Wednesday, May 01, 2013

The Tyrant's Headache

When the doctors couldn’t cure the Tyrant’s headache, he called upon the philosophers. “Show me some necessary condition for having a headache, which I can defeat!”

The philosophers sent forth the great David K. Lewis in magician’s robes....

[See the remainder of this story here. Hint: It doesn't have a happy ending.]


Anonymous said...

If I read your fable as a way of making a complaint against Lewis' theory, it seems to be the complaint that he defines pain as a natural kind satisfying a description (pain-behaviour source in beings appropriately like the tyrant), but we can eliminate all things satisfying the description without eliminating all instances of pain. How is this different from complaining that 'x is a mountain summit iff X is the highest point on an appropriate region of the crust' is false because we could eliminate all things satisfying the part after 'iff' (e.g. by filling in all the valleys, so everything is at the same altitude and there are no appropriate mountain-like regions) without actually destroying any summits?

However, I can instead read the fable as making exactly the point that this kind of complaint is misguided, so I won't press too hard.

Also, 'drifted socially across the sea' is brilliant.

Jorge A. said...

In the story:

"The Tyrant turned his attention to conditions (2)-(4). Though he humbly knew that he was in no way exceptional, that could be changed! There were, he had learned, a small group of people in whom Brain State #2324B rather than Brain State #1117A played the causal role of pain. In them, Brain State #1117A played the causal role of mild annoyance at someone else’s bad jokes – a far preferable state! The Tyrant might kill everyone except for this small group of people. He would then belong to a species in which Brain State #1117A normally played the causal role of annoyance at bad jokes, not the causal role of pain."

Although this may be generally conceivable, it is not conceivable (to me) if I assume a materialist position.

We're talking about two groups of people with "identical" brain states who have different subjective phenomenal states.

(The same objection can be raised against the classic p-zombi parallel Earth, where philosophers debate endlessly about qualia and mental phenomenology without actually having them.)

Nonetheless the question remains: why should a piece of electric (or pneumatic) meat "have" associated qualia, even if it's not possible to conceive otherwise from a materialist framework?

Anon 4:42 said...

Re: first comment

It would probably have been better to say 'make the number of things satisfying the description zero' rather than 'eliminate all things satisfying the description', because the whole point is that you can do the former without actually destroying anything.

Arnold Trehub said...

Jorge: "Nonetheless the question remains: why should a piece of electric (or pneumatic) meat "have" associated qualia, even if it's not possible to conceive otherwise from a materialist framework?"

If the "piece of meat" is a particular kind of neuronal mechanism that provides a creature with a *subjective* representation of events in the world, then one can claim that this kind of "meat" will have qualia. The only scientific basis on which we might believe this claim is if a theoretical model of such a mechanism can successfully predict relevant instances of qualitative experience.

Eric Schwitzgebel said...

Thanks for the interesting comments, folks, and sorry for the slow reply!

@ Anons May 2 and May 3: My intended point is skeptical. I think there is something very odd about Lewis's theory, which seems to include something relational (what is normal in your species or population) among the conditions of pain; but I know no *better* theory, and I don't know where to get off the train.

Eric Schwitzgebel said...

@ Jorge and Arnold: It doesn't seem to me inconceivable that the same physical state, intrinsically/narrowly defined, could play different causal/functional roles in different groups of beings; and then the question arises what the phenomenology tracks. In the story, I try to be careful not to assume that it does track the brain state rather than the functional role -- keeping my description neutral on that point.

Susan Hurley and Alva Noe's discussion of cortical dominance and deference is interesting here, though I'm not sure how far I want to go with them.

Anonymous said...

But is the relationality really in the proposed referent for 'pain', or in the way of picking it out? If you pick out Charlie as the tallest player on the basketball team, you've picked out something non-relational (Charlie) with a relational description. Changing the composition of the team, or even the state of the English language, won't alter, let alone destroy, Charlie, even if it changes what 'the tallest' picks out, or makes it the case that there is no tallest player. Doesn't Lewis think that what is picked out by things like 'the state that causes pain-behaviour in an appropriate species' is always a natural kind, and not relational? It's just that the reason we are interested in that natural kind rather than others for these purposes is that it happens to possess that particular relational property (causing pain-behaviour in our species).

Eric Schwitzgebel said...

Anon: Doesn't the reference to what is normal in one's population make pain, at least in a way, relational -- i.e., dependent on what is going on in other people? It doesn't seem on his theory to be just a matter of "picking something non-relational out" by means of a relational description. But the theory isn't developed in a lot of detail, so maybe there's some room for maneuvering here.

A related issue: "belonging to species X" seems also to be a relational property.