Saturday, June 30, 2007

Why do you think you're conscious?

If consciousness is epiphenomenal, then it plays no part in the actual causal explanation of why we believe ourselves to be conscious beings. That seems problematic, or so Peli Grietzer argues in an interesting email [quoted with permission]:
There's one aspect of the Zombie argument that I always felt was a tough bullet to bite, though, and I wonder if you have any thoughts: While various non-causal accounts of knowledge can handle the core-analytic kind of epistemological issues about phenomenal beliefs that arise from epiphenomenalism, I often fear that it leaves us with no decent scientific hypotheses as to the evolution of phenomenal beliefs, other than the amazing luck of humans evolving to believe they are not zombies coinciding with humans really not being zombies.

The obvious counter-argument is that it's no different than the "lucky coincidence" we assume when we reject brain-in-a-vat type sceptical scenarios despite lacking any good probabilistic reasons to do so, but there's a possibly critical difference. At least if our beliefs about the external world are true there is a natural story about how we came to have such beliefs and that story involves the external world, while if epiphenomenalism is true we still have no natural story about how we came to have beliefs about phenomenal consciousness, that [the story] involves phenomenal consciousness.

The technical version of the Zombie argument doesn't commit to epiphenomenalism in the same way the more vivid "zombie world" way of telling it does, though, so maybe it's more of a problem for epiphenomenalism than for the zombie argument.

Chalmers offers a neat answer: "The content of a conscious being's direct phenomenal beliefs is partly constituted by underlying phenomenal qualities. A zombie lacks those qualities, so it cannot have a phenomenal belief with the same content." For example, my concept of phenomenal 'redness' is grounded in the phenomenal quality of redness that I experience. My Zombie twin talks about 'redness', but in actual fact his concept is empty, ungrounded. So he doesn't mean what I do by the term.

On this account, phenomenal qualities (consciousness) can influence what we believe after all. Not causally, of course, but more directly, through constitution. The physical facts alone do not suffice to fix the intentional facts (i.e. what our thoughts are about). Phenomenal properties are part of what it is to have a phenomenal belief -- a belief that's truly "about" those very qualities. So, although a zombie would make all the same noises, their words and cognitive processes (arguably not really "thoughts") wouldn't have the meaning that ours do.

So much for particular phenomenal beliefs like 'this is red'. The original challenge was to explain our general self-attributions of consciousness. Can "phenomenal properties and the capacity to have them [still] play a crucial role in constituting its content", as Chalmers suggests? Seems plausible enough; there's nothing in the zombie's mind for his alleged concept of 'consciousness' to latch on to, for example. Surely it must be empty, again. The upshot: if we weren't conscious, we wouldn't believe it after all. Sure, we would still utter things like "of course I'm conscious!" -- but they would just be so much meaningless babble.

So, zombies are incapable of any positive conception of consciousness (and thus derivative concepts such as zombiehood). But I can think of a residual problem: what of the even more general claim that physicalism is false (i.e. a minimal physical duplicate of our world is not a full duplicate of our world)? This claim involves no phenomenal concepts, and so presumably can be thought without any need for phenomenal properties -- they aren't constitutive of this belief, at least. But that leaves us with only physical factors to explain why we disbelieve in physicalism! Odd, no?

The only way out that I can see is for the epiphenomenalist to adopt the strong position that consciousness grounds all genuine intentionality, so that zombies can't have any real concepts, beliefs, or mean anything at all. This way, even our non-phenomenal beliefs (e.g. about whether physicalism is true) are partly constituted by -- or otherwise depend upon -- phenomenal properties.

[Correction: an alternative option has been pointed out to me: the belief against physicalism inherits its epistemic warrant from the phenomenal beliefs from which it is inferred. In case of zombies, there is no such warrant to be transmitted.]

Consciousness explains why we have the beliefs we do, because without it, we wouldn't have any genuine beliefs at all.


  1. if what you mean by red is shaped by "your phenomenal quality of redness" then unless that is a empty statement how can the zombie act identically to you without that input? Or does he have a similar input that happens to be identical to your phenomenal quality of redness and yet isn't.

    also i presume we take it that my phenomenal quality of redness is different from yours?

    >phenomenal qualities (consciousness) can influence what we believe after all.

    and at the same time what the zombie says he believes?

    ie a zombie that never had any beliefs says he has beliefs identical to if he was influenced by somthing he has never been influenced by.

    its a bit like a world where everyone flies around exactly as if planes had been invented (down to sitting in the seats and watching the windows) and yet there arent actually any planes.


  2. Thanks for that strong account of the epistemology of phenomenal beliefs. I think, however, that the issue might be better phrased a little differently, since it's not a pure epistemological issue but one about the coherence of our hypotheses about the evolution of phenomenal beliefs. Consider not Zombie World, but Dennettt World. In Dennett World, people have non-physical epiphenomenal consciousness, but are all natural Dennettians, who don't believe they have phenomenal consciousness in any sense that isn't' functionally and representationally analyzable a-priori. Let's stick to the highest order beliefs here, I.E "physicalism is true", to avoid the issue of whether they have particular phenomenal beliefs despite themselves.
    The Dennett World people evolved somewhat differently than us, and in a way that made them have false beliefs [the belief that physicalism is true].
    Any explanation of why we evolved to be different than Dennett World People would involve astounding luck, because at no point will different ways our coginition\evolution reacted to the existence of phenomenal consciousness be mentioned in this explanation.
    The difference between our evolution and that of Denett World people is like the difference between us and a humanity that evolved with innate true beliefs about the geology of the moon - we're just luckier.

    If you want to avoid evolution, the issue can be presented like this: What is Dennett's mistake or cognitive limitation in having no Qualia intuitions? Is it possible to point out a disfunciton in his rationallity or in his perceptual faculties? If not, Dennett is a) justified b) making use of all possible data and c) wrong. This seems strange.

  3. GNZ - Meaning need not impact on behaviour, according to this view. My zombie twin and I are physically and behaviourally identical. It's just that my thoughts and utterances are meaningful, whereas his are more like leaves in the wind that appear to spell out a message (but don't really, due to the absence of an intentional mind).

    "a zombie that never had any beliefs says he has beliefs identical to if he was influenced by somthing he has never been influenced by."

    That would indeed be strange if we were talking about a causal influence. But we're not. So it should come as no surprise, according to epiphenomenalism, that the zombie's behaviour is indiscernible from our own. (Note also that the zombie doesn't really "say" anything, he just makes noises that sound like words that would be meaningful if a conscious person said them. Again, cf. leaves in the wind.)

    Peli - I think our philosophical beliefs are a response to proximate rational pressures, rather than distant evolutionary ones, so I'll focus on your last paragraph:

    "What is Dennett's mistake or cognitive limitation in having no Qualia intuitions?"

    Tough question! In this case, the denials look ideologically motivated to me (though perhaps it's just a failure of imagination on my part, to see how anyone could honestly fail to have the intuition). But note that the problem generalizes to all philosophical disagreements. One or other party must be making some rational error, but it sure can be difficult to diagnose just what it is!

  4. So on your account the belief in Qualia is a-priori true, and can be derived from rational reflection alone? [because obviously it can't be that any empirical data rationally motivates it, if it's epiphenomenal]
    That either denies Zombie-Worlds, or folds into a "cogito ergo sum" argument (which I suppose is legit).

  5. Yes, like a cogito argument. If qualia are necessary for meaningful thought, the mere fact that we're thinking suffices to guarantee that we're also conscious.

    (Whether this sort of inference is best classified as 'a priori' or 'a posteriori' is a matter of some contention, but doesn't really matter for present purposes.)

  6. > That would indeed be strange if we were talking about a causal influence. But we're not.

    yes I realise that we are not talking about a causal relationship but doesnt that make it MORE of a problem?

    Almost all of science is based on looking at two things and presuming if they behave in related ways then they are related.

    If we found a zombie that behaved absolutely identicaly to you then the first thing any smart scientist would do is try and see if there was a connection to you. and a little research would probably make them as certain that there was a connection as they are certain of anything even if they could not find the mechanism.

    besides imagine the situation. I (a non zombie) might be debating "the redness of red" with you (in this case a zombie) and you have just as much insight (maybe even more) as me despite not actualy having the aditional data input of actualy experiencing red to put into the mindless calculation that drives your zombie words.

    or are we saying that there wuld be a theoretical debate where we could catch out a zombie and go "ahah! your a zombie!" and the zombie would go - "er damn your right!"


  7. So in that case Dennett's problem on your account is pretty clear, I think: Dennett believes that intentionality can be defined teleo-functionalistically, or at least is a conditional of some sort that computation+species-history can satisfy, while in fact only phenomenological states can ground intentionality.

    What's provocative here I think is your idea that the manifest "qualia intution" is just a bonus, and qualia can inferred strictly from the unavoidable acknowledgement of the existence of intentionality and the inadequacy of functionalist and teleo-functionalist attempts to ground it. I'm cool with that, but that's worth noting as an unusual position.

  8. Peli - yes, that's a better way to put it - thanks!

    GNZ - right, it is kind of funny that we could learn just as much about consciousness from talking to a zombie or a conscious person. They would seem to us to be saying the exact same things.

    One way to look at this is to take the zombie as a baseline, and then marvel at how adding genuine consciousness doesn't make the third party seem any more insightful about the topic.

    I'm more inclined to take our state as a baseline, and then note (without such surprise) that if you remove consciousness whilst leaving all the corresponding physical mechanisms in place, the zombie will still vocalize all the same "words", etc. He appears to have insights, simply because he's behaviourally identical to someone who really does! Looked at this way, then, it no longer seems so mysterious.

  9. marveling is usually a bad sign. It implies you haven't got your assumptions right.

  10. that was me of course


  11. Richard,

    I have always been perplexed by this thought that bits of phenomenality are constituents in our concepts and thoughts.

    Here's one way to point at my befuddlement. Take Mary. She's never encountered blue, but knows a lot about it. She says that she believes that the sky is blue today, or that one of the crayons in the crayon box is blue (and it appears that she's right). In conceptual terms, it seems that she is employing the concept of blueness. The same one that I do when I say that I think that the sky is blue today (before I walk outside). On your account Mary and I are having radically different thoughts. My question: What is the alternative suggestion, what concept is a constituent of her thoughts if not blueness?

  12. Hi Jack, I'm not sure how best to resolve this, but the first thing I'd want to do is ask Mary to clarify precisely what she means. Maybe she's just claiming that the crayon has a particular surface reflectance property, i.e. a disposition to reflect certain wavelengths of light. Or perhaps her concept is derivative of our phenomenology, i.e. she's claiming that the sky is disposed to cause in us a certain visual sensation (what we call 'blue'). Something like that. In any case, she can't be claiming the sky has a [insert raw visual image of blue] look to it, which is what I'd take you to be claiming when you utter those words.

    What do you reckon?

  13. Richard,

    Maybe I'm slow, but I still don't see what I take to be the Grietzer main issue from the email to be addressed. Though I see Grietzer seems to be satisfied with your answer, so maybe I don't understand.

    The issue is... what is the reason for zombie-Chalmers or zombie-Richard to pronounce or write down any of the things Chalmers or Richard have written?

    Let's forget non-zombie world, and concentrate on this question. Is there *any* plausible account why this would happen in zombie world?

  14. 'Reason' is ambiguous between causal explanation and normative justification. The causal explanation is obvious enough: writing is a physical act, ultimately reducible to fundamental physics. Let's suppose determinism for sake of argument. A Laplacean demon could then look at the initial state of the world, and use the laws of physics to infer what all future states will be (in terms of the physics). One of these will involve that arrangement of atoms which we describe as 'writing', etc.

    N.B. The causal closure of the physical ensures that there is a purely physical explanation for the current arrangement of atoms.

  15. Yes, it is a given that what we can refer to in our causal explanation is eventually the fundamental physics.

    However, on the face of it, it seems highly implausible that given the laws of physics in a zombie-world what you would get is a bunch of human-zombies writing and talking (or show a behavior as of writing and talking). This - "showing behavior as of writing and talking about the dualism" is a phenomenon in that zombie-world, same as the other "purely physical macro phenomena" like evolution, different traits of species, tides, erosion, and so on, and so on... Does epiphenomenalist have any explanation of why would some such thing happen in the zombie-world, except point that there is a chance that it will happen? (And not just that it could, but it does happen now, as the same causal explanation needs to be true here in the non-zombie world)

  16. Brods,
    in zombie world there would be the same evolutionary pressures etc.

    Everything consequential is identical except a particular item that is defined as having no effect on the rest of the world so all the lines of causality hold.

  17. ie qualia world has exactly the same probabilistic distribution of results as zombie world.

    qualia world just happens to also have a unique structure* which that is cut off from interaction with the real world except in how it is effected by "first person experience".

    *its unusual properties seem to be one way "cause/effect", ability to become more ordered and more numerous over time, and otherwise not obeying any specific rule.

  18. Hey genius,

    So, are you saying that evolutionary pressure makes zombie-Chalmers'-twin write the books he does?

    In non-zombie world we have good explanation using psychological terms, like beliefs, awareness of qualia, etc...

    But in zombie world, we can't use those terms to explain why zombie-Chalmers wrote those books because those things are not there. However there is still that phenomenon of "zombie-humans who behave as if they write books about epiphenomenalism". So, I'm asking if epiphenomenalist has any account of why such thing happens in zombie-world?

    You mention evolutionary pressure. I don't see why would evolutionary pressure select for a people that behave as if they speak or write about qualia, epiphenomenalism, etc...

    Tanasije Gjorgoski
    (sorry for the ident, I'm using the OpenID option for comments, and that's the id I have)

  19. Tanasije - I don't think you fully appreciate that the zombie world is physically identical to our own. This means that zombie-Chalmers has all the same brain states and dispositional states as the actual Chalmers. Third-personal "psychological" explanations (i.e. those offered by cognitive science) remain unchanged. The dispositional component of beliefs and desires is fully present. Just not the phenomenal component (which is behaviourally inert in any case, and plays no role in causal explanation even in the actual world). See Dennett on The Intentional Stance, "real patterns", etc. This is all we need to explain the third-personal (physical) facts.

  20. Richard,

    Of course the zombie world is physically identical to ours.

    But this is what you have in that zombie-world. You have zombie-epiphenomenalists that write *whole books* (or behave as if they write whole books) on consciousness, qualia, and so on. You have further a big number of zombies that buy those books (or behave as if buying those books), and behave as if reading those books.

    Doesn't anything here strike you as strange? This is supposed to be purely physical phenomenon (or dispositional-psychology phenomenon, or call it as you like). The question is... why do those zombies behave this way? You are implying that there is a dispositional account of this, and you say this is all we need to explain the third-person facts in general. But here we have a concrete behavior. Do you have any plausible explanation of this behavior?

    (And yes, of course this is what is supposed to be happening here. We are talking about the zombie world just to remove the consciousness out of the picture)

  21. No, it's no more strange than the chess-playing behaviour of a computer. Certain behavioural outputs predictably result from certain computations, which we can understand (somewhat metaphorically) by attributing 'belief-like' and 'goal-like' states to the computing entity. None of this requires that the computer/zombie actually has phenomenal states that correspond to their computational states. Physical behaviour obviously has physical causes. If you can't see this, I don't think further conversation will be fruitful.

  22. The problem is not the possibility of there to be zombie which would behave as if conscious.

    What I'm asking you to do is give a plausible third-person story of why there happen to appear such zombies (which spend lot of time on behaving as if writing and talking about epiphenomenalism) in the zombie-world.

    For the chess-playing computer, I can point to the creator of the chess program. So, I can give the story, of how such program came to be.
    So, given this example, what would be your story about the behavior of zombie-epiphenomenalists? Why do they behave as writing the books about epiphenomenalism? Just not to be misunderstood, I'm not discussing general behavior of zombie-people, like eating, communicating, etc.., we could give stories based on evolution about this. This is just about this specific behavior zombie-people show. And, yes, I ask for an explanation in terms of physical causes.

    I hope I'm more clear in the request now. Do you have any kind of explanation for that zombie-behavior?


  23. Right, evolution explains our general capacities. Social sciences (esp. psychology) will presumably explain our more particular behaviours. There's no philosophical question here.

  24. The philosophical issue, is that both physicalism and interactionism can give causal explanation for people writing books about consciousness (namely, because we are aware that we are conscious), but epiphenomenalists can't , as whatever is causal explanation in our world, it has to be also causal explanation in the zombie world. So "because we are aware that we are conscious" doesn't work, and another explanation need to be given.

    However, as far as I can see, one will have to be conclude that this behavior (that zombie-Chalmers have written bunch of books with hundreds of pages), is just a result of pure chance. Anyway, sorry if I come out as stubborn or something, I just find the topic interesting. Anyway, I won't comment any more on this topic.

  25. One last quick note: epiphenomenalists can give all the same explanations that physicalists can. It's not as though the physicalist posits anything that epiphenomenalists don't. We both agree that there are brain states that are at least associated with consciousness. The physicalist thinks that that's all there is, whereas the epiphenomenalist thinks that these are 'neural correlates' to some additional thing. But the neural correlates do exactly what physicalists think consciousness does. (Cf. the "easy problem" of consciousness.) It's certainly not "pure chance" -- a fully informed scientist looking at the zombie world could predict the outcomes easily enough, and make sense of the patterns of behaviour found therein. Just like real world scientists do (you won't find 'qualia' mentioned in scientific explanations).


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.