Comments on Brown’s “Turning the Tables on Dualism”

Consciousness Online Conference, Feb. 20-28

First of all, I’d like to thank Richard Brown for putting this conference together and for inviting me to participate. He’s collected an excellent group of participants and the whole thing promises to be a great success. Thank you Richard.

Second, I’d like to beg your tolerance with what might be a fairly shabby power-point presentation. It is only the second one I’ve ever done, and the first online, so please bear with me.

Brown’s paper falls into three main parts which correspond to three related attacks on the anti-materialist arguments. There are the two reverse zombie arguments, the two reverse knowledge arguments, and there is, throughout the paper, the claim that the original anti-materialist arguments beg the question. It is this latter claim that Brown thinks his reverse arguments prove. In response to Brown, I’ll begin with some general words about begging the question and philosophical. Then with that discussion as a sort of metaphilosophical backdrop, I wish to consider the individual arguments in order.

Accusations that arguments beg the question can become the “I am rubber you are glue” of the philosophical world: everyone can feel that the accusation bounces off of them and sticks on their opponent. There is a reason for this which is also the reason that one philosopher’s modus ponens is another’s modus tollens. It is true of every deductively valid argument that in some sense one should not accept the premises unless one accepts the conclusion. So, it is also true that in a valid argument with four premises, say, if one grants three of the premises, whether or not one should grant the fourth has a lot to do with one’s stance on the conclusion. So it is not really a mark against the conceivability argument, for example, that if a physicalist grants the connection between conceivability and possibility, she should not grant the conceivability of zombies. This is true because the conceivability argument is valid, not because any questions are being begged. Brown says in several places that the anti-materialist arguments do not defeat physicalism, rather they simply divide us into the camps to which we are antecedently committed—physicalists will deny the conceivability of zombies and will prefer the reverse zombie arguments, and dualists will be their mirror image. I think this ignores the possibility that we often have divided commitments and can suffer from conceptual confusions. The purpose of a priori philosophical arguments is sometimes to iron out those confusions. The fact that there are many physicalists, myself included, who insist they can conceive of zombies suggests that this is not a simple case of begging the question, but is a case where conceptual confusion is lurking somewhere—and hopefully not in me!

I think the anti-materialist arguments are far preferable to the reverse arguments Brown proposes.

Consider the argument from Zoombies. A zoombie is a creature identical to me in every non-physical respect but which lacks any conscious experience. Brown argues that zoombies are conceivable, so possible, and therefore dualism is false. There are two answers to this. First of all, this argument is not valid. Dualism doesn’t have to claim that non-physical properties are sufficient for a particular global conscious experience. They only hold they are necessary. So the existence of creatures that are indiscernible with respect to non-physical properties but are phenomenally discernible is no problem for dualism. The physicalist, on the other hand, does claim that physical properties are sufficient for conscious experience, so zombies are a problem for him. Second, even against a form of dualism that claimed non-physical properties were sufficient for conscious experience, it’s not obvious what I am conceiving when I am conceiving that a creature is my non-physical duplicate. If there are non-physical properties that are sufficient for conscious experience, even if they don’t exist in this world but are merely possible—which many physicalists wish to allow—then it is simply a contradiction to imagine that two things can be identical with respect to these properties and yet be phenomenally discernable. It is just part of what it is to be these properties that they endow their bearers with certain conscious experiences. We don’t need to do much conceiving to see that this is the case. That is the disanalogy with physical properties—physical properties do not have written in their essence anything that would seem to guarantee phenomenal experience. So, it seems the zombie argument is either invalid, or zoombies are not conceivable—and their inconceivability is not just a matter of my being a dualist. I’m not. It is because they are by definition impossible: I’m supposed to imagine there being two things which have the same properties which are sufficient for a certain global phenomenal state differing in their global phenomenal states. That’s simply round squarish.

Now consider shombies. According to Brown, Shombies are creatures that are phenomenally indistinguishable from me but have no non-physical properties. Shombies are supposedly conceivable, and therefore possible, and so dualism is false. This argument is valid, but it is not persuasive. It is the same argument as the following: It is conceivable that physicalism is true. Since physicalism is a modal thesis about this world, if it is possibly true it is actually true. If p is conceivable p is possible. So, physicalism is true. Now what does it mean for physicalism to be conceivably true? It means that we can conceive of our physical properties being sufficient for our phenomenal experience, which means that for every world that we conceive where our physical properties are instantiated, we are forced to conceive of our phenomenal properties being instantiated as well. In other words, to say that shombies are conceivable is precisely to say that zombies are not conceivable. In other words, it is not as though Brown has shown us something that is conceivable which happens to entail that zombies are not possible. That would be an interesting counterargument. Instead, Brown is insisting that something is conceivable such that the very question of its conceivability amounts to the inconceivability of zombies. In other words, Brown is just responding to the premise that says zombies are conceivable by saying “no they aren’t.” But where is the argument?

Brown’s response will probably be to say “well, where is the argument that zombies are conceivable?” But there are such arguments, and they’re pretty good. A crude version of the argument goes something like this. Physical properties are defined by structure and dynamics. In other words, what physical properties a thing has is defined by what a thing can do. Phenomenal properties are defined by how they feel. Nothing about what a property can make its bearer do seems to say anything about how the bearer must feel. So one can conceive of physical indiscernibles that are phenomenally discernable.

I anticipate two responses from Brown. One is to point to something he says at the beginning of his paper. He says when we think we are conceiving of zombies, we are really conceiving of something that is physically very similar to us, but is actually physically different, and that is therefore phenomenologically different. Perhaps, but to make that stick it would be nice to point to what physical properties our alleged zombies are missing. Am I failing to imagine it’s having a certain sort of neuron? Ok, let me add that to my conception, and, nope…it still seems perfectly consistent to imagine the zombie has this neuron and yet lacks consciousness. The problem is, that its hard to imagine any physically specified candidate that will make the conception inconsistent. Simply saying “no it’s not” is not good enough. We need to hear the candidate, or at least a story about what such a candidate would look like.

The second response I anticipate from Brown is that somehow I am focusing upon reduction, and not on the ontology of physicalism. There is much to say here, but the long story short is that by opening the gap between conceptual reduction and ontology, one is really denying the premise Brown was meant to be granting—that the conceivability of zombies entails their possibility. That is, relying on this gap amounts to answering the anti-materialist arguments in a different way, one that acknowledges that zombies are conceivable, but that this conceivability doesn’t entail possibility.

Now for the reverse knowledge arguments. The first was given by Churchland. If you give Mary all the ectoplasmic or otherwise non-physical knowledge from within her room she still wouldn’t be able to deduce what it was like to see red, so when she leaves the room she will learn something. Of course so far this is no problem for the dualist, since she could be learning physical facts about conscious experience, or anything else physical for that matter. To make the argument work, we can suppose that when in the room Mary learns all the physical and non-physical facts. It still seems that when she leaves the room she learns something, so the physicalist and the non-physicalist facts together are still incomplete. I have some sympathies here. I discuss this argument in my Phil Studies paper “The Knowledge Argument and Objectivity” where I contend that the knowledge argument is best aimed at the claim that no objective theory is complete, where a theory is objective if it does not require that one actually enter a state in order to fully grasp that state. If dualism is meant to be an objective theory of this sort, Churchland and Brown are right—the knowledge argument works against it. What this shows, though, is that dualists should not claim there is some complete objective dualist theory either. Rather, they should say there are some properties that must be possessed in order to be fully understood, and since the physical sciences describe properties in objective or intersubjective terms, those sciences will necessarily miss these properties. This “subjective dualism” does not fall afoul of the reverse knowledge argument. “Subjective Physicalism” doesn’t fall afoul of either one, but for a defense of the coherence of that notion I refer the reader to my work elsewhere. If Brown wants to join me on that wagon, my arms are open.

Finally, the second reverse knowledge argument. Here, Maria is a perfect phenomenologist but lacks an understanding of science. When she gains the understanding of a completed science, Brown says, she will learn something new: namely that phenomenal states are identical with physical states. So, phenomenal states are not all there is. That is how the reverse knowledge argument would go if it were really to parallel the knowledge argument, but notice that the conclusion is not something the dualist will disagree with! Of course Maria learns something new when she learns physics! The key to the argument is what specifically Brown says she learns: that brain states are phenomenal states. In fact, that premise is all the argument needs. Since learning is factive, this premise alone entails the falsity of dualism. Now not to play the rubber/glue game, but surely a one premise argument stands a pretty good chance of being question begging. The original knowledge argument doesn’t include a premise like “Mary learned that physicalism was false,” and if it did, I think it would be question begging. It says she learns something, and given her circumstances this is supposed to entail that physicalism is false. That’s all the difference in the world, dialectically speaking. The anti-materialist arguments start with a specification of the type of information Mary has, physical information about structure and dynamics, and then says it doesn’t look like that sort of information can lead Mary to know anything about how things feel. It’s not enough to say “yes it can.” A response needs to either give some indication of how it can do that, or explain why it doesn’t need to in order for physicalism to be true. Brown’s reverse arguments do neither.