Small things

[davv]

Apparently there are such things as physicalist dragons. I was really surprised when I first found one, but they do exist!

I'm not one, but I know the evidence, so to call it, I have is incommunicable. One could make a robot that would utter the same sentences, and so a third person hearing the claims could not know whether they were true just by hearing them.

(Jaron Lanier had an amusing thought that the way to distinguish "robots" like the above from real people was to check whether they could hold a dualist position. But even if he were right, making use of it could lead us on that old dangerous path of "us vs them".)

Comments

[lhexa]

For me, the AD&D dragon scheme started to make much more sense when I started thinking of the categories in terms of smell rather than color. Doing so, the evil dragons would correspond to physicalist or naturalist odors. The weakest, white odor would be a cold severe enough to eliminate most smells, and the second weakest, black odor would be a chemical smell severe enough to drown out all else. These would be the physicalist ones. The next three, in order of strength, are the green odor, which would be the scent of a dying temperate biome (think fetid swamp), the blue odor, which would be the scent of unending deluge (and nasty sinus headaches), and the red odor, which would be the scent of burning things. But then you switch over to the metallic schema, and you have more precise biological odors -- copper being roughly the smell of blood, silver the smell of changing weather (when heavy ions can temporarily be suspended in air), and so on. A person in a different field might call these "auras", but if I'm going to be overloading my senses, I'd rather use the ones which human beings don't normally train. Sight is, in my opinion, much too important to use such self-programming tricks on.

[lhexa]

Also, if you have a robot sophisticated enough to process natural language, it could fake dualism quite easily. Vagueness (of all kinds) and term-based ambiguity are both much more difficult logical problems than truth-value paradoxes. Virtually all truth-value paradoxes can be handled by shifting to a higher order logic (zeroth-order to first-order to second-order, corresponding in computing terms to the bounded-finite-automata -> Turing-machine -> Hypercomputing progression), but those same shifts make ambiguity and vagueness less tractable.

Mathematically speaking, dualism isn't even graduate-level.

[davv]

Sure you (or the robot) could fake dualism if you wanted to. That's the whole point of the thought experiment, and to make that clear, consider another variant. Say that neuroscience has arrived to the point where you could produce any experience by tinkering with the brain in an appropriate manner. Then if people have subjective consciousness, anything they experience can be manufactured, but the act of experiencing, itself, couldn't; whereas if they don't, there's no act of experiencing to fake.

In either case, the not-aware would produce the same account of what was experienced as the aware (since "experience" would mean something different to the not-aware), and this is the point where physicalists say "a difference that makes no difference is no difference". The problem is that there is a difference, but only to the person experiencing, and he can't communicate whether he's really experiencing.

I then think Jaron's idea is that if people aren't externally compelled to be dualists or physicalists, then they would pick that which makes most sense with their state. If there is no "there" there, then they pick an everything-is-physical view[1] because that's how it is to them. To quote, "One gets the same sense reading Dennett's appraisals of thinkers like Searle and Nagel. He doesn't just disagree, he doesn't even agree that anything's been said". Conversely, those who are aware would then find dualism obvious. Because they can't communicate their experience, it's easy for others to chalk it up as "their intuition fooling them". That explanation is consistent with the observed data - but so is their own explanation.

(Jaron gets a bit confused later, I think - or at least it seems he lets the definition of "people who are conscious but don't think they are" gradually seep into his usage of the word "zombie" so that it's not clear whether he means actual philosophical zombies or people who don't know they are conscious.)

So yes, the robot could fake dualism, but he probably wouldn't believe in it, Jaron thinks.

I hope the meaning is understandable, because I'm really bending the words here. Language is limited.

-

[1] And let's use "physical" as a shorthand for "physics as currently considered" so as to avoid Hempel's dilemma. As any dualist system with connections between the realms can be collapsed to a monist system with certain laws, you could collapse ordinary dualism to a monism where some now-"physical" things have consciousness (in the intentional sense) and then call yourself a physicalist. That's not what I'm talking about :)

[lhexa]

To quote, "One gets the same sense reading Dennett's appraisals of thinkers like Searle and Nagel. He doesn't just disagree, he doesn't even agree that anything's been said".

Well, there's at least a third of your problem right there. Searle regards himself as Austin's heir, when in reality he has little to no sense of his master's accomplishments.

Anyhow, I've got a simpler test for robothood: ask it/he/she/whatever to have sex with you. Purely linguistics-based algorithms would not be able to handle that scenario, to put it mildly.

Also, what I meant by "physicalist" was "explains and justifies things in terms of laws of physics", with a similar but more holistic definition for "naturalist". Provided with enough knowledge, it becomes very easy for an unethical person to disregard or redirect responsibility.