Blog powered by Typepad

« Russell, Newman, and the mind-body problem | Main | Revelation and Humility »

October 24, 2005


David Papineau

Thanks for your useful comments on my paper. I’ve been thinking about them. I agree the interesting questions are to do with whether zombies share our epistemic situation. But let me first say something about conceivability and explanation.

I. Some of the things I say about your paper stem from my take on the ‘explanatory gap’. I don’t accept that the conceivability gap between the physical and the phenomenal constitutes any kind of epistemological or explanatory gap. In my view, when people say they feel an ‘explanatory gap’, they are simply expressing a deep-seated dualist intuition whose source lies elsewhere. In the absence of this independent dualist intuition, we wouldn’t feel that physical-phenomenal identities were explanatorily deficient or epistemologically inaccessible, simply on account of their conceivable falsity.

So in general I don’t accept that we are stuck with residual explanatory problems, whenever the ‘phenomenal concepts’ strategy turns out still to leave us with physical duplicates that are conceivably phenomenally different.

(OK—I know I’ve never persuaded anybody much of my ‘antipathetic’ account of the source of my posited prior dualist intuition. But wait—I’m working on a nice paper which steps back from the question of why we have this supposed intuition, and simply aims to show that everybody, including paid-up physicalists, are in its grip.)

You worry about what dialectical work phenomenal concepts are doing, for someone who denies the conceivability-explanation connection, and so don’t use them to deal with explanatory/epistemological problems supposedly occasioned by the conceivability gap. My answer: they are crucial to explaining the prior intuition of dualism (and also, if the question is raised, to explaining why physical-phenomenal identities are arguably peculiar among a posteriori necessities in involving no semantic instability).)

This take on the ‘explanatory gap’ explains why I’m happy to embrace the first horn of your dilemma as well as the second. In discussing the first horn, you specify that C shouldn’t be understood so as to conceptually require phenomenality, so as to give us Type-Bs a fair chance of avoiding the first horn. That’s a generous offer, but since I don’t accept that the conceivability of P&-C automatically generates explanatory or epistemological gaps, I’m happy to understand C as conceptually requiring phenomenality, and not particularly fussed that this puts me on the first horn.

II. Now for the epistemic situation of zombies. I agree, as you observe towards the end of your paper, that even apart from conceivability-explanation issues, the Type-B physicalist can be faced with a directly explanatory question: how can a physically explicable C explain our epistemic situation, given that zombies will have C but not share that epistemic situation?

(Though doesn’t this question need to be focussed specifically on silicon zombies, rather than full-on ones, once the conceivability-explanation connection is denied? That’s how it seemed to me in my paper. I don’t see that any explanatory challenge is raised by the mere conceivability of full-on zombies with C who lack our epistemic situation, since I think that’s consistent with C explaining the epistemic situation that we necessarily share with our physical duplicates.)

So the real issue is with silicon zombies, who look as if they would share C yet really lack our epistemic situation. They would seem to lack the intimate epistemic connection with their ‘schenomenal’ states that we have to our phenomenal states.

My line, as you observe, is to respond that in truth silicon zombies would share our epistemic situation. In this connection, you press me for a fuller account of the special semantic stability of our phenomenal concepts, which you take to require that we have some especially intimate epistemic acquaintance with our own phenomenal states.

OK. I say that phenomenal concepts strike us as especially semantically stable simply because they ‘quote’ their referents (their exercises are inevitably accompanied by (a version of) the phenomenal states they mention). In so far as we conceive of phenomenal concepts as essentially involving this feature, we will suppose that the phenomenal concept pain, say, will inevitably refer to the property pain, however the actual facts turn out—otherwise it wouldn’t be that concept. I don’t see that this story demands any especially intimate epistemic acquaintance with the phenomenal property pain, certainly not some kind of acquaintance that would guarantee knowledge of all its essential features. Why you can’t refer to pain by so quoting it, and yet remain ignorant of the fact that it is identical to C-fibres firing? (NB—take ‘quotation’ with a pinch of salt here—I don’t really think there is any good mental analogue of linguistic quotation—my thought is solely that exercises of phenomenal concepts are accompanied by versions of their referents.)

You feel that a thin account of phenomenal concepts, that allows them to be shared by silicon zombies, will only yield some ‘blind’ indexical/demonstrative knowledge, which falls short of own epistemic situation. I now agree (in my new paper) that ‘blind’ demonstrative knowledge is a lousy account of what Mary learns. But I don’t see why a thin bottom-up account should leave the silicon zombies with only blind demonstration. On my current account, phenomenal reference requires activation of sensory patterns, which activation will then (‘quotationally’) instance the states referred to. Even for zombies, that kind of reference is quite different from mere ‘blind’ demonstration, which they (like us) can do without any sensory activation.

Still, you will say, won’t their reference still be ‘blind’ in the sense that it would be epistemically just the same for them if we substituted a different neural state as the target of their ‘schenomenal’ thinking? I say no—given the ‘quotational’ feature of their ‘schenomenal’ concepts, that would require that they be thinking with a different concept, which I take it would make a real epistemic difference.

Maybe the silicon zombies don’t have, as you put it, ‘substantive knowledge of the state’s intrinsic character’. But I dispute that we have this ourselves, in the relevant sense. Phenomenal thinking connects us (and the silicon zombies) intimately to phenomenal (‘schenomenal’) states only in the sense that the thinking will inevitably be accompanied by the states themselves—but, as I said earlier, it seems to me unwarranted to infer from this that phenomenal thinking must acquaint us with all the essential features of phenomenal states.

Ram Lakhan Pandey Vimal

If I understand correctly, according to Papineau, the prior intuition of dualism (such as property dualism, subjective experiences are irreducible) is the major cause of Levine’s explanatory gap. This means that Type-B materialists who lack this intuition think fundamentally different from the anti-materialists who are unable to eliminate it. Thus, the prospects for reaching any single, agreed, view independent framework appears remote unless some criterion for an optimal framework (such as with the least number of problems) is accepted by both views. Furthermore, in the second horn, it is still not clear to me that how zombies can fully share our epistemic situation.

[Note: please delete my previous comment.]

The comments to this entry are closed.