So I am finally done teaching summer school and am ready to settle in to my two weeks of ‘vacation’ before the Fall semester begins. Just as I am about to switch on the PS3 I am struck by the following line of argument…let me know what you think of it…
Those who know me know that I am fond of an argumentative strategy that I call ‘deprioritizing’ when it comes to a priori arguments against (or for) materialism. The idea is taken from the police. When something is deprioritized we still recognize it is a legitimate thing but also recognize that it is not a high priority. So if we are deprioritize the a priori arguments we can still acknowledge that in principle we can tell a priori what is what but for us it will be an empirical discovery. By the time a priori methods will be useful it will be too late. I do this by introducing shombies and zoombies. A shombie is a physical duplicate of me that has consciousness in the absence of any non-material properties. I have claimed that when we are conceiving of a shombie world we are NOT conceiving a a zombie world. But how do we know that it is not? I tend to think of the shombie world as the close possible world where some kind of higher-order theory is true and we have consciousness just like we do in the actual worlds.
This got me to thinking. How does the other side know that consciousness is absent at the zombie world? According to them to know that one is consciously seeing red is to be acquainted with a red quale in such a way as to have it partly constituting my belief or judgment. So to know that we have consciousness, or to know that it isn’t lacking at the actual world, requires being acquainted with it. So how do we know that it is lacking at the zombie world? Sure can conceive of a word with our physics at some future date but all we can ‘see’ is that there are beings there who look like us, talk like us, etc. It would seem that we have no way to tell from the third-person whether these ‘zombies’ really do lack consciousness and since that is the only way for us to know about zombies we are led to a contradiction. In order to conceive of zombies we must know that they lack consciousness, but it is impossible for us to know that they lack consciousness, thus zombies are inconceivable. We can sum this up in the following argument.
1. If zombies are ideally conceivable then we can know that they (the zombies) lack consciousness
2. We cannot know that they lack consciousness
3. Therefore zombies are not ideally conceivable
An opponent might respond that it is just stipulated that there is no consciousness at the zombie world but this is exactly the reason why physicalist claim that the zombie argument is question begging or that it builds into the very concept of consciousness that it is non-physical.
I just noticed that the Stanford Encyclopedia of Philosophy entry on Zombies was recently updated (authored by Robert Kirk, who’s book I reviewed for phil. psych). I was pleased to see that my JCS paper was mentioned in the “anti-zombie argument for physicalism” section. But Kirk cites my paper as arguing that “we should reject the inference from conceivability to possibility”. It is true that others that have pressed versions of the ‘anti-zombie’ argument for this conclusion, I am not one of them. I want to grant the link between conceivability and possibility. It is true that I harbor empiricist leanings but if I were a rationalist I would find Chalmers’ CP thesis very attractive; but even so the zombie argument is inconclusive because we cannot simply assert that zombies are conceivable.
My complaint against the zombie argument has always been that the move from (1) ‘zombies seem conceivable to me’ to (2) ‘zombies are ideally conceivable’ is question begging. The only thing we really have evidence for is (1) but it is (2) that is actually used in the zombie argument. That this move is illegitimate is shown by the fact that shombies and zoombies seem conceivable to me (and others it turns out) but if I were to then say that they were ideally conceivable I would be accused of begging the question. Both zombies and shombies seem conceivable but only one of them can actually be ideally conceivable and importantly we have no a priori reasons that can decide which is which. Rather what seems to be happening is that one’s intuitions are tracking the theory that one accepts, perhaps implicitly. Thus we don’t know if zombies are ideally conceivable at this point. Nor do we know if shombies are. Both seem to be conceivable to various people but we don’t have enough empirical knowledge of the brain to decide. From this I draw the meta-lesson that we should deprioritize the a priori arguments for and against physicalism. What we need to do now is focus on specific theories of consciousness (like higher-order theories, say 🙂 ) and brain science. Even if we can in principle know a priori that the mind is just the brain, or that it isn’t, the way that we will come to know is empirical (just like water and H2O: even if it is in principle knowable a priori that water is H2O (because on can deduce one set of facts from the other) we discovered it empirically. A priori arguments played no positive role in the discovery).
Via Leiter’s blog I happened to be looking at the list of recent SEP entries. I read this entry on impossible worlds that got me thinking.
The response to the zombie argument that I have been developing over the last couple of year appeals to the distinction between prima facie and ideal conceivability. Something is prima facie conceivable, roughly, if there is no obvious contradiction in the imagined scenario. Something is ideally conceivable if, roughly, there is no contradiction in the imagined scenario even upon ideal reflection. I have tried to argue that zombies are merely prima facie conceivable and may not turn out to be ideally conceivable (another way of putting it that is roughly equivalent is that zombies are epistemically possible but not metaphysically possible) since there are equally plausible parity arguments (zoombies and shombies). As a corollary of this line of defense I have argued that what people like Dave actually succeed in imagining when they *think* they imagine the zombie world is really just a world that is very similar to the actual world. Just as a point of clarification I have always meant this to be a different claim than the Russellian response that the zombie world may have different ‘inscrutable’ fundamental physical properties. What I mean is that since we do not yet know all of the facts about the brain, physics, or theories of consciousness, we may be inadvertently failing to include some crucial physical law, property, or theory of consciousness. So it is very easy, I claim, to imagine a world that is physical in roughly the same sense that ours is but where there is no consciousness. For instance, if the higher-order thought theory of consciousness is right then the ‘zombie’ world is really just a world like ours that lacks higher-order thoughts.
Now people like Dave often claim that they can conceive that we add this feature and yet still it is intuitive that those creature could lack consciousness. If this is really the case and the higher-order theory is true then Dave has imagined an impossible world. But it seems to me that we can at this point admit that the traditional zombie world is conceivable and go on to argue for a restriction on the second premise of zombie argument, which to remind us, is the claim that if zombie are conceivable then they are possible. This premise becomes possibly false since it may be the case that zombies are conceivable but not metaphysically possible, where this means that they inhabit an impossible world.
One response to this line of thought might be that the use of ‘conceivability’ here isn’t the same as that employed by the zombie argument. As used by Dave ‘conceivable’ means roughly imaginable without contradiction but in these impossible worlds we conceive of a world with a contradiction (by stipulation it contains a contradiction). But, of course, the point here is that one may not notice or be in a position to spot the contradiction, which is exactly one of the reasons for postulating impossible worlds (or in this case impossible scenarios in Dave’s sense). If one takes this line, as I am inclined to do myself, then the issue reduces to the original one of the difference between prima facie conceivability and ideal conceivability. But if one has a more generic version of conceivability one can argue that zombies are conceivable and impossible in way that seems different from the usual type-b line…
Tomorrow marks the third anniversary of my starting Philosophy Sucks! I started my blogging career over at Brains and had my first post on April 12, 2007. I had several posts there before I was compelled to start my own blog and as people may know I continue to contribute to Brains and am very pleased to have seen it grow in recent times. I continue to post here as well and limit my posts at Brains to ones that directly relate to philosophy of mind and consciousness.
In these three years I have had over 100,000 hits, nearly 350 posts, and almost 2,000 comments…and next week I will be hosting my third Philosopher’s Carnival (I hosted the 58th and the 50th); not bad! I have had some rough experiences adapting to online discussion (there are some crazies out there as people well know) but all in all the discussion has been extremely helpful and challenging. I have had two papers and numerous presentations (two at the apa Pacific) develop out of discussions that started here. So thanks to everyone and I hope it continues in the future!
The year is still young but here are the most viewed posts so far (see also the best of all time).
10. HOT Qualia Realism
9. Am I a Type-Q Materialist?
8. Why I am not a Type-Z Materialist
7. Consciousness, Consciousness, and More Consciousness
6. More on Identity
5. The Singularity, Again
4. HOT Damn! It’s a HO Down-Showdown
3. Attention & Mental Paint
2. Part-Time Zombies
1. The Identity Theory in 2-D
I listened to the first lecture in David Chalmers’ Locke Lectures currently taking place at Oxford and I was intrigued by the argument he gave in defense of the claim that we can have a priori knowledge and do conceptual analysis even if we cannot give definitions of the concepts that we are analyzing. The argument appealed to the claim that any counter-example to a definition involved reasoning about possible cases and so we could give an account of the a priori in terms of our capacity to think about possible scenarios and our judgments about whether certain sentences are true in those scenarios.
I wanted to find the text of the talk to check on the details of the argument and in the lecure Dave mentioend that he was putting manuscripts up online and I went to his website to see if I could find them…sadly I couldn’t. But I did find this paper which if I am right is probably the text that the fourth lecture will center on. Anyways, I read the paper and now want to say something about it. As I read it the central point is very simple: one can accept Quinian arguments about conceptual revisibility and still have a robust a priori/a posteriori and analytic/synthetic distinction. One does this by simply stipulating that something is a priori if it is knowable independently of experience without conceptual change. That is given that we hold the conceptual meanings fixed is the statement knowable a priori? Much of the paper is spent fleshing out a suggestion made by Carnap updated with 2-d semantics and Bayesian probability theory aimed at giving an account of conceptual change.
So to put it overly simply one can say to Quine “sure, my concept may change and if so this wouldn’t be true but given that my concepts don’t change we can see that this would be the case.” So to take pain as an example. When we are reasoning a priori about what we would say about pain (can there be pain/pleasure inversion for instance) we can admit that if we change what we mean by pain this or that will be different. But as long as our concept of pain doesn’t change we can say this or that would be true in this or that scenario and therefore bypass the entire Quinian argument altogether. This would seem to give Dave a response to the type-q materialist who has been getting so much attention around here lately. This is because they seem to be saying that since our concept of pain might change we cannot know a priori whether zombies are conscious or not. Dave responds by saying that as long as we do not have to change our concept of pain we can see that zombies are not conscious. I think that this response to the Quinian argument is quite good but I would respond to it differently. I would argue that as of right now we do not know which scenarios are ideally conceivable because we have cases of disagreement about decisive scenarios.
To fill this in with a particular example that I have talked about before let us focus on the notion of pain and Pain Asymbolia. Now many philosophers hold that it is a priori that if something is a pain then it will be painful (and that conversely if something is painful then it will be a pain). Now suppose that one of these philosophers finds out about pain asymbolia and denies that these people are in pain. Now suppose that this person comes to change their mind and instead thinks that they are in pain but that pain and painfulness are (contrary to appearances) only contingently related. What are we to say? In the paper Dave says,
A ﬁfth issue is the worry that subjects might change their mind about a possible case without a change of meaning. Here, one can respond by requiring, as above, that the speciﬁcations of a scenario are rich enough that judgments about the scenario are determined by its speciﬁcation and by ideal reasoning. If so, then if the subject is given such a speciﬁcation and is reasoning ideally throughout, then there will not be room for them to change their mind in this way. Changes of mind about a fully speciﬁed scenario will always involve either a failure of ideal reasoning or a change in meaning.
I can agree with this in principle but since I can clearly conceive pain and painfulness being only contingently related it cannot be the case that we are in a position to determine which concept of pain is the one which will be employed in ideal reasoning. We may have our favorite but there are arguments on both sides and it is not clear where the truth lies. So though we can know a priori that either pain is necessarily painful or that it is contingently painful but we cannot know which is true now. To know that we would have to settle the pain asymbolia case; but that case it hotly contested (pun sadly intended :()
The upshot then is whether or not Dave has a response to Quinian worries about the a priori in principle he has not done enough to show that we are currently in a position to make use of this apparatus and so we are forbidden any of its fruits.
Thanks to everyone who came out to the Parkside Lounge last night! It was a weird and wonderful night! For those of you who couldn’t make it here is some video recorded by Jennifer on my iPhone set to our version of Freddie Freeloader…We’ll be back @ the Parkside April 26th and May 31st…Let me know if you are in town!
This video doesn’t exist
I will be presenting Deprioritizing the A Priori Arguments against Physicalism at Columbia University tonight as a guest lecture in Hakwan Lau’s seminar on consciousness (other guest lecturers include Uriah Kriegel, David Rosenthal, & Jessee Prinz). I have made a podcast of the talk (a video) available here.