Saturday, April 12, 2008

Intuition-pumps: on zombies, amoralists, counterfactual demons, and direction of fit.

I'm bored, so I'm coming out of the woodwork rather than making a dent in my pile of work. Deal.

There's a little bloggy brouhaha over at Richard Chappell's place (and continued at, amongst others, Richard Brown's) over the issue of the philosophical zombie. As far as I can tell, Chapell is making a mistake common to many philosophers, and I want to tease out, first, what the mistake is and, second, why it's a mistake (instead of an endearing quirk or brilliant insight or some such).

Let me give four cases which, I think, share the error. The first is the zombie case. We are asked to imagine a race of beings physically identical to us who live in a physically identical world to our own. To all external observations, these beings react exactly as humans do: when hit, they cry out in pain and try to escape; they seek out food and drink at periodic intervals; they appear to accurately and consistently report emotional states; and so on. However, ex hypothesi, there is nothing "inside" these beings. They lack subjective/phenomenal/personal/inner (pick your favourite adjective) experience. All the "stuff" that goes on in our minds -- feelings of pain and pleasure, pure sensations of objects, and so on -- doesn't exist for these beings. Hence, they are called "zombies". The conclusion drawn from this argument is that, since these creatures are conceivable, they are logically possible; and, since they are logically possible, the claim that all there is to the mind is physical (i.e., brain and/or CNS) stuff is false.

The second is the amoralist. We are asked to imagine someone who understands morality perfectly, perhaps better than anyone who has ever existed. However, this person never acts in accordance with morality as such. Whenever this person does the right thing, it is for purely self-interested reasons. Moral reasons -- such as empathy or altruism -- have absolutely no valence for this person. This person, the amoralist, has perfect moral knowledge but sees no reason to be moral qua moral. The conclusion drawn here is that, since the amoralist is conceivable and thus logically possible, the connection between moral reasons and actions must come from somewhere other than the moral reasons.

The third is drawn from Harry Frankfurt's famous example(s) regarding the freedom of the will and moral responsibility (see here). This isn't Frankfurt's original example, but it's a striking way of making the same point. Let us imagine a variant on Descartes' evil demon. Descartes has us imagine an all-powerful and malicious being who can intervene and alter our thoughts at will, and thus deceive us fully regarding the nature of the external world. Instead of this kind of demon, let's conceive of a demon that is a counter-factual actor: the demon will only change our thoughts and actions if we were going to do something the demon didn't want us to do. From this, though, it follows that as long as we're doing what the demon wants us to, then whatever we do is a result of our own choices -- even though, since the demon would interfere if we went off-track, we really couldn't have done otherwise. The conclusion that's supposed to follow is that, since the counterfactual demon is conceivable and thus logically possible, the connection between moral responsibility and the possibility of acting otherwise does not exist.

The fourth is allegedly drawn from Hume, or even Plato. Personally, I think this is pretty tendentious; it seems to me that people are reading what they want into these figures. I'm honestly not sure who came up with it: Anscombe and J. L. Austin are possibilities. In any event, Michael Smith makes it do a lot of work for him. The idea is that there must be two, and only two, kinds of mental states involved in reasons for action: beliefs and desires. This is because the two have opposite directions of fit. Beliefs are supposed to fit the world; if a belief fails to accurately capture the way the world is, then the belief tends to go out of existence. Desires, by contrast, are supposed to have the world fit them; if a desire fails to accurately capture the way the world is, then that is reason to change the world -- and the desire persists. Since we can conceive of these two directions of fit, they are logically possible and, indeed, according to Smith, actual.

Of course, I'm overlooking subtleties. (It's a blog post, not a dissertation, or even a reasonably rigorous essay.) But I think the sketches are clear enough to expose the common thread. We start with some conceivable possibility. Since it is conceivable, it is concluded the possibility is logically possible. (A suspect move which assumes we cannot conceive of contradictions, amongst other problems, but I'll grant it.) Then some interesting philosophical consequence is supposed to follow from the logical possibility.

Here's the thing, though: all four of these cases -- and every other case like them -- is hopelessly question-begging. The zombie scenario just assumes, without argument, that a fully-specified physical world contains no consciousness. The amoralist case just assumes, without argument, that full knowledge of morality will not give sufficient motivation to act. The Frankfurt-style examples just assume, without argument, that being under the control of a counterfactual demon is still a case of genuinely free choice in action. And the direction of fit metaphor just assumes, without argument, that these two directions of fit must both exist. In other words, the cases are structured in such a way as to appeal solely to intuitions favourable to one side of a philosophical dispute, bypassing the inconvenient "giving an argument" stage completely.

This is not to say you can't make a good argument from these cases. Zombie cases can lead us to investigate whether various animals that appear to have minds similar to ours really can be found to have phenomenal consciousness, and to investigate whether any good story can be told connecting physical structures in the brain and/or CNS to subjective experience. The amoralist can lead us to investigate whether genuine amoralists actually exist, and also forces us to consider more carefully why moral knowledge gives reason to act. Frankfurt-style examples force us to reconsider whether freedom really requires the possibility of acting otherwise, or if freedom consists in something more like authorship or ownership of actions. And the direction of fit metaphor... well, for that one, I've got nothing. I'm with GF Schueler: it's just shit.

The point, though, is that the cases in themselves are not arguments. They're intuition-pumps. They're good for getting clear on what convictions are motivating certain philosophical positions, and which convictions are being rejected by other positions, but they do nothing more than this. They're the start of an argument (in all four cases, an argument which has to be exceptionally well-informed by science); not an argument in and of themselves.

The examples, all on their own, do no interesting philosophical work at all.


Richard said...

Interesting! I've responded here.

undergroundman said...

Nice points. This could make a damn good paper. It's tiresome to hear arguments (or rather, as you point out, intuition pumps) like this get passed over as "convincing".

ADHR said...

I'm not entirely sure how this could be worked up into a paper. Maybe if I canvassed several prominent thought-experiments and took a shot at them?

undergroundman said...

I don't know. But if you make a decent point, even if it is somewhat vague, it should be accepted.

It doesn't seem to take much to become noted for a philosophical point. If you think about it, most of the major philosophical breakthroughs appear trivial in retrospect.

According to WP Peter van Inwagen made a big splash in philosophy with his An Essay on Free Will -- his major (and perhaps only) argument is the Consequence Argument, which is, as he notes, completely obvious and completely implicit in every hard determinist's argument.

Similarly, Frankfurt is famous for his thought-experiment, but it's really just a question-begging argument tailored to pump up inuition. He should not get a pass on it.

Incidentally, I replied on that Hume's induction thing. Also, do you have a Wikipedia username?

ADHR said...

Well, true to both cases, but: van Inwagen wrote a book, which is a whole different thing than an essay; and Frankfurt was already established (and, if I've got my timing right, tenured) when he came up with the above thought-experiment. So, neither really helps. ;)

I'm gettin' to the Hume thing. Eventually! And, no, I don't have a Wikipedia username. I get too wrapped up in stupid internet wars sometimes, and I need to control that tendency.