Given these options, I think I have to reluctantly choose “metaphysically possible.” A decent respect to the opinions of Less Wrong requires that I should declare the causes which impel me to this.
I have never been able to conceive of a zombie and a physically identical non-zombie without subsequently concluding that my conception leads to positing causeless effects (e.g., a zombie spontaneously talking about its nonexistent phenomenal experience while using all and only the cognitive structures that the non-zombie uses to talk about actual phenomenal experience).
But, well… causeless effects aren’t incoherent, or metaphysically impossible, they’re just vanishingly unlikely. . (Not even necessarily that. E.g., in a world where there are only a few possible utterances, one of which, X, is conventionally understood to be a reference to the speaker’s phenomenal experience, it’s not even all that unlikely for a zombie to just kind of coincidentally happen to utter X in exactly those circumstances that a non-zombie analog utters them due to phenomenal experience.)
Further, nothing in these definitions asserts that humans do have phenomenal experience, which allows for Dennett’s answer (“We’re all zombies.”) In which case cause and effect aren’t even at issue… all of us zombies just happen to talk about phenomenal experience despite not having any, because our brains are wired to do so, presumably for signalling reasons but possibly because the Matrix Lords think it’s funny, or for other reasons.
This strikes me as implausible… I mean, it sure does seem to me that I have phenomenal experience, and I can’t imagine being wrong about that. But that might just be a failure of my imagination.
Which seems to add up to “zombies are metaphysically possible” on this account.
How did you answer the physicalism question? If you think any physical duplicate of our world must also be a mental duplicate then I think you’re committed to the impossibility of zombies.
Think about it this way: Do you think there can be a world where the distribution of micro-physical properties is identical to ours but which does not have tables? Presumably no, because you believe that once microscopic properties are distributed a certain way in some location, you can’t avoid having a table in that location. But you don’t think the same is true of phenomenal experience if you believe zombies are possible. You think that there needs to be something more than just the right distribution of physical properties for consciousness to appear.
I suspect you’re conflating epistemic and metaphysical possibility when you say causeless effects are vanishingly unlikely. Take the sentence “2 + 2 = 5”. I think its metaphysically impossible for that to be true (assuming the meanings of the terms are kept constant) -- there is no possible world in which it is true—but I don’t assign it probability 0 because (as Eliezer has pointed out) I acknowledge that there is some possible sequence of experiences that might lead me to believe it (or increase my credence in it). Thinking something is impossible does not mean assigning it probability 0.
Dennett doesn’t think we’re all zombies, BTW. He’s skeptical of qualia as a coherent concept, but he doesn’t deny that we have phenomenal experience.
I don’t believe mental events have nonphysical causes, which is precisely why I consider zombies to entail causeless effects.
And yeah, perhaps I’m just reading too much into “metaphysically impossible” as distinct from “impossible”; the truth is I really don’t know how to think cogently about what would be impossible if the laws of nature were different.
The “we’re all zombies” bit was intended with tongue in cheek; it’s actually a direct quote of his from some book or another—Consciousness Explained, probably—that I read like 20 years ago and stuck with me. He isn’t entirely serious when he says it, of course, and IIRC has a little footnote that says “To quote this phrase out of context would be the height of intellectual dishonesty.” or words to that effect.
EDIT: That said, another relevant Dennetism that stuck with me was his response to an undergrad at a seminar I was listening in on years back. The undergrad said, in effect, “But I don’t feel like a merely computational process!” and he replied “How do you know? Maybe this is exactly what a merely computational process feels like!”
“Metaphysically impossible” is a rather strong requirement if taken strictly, and I think strictly is the right way to take it when talking about metaphysical possibility.
So if you think there’s a possible world where human actions just happen by chance to line up with some mysterious experiences and zombie actions don’t, then it’s metaphysically possible.
I’m not sure whether Dennett’s answer should really be read as “zombies are metaphysically possible” but I’m now convinced this whole question is silly.
Given these options, I think I have to reluctantly choose “metaphysically possible.”
A decent respect to the opinions of Less Wrong requires that I should declare the causes which impel me to this.
I have never been able to conceive of a zombie and a physically identical non-zombie without subsequently concluding that my conception leads to positing causeless effects (e.g., a zombie spontaneously talking about its nonexistent phenomenal experience while using all and only the cognitive structures that the non-zombie uses to talk about actual phenomenal experience).
But, well… causeless effects aren’t incoherent, or metaphysically impossible, they’re just vanishingly unlikely. . (Not even necessarily that. E.g., in a world where there are only a few possible utterances, one of which, X, is conventionally understood to be a reference to the speaker’s phenomenal experience, it’s not even all that unlikely for a zombie to just kind of coincidentally happen to utter X in exactly those circumstances that a non-zombie analog utters them due to phenomenal experience.)
Further, nothing in these definitions asserts that humans do have phenomenal experience, which allows for Dennett’s answer (“We’re all zombies.”) In which case cause and effect aren’t even at issue… all of us zombies just happen to talk about phenomenal experience despite not having any, because our brains are wired to do so, presumably for signalling reasons but possibly because the Matrix Lords think it’s funny, or for other reasons.
This strikes me as implausible… I mean, it sure does seem to me that I have phenomenal experience, and I can’t imagine being wrong about that. But that might just be a failure of my imagination.
Which seems to add up to “zombies are metaphysically possible” on this account.
I’m not quite sure.
How did you answer the physicalism question? If you think any physical duplicate of our world must also be a mental duplicate then I think you’re committed to the impossibility of zombies.
Think about it this way: Do you think there can be a world where the distribution of micro-physical properties is identical to ours but which does not have tables? Presumably no, because you believe that once microscopic properties are distributed a certain way in some location, you can’t avoid having a table in that location. But you don’t think the same is true of phenomenal experience if you believe zombies are possible. You think that there needs to be something more than just the right distribution of physical properties for consciousness to appear.
I suspect you’re conflating epistemic and metaphysical possibility when you say causeless effects are vanishingly unlikely. Take the sentence “2 + 2 = 5”. I think its metaphysically impossible for that to be true (assuming the meanings of the terms are kept constant) -- there is no possible world in which it is true—but I don’t assign it probability 0 because (as Eliezer has pointed out) I acknowledge that there is some possible sequence of experiences that might lead me to believe it (or increase my credence in it). Thinking something is impossible does not mean assigning it probability 0.
Dennett doesn’t think we’re all zombies, BTW. He’s skeptical of qualia as a coherent concept, but he doesn’t deny that we have phenomenal experience.
I don’t believe mental events have nonphysical causes, which is precisely why I consider zombies to entail causeless effects.
And yeah, perhaps I’m just reading too much into “metaphysically impossible” as distinct from “impossible”; the truth is I really don’t know how to think cogently about what would be impossible if the laws of nature were different.
The “we’re all zombies” bit was intended with tongue in cheek; it’s actually a direct quote of his from some book or another—Consciousness Explained, probably—that I read like 20 years ago and stuck with me. He isn’t entirely serious when he says it, of course, and IIRC has a little footnote that says “To quote this phrase out of context would be the height of intellectual dishonesty.” or words to that effect.
EDIT: That said, another relevant Dennetism that stuck with me was his response to an undergrad at a seminar I was listening in on years back. The undergrad said, in effect, “But I don’t feel like a merely computational process!” and he replied “How do you know? Maybe this is exactly what a merely computational process feels like!”
On reflection I agree.
“Metaphysically impossible” is a rather strong requirement if taken strictly, and I think strictly is the right way to take it when talking about metaphysical possibility.
So if you think there’s a possible world where human actions just happen by chance to line up with some mysterious experiences and zombie actions don’t, then it’s metaphysically possible.
I’m not sure whether Dennett’s answer should really be read as “zombies are metaphysically possible” but I’m now convinced this whole question is silly.
Well, yes.