychological (i.e. functional states), although I do assume that there's at least a correlative sort of functional state, which I call judgment. I'm now more inclined than I was then to think that phenomenology plays a constitutive role in belief, in which case if phenomenal states aren't functional states, belief states won't be either. This doesn't yet entail that a zombie can't have beliefs, but it does suggest that a zombie won't have the same beliefs as its conscious duplicate, and it might provide some motivation for the stronger thesis. Of course the matter is complicated by the mixing of substantive and terminological issues -- at least some part of the issue turns on terminological issues about how to use 'belief'.

For the purposes of the paper on phenomenal concepts (as I say in the paper), I assume for the sake of argument that a zombie can have concepts and beliefs. In context that assumption is a concession, as if zombie's can't have concepts, then the phenomenal concepts strategy will suffer from an even worse version of the problem I mentioned, involving an explanatory gap between physical processes and phenomenal concepts.

On some previous comments. To QM skeptic: See "Conceptual Analysis and Reductive Explanation" for some discussion of what P is, and "Consciousness and its Place in Nature" (especially the section on type C) for discussion of why the argument that P&~Q is conceivable is robust over different choices of P, even before we know in advance about what P is. Of course there is also a serious issue about whether QP entails CP, where QP is a quantum-mechanical description of the world and CP is a classical description of some part of the world. If not, then we must either (i) say that P goes beyond QP, (ii) deny CP, or (iii) admit fundamental facts over and above P. I sometimes bypass this issue by assuming something like a Bohmian interpretation of QM for the sake of argument, in which case the gap between QP and CP goes away. But even on options (i)-(iii), the argument I give about a gap between P and Q will still go through. Your own preferred option seems to be (iii) with a role for Q (over and above P) in constituting CP. Of course that view itself is a version of dualism, and is quite compatible with my overall picture.

Jplate: actually, in practice I've found that type-B materialists are divided between those who want to take the first horn (e.g. Tye, Block) and those who want to take the second horn (e.g. Papineau, Aydede). I discuss the option of narrowing the "relevant" epistemic situation under option 2 in the objections section. There I give reasons for why it's crucial to the type-B strategy that truth be included in the relevant epistemic situation.

Re deletions: Note that I said in my opening message on this weblog that I will moderate comments for quality and topicality.


djc, you may want to make that fact i.e. discretion to moderate, prominent. The blog's first post is no longer on the front page, and there's no "About" link. I haven't followed this blog for long, so I didn't catch that. I thought my posts vanished because of some erratic software behaviour. Maybe you could post a short comment to the effect that "daksya's comment at 3:25 PM on 08/20 was deleted/edited...etc".


Dave: thanks for pointing me to that section of your paper. I had actually skipped it, since to "Accept that C does not explain E" (the section's heading) is precisely what I would not do. But then, I should of course have taken into account that you have a different understanding of 'epistemic situation' (or 'E') than I have.

The central passage where you explain why you think the truth of phenomenal beliefs is a crucial component of the 'explanatory gap' problem seems to be this:

In the anti-physicalist's arguments, the relevant epistemic gap (from which an ontological gap is inferred) is characterized in such a way that truth and knowledge are essential. For example, it is crucial to the knowledge argument that Mary gains new factual knowledge, or at least that she gains new true beliefs. It is crucial to the conceivability argument that one can conceive beings that lack phenomenal states that one actually has. And it is crucial to the explanatory gap that one has substantive knowledge of the states that we cannot explain. [...] So truth-value is essential to the relevant epistemic gaps. If so, then to undercut the inference from these gaps to an ontological gap, the phenomenal concept strategy needs to show how the relevant truth-involving epistemic gaps are consistent with physicalism.

I'm not sure a physicalist would be very convinced by this. To begin with, your take on the Mary example may seem question-begging. That Mary should gain "new factual knowledge" is surely one of the stronger premises of the knowledge argument (if it is really a premise at all), and a physicalist might rather want to describe the situation by saying that Mary merely starts to exercise a new concept (in the sense of: a new type of mental representation). As for the other examples, however, I'd agree that it's crucial to the problem - and uncontested by most physicalists - that (a) we do actually have phenomenal states and (b) that we have substantive knowledge of them.

That we have phenomenal states and know about them is, of course, crucial to the puzzle insofar as it would be nonsensical to demand an explanation for something that doesn't exist or cannot be known about. But is an explanation of why we can have knowledge of our phenomenal states really part of the physicalist's task, as far as the 'explanatory gap' is concerned? From the paragraph following the passage just quoted, it seems that your answer would be 'yes':

Perhaps proponents could augment their explanation of the narrow epistemic situation with an additional element that explains why the relevant beliefs are true and qualify as knowledge. For example, one might augment it with an explanation (perhaps via a causal theory of reference?) of why phenomenal beliefs refer to physical states, and an explanation (perhaps via a reliabilist theory of knowledge?) of why such beliefs constitute knowledge.

However, this seems to me to give a somewhat distorted view of the physicalist project. On the one hand, there is an "explanation of the narrow epistemic situation", which I take to be the physicalist's explanation of why we have certain beliefs about our phenomenal states that make it seem strange how they should be identical to physico-functional states, or in other words: (part of) the physicalist's reaction to the 'explanatory gap' puzzle. And on the other hand, there are also two further explanations, namely of "why phenomenal beliefs refer to physical states" (i.e., regarding the content of phenomenal beliefs) and of "why such beliefs constitute knowledge" (regarding, among other things, their justification), and you say that these latter explanations "augment" the first one. The reason why this seems a distortion to me is that to talk of the two latter explanations as an augmentation makes it seem as if they were just another part of the physicalist's reaction to the explanatory gap puzzle, when in fact, their role is much more fundamental. Concerns about the content and justification of phenomenal beliefs obviously have to enter at a very early stage in the physicalist's considerations regarding the identification of phenomenal states with physico-functional ones, since no such identification would be any good if the states with which phenomenal states are identified were such that the system in question had no way of knowing about them. In fact, it seems that this concern is in many cases precisely what motivates such an identification. For, it is certainly not for nothing that HOT theories of consciousness practically equate the consciousness of a given state (in the sense of: whether it is a conscious state or not) with (at least certain aspects of) its introspective accessibility. So, far from being an 'augmentation' of the physicalist solution of the explanatory gap puzzle, those explanations form an important and much more fundamental physicalist concern and should therefore be cleanly separated from that solution.

Be that as it may, you go on:

However, such a augmented explanation is now subject to the original dilemma. If such an account applies equally to a zombie (as might be the case for simple causal and reliabilist theories, for example), then it cannot account for the crucial epistemic differences between conscious beings and zombies. And if it does not apply equally to a zombie (if it relies on a notion of acquaintance, for example), then crucial explanatory elements in the account will not be physically explainable.

Now this is something that I'm not sure I understand correctly. "If such an account applies equally to a zombie..." - well, of course it (i.e., the physicalist account of the mind) applies equally to a zombie, since, by definition, a zombie is a physical duplicate of a conscious human being - "... then it cannot account for the crucial epistemic differences between conscious beings and zombies." By this, I presume you primarily mean that it cannot account for the fact that zombies are not conscious, whereas conscious beings obviously are. Well, but what if zombies don't exist, and are not even metaphysically possible? After all, if physicalism is right and phenomenal states are physico-functional states, no physical duplicate of a conscious human being could ever fail to be conscious, and that's by metaphysical necessity. Suppose someone draws up a theory of heat according to which heat does not supervene on molecular motion. According to this theory, it is possible that there are beings kinetically identical to us but whose body temperature would nevertheless be 0 K (call them 'heat-zombies'). Now, the kinetic theory of heat applies of course equally well to heat-zombies as it applies to us, but it doesn't account for the crucial caloric differences between them and us. Does this mean the kinetic theory of heat is wrong?

The point of this example is, of course, a very old one, but I found it useful to illustrate what seems to be the flaw in your above reasoning. Also, I'm aware of the Kripkean objection to this example, but I don't find that objection very convincing (cf. Loar, Phenomenal States; Hill, Imaginability, Conceivability, Possibility and the Mind-Body Problem). So, all in all, I still don't see what's wrong with the physicalist reaction to the explanatory gap puzzle.

Finally, regarding the question of whether a physicalist is more likely to take either the first or second horn of the dilemma: It seems that that depends on what the respective physicalist would say that "C" is, i.e., by what 'key features' he would propose to explain our epistemic situation with respect to consciousness. And this, in turn, depends on what he would think this epistemic situation consists in. In particular, if he thinks that this situation actually encompasses being phenomenally conscious, he may find it much more conceivable that a physical duplicate of a human being might not have those 'key features', thus inclining him towards the first horn of the dilemma. But if he thinks that the epistemic situation only consists in what beliefs one has and how they are justified, consciousness itself would perhaps not be among the features deemed necessary to explain that situation. So, whether a given physicalist would take either the one or the other horn of the dilemma seems to depend on the question of how that physicalist construes our epistemic situation with respect to consciousness. I'm not sure as to how Block and Tye would answer this question, but if they construe the epistemic situation the way I would do it (i.e., leaving consciousness itself out of the picture), I think they might also prefer the second horn.

Simon Floth

The paper has a grammatical error in the third paragraph under the heading 'Option 1: Accept that P cannot explain C':

"Because of this, while the strategy may make some progress in diagnosing the explanatory gap, but it will do little to deflate the gap."

Note that the choice between 'but' and 'while' might be perilous in a Buridan's ass kind of way.

free dating services

Your rough piece of blog is much appreciated. Hope to see the new one very soon.

The comments to this entry are closed.