Saturday, August 7, 2010

Conceiving and hallucinating

The “conceivability argument” for dualism holds that (1) it is conceivable that I might exist apart from by body, and therefore that (2) it is metaphysically possible for me to exist apart from my body, so that (3) I am not identical to my body. I put forward a detailed and sympathetic exposition of this argument in chapter 2 of Philosophy of Mind – for it is more interesting and defensible than some philosophers give it credit for – but I do not in fact endorse it. It is essentially a variation on Descartes’ “clear and distinct perception” argument for dualism, and as I’ve argued elsewhere, what’s true in that argument isn’t new and what’s new isn’t true. (As an Aristotelico-Thomist I reject the rationalist assumptions inherent in Descartes’ version, and the Kripke-style modal assumptions inherent in modern versions. But demonstrating the mind’s immateriality does not require committing oneself to these assumptions.)

But as I say, the argument is more defensible than it is often given credit for. Take the thought experiment I discussed in Philosophy of Mind, W. D. Hart’s “seeing without a body” example. (Hart develops this example in his book The Engines of the Soul and in his article “Dualism” in Samuel Guttenplan, ed., A Companion to the Philosophy of Mind. Follow the links and you can read the relevant sections online via Google books.) Hart asks you to imagine that you wake up one morning, look in the bathroom mirror, and see staring back at you two empty eye sockets where your eyeballs used to be. This is no doubt physically impossible, but it does seem that we can at least conceive of having such an experience. There is no apparent contradiction or incoherence in the idea. Seeing without eyeballs is thus arguably metaphysically possible even if it is not physically possible.

Hart then asks us to extend the thought experiment in various ways. For example (and as I suggested extending it in the book) you might imagine that instead of seeing two empty eye sockets staring back at you, what you see in the mirror is the stump of your neck where your head used to be – a headless body, arms raised in horror. Seeing without a head, and thus without a brain, seems perfectly conceivable, then, and therefore (by the reasoning of the conceivability argument) at least metaphysically possible even if not physically possible. And if that is conceivable, so too is it conceivable to have the experience of seeing without arms, legs, torso, or any body at all; in which case (the argument claims) it is metaphysically possible to see without a body. But since seeing is a kind of mental state, it is therefore metaphysically possible to have a mind without a body.

Now, to evaluate the conceivability argument, we have to determine (a) whether what seems to be conceivable in this case really is conceivable, (b) whether conceivability really does entail metaphysical possibility, and (c) whether the metaphysical possibility of A existing without B really does entail that A and B are non-identical. I commend Hart’s discussion of these various issues (and also my discussion in Philosophy of Mind) to the interested reader. What I want to focus on here is just (a), and in particular on one possible challenge to (a).

A friend of mine sometimes reads this blog at work. (“I’m that bored,” he explains.) He’s been reading Philosophy of Mind, and when I saw him the other day he asked vis-à-vis the Hart example: “What does that prove? Isn’t it just obvious that I would be hallucinating?” That’s a fair question. We might put the objection this way: For the conceivability argument to work, we have to be able to show that (among other things) it really is conceivable to see without a body. But Hart’s example shows at most only that it is conceivable that we might have an experience that seems like the experience of seeing without a body. And that’s not good enough, because such an experience might be hallucinatory. “Seems like” does not entail “really is.” Therefore, “We can conceive of it seeming like such-and-such is the case” does not entail “We can conceive of such-and-such really being the case.”

But I think the objection fails. The reason is that hallucination seems parasitic on veridical experience in a way that ensures that if we can conceive of hallucinating something, it follows that we can conceive of really experiencing it. Consider that it isn’t just impossible to conceive of seeing a round square. It is also impossible to conceive of hallucinating a round square. And this is so, surely, precisely because it is impossible to conceive of seeing one. Hallucinating (or visually hallucinating, anyway) is just a kind of defective seeing. So, what’s impossible for seeing in general is impossible for hallucinations in particular. But then, if it were impossible for us to see without a body, it would also be impossible for us to hallucinate seeing without a body. So, since the Hart example at least shows (by the objector’s own admission) that the latter is possible, that suffices to show that the former is possible as well.

Or so it seems to me. The conceivability argument may or may not fail on other grounds, but I think that at least to this extent it is on solid ground.


  1. Can Animals think?,9171,978023,00.html

  2. One question: Is the argument meant to show that I am not identical to *this* body, or that I am not identical to *any* body?

  3. Great topic. I'm looking forward to giving it a thorough reading.

  4. This is a very interesting case for the dualism of the person. I like the method of trying to connect conceivability to metaphysical possibility. I like how Feser showed the direct proportion between the conceivability of hallucinations and true seeing. But I wish that he would have continued by showing that conceivability might imply metaphysical possibility. If a solid argument for the connection of conceivability and actual possibility could be shown it would build and excellent case for dualism.

  5. Edward:

    I'm new to your blog and your work. All excellent thus far. I've been reading philosophy as an amateur for years (BA in philosophy from USC back in 1980) and your stuff is exceptionally clear, cogent, and readable. I just purchased Aquinas, The Last Superstition, and Philosophy of Mind -- finished with the first, half way through the second, looking forward to the third. Again, all top notch.

    Re this particular post, I think you've missed the strongest version. (Maybe you hit it in Phil of Mind -- we'll see in a week or so.) In any event, I'm a fan of the model version, a la Plantinga and others. What's your opinion of it? From what I can see, it runs like this:

    1. A=B iff all properties predicated truthfully of A can be predicated truthfully of B, and vice versa. (Leibniz's Law)

    2. Modal properties (e.g., possibly x, actually y, and so forth) are properties.

    3. I can imagine myself as existing when my body does not (e.g., the examples in your post.)

    4. Therefore, it is possible (though not actual) that I am no identical with my body.

    5. But it is not possible that my body not be identical with my body. (Law of Identity)

    6. Therefore, I have a (modal) property -- i.e., possibly existing without my body -- that my body does not have.

    7. Therefore, I am not identical with my body (because I have a modal property my body does not have), from premise 1.

    I find that pretty persuasive.

  6. tr:

    It's good stuff.

    Someone (the Japanese blogger interested in paranormal research) posted that video here some months ago but I don't recall it getting much discussion.

  7. Anon217:

    Please elaborate on how the article to which you linked relates to the argument of this post. Not a snarky request, mind you.

    I suspect your implication is that, insofar as we take humans to have evolved from nonhuman ancestors, and insofar as such animals have "mentation" as well, ergo there is no principled reason we, as animals, could not think, and therefore no reason to argue for abstruse doctrines like the immateriality of the intellect. If animality, in other words, sufficiently grounds, 'thinking', then there is no reason to argue for anything beyond animality in a theory of thought.

    Or maybe you were just posting a link out of sheer personal interest.

    As for whether animals can think, a lot hinges on what you mean by "think."