Richard, I’m pretty sure I remember you treating the apparent conceivability of zombies as a primary fact about the conceivability of zombies to which you have direct access, rather than treating it as an output of some cognitive algorithm in your brain and asking what sort of thought process might have produced it.
It seems like some people are using “conceivable” to mean “imaginable at some resolution”, and some to mean “coherently imaginable at any resolution”, or something. By which I mean, the first group would say that they could conceive of “America lost the Revolutionary War” or “heavier objects fall faster” or “we are composed of sentient superstrings, and the properties of matter are their tiny, tiny emotions” or “the president has been kidnapped by ninjas”; whereas the second group would say these things are not conceivable.
As a result, group A wouldn’t really consider the conceivability of p-zombies as evidence of their possibility (well, it’d technically be extremely weak evidence), whereas group B would consider the problem of the conceivability of p-zombies as essentially equivalent to the actuality of p-zombies. (There may be other groups, such as those who think “If it’s imaginable, then it’s coherent,” but based on my brief glance the discussion hasn’t actually made it that far yet.)
Is this right? I’d think the whole thing could be resolved if you taboo’d “conceivable”...?
Talking about “the” possibility of p-zombies is pretty pointless, because of the important
difference between logical and physical impossibility. Even Chalmers thinks PZs are
physically/naturally impossible.
I don’t think the coherent/incoherent distinction you are making is clear. Of course, in a universe where everything is exactly the same, heavier objects would not fall faster in vacuo. But then we understand gravity and acceleration, so we can say what the contradictions would be. We don’t understand what the contradictions would be in the case of p-zombies, because we don’t have the psychophysical laws.Physicalism is Not An Explanation.
By ‘coherent’, I mean something like ‘consistent’ (to make an analogy to logic) - given all our observations, and extrapolating the concept as needed, there are no contradictions. “Heavier objects fall faster” leads to contradictions pretty quickly. Some people believe that “p-zombies are possible” (in some sense, which might match up with what you mean by either logical or physical) also leads to a contradiction, though we of course don’t understand the laws that would cause this.
This is beside the point! I’m not arguing for or against p-zombies (here), I’m saying I think the people in this argument are talking past each other because they have diverging definitions.
“Heavier objects fall faster” is imaginable at a particular resolution. Once you ask, say, “what happens if you glue two stones together?”, it contradicts more deeply-held notions, and the concept falls apart at that resolution.
Some people believe that p-zombies are incoherent if analyzed sufficiently, or expect that they necessitate a severe contradiction of much more deeply-held beliefs.
Moreover, it is possible to hold that we don’t know the laws that would contradict p-zombies but that they are nevertheless contradicted—as it is possible to hold that things should not fall up without knowing the laws of gravitation (leaving aside that some things do fall up).
Do you disagree with my central assertion, or just my definition of coherence?
The stone-gluing can be worked around with auxilliary laws. To assume those laws
are absent is to assume some other laws.
People can believe what they like. If you are going to stake a claim that there
is a literal self contradiction in p-zombies, you need to say what it is. However
most cases of aleged self contradiction turn out to be contradiction with unexamined
background assumptions—laws, again. Talk of “resolution” is misleading: this is cognitive, not pictorial.
It is in fact the philosopher’s point that p-zombies are really, for unknown
reasons, impossible. They are not arguing zombies in order to argue zombies!
Non-philosophers keep misunderstanding that.
That’s a difference that doesn’t make a difference. That I can (not) conceive p-zombies can only mean that my cognitive processes produce a certain output. Whether is is somehow a mistaken output is another matter entirely.
Distinguish two questions:
(1) Are zombies logically coherent / conceivable?
(2) What cognitive processes make it seem plausible that the answer to Q1 is ‘yes’?
The first question should really be: what does the apparent conceivability of zombies by humans imply about their possibility?
Philosophers on your side of the debate seem to take it for granted (or at least end up believing) that it implies a lot, but those of us on the other side think that the answer to the cogsci question undermines that implication considerably, since it shows how we might think zombies are conceivable even when they are not.
It’s been quite a while since I was actively reading philosophy, so maybe you can tell me: are there any reasons to believe zombies are logically possible other than people’s intuitions?
The first question should really be: what does the apparent conceivability of zombies by humans imply about their possibility?
I’m aware that the LW community believes this, but I think it is incorrect. We have an epistemological dispute here about whether non-psychological facts (e.g. the fact that zombies are coherently conceivable, and not just that it seems so to me) can count as evidence. Which, again, reinforces my point that the disagreement between me and Eliezer/Lukeprog concerns epistemological principles, and not matters of empirical fact.
We have an epistemological dispute here about whether non-psychological facts (e.g. the fact that zombies are coherently conceivable, and not just that it seems so to me) can count as evidence
At least around here, “evidence (for X)” is anything which is more likely to be the case under the assumption that X is true than under the assumption that X is false. So if zombies are more likely to be conceivable if non-physicalism is true than if physicalism is true, then I for one am happy to count the conceivability of zombies as evidence for non-physicalism.
But again, the question is: how do you know that zombies are conceivable? You say that this is a non-psychological fact; that’s fine perhaps, but the only evidence for this fact that I’m aware of is psychological in nature, and this is the very psychological evidence that is undermined by cognitive science. In other words, the chain of inference still seems to be
people think zombies are conceivable ⇒ zombies are conceivable ⇒ physicalism is false
so that you still ultimately have the “work” being done by people’s intuitions.
How do you know that “people think zombies are conceivable”? Perhaps you will respond that we can know our own beliefs through introspection, and the inferential chain must stop somewhere. My view is that the relevant chain is merely like so:
zombies are conceivable ⇒ physicalism is false
I claim that we may non-inferentially know some non-psychological facts, when our beliefs in said facts meet the conditions for knowledge (exactly what these are is of course controversial, and not something we can settle in this comment thread).
I know that people think zombies are conceivable because they say they think zombies are conceivable (including, in some cases, saying “zombies are conceivable”).
To say that we may “non-inferentially know” something appears to violate the principle that beliefs require justification in order to be rational. By removing “people think zombies are conceivable”, you’ve made the argument weaker rather than stronger, because now the proposition “zombies are conceivable” has no support.
In any case, you now seem as eminently vulnerable to Eliezer’s original criticism as ever: you indeed appear to think that one can have some sort of “direct access” to the knowledge that zombies are conceivable that bypasses the cognitive processes in your brain. Or have I misunderstood?
Depending on what you mean by ‘direct access’, I suspect that you’ve probably misunderstood. But judging by the relatively low karma levels of my recent comments, going into further detail would not be of sufficient value to the LW community to be worth the time.
You’re still getting voted up on net, despite not explaining how, as you’ve claimed, the psychological fact of p-zombie plausibility is evidence for it (at least beyond references to long descriptions of your general beliefs).
I believe he’s trying to draw a distinction between two potential sources of evidence:
The factual claim that people believe zombies are conceivable, and
The actual private act of conceiving of zombies.
Richard is saying that his justification for his belief that p-zombies are conceivable lies in his successful conception of p-zombies. So what licenses him to believe that he’s successfully conceived of zombies after all? His answer is that he has direct access to the contents of his conception, in the same way that he has access to the contents of his perception. You don’t need to ask, “How do I know I’m really seeing blue right now, and not red?” Your justification for your belief that you’re seeing blue just is your phenomenal act of noticing a real, bluish sensation. This justification is “direct” insofar as it comes directly from the sensation, and not via some intermediate process of reasoning which involve inferences (which can be valid or invalid) or premises (which can be true or false). Similarly, he thinks his justification for his belief that p-zombies are conceivable just is his p-zombie-ish conception.
A couple of things to note. One is that this evidence is wholly private. You don’t have direct access to his conceptions, just as you don’t have direct access to his perceptions. The only evidence Richard can give you is testimony. Moreover, he agrees that testimony of this sort is extremely weak evidence. But it’s not the evidence he claims that his belief rests on. The evidence that Richard appeals to can be evidence-for-Richard only.
Another thing is that the direct evidence he appeals to is not “neutral.” If p-zombies really are inconceivable, then he’s in fact not conceiving of p-zombies at all, and so his conception, whatever it was, was never evidence for the conceivability of p-zombies in the first place (in just the same way that seeing red isn’t evidence that you’re seeing blue). So there’s no easy way to set aside the question of whether Richard’s conception is evidence-for-him from the question of whether p-zombies are in general conceivable. The worthiness of Richard’s source of evidence is inextricable from the actual truth or falsehood of the claim in contention, viz., that p-zombies are conceivable. But he thinks this isn’t a problem.
If you want to move ahead in the discussion, then the following are your options:
You simply deny that Richard is in fact conceiving of p-zombies. This isn’t illegitimate, but it’s going to be a conversation-stopper, since he’ll insist that he does have them but that they’re private.
You accept that Richard can successfully conceive of p-zombies, but that this isn’t good evidence for their possibility (or that the very notion of “possibility” in this context is far too problematic to be useful).
You deny that we have direct access to anything, or that access to conceptions in particular is direct, or that one can ever have private knowledge. If you go this route, you have to be careful not to set yourself up for easy reductio. Specifically, you’d better not be led to deny the rationality of believing that you’re seeing blue when, e.g., you highlight this text.
I hope this helps clear things up. It pains me when people interpret their own confusion as evidence of some deep flaw in academic philosophy.
I believe he’s trying to draw a distinction between two potential sources of evidence:
The factual claim that people believe zombies are conceivable, and
The actual private act of conceiving of zombies.
I was very deliberately ignoring this distinction: “people” includes Richard, even for Richard. The point is that Richard cannot simply trust his intuition; he has to weigh his apparent successful conception of zombies against the other evidence, such as the scientific success of reductionism, the findings from cognitive science that show how untrustworthy our intuitions are, and in particular specific arguments showing how we might fool ourselves into thinking zombies are conceivable.
The evidence that Richard appeals to can be evidence-for-Richard only
If p-zombies really are inconceivable, then he’s in fact not conceiving of p-zombies at all, and so his conception, whatever it was, was never evidence for the conceivability of p-zombies in the first place...The worthiness of Richard’s source of evidence is inextricable from the actual truth or falsehood of the claim in contention
This is a confusion of map and territory. It is possible to be rationally uncertain about logical truths; and probability estimates (which include the extent to which a datum is evidence for a proposition) are determined by the information available to the agent, not the truth or falsehood of the proposition (otherwise, the only possible probability estimates would be 1 and 0). It may be rational to assign a probability of 75% to the truth of the Riemann Hypothesis given the information we currently have, even if the Riemann Hypothesis turns out to be false (we may have misleading information).
If you want to move ahead in the discussion, then the following are your options:
My position could be described by any of those three options—in other words, they seem to differ only in the interpretation of terms like “conceivable”, and don’t properly hug the query.
1.You simply deny that Richard is in fact conceiving of p-zombies.
I must do so to the extent I believe zombies are in fact inconceivable. But I don’t see why it should be a conversation-stopper: if Richard is right and I am wrong, Richard should be able to offer evidence that he is unusually capable of determining whether his apparent conception is in fact successful (if he can’t, then he should be doubting his own successful conception himself).
2.You accept that Richard can successfully conceive of p-zombies, but that this isn’t good evidence for their possibility
I can assent to this if “conceive” is interpreted in such a way that it is possible to conceive of something that is logically impossible (i.e. if it is granted that I can conceive of Fermat’s Last Theorem being false).
3. You deny that we have direct access to anything, or that access to conceptions in particular is direct, or that one can ever have private knowledge.
“Private knowledge” in this sense is ruled out by Aumann, as far as I can tell. As for “direct access”, well, that was Eliezer’s original point, which I agree with: all knowledge is subject to some uncertainty due to the flaws in human psychology, and in particular all knowledge claims are subject to being undermined by arguments showing how the brain could generate them independently of the truth of the proposition in question. (In other words, the “genetic fallacy” is no fallacy, at least not necessarily.)
Specifically, you’d better not be led to deny the rationality of believing that you’re seeing blue when, e.g., you highlight this text.
I was very deliberately ignoring this distinction: “people” includes Richard, even for Richard. The point is that Richard cannot simply trust his intuition; he has to weigh his apparent successful conception of zombies against the other evidence, such as the scientific success of reductionism, the findings from cognitive science that show how untrustworthy our intuitions are, and in particular specific arguments showing how we might fool ourselves into thinking zombies are conceivable.
I don’t think Richard said anything to dispute this. He never said that his direct access to the conceivability of zombies renders his justification indefeasible.
This would appear to violate Aumann’s agreement theorem.
“Private knowledge” in this sense is ruled out by Aumann, as far as I can tell.
This is not a case in which you share common priors, so the theorem doesn’t apply. You don’t have, and in fact can never have, the information Richard (thinks he) has. Aumann’s theorem does not imply that everyone is capable of accessing the same evidence.
This is a confusion of map and territory. It is possible to be rationally uncertain about logical truths; and probability estimates (which include the extent to which a datum is evidence for a proposition) are determined by the information available to the agent, not the truth or falsehood of the proposition (otherwise, the only possible probability estimates would be 1 and 0). It may be rational to assign a probability of 75% to the truth of the Riemann Hypothesis given the information we currently have, even if the Riemann Hypothesis turns out to be false (we may have misleading information).
That’s certainly true, but I can’t see its relevance to what I said. In part because of some of the very reasons you name here, we can be mistaken about whether an observation O confirms a hypothesis H or not, hence whether an observation is evidence for a hypothesis or not. If the hypothesis in question concerns whether O is in fact even observable, and my evidence for ~H is that I’ve made O, then someone who strongly disagrees with me about H will conclude that I made some other observation O’ and have been mistaking it for O. And since the observability of O’ doesn’t have any evidentiary bearing on H, he’ll say, my observation wasn’t actually the evidence that I took it to be. That’s the point I was trying to illustrate: we may not be able to agree about whether my purported evidence should confirm H if we antecedently disagree about H. [Edited this sentence to make it clearer.]
But I don’t see why it should be a conversation-stopper: if Richard is right and I am wrong, Richard should be able to offer evidence that he is unusually capable of determining whether his apparent conception is in fact successful (if he can’t, then he should be doubting his own successful conception himself).
I don’t really see what this could mean.
As for “direct access”, well, that was Eliezer’s original point, which I agree with: all knowledge is subject to some uncertainty due to the flaws in human psychology, and in particular all knowledge claims are subject to being undermined by arguments showing how the brain could generate them independently of the truth of the proposition in question. (In other words, the “genetic fallacy” is no fallacy, at least not necessarily.)
Richard didn’t state that his evidence for the conceivability of zombies is absolutely incontrovertible. He just said he had direct access to it, i.e., he has extremely strong evidence for it that doesn’t follow from some intermediary inference.
This is not a case in which you share common priors
Why not?
Postulating uncommon priors is not to be done lightly: it imposes specific constraints on beliefs about priors. See Robin Hanson’s paper “Uncommon Priors Require Origin Disputes”.
In any case, what I want to know is how I should update my beliefs in light of Richard’s statements. Does he have information about the conceivability of zombies that I don’t, or is he just making a mistake?
If the hypothesis in question concerns whether O is in fact even observable, and my evidence for ~H is that I’ve made O, then someone who strongly disagrees with me about H will conclude that I made some other observation O’ and have been mistaking it for O. And since the observability of O’ doesn’t have any evidentiary bearing on H, he’ll say, my observation wasn’t actually the evidence that I took it to be. That’s the point I was trying to illustrate: we may not be able to agree about whether my purported evidence should confirm H if we antecedently disagree about H.
In such a dispute, there is some observation O″ that (both parties can agree) you made, which is equal to (or implies) either O or O’, and the dispute is about which one of these it is the same as (or implies). But since O implies H and O’ doesn’t, the dispute reduces to the question of whether O″ implies H or not, and so you may as well discuss that directly.
In the case at hand, O is “Richard has conceived of zombies”, O’ is “Richard mistakenly believes he has conceived of zombies”, and O″ is “Richard believes he has conceived of zombies”. But in the discussion so far, Richard has been resisting attempts to switch from discussing O (the subject of dispute) to discussing O″, which obviously prevents the discussion from proceeding.
Because, again, you do not have access to the same evidence (if Richard is right about the conceivability of zombies, that is!). Robin’s paper is unfortunately not going to avail you here. It applies to cases where Bayesians share all the same information but nevertheless disagree. To reiterate, Richard (as I understand him) believes that you and he do not share the same information.
In any case, what I want to know is how I should update my beliefs in light of Richard’s statements.
Well, you shouldn’t take his testimony of zombie conceivability as very good evidence of zombie conceivability. In that sense, you don’t have to sweat this conversation very much at all. This is less a debate about the conceivability of zombies and more a debate about the various dialectical positions of the parties involved in the conceivability debate. Do people who feel they can “robustly” conceive of p-zombies necessarily have to found their beliefs on publicly evaluable, “third-person” evidence? That seems to me the cornerstone of this particular discussion, rather than: Is the evidence for the conceivability of p-zombies any good?
In such a dispute, there is some observation O″ that (both parties can agree) you made, which is equal to (or implies) either O or O’, and the dispute is about which one of these it is the same as (or implies). But since O implies H and O’ doesn’t, the dispute reduces to the question of whether O″ implies H or not, and so you may as well discuss that directly.
Yes, that’s the “neutral” view of evidence Richard professed to deny.
The actual values of O and O’ at hand are “That one particular mental event which occurred in Richard’s mind at time t [when he was trying to conceive of zombies] was a conception of zombies,” and “That one particular mental event which occurred in Richard’s mind at time t was a conception of something other than zombies, or a non-conception.” The truth-value value of the O″ you provide has little bearing on either of these.
EDIT: Here’s a thought experiment that might illuminate my argument a bit. Imagine a group of evil scientists kidnaps you and implants special contact lenses which stream red light directly into your retina constantly. Your visual field is a uniformly red canvas, and you can never shut it off. The scientists then strand you on an island full of Bayesian tribespeople who are congenitally blind. The tribespeople consider the existence of visual experience ridiculous and point to all sorts of icky human biases tainting our judgment. How do you update your belief that you’re experiencing red?
EDIT 2: Looking over this once again, I think I should be less glib in my first paragraph. Note that I’m denying that you share common priors, but then appealing to different evidence that you have to explain why this can be rational. If the difference in priors is a result of the difference in evidence, aren’t they just posteriors?
The answer I personally would give is that there are different kinds of evidence. Posteriors are the result of conditionalizing on propositional evidence, such as “Snow is white.” But not all evidence is propositional. In particular, many of our introspective beliefs are justified (when they are justified at all) by the direct access we have to our own experiences. Experiences are not propositions! You cannot conditionalize on an experience. You can conditionalize on a sentence like “I am having experience E,” of course, but the evidence for that sentence is going to come from E itself, not another proposition.
Robin’s paper is unfortunately not going to avail you here. It applies to cases where Bayesians share all the same information but nevertheless disagree.
This is not correct. Even the original Aumann theorem only assumes that the Bayesians have (besides common priors) common knowledge of each other’s probability estimates—not that they share all the same information! (In fact, if they have common priors and the same information, then their posteriors are trivially equal.)
Robin’s paper imposes restrictions on being able to postulate uncommon priors as a way of escaping Aumann’s theorem: if you want to assume uncommon priors, certain consequences follow. (Roughly speaking, if Richard and I have differing priors, then we must also disagree about the origin of our priors.)
In any event, you do get closer to what I regard as the point here:
Experiences are not propositions! You cannot conditionalize on an experience.
Another term for “conditionalize” is “update”. Why can’t you update on an experience?
The sense I get is that you’re not wanting to apply the Bayesian model of belief to “experiences”. But if our “experiences” affect our beliefs, then I see no reason not to.
The actual values of O and O’ at hand are “That one particular mental event which occurred in Richard’s mind at time t [when he was trying to conceive of zombies] was a conception of zombies,” and “That one particular mental event which occurred in Richard’s mind at time t was a conception of something other than zombies, or a non-conception.” The truth-value value of the O″ you provide has little bearing on either of these.
In these terms, O″ is simply “that one particular mental event occurred in Richard’s mind”—so again, the question is what the occurrence of that mental event implies, and we should be able to bypass the dispute about whether to classify it as O or O’ by analyzing its implications directly. (The truth-value of O″ isn’t a subject of dispute; in fact O″ is chosen that way.)
Here’s a thought experiment that might illuminate my argument a bit. Imagine a group of evil scientists kidnaps you and implants special contact lenses which stream red light directly into your retina constantly. Your visual field is a uniformly red canvas, and you can never shut it off. The scientists then strand you on an island full of Bayesian tribespeople who are congenitally blind. The tribespeople consider the existence of visual experience ridiculous and point to all sorts of icky human biases tainting our judgment. How do you update your belief that you’re experiencing red?
It goes down, since the tribespeople would be more likely to say that if there is no visual experience than if there is. Of course, the amount it goes down by will depend on my other information (in particular, if I know they’re congenitally blind, that significantly weakens this evidence).
I would categorize my position as somewhere between 1 and 2, depending on what you mean by “conceiving”. I think he has a name attached to some properties associated with p-zombies and a world in which they exist, but this doesn’t mean a coherent model of such a world is possible, nor that he has one. That is, I believe that following out the necessary implications will eventually lead to contradiction. My evidence for this is quite weak, of course.
I can certainly talk about an even integer larger than two that is not expressible as the sum of two primes. But that doesn’t mean it’s logically possible. It might be, or it might not. Does a name without a full-fledged model count as conceiving, or not? Either way, it doesn’t appear to be significant evidence for.
I think they were stuck on the task of getting him to explain what that evidence was (and what evidence the access he does have gives him), which in turn was complicated by his insistence that he wasn’t referring to a psychological fact of ease of conceivability.
If it helps (which I don’t expect it does), I’ve been pursuing the trail of this (and related things) here.
Thus far his response seems to be that certain beliefs don’t require evidence (or, at least, don’t require “independent justification,” which may not be the same thing), and that his beliefs about zombies “cohere well” with his other beliefs (though I’m not sure which beliefs they cohere well with, or whether they coheres better with them than their negation does), and that there’s no reason to believe it’s false (though it’s not clear what role reasons for belief play in his decision-making in the first place).
So, the Bayesian translation of his position would seem to be that he has a high prior on zombies being conceivable. But of course, that in turn translates to “zombies are conceivable for reasons I’m not being explicit about”. Which is, naturally, the point: I’d like to know what he thinks he knows that I don’t.
Regarding coherence, and reasons to believe it’s false: the historical success of reductionism is a very good reason to believe it’s false, it seems to me. Despite Richard’s protestations, it really does appear to me that this is a case of undue reluctance on the part of philosophers to update their intuitions, or at least to let them be outweighed by something else.
Good point. I think my biggest frustration is that I can’t tell what point Richard Chappell is actually making so I can know whether I agree with it. It’s one thing to make a bad argument; it’s quite another to have a devastating argument that you keep secret.
You would probably have had more opportunity to draw it out of him if it weren’t for the karma system discouraging him from posting further on the topic. Remember that next time you’re tallying the positives and negatives of the karma system.
I don’t follow: he’s getting positive net karma from this discussion, just not as much as other posters. Very few of his comments, if any, actually went negative. In what sense in the karma system discouraging him?
Yes, slightly positive. Whether something encourages or discourages a person is a fact, not about the thing considered in itself, but about its effect on the person. The fact that the karma is slightly net positive is a fact about the thing considered in itself. The fact that he himself wrote:
But judging by the relatively low karma levels of my recent comments, going into further detail would not be of sufficient value to the LW community to be worth the time.
tells us something about its effect on the person.
Yes, he’s taking that as evidence that his posts are not valued. And indeed, like most posts that don’t (as komponisto and I noted) clearly articulate what their argument is, his posts aren’t valued (relative to others in the discussion). And he is correctly reading the evidence.
I was interpreting the concerns about “low karma being discouraging” as saying that if your karma goes negative, you actually get posting restrictions. But that’s not happening here; it’s just that Richard Chappell is being informed that his posts aren’t as valued as the others on this topic. Still positive value, mind you—just not as high as others.
In the absence of a karma system, he would either be less informed about his unhelpfulness in articulating his position, or be informed through other means. I don’t understand what your complaint is.
Yes, people who cannot articulate their position rigorously are going to have their feelings hurt at some level when people aren’t satisfied with their explanations. What does that have to do with the merits of the karma system?
Yes, people who cannot articulate their position rigorously are going to have their feelings hurt at some level when people aren’t satisfied with their explanations
You are speculating about possible reasons that people might have had for faling to award karma points.
What does that have to do with the merits of the karma system?
The position of your sentence implies that “that” refers to your speculation about the reasons that people might have had for withholding karma points. But my statement concerning the merits of the karma system had not referred to that speculation. Here is my statement again:
You would probably have had more opportunity to draw it out of him if it weren’t for the karma system discouraging him from posting further on the topic.
I am pointing out that had he not been discouraged as early as he was in the exchange, then you would probably have had more opportunity to draw him out. Do you dispute this? And then I wrote:
Remember that next time you’re tallying the positives and negatives of the karma system.
I have left it up to you to decide whether your loss of this opportunity is on the whole a positive or a negative.
You are speculating about possible reasons that people might have had for faling to award karma points.
Kind of. I was drawing on my observations about how the karma system is used. I’ve generally noticed (as have others) that people with outlier views do get modded up very highly, so long as they articulate their position clearly. For example: Mitchell Porter on QM, pjeby on PCT, lukeprog on certain matters of mainstream philosophy, Alicorn on deontology and (some) feminism, byrnema on theism, XiXiDu on LW groupthink.
Given that history, I felt safe in chalking up his “insufficiently” high karma to inscrutability rather than “He’s deviating from the party line—get him!” And you don’t get to ignore that factor (of controversial, well-articulated positions being voted up) by saying you “weren’t referring to that speculation”.
I am pointing out that had he not been discouraged as early as he was in the exchange, then you would probably have had more opportunity to draw him out. Do you dispute this?
My response is that, to the extent that convoluted, error-obscuring posting is discouraged, I’m perfectly fine with such discouragement, and I don’t want to change the karma system to be more favoring of that kind of posting.
If Richard couldn’t communicate his insight about “p-zombies being so easy to conceive of” on the first three tries, we’re probably not missing out on much by him being discouraged to post the fifty-third.
My most recent comment directed toward him was not saying, “No! Please don’t leave us! I love your deep insights!” Rather, it was saying, “Hold on—there’s an easy way to dig yourself out of this hole, as there has been the whole time. Just tell us why [...].”
Moreover, to the extent that the karma system doesn’t communicate to him what it did, that just means we’d have to do it another way, or fail to communicate it at all, neither of which is particularly appealing to me.
Thanks for laying this out. I’m one of the people who thinks philosophical zombies don’t make sense, and now I understand why—they seem like insisting that a result is possible while eliminating the process which leads to the result.
This doesn’t explain why it’s so obvious to me that pz are unfeasible and so obvious to many other people that pz at least make enough sense to be a basis for argument. Does the belief or non-belief in pz correlate with anything else?
It’s hard to be sure that I’m using the right words, but I am inclined to say that it’s actually the connection between epistemic conceivability and metaphysical possibility that I have trouble with. To illustrate the difference as I understand it, someone who does not know better can epistemically conceive that H2O is not water, but nevertheless it is metaphysically impossible that H2O is not water. I am not confident I know the meanings of the philosophical terms of the preceding comment, but employing mathematics-based meanings of the words “logic” and “coherent”, then it is perfectly logically coherent for someone who happens to be ignorant of the truth to conceive that H2O is not water, but this of course tells us very little of any significant interest about the world. It is logically coherent because try as he might, there is no way for someone ignorant of the facts to purely logically derive a contradiction from the claim that H2O is not water, and therefore reveal any logical incoherence in the claim. To my way of understanding the words, there simply is no logical incoherence in a claim considered against the background of your (incomplete) knowledge unless you can logically deduce a contradiction from inside the bounds of your own knowledge. But that’s simply not a very interesting fact if what you’re interested in is not the limitations of logic or of your knowledge but rather the nature of the world.
I know Chalmers tries to bridge the gap between epistemic conceivability and metaphysical possibility in some way, but at critical points in his argument (particularly right around where he claims to “rescue” the zombie argument and brings up “panprotopsychism”) he loses me.
My view on this question is similar to that of Eric Marcus (pdf).
When you think you’re imagining a p-zombie, all that’s happening is that you’re imagining an ordinary person and neglecting to imagine their experiences, rather than (impossibly) imagining the absence of any experience. (You can tell yourself “this person has no experiences” and then it will be true in your model that HasNoExperiences(ThisPerson) but there’s no necessary reason why a predicate called “HasNoExperiences” must track whether or not people have experiences.)
Here, I think, is how Chalmers might drive a wedge between the zombie example and the “water = H2O” example:
Imagine that we’re prescientific people familiar with a water-like substance by its everyday properties. Suppose we’re shown two theories of chemistry—the correct one under which water is H2O and another under which it’s “XYZ”—but as yet have no way of empirically distinguishing them. Then when we epistemically conceive of water being XYZ, we have a coherent picture in our minds of ‘that wet stuff we all know’ turning out to be XYZ. It isn’t water, but it’s still wet.
To epistemically but not metaphysically conceive of p-zombies would be to imagine a scenario where some physically normal people lack ‘that first-person experience thing we all know’ and yet turn out to be conscious after all. But whereas there’s a semantic gap between “wet stuff” and “real water” (such that only the latter is necessarily H2O), there doesn’t seem to be any semantic gap between “that first-person thing” and “real consciousness”. Consciousness just is that first-person thing.
Perhaps you can hear the sound of some hairs being split. I don’t think we have much difference of opinion, it’s just that the idea of “conceiving of something” is woolly and incapable of precision.
Thanks, I like the paper. I understand the core idea is that to imagine a zombie (in the relevant sense of imagine) you would have to do it first person—which you can’t do, because there is nothing first person to imagine. I find the argument for this persuasive.
And this is just what I have been thinking:
the idea of “conceiving of something” is woolly and incapable of precision.
When you think you’re imagining a p-zombie, all that’s happening is that you’re imagining an ordinary person and neglecting to imagine their experiences, rather than (impossibly) imagining the absence of any experience. (You can tell yourself “this person has no experiences” and then it will be true in your model that HasNoExperiences(ThisPerson) but there’s no necessary reason why a predicate called “HasNoExperiences” must track whether or not people have experiences.)
This is an interesting proposal, but we might ask why, if consciousness is not really distinct from the physical properties, is it so easy to imagine the physical properties without imagining consciousness? It’s not like we can imagine a microphysical duplicate of our world that’s lacking chairs. Once we’ve imagined the atoms-arranged-chairwise, that’s all it is to be a chair. It’s analytic. But there’s no such conceptual connection between neurons-instantiating-computations and consciousness, which arguably precludes identifying the two.
But there’s no such conceptual connection between neurons-instantiating-computations and consciousness
Only for people who haven’t properly internalized that they are brains. Just like people who haven’t internalized that heat is molecular motion could imagine a cold object with molecules vibrating just as fast as in a hot object.
Distinguish physical coldness from phenomenal coldness. We can imagine phenomenal coldness (i.e. the sensation) being caused by different physical states—and indeed I think this is metaphysically possible. But what’s the analogue of a zombie world in case of physical heat (as defined in terms of its functional role)? We can’t coherently imagine such a thing, because physical heat is a functional concept; anything with the same microphysical behaviour as an actual hot (cold) object would thereby be physically hot (cold). Phenomenal consciousness is not a functional concept, which makes all the difference here.
You are simply begging the question. For me philosophical zombies make exactly as much sense as cold objects that behave like hot objects in every way. I can even imagine someone accepting that molecular movement explains all observable heat phenomena, but still confused enough to ask where hot and cold come from, and whether it’s metaphysically possible for an object with a lot of molecular movement to be cold anyway. The only important difference between that sort of confusion and the whole philosophical zombie business in my eyes is that heat is a lot simpler so people are far, far less likely to be in that state of confusion.
This comment is unclear. I noted that out heat concepts are ambiguous, between what we can call physical heat (as defined by its causal-functional role) and phenomenal heat (the conscious sensations). Now you write:
I can even imagine someone accepting that molecular movement explains all observable heat phenomena, but still confused enough to ask where hot and cold come from...
Which concept of ‘hot’ and ‘cold’ are you imagining this person to be employing? If the phenomenal one, then they are (in my view) correct to see a further issue here: this is simply the consciousness debate all over again. If the physical-functional concept, then they are transparently incoherent.
Now, perhaps you are suggesting that you only have a physical-function conception of consciousness, and no essentially first-personal (phenomenal) concepts at all. In that case, we are talking past each other, because you do not have the concepts necessary to understand what I am talking about.
You are over-extending the analogy. The heat case has an analogous dichotomy (heat vs. molecular movement) to first and third person, but if you try to replace it with the very same dichotomy the analogy breaks. The people I imagine are thinking about heat as a property of the objects themselves, so non-phenomenal, but using words like functional or physical would imply accepting molecular movement as the thing itself, which they are not doing. They are talking about the same thing as physical heat, but conceptionalize it differently.
Now, perhaps you are suggesting that you only have a physical-function conception of consciousness, and no essentially first-personal (phenomenal) concepts at all.
No, and I imagine you also have some degree of separation between the concepts of physical heat and molecular movement even though you know them to be the same, so you can e. g. make sense of cartoons with freeze rays fueled by “cold energy”. The fact that I understand “first-” and “third person consciousness” to be the same thing doesn’t mean I have no idea at all what people who (IMO confusedly) treat them as different things mean when they are talking about first person consciousness.
but using words like functional or physical would imply accepting molecular movement as the thing itself, which they are not doing.
Yes and no. It’s a superficially open question what microphysical phenomena fills the macro-level functional role used to define physical heat (causing state changes, making mercury expand in the thermometer, or whatever criteria we use to identify ‘heat’ in the world). So they can have a (transparently) functional concept of heat without immediately recognizing what fills the role. But once they have all the microphysical facts—the Laplacean demon, say—it would clearly be incoherent for them to continue to see a micro-macrophysical “gap” the way that we (putatively) find a physical-phenomenal gap.
(knowledge that molecular movement is sufficient to explain observable macro-phenomena was assumed so the first half of the reply does not apply)
But once they have all the microphysical facts—the Laplacean demon, say—it would clearly be incoherent for them to continue to see a micro-macrophysical “gap” the way that we (putatively) find a physical-phenomenal gap.
You and I would agree on that, but presumably they would disagree on being incoherent. And I see no important distinction between their claim to coherence and that of philosophical zombies, other than simplicity of the subject matter.
You can show that they’re incoherent by (i) explicating their macro-level functional conception of heat, and then (ii) showing how the micro functional facts entail the macro functional facts.
The challenge posed by the zombie argument is to get the physicalist to offer an analogous response. This requires either (i) explicating our concept of phenomenal consciousness in functional terms, or else (ii) showing how functional-physical facts can entail non-functional phenomenal facts (by which I mean, facts that are expressible using non-functional phenomenal concepts).
Do you think you can do one of these? If so, which one?
You can show that they’re incoherent by (i) explicating their macro-level functional conception of heat, and then (ii) showing how the micro functional facts entail the macro functional facts.
Okay, let’s imagine this. First, to explicate “macro functional facts”, we have the examples:
causing state changes, making mercury expand in the thermometer, or whatever criteria we use to identify ‘heat’ in the world
So, you try to show someone that jiggling around the molecules of mercury will cause the mercury to expand. How exactly would you do this? I’ll try to imagine it. You present them with some mercury. You lend them an instrument which lets them see the individual molecules of the mercury. Then you start jiggling the molecules directly by some means (demonic powers maybe), and the mercury expands. Or, alternatively, you apply what they recognize as heat to mercury, and you show them that the molecules are jiggling faster. So, in experience after experience, you show them that what they recognize as heat rises if and only if the molecules jiggle faster.
This is not mere observation of correlation, because you are manipulating the molecules and the mercury by one means or another rather than passively observing.
But what they can say to you is, “I accept that there seems to be some sort of very tight relationship between the jiggling and the heat, but this doesn’t mean that the jiggling is the heat. After all, we already know that there is a tight relationship between manipulations of the brain and conscious experiences, but that doesn’t disprove dualism.”
What could you say in response? Maybe: “if you jiggle the molecules, the molecules spread apart, i.e., the mercury expands.” They could reply, “you are assuming that the molecules are identical with the mercury. But all I see is nothing but a tight correlation between where the molecules are and where the mercury is—similar to the tight correlation between where the brain is and where the conscious mind finds itself, but that doesn’t disprove dualism.”
How do you force a reluctant person to accept the identification of certain macro facts with certain micro facts?
But of course, you don’t really have to, because when people see such strong correlations, their natural inclination is to stop seeing two things and start seeing one thing. They might even lose the ability to see two things—for example, when we look at the world with our two eyes, what we see is one image with depth, rather than two flat images (though we can see the individual images by closing one eye). So of course, someone who has experienced the correlation between a micro fact and macro fact will have no trouble merging them into one fact merely seen from two perspectives (micro versus macro).
In principle, the brain could be manipulated in all sorts of ways. Nobody would be willing to submit to arbitrary manipulations, but in principle it could be done, and someone who had undergone such manipulations might develop a strong identification with his physical brain.
You can show that they’re incoherent by (i) explicating their macro-level functional conception of heat, and then (ii) showing how the micro functional facts entail the macro functional facts.
They already agree on that, just like zombie postulators will (usually?) grant that a functional view will be sufficient to explain all outward signs of consciousness. Their postulated opinion that there is something more to the question is IMO only more transparently incoherent than the equivalent. If you were claiming that the functional view was insufficient to explain people writing about conscious experience that would mean not sharing the same incoherence.
For example, assume I stubbed by toe. From my first person perspective I feel pain. From a third person perspective a nerve signal is sent to the brain and causes various parts of the neural machinery to do things. If I look at what I call “pain” from my first person perspective I can discriminate various, but perhaps not all parts of the sensation. I can feel where it comes from, spatially, and that the part of my body it comes from is that toe. From a third person perspective this information must be encoded somewhere since the person can answer the corresponding questions, or simply point, and perhaps we can already tell form neuroimaging? From an evolutionary perspective it’s obvious why that information is present.
Back to first person, I strongly want it to stop. Also verifiable and explainable. I have difficulty averting my attention, find myself physically reacting in various ways unless I consciously stop it, I have pain related associations like the word “ouch” or the color red, and so on. Nothing I can observe first person except the base signal and baggage I can deduce to have a correlate third person stands out.
The signal itself seems uninteresting enough that I’m not sure if I would even notice if it was replaced with a different signal as long as all baggage was kept the same (and that didn’t imply my memories changed to match). I’m not even completely sure that I really perceive such a base signal and it’s not just the various types of baggage bleeding together. If such a base signal is there for me to perceive and is what made me write this it obviously also must also be part of the functional side. if it isn’t it doesn’t require any explanation.
still confused enough to ask … whether it’s metaphysically possible for an object with a lot of molecular movement to be cold anyway.
Not so fast! That is possible, and that was EY’s point here:
Suppose there was a glass of water, about which, initially, you knew only that its temperature was 72 degrees. Then, suddenly, Saint Laplace reveals to you the exact locations and velocities of all the atoms in the water. You now know perfectly the state of the water, so, by the information-theoretic definition of entropy, its entropy is zero. Does that make its thermodynamic entropy zero? Is the water colder, because we know more about it?
Ignoring quantumness for the moment, the answer is: Yes! Yes it is!
And then he gave the later example of the flywheel, which we see as cooler than a set of metal atoms with the same velocity profile but which is not constrained to move in a circle:
But the more important point: Suppose you’ve got an iron flywheel that’s spinning very rapidly. That’s definitely kinetic energy, so the average kinetic energy per molecule is high. Is it heat? That particular kinetic energy, of a spinning flywheel, doesn’t look to you like heat, because you know how to extract most of it as useful work, and leave behind something colder (that is, with less mean kinetic energy per degree of freedom).
I think it does. Richard was making the point that your analogy blurs an important distinction between phenomenal heat and physical heat (thereby regressing to the original dilemma).
And it turns out this is important even in the LW perspective: the physical facts about the molecular motion are not enough to determine how hot you experience it to be (i.e. the phenomenal heat); it’s also a function of how much you know about the molecular motion.
If you met someone who said with a straight face “Of course I can imagine something that is physically identical to a chair, but lacks the fundamental chairness that chairs in our experience partake of… and is therefore merely a fake chair, although it will pass all our physical tests of being-a-chair nevertheless,” would you consider that claim sufficient evidence for the existence of a non-physical chairness?
Or would you consider other explanations for that claim more likely?
Would you change your mind if a lot of people started making that claim?
You misunderstand my position. I don’t think that people’s claims are evidence for anything.
When I invite people to imagine the zombie world, this is not because once they believe that they can do so, this belief (about their imaginative capabilities) is evidence for anything. Rather, it’s the fact that the zombie world is coherently conceivable that is the evidence, and engaging in the appropriate act of imagination is simply a psychological precondition for grasping this evidence.
That’s not to say that whenever you believe that you’ve coherently imagined X, you thereby have the fact that X is coherently conceivable amongst your evidence. For this may not be a fact at all.
(This probably won’t make sense to anyone who doesn’t know any epistemology. Basically I’m rejecting the dialectical or “neutral” view of evidence. Two participants in a debate may be unable to agree even about what the evidence is, because sometimes whether something qualifies as evidence or not will depend on which of the contending views is actually correct. Which is to reiterate that the disagreement between me and Lukeprog, say, is about epistemological principles, and not any empirical matter of fact.)
Are you assuming that in order for me to be able to justifiedly believe and reason from the premise that the zombie world is conceivable, I need to be able to give some independent justification for this belief? That way lies global skepticism.
I can tell you that the belief coheres well with my other beliefs, which is a necessary but not sufficient condition for my being justified in believing it. There’s no good reason to think that it’s false. (Though again, I don’t mean to suggest that this fact suffices to make it reasonable to believe.) Whether it’s reasonable to believe depends, in part, on facts that cannot be agreed upon within this dialectic: namely, whether there really is any contradiction in the idea.
At the moment, I’m asking you what your reasons are for believing that the zombie world is coherently conceivable; I will defer passing judgment on them until I’m confident that I understand them, as I try to avoid judging things I don’t understand.
So, no, I’m not making that assumption, though I’m not rejecting that assumption either.
Which of your other beliefs cohere better with a belief that the zombie world is coherently conceivable than with a belief that it isn’t?
When I invite people to imagine the zombie world, this is not because once they believe that they can do so, this belief (about their imaginative capabilities) is evidence for anything. Rather, it’s the fact that the zombie world is coherently conceivable that is the evidence, and engaging in the appropriate act of imagination is simply a psychological precondition for grasping this evidence.
If someone were to claim the following, would they be making the same point as you are making?
“The non-psychological fact that ‘SS0 + SS0 = SSSS0’ is a theorem of Peano arithmetic is evidence that 2 added to 2 indeed yields 4. A psychological precondition for grasping this evidence is to go through the process of mentally verifying the steps in a proof of ‘SS0 + SS0 = SSSS0’ within Peano arithmetic.
“This line of inquiry would provide evidence to the verifier that 2+2 = 4. However, properly speaking, the evidence would not be the psychological fact of the occurrence of this mental verification. Rather, the evidence is the logical fact that ‘SS0 + SS0 = SSSS0’ is a theorem of Peano arithmetic.”
Yes, when you make statements about how easy it is to imagine this thing or that thing, you do indeed seem to me to be presenting those statements as evidence of something.
If I’ve misunderstood that, then I’ll drop the subject here.
This is an interesting proposal, but we might ask why, if consciousness is not really distinct from the physical properties, is it so easy to imagine the physical properties without imagining consciousness? It’s not like we can imagine a microphysical duplicate of our world that’s lacking chairs.
But these kinds of imagining are importantly dissimilar. Compare:
1) imagine the physical properties without imagining consciousness
2) imagine a microphysical duplicate of our world that’s lacking chairs
The key phrases are: “without imagining” and “that’s lacking”. It is one thing to imagine one thing without imagining another, and quite another to imagine one thing that’s lacking another. For example, I can imagine a ball without imagining its color (indeed, as experiments have shown, we can see a ball without seeing its color), but I may not be able to imagine a ball that’s lacking color.
This is no small distinction.
To bring (2) into line with (1) we would need to change it to this:
2a) imagine a microphysical duplicate of our world without imagining chairs
And this, I submit, is possible. In fact it is possible not only to imagine a physical duplicate of our world without imagining chairs, it is (in parallel to the ball example above) possible to see a duplicate of our world (namely the world itself) without seeing (i.e. perceiving, recognizing) chairs. It’s a regular occurrence that we fail to see (to recognize) what’s right in front of us in plain view. It is furthermore possible for a creature like Laplace’s Demon to imagine every particle in the universe and all their relations to each other without recognizing, in its own imagined picture, that a certain group of particles make up a chair, etc. The Demon can in other words fail to see the forest for the trees in its own imagined world.
Now, if instead of changing (2) to bring it into line with (1), we change (1) to bring it into line with (2), we get:
1a) imagine a microphysical duplicate of our world that’s lacking consciousness
Now, your reason for denying (2) was:
Once we’ve imagined the atoms-arranged-chairwise, that’s all it is to be a chair.
Converting this, we have the following proposition:
Once we’ve imagined the atoms-arranged-personwise, that’s all it is to be a person.
But this seems to be nothing other than the issue in question, namely, the issue of whether there is anything more to being a person than atoms-arranged-personwise. If you assume that there is, then you are assuming the possibility of philosophical zombies. In other words this particular piece in the argument for the possibility of philosophical zombies assumes the possibility of philosophical zombies.
we get: 1a) imagine a microphysical duplicate of our world that’s lacking consciousness
Right, that’s the claim. I explain why I don’t think it’s question-begging here and here
How can can you perform that step unless you’ve first defined consciousness as something that’s other-than-physical?
If the “consciousness” to be imagined were something we could point to and measure, then it would be a physical property, and would thus be duplicated in our imagining. Conversely, if it is not something that we can point to and measure, then where does it exist, except in our imagination?
The logical error in the zombie argument comes from failing to realize that the mental models we build in our minds do not include a term for the mind that is building the model. When I think, “Richard is conscious”, I am describing a property of my map of the world, not a property of the world. “Conscious” is a label that I apply, to describe a collection of physical properties.
If I choose to then imagine that “Zombie Richard is not conscious”, then I am saying, “Zombie Richard has all the same properties, but is not conscious.” I can imagine this in a non-contradictory way, because “conscious” is just a label in my brain, which I can choose to apply or not apply.
All this is fine so far, until I try to apply the results of this model to the outside world, which contains no label “conscious” in the first place. The label “conscious” (like “sound” in the famous tree-forest-hearing question) is strictly something tacked on to the physical events to describe a common grouping.
In other words, my in-brain model is richer than the physical world—I can imagine things that do not correspond to the world, without contradiction in that more-expressive model.
For example, I can label Charlie Sheen as “brilliant” or “lunatic”, and ascribe these properties to the exact same behaviors. I can imagine a world in which he is a genius, and one in which he is an idiot, and yet, he remains exactly the same and does the same things. I can do this because it’s just my label—my opinion—that changes from one world to the other.
The zombie world is no different: in one world, you have the opinion that I’m conscious, and in the other, you have the opinion that I’m not. It’s your failure to notice that “conscious” is an opinion or judgment—specifically, your opinion or judgment—that makes it appear as though it is proving something more profound than the proposition that people can hold contradictory opinions about the same thing.
If you map the argument from your imagination to the real world, then you can imagine/opine that people are conscious or zombies, while the physical world remains the same. This isn’t contradictory, because it’s just an opinion, and you can change your opinion whenever you like.
The reason the zombie world doesn’t then work as an argument for non-materialism, is that it cheats by dropping out the part where the person doing the experiment is the one holding the opinion of consciousness. In your imagined world, you are implicitly holding the opinion, then when you switch to thinking about the real world, you’re ignoring the part that it’s still just you, holding an opinion about something.
we might ask why, if consciousness is not really distinct from the physical properties, is it so easy to imagine the physical properties without imagining consciousness?
And that is a question of cognitive science, is it not?
What was poor about it? The rest of your point is consistent with that wording. What would you put there instead so as to make your point more plausible?
Good question. It really needed to be stated in more objective terms (which will make the claim less plausible to you, but more logically relevant):
It’s a fact that a scenario containing a microphysical duplicate of our world but lacking chairs is incoherent. It’s not a fact that the zombie world is incoherent. (I know, we dispute this, but I’m just explaining my view here.)
With the talk of what’s easily imaginable, I invite the reader to occupy my dialectical perspective, and thus to grasp the (putative) fact under dispute; but I certainly don’t think that anything I’m saying here forces you to take my position seriously. (I agree, for example, that the psychological facts are not sufficient justification.)
Okay, but there was some evidence you were trying to draw on that you previously phrased as “it’s easy to imagine p-zombies...”—and presumably that evidence can be concisely stated, without having to learn your full dialectic perspective. Whether or not you think it’s “not a fact that the zombie world is incoherent”, there was something you thought was relevant, and that something was related (though not equivalent!) to the ease of imagining p-zombies. What was that?
(And FWIW, I do notice you are replying to many different people here and appreciate your engagement.)
This is an interesting proposal, but we might ask why, if consciousness is not really distinct from the physical properties, is it so easy to imagine the physical properties without imagining consciousness?
World-models that are deficient at this aspect of world representation in ape brains.
Once we’ve imagined the atoms-arranged-chairwise, that’s all it is to be a chair. It’s analytic. But there’s no such conceptual connection between neurons-instantiating-computations and consciousness, which arguably precludes identifying the two.
That’s true. The difference between chairs and consciousness is that chair is a 3rd person concept, whereas consciousness is a 1st person concept. Imagining a world without consciousness is easy, because we never know if there are consciousnesses or not in the world—consciousness is not an empirical data, it’s something we speculate other have by analogy with ourselves.
Or in simpler terms: we can’t see how particular physics produces particular consciousness, even if we accept in general that physics produces consciousness.
The conceivability of p-zombies doesn’t mean they are really possible, or that physicalism is false, but it does mean that our explanations are inadequate. Reductivism
is not, as it stands, an explanation of consciousness, but only a proposal of the form an explanation would have.
My view on this question is similar to that of Eric Marcus (pdf).
When you think you’re imagining a p-zombie, all that’s happening is that you’re imagining an ordinary person and neglecting to imagine their experiences, rather than (impossibly) imagining the absence of any experience. (You can tell yourself “this person has no experiences” and then it will be true in your model that “HasNoExperiences(ThisPerson)” but there’s no necessary reason why an inert predicate called “HasNoExperiences” must track whether or not people have no experiences.)
Aleph basically has it right in his reply: ‘water’ is a special case because it’s a rigid designator, picking out the actual watery stuff in all counterfactual worlds (even when some other stuff, XYZ, is the watery stuff instead of our water).
Conceiving of the “twin earth” world (where the watery stuff isn’t H2O) is indeed informative, since if this really is a coherent scenario then there really is a metaphysically possible world where the watery stuff isn’t H2O. It happens that we shouldn’t call that stuff “water”, if it differs from the watery stuff in our world, but that’s mere semantics. The reality is that there is a possible world corresponding to the one we’re (coherently) conceiving of.
Richard, I’m pretty sure I remember you treating the apparent conceivability of zombies as a primary fact about the conceivability of zombies to which you have direct access, rather than treating it as an output of some cognitive algorithm in your brain and asking what sort of thought process might have produced it.
It seems like some people are using “conceivable” to mean “imaginable at some resolution”, and some to mean “coherently imaginable at any resolution”, or something. By which I mean, the first group would say that they could conceive of “America lost the Revolutionary War” or “heavier objects fall faster” or “we are composed of sentient superstrings, and the properties of matter are their tiny, tiny emotions” or “the president has been kidnapped by ninjas”; whereas the second group would say these things are not conceivable.
As a result, group A wouldn’t really consider the conceivability of p-zombies as evidence of their possibility (well, it’d technically be extremely weak evidence), whereas group B would consider the problem of the conceivability of p-zombies as essentially equivalent to the actuality of p-zombies. (There may be other groups, such as those who think “If it’s imaginable, then it’s coherent,” but based on my brief glance the discussion hasn’t actually made it that far yet.)
Is this right? I’d think the whole thing could be resolved if you taboo’d “conceivable”...?
Talking about “the” possibility of p-zombies is pretty pointless, because of the important difference between logical and physical impossibility. Even Chalmers thinks PZs are physically/naturally impossible.
I don’t think the coherent/incoherent distinction you are making is clear. Of course, in a universe where everything is exactly the same, heavier objects would not fall faster in vacuo. But then we understand gravity and acceleration, so we can say what the contradictions would be. We don’t understand what the contradictions would be in the case of p-zombies, because we don’t have the psychophysical laws.Physicalism is Not An Explanation.
By ‘coherent’, I mean something like ‘consistent’ (to make an analogy to logic) - given all our observations, and extrapolating the concept as needed, there are no contradictions. “Heavier objects fall faster” leads to contradictions pretty quickly. Some people believe that “p-zombies are possible” (in some sense, which might match up with what you mean by either logical or physical) also leads to a contradiction, though we of course don’t understand the laws that would cause this.
This is beside the point! I’m not arguing for or against p-zombies (here), I’m saying I think the people in this argument are talking past each other because they have diverging definitions.
“Heavier objects fall faster” leads to contradictions with a theory,
If we don’t know the laws that would contradict p-zombies, there is no see-able contradiction in them, and conceivability=logical possibility follows.
“Heavier objects fall faster” is imaginable at a particular resolution. Once you ask, say, “what happens if you glue two stones together?”, it contradicts more deeply-held notions, and the concept falls apart at that resolution.
Some people believe that p-zombies are incoherent if analyzed sufficiently, or expect that they necessitate a severe contradiction of much more deeply-held beliefs.
Moreover, it is possible to hold that we don’t know the laws that would contradict p-zombies but that they are nevertheless contradicted—as it is possible to hold that things should not fall up without knowing the laws of gravitation (leaving aside that some things do fall up).
Do you disagree with my central assertion, or just my definition of coherence?
The stone-gluing can be worked around with auxilliary laws. To assume those laws are absent is to assume some other laws.
People can believe what they like. If you are going to stake a claim that there is a literal self contradiction in p-zombies, you need to say what it is. However most cases of aleged self contradiction turn out to be contradiction with unexamined background assumptions—laws, again. Talk of “resolution” is misleading: this is cognitive, not pictorial.
It is in fact the philosopher’s point that p-zombies are really, for unknown reasons, impossible. They are not arguing zombies in order to argue zombies! Non-philosophers keep misunderstanding that.
So, ah, just the latter then?
That’s all right, and I admit it’s a fuzzy term. But if you want to make any progress, I suggest you consider the former point instead.
Can you make the connection between Richard’s comment and yours clearer?
That’s a difference that doesn’t make a difference. That I can (not) conceive p-zombies can only mean that my cognitive processes produce a certain output. Whether is is somehow a mistaken output is another matter entirely.
Distinguish two questions: (1) Are zombies logically coherent / conceivable? (2) What cognitive processes make it seem plausible that the answer to Q1 is ‘yes’?
I’m fully aware that one can ask the second, cogsci question. But I don’t believe that cogsci answers the first question.
The first question should really be: what does the apparent conceivability of zombies by humans imply about their possibility?
Philosophers on your side of the debate seem to take it for granted (or at least end up believing) that it implies a lot, but those of us on the other side think that the answer to the cogsci question undermines that implication considerably, since it shows how we might think zombies are conceivable even when they are not.
It’s been quite a while since I was actively reading philosophy, so maybe you can tell me: are there any reasons to believe zombies are logically possible other than people’s intuitions?
I’m aware that the LW community believes this, but I think it is incorrect. We have an epistemological dispute here about whether non-psychological facts (e.g. the fact that zombies are coherently conceivable, and not just that it seems so to me) can count as evidence. Which, again, reinforces my point that the disagreement between me and Eliezer/Lukeprog concerns epistemological principles, and not matters of empirical fact.
For more detail, see my response to TheOtherDave downthread.
At least around here, “evidence (for X)” is anything which is more likely to be the case under the assumption that X is true than under the assumption that X is false. So if zombies are more likely to be conceivable if non-physicalism is true than if physicalism is true, then I for one am happy to count the conceivability of zombies as evidence for non-physicalism.
But again, the question is: how do you know that zombies are conceivable? You say that this is a non-psychological fact; that’s fine perhaps, but the only evidence for this fact that I’m aware of is psychological in nature, and this is the very psychological evidence that is undermined by cognitive science. In other words, the chain of inference still seems to be
people think zombies are conceivable ⇒ zombies are conceivable ⇒ physicalism is false
so that you still ultimately have the “work” being done by people’s intuitions.
How do you know that “people think zombies are conceivable”? Perhaps you will respond that we can know our own beliefs through introspection, and the inferential chain must stop somewhere. My view is that the relevant chain is merely like so:
zombies are conceivable ⇒ physicalism is false
I claim that we may non-inferentially know some non-psychological facts, when our beliefs in said facts meet the conditions for knowledge (exactly what these are is of course controversial, and not something we can settle in this comment thread).
I know that people think zombies are conceivable because they say they think zombies are conceivable (including, in some cases, saying “zombies are conceivable”).
To say that we may “non-inferentially know” something appears to violate the principle that beliefs require justification in order to be rational. By removing “people think zombies are conceivable”, you’ve made the argument weaker rather than stronger, because now the proposition “zombies are conceivable” has no support.
In any case, you now seem as eminently vulnerable to Eliezer’s original criticism as ever: you indeed appear to think that one can have some sort of “direct access” to the knowledge that zombies are conceivable that bypasses the cognitive processes in your brain. Or have I misunderstood?
Depending on what you mean by ‘direct access’, I suspect that you’ve probably misunderstood. But judging by the relatively low karma levels of my recent comments, going into further detail would not be of sufficient value to the LW community to be worth the time.
You’re still getting voted up on net, despite not explaining how, as you’ve claimed, the psychological fact of p-zombie plausibility is evidence for it (at least beyond references to long descriptions of your general beliefs).
Actually he seems to have denied this here, so at this point I’m stuck wondering what the evidence for zombie-conceivability is.
I believe he’s trying to draw a distinction between two potential sources of evidence:
The factual claim that people believe zombies are conceivable, and
The actual private act of conceiving of zombies.
Richard is saying that his justification for his belief that p-zombies are conceivable lies in his successful conception of p-zombies. So what licenses him to believe that he’s successfully conceived of zombies after all? His answer is that he has direct access to the contents of his conception, in the same way that he has access to the contents of his perception. You don’t need to ask, “How do I know I’m really seeing blue right now, and not red?” Your justification for your belief that you’re seeing blue just is your phenomenal act of noticing a real, bluish sensation. This justification is “direct” insofar as it comes directly from the sensation, and not via some intermediate process of reasoning which involve inferences (which can be valid or invalid) or premises (which can be true or false). Similarly, he thinks his justification for his belief that p-zombies are conceivable just is his p-zombie-ish conception.
A couple of things to note. One is that this evidence is wholly private. You don’t have direct access to his conceptions, just as you don’t have direct access to his perceptions. The only evidence Richard can give you is testimony. Moreover, he agrees that testimony of this sort is extremely weak evidence. But it’s not the evidence he claims that his belief rests on. The evidence that Richard appeals to can be evidence-for-Richard only.
Another thing is that the direct evidence he appeals to is not “neutral.” If p-zombies really are inconceivable, then he’s in fact not conceiving of p-zombies at all, and so his conception, whatever it was, was never evidence for the conceivability of p-zombies in the first place (in just the same way that seeing red isn’t evidence that you’re seeing blue). So there’s no easy way to set aside the question of whether Richard’s conception is evidence-for-him from the question of whether p-zombies are in general conceivable. The worthiness of Richard’s source of evidence is inextricable from the actual truth or falsehood of the claim in contention, viz., that p-zombies are conceivable. But he thinks this isn’t a problem.
If you want to move ahead in the discussion, then the following are your options:
You simply deny that Richard is in fact conceiving of p-zombies. This isn’t illegitimate, but it’s going to be a conversation-stopper, since he’ll insist that he does have them but that they’re private.
You accept that Richard can successfully conceive of p-zombies, but that this isn’t good evidence for their possibility (or that the very notion of “possibility” in this context is far too problematic to be useful).
You deny that we have direct access to anything, or that access to conceptions in particular is direct, or that one can ever have private knowledge. If you go this route, you have to be careful not to set yourself up for easy reductio. Specifically, you’d better not be led to deny the rationality of believing that you’re seeing blue when, e.g., you highlight this text.
I hope this helps clear things up. It pains me when people interpret their own confusion as evidence of some deep flaw in academic philosophy.
I was very deliberately ignoring this distinction: “people” includes Richard, even for Richard. The point is that Richard cannot simply trust his intuition; he has to weigh his apparent successful conception of zombies against the other evidence, such as the scientific success of reductionism, the findings from cognitive science that show how untrustworthy our intuitions are, and in particular specific arguments showing how we might fool ourselves into thinking zombies are conceivable.
This would appear to violate Aumann’s agreement theorem.
This is a confusion of map and territory. It is possible to be rationally uncertain about logical truths; and probability estimates (which include the extent to which a datum is evidence for a proposition) are determined by the information available to the agent, not the truth or falsehood of the proposition (otherwise, the only possible probability estimates would be 1 and 0). It may be rational to assign a probability of 75% to the truth of the Riemann Hypothesis given the information we currently have, even if the Riemann Hypothesis turns out to be false (we may have misleading information).
My position could be described by any of those three options—in other words, they seem to differ only in the interpretation of terms like “conceivable”, and don’t properly hug the query.
I must do so to the extent I believe zombies are in fact inconceivable. But I don’t see why it should be a conversation-stopper: if Richard is right and I am wrong, Richard should be able to offer evidence that he is unusually capable of determining whether his apparent conception is in fact successful (if he can’t, then he should be doubting his own successful conception himself).
I can assent to this if “conceive” is interpreted in such a way that it is possible to conceive of something that is logically impossible (i.e. if it is granted that I can conceive of Fermat’s Last Theorem being false).
“Private knowledge” in this sense is ruled out by Aumann, as far as I can tell. As for “direct access”, well, that was Eliezer’s original point, which I agree with: all knowledge is subject to some uncertainty due to the flaws in human psychology, and in particular all knowledge claims are subject to being undermined by arguments showing how the brain could generate them independently of the truth of the proposition in question. (In other words, the “genetic fallacy” is no fallacy, at least not necessarily.)
I think it’s overwhelmingly likely that I’m seeing blue, but I could turn out to be mistaken.
I don’t think Richard said anything to dispute this. He never said that his direct access to the conceivability of zombies renders his justification indefeasible.
This is not a case in which you share common priors, so the theorem doesn’t apply. You don’t have, and in fact can never have, the information Richard (thinks he) has. Aumann’s theorem does not imply that everyone is capable of accessing the same evidence.
That’s certainly true, but I can’t see its relevance to what I said. In part because of some of the very reasons you name here, we can be mistaken about whether an observation O confirms a hypothesis H or not, hence whether an observation is evidence for a hypothesis or not. If the hypothesis in question concerns whether O is in fact even observable, and my evidence for ~H is that I’ve made O, then someone who strongly disagrees with me about H will conclude that I made some other observation O’ and have been mistaking it for O. And since the observability of O’ doesn’t have any evidentiary bearing on H, he’ll say, my observation wasn’t actually the evidence that I took it to be. That’s the point I was trying to illustrate: we may not be able to agree about whether my purported evidence should confirm H if we antecedently disagree about H. [Edited this sentence to make it clearer.]
I don’t really see what this could mean.
Richard didn’t state that his evidence for the conceivability of zombies is absolutely incontrovertible. He just said he had direct access to it, i.e., he has extremely strong evidence for it that doesn’t follow from some intermediary inference.
Why not?
Postulating uncommon priors is not to be done lightly: it imposes specific constraints on beliefs about priors. See Robin Hanson’s paper “Uncommon Priors Require Origin Disputes”.
In any case, what I want to know is how I should update my beliefs in light of Richard’s statements. Does he have information about the conceivability of zombies that I don’t, or is he just making a mistake?
In such a dispute, there is some observation O″ that (both parties can agree) you made, which is equal to (or implies) either O or O’, and the dispute is about which one of these it is the same as (or implies). But since O implies H and O’ doesn’t, the dispute reduces to the question of whether O″ implies H or not, and so you may as well discuss that directly.
In the case at hand, O is “Richard has conceived of zombies”, O’ is “Richard mistakenly believes he has conceived of zombies”, and O″ is “Richard believes he has conceived of zombies”. But in the discussion so far, Richard has been resisting attempts to switch from discussing O (the subject of dispute) to discussing O″, which obviously prevents the discussion from proceeding.
Because, again, you do not have access to the same evidence (if Richard is right about the conceivability of zombies, that is!). Robin’s paper is unfortunately not going to avail you here. It applies to cases where Bayesians share all the same information but nevertheless disagree. To reiterate, Richard (as I understand him) believes that you and he do not share the same information.
Well, you shouldn’t take his testimony of zombie conceivability as very good evidence of zombie conceivability. In that sense, you don’t have to sweat this conversation very much at all. This is less a debate about the conceivability of zombies and more a debate about the various dialectical positions of the parties involved in the conceivability debate. Do people who feel they can “robustly” conceive of p-zombies necessarily have to found their beliefs on publicly evaluable, “third-person” evidence? That seems to me the cornerstone of this particular discussion, rather than: Is the evidence for the conceivability of p-zombies any good?
Yes, that’s the “neutral” view of evidence Richard professed to deny.
The actual values of O and O’ at hand are “That one particular mental event which occurred in Richard’s mind at time t [when he was trying to conceive of zombies] was a conception of zombies,” and “That one particular mental event which occurred in Richard’s mind at time t was a conception of something other than zombies, or a non-conception.” The truth-value value of the O″ you provide has little bearing on either of these.
EDIT: Here’s a thought experiment that might illuminate my argument a bit. Imagine a group of evil scientists kidnaps you and implants special contact lenses which stream red light directly into your retina constantly. Your visual field is a uniformly red canvas, and you can never shut it off. The scientists then strand you on an island full of Bayesian tribespeople who are congenitally blind. The tribespeople consider the existence of visual experience ridiculous and point to all sorts of icky human biases tainting our judgment. How do you update your belief that you’re experiencing red?
EDIT 2: Looking over this once again, I think I should be less glib in my first paragraph. Note that I’m denying that you share common priors, but then appealing to different evidence that you have to explain why this can be rational. If the difference in priors is a result of the difference in evidence, aren’t they just posteriors?
The answer I personally would give is that there are different kinds of evidence. Posteriors are the result of conditionalizing on propositional evidence, such as “Snow is white.” But not all evidence is propositional. In particular, many of our introspective beliefs are justified (when they are justified at all) by the direct access we have to our own experiences. Experiences are not propositions! You cannot conditionalize on an experience. You can conditionalize on a sentence like “I am having experience E,” of course, but the evidence for that sentence is going to come from E itself, not another proposition.
This is not correct. Even the original Aumann theorem only assumes that the Bayesians have (besides common priors) common knowledge of each other’s probability estimates—not that they share all the same information! (In fact, if they have common priors and the same information, then their posteriors are trivially equal.)
Robin’s paper imposes restrictions on being able to postulate uncommon priors as a way of escaping Aumann’s theorem: if you want to assume uncommon priors, certain consequences follow. (Roughly speaking, if Richard and I have differing priors, then we must also disagree about the origin of our priors.)
In any event, you do get closer to what I regard as the point here:
Another term for “conditionalize” is “update”. Why can’t you update on an experience?
The sense I get is that you’re not wanting to apply the Bayesian model of belief to “experiences”. But if our “experiences” affect our beliefs, then I see no reason not to.
In these terms, O″ is simply “that one particular mental event occurred in Richard’s mind”—so again, the question is what the occurrence of that mental event implies, and we should be able to bypass the dispute about whether to classify it as O or O’ by analyzing its implications directly. (The truth-value of O″ isn’t a subject of dispute; in fact O″ is chosen that way.)
It goes down, since the tribespeople would be more likely to say that if there is no visual experience than if there is. Of course, the amount it goes down by will depend on my other information (in particular, if I know they’re congenitally blind, that significantly weakens this evidence).
I would categorize my position as somewhere between 1 and 2, depending on what you mean by “conceiving”. I think he has a name attached to some properties associated with p-zombies and a world in which they exist, but this doesn’t mean a coherent model of such a world is possible, nor that he has one. That is, I believe that following out the necessary implications will eventually lead to contradiction. My evidence for this is quite weak, of course.
I can certainly talk about an even integer larger than two that is not expressible as the sum of two primes. But that doesn’t mean it’s logically possible. It might be, or it might not. Does a name without a full-fledged model count as conceiving, or not? Either way, it doesn’t appear to be significant evidence for.
I think the critics of Richard Chappell here are taking route 2 in your categorization.
komponisto and TheOtherDave appear to have been taking route 3. (challenging Richard’s purported access to evidence for zombie conceivaiblity).
I think they were stuck on the task of getting him to explain what that evidence was (and what evidence the access he does have gives him), which in turn was complicated by his insistence that he wasn’t referring to a psychological fact of ease of conceivability.
If it helps (which I don’t expect it does), I’ve been pursuing the trail of this (and related things) here.
Thus far his response seems to be that certain beliefs don’t require evidence (or, at least, don’t require “independent justification,” which may not be the same thing), and that his beliefs about zombies “cohere well” with his other beliefs (though I’m not sure which beliefs they cohere well with, or whether they coheres better with them than their negation does), and that there’s no reason to believe it’s false (though it’s not clear what role reasons for belief play in his decision-making in the first place).
So, the Bayesian translation of his position would seem to be that he has a high prior on zombies being conceivable. But of course, that in turn translates to “zombies are conceivable for reasons I’m not being explicit about”. Which is, naturally, the point: I’d like to know what he thinks he knows that I don’t.
Regarding coherence, and reasons to believe it’s false: the historical success of reductionism is a very good reason to believe it’s false, it seems to me. Despite Richard’s protestations, it really does appear to me that this is a case of undue reluctance on the part of philosophers to update their intuitions, or at least to let them be outweighed by something else.
Good point. I think my biggest frustration is that I can’t tell what point Richard Chappell is actually making so I can know whether I agree with it. It’s one thing to make a bad argument; it’s quite another to have a devastating argument that you keep secret.
You would probably have had more opportunity to draw it out of him if it weren’t for the karma system discouraging him from posting further on the topic. Remember that next time you’re tallying the positives and negatives of the karma system.
I don’t follow: he’s getting positive net karma from this discussion, just not as much as other posters. Very few of his comments, if any, actually went negative. In what sense in the karma system discouraging him?
Yes, slightly positive. Whether something encourages or discourages a person is a fact, not about the thing considered in itself, but about its effect on the person. The fact that the karma is slightly net positive is a fact about the thing considered in itself. The fact that he himself wrote:
tells us something about its effect on the person.
Yes, he’s taking that as evidence that his posts are not valued. And indeed, like most posts that don’t (as komponisto and I noted) clearly articulate what their argument is, his posts aren’t valued (relative to others in the discussion). And he is correctly reading the evidence.
I was interpreting the concerns about “low karma being discouraging” as saying that if your karma goes negative, you actually get posting restrictions. But that’s not happening here; it’s just that Richard Chappell is being informed that his posts aren’t as valued as the others on this topic. Still positive value, mind you—just not as high as others.
In the absence of a karma system, he would either be less informed about his unhelpfulness in articulating his position, or be informed through other means. I don’t understand what your complaint is.
Yes, people who cannot articulate their position rigorously are going to have their feelings hurt at some level when people aren’t satisfied with their explanations. What does that have to do with the merits of the karma system?
You are speculating about possible reasons that people might have had for faling to award karma points.
The position of your sentence implies that “that” refers to your speculation about the reasons that people might have had for withholding karma points. But my statement concerning the merits of the karma system had not referred to that speculation. Here is my statement again:
I am pointing out that had he not been discouraged as early as he was in the exchange, then you would probably have had more opportunity to draw him out. Do you dispute this? And then I wrote:
I have left it up to you to decide whether your loss of this opportunity is on the whole a positive or a negative.
Kind of. I was drawing on my observations about how the karma system is used. I’ve generally noticed (as have others) that people with outlier views do get modded up very highly, so long as they articulate their position clearly. For example: Mitchell Porter on QM, pjeby on PCT, lukeprog on certain matters of mainstream philosophy, Alicorn on deontology and (some) feminism, byrnema on theism, XiXiDu on LW groupthink.
Given that history, I felt safe in chalking up his “insufficiently” high karma to inscrutability rather than “He’s deviating from the party line—get him!” And you don’t get to ignore that factor (of controversial, well-articulated positions being voted up) by saying you “weren’t referring to that speculation”.
My response is that, to the extent that convoluted, error-obscuring posting is discouraged, I’m perfectly fine with such discouragement, and I don’t want to change the karma system to be more favoring of that kind of posting.
If Richard couldn’t communicate his insight about “p-zombies being so easy to conceive of” on the first three tries, we’re probably not missing out on much by him being discouraged to post the fifty-third.
My most recent comment directed toward him was not saying, “No! Please don’t leave us! I love your deep insights!” Rather, it was saying, “Hold on—there’s an easy way to dig yourself out of this hole, as there has been the whole time. Just tell us why [...].”
Moreover, to the extent that the karma system doesn’t communicate to him what it did, that just means we’d have to do it another way, or fail to communicate it at all, neither of which is particularly appealing to me.
Thanks for laying this out. I’m one of the people who thinks philosophical zombies don’t make sense, and now I understand why—they seem like insisting that a result is possible while eliminating the process which leads to the result.
This doesn’t explain why it’s so obvious to me that pz are unfeasible and so obvious to many other people that pz at least make enough sense to be a basis for argument. Does the belief or non-belief in pz correlate with anything else?
Since no physical law is logically necessary, it is always logically possible that an effect could fail to follow from a cause.
since “logically possible” just means “conceviable” there doesn’t need to be.
It’s hard to be sure that I’m using the right words, but I am inclined to say that it’s actually the connection between epistemic conceivability and metaphysical possibility that I have trouble with. To illustrate the difference as I understand it, someone who does not know better can epistemically conceive that H2O is not water, but nevertheless it is metaphysically impossible that H2O is not water. I am not confident I know the meanings of the philosophical terms of the preceding comment, but employing mathematics-based meanings of the words “logic” and “coherent”, then it is perfectly logically coherent for someone who happens to be ignorant of the truth to conceive that H2O is not water, but this of course tells us very little of any significant interest about the world. It is logically coherent because try as he might, there is no way for someone ignorant of the facts to purely logically derive a contradiction from the claim that H2O is not water, and therefore reveal any logical incoherence in the claim. To my way of understanding the words, there simply is no logical incoherence in a claim considered against the background of your (incomplete) knowledge unless you can logically deduce a contradiction from inside the bounds of your own knowledge. But that’s simply not a very interesting fact if what you’re interested in is not the limitations of logic or of your knowledge but rather the nature of the world.
I know Chalmers tries to bridge the gap between epistemic conceivability and metaphysical possibility in some way, but at critical points in his argument (particularly right around where he claims to “rescue” the zombie argument and brings up “panprotopsychism”) he loses me.
My view on this question is similar to that of Eric Marcus (pdf).
When you think you’re imagining a p-zombie, all that’s happening is that you’re imagining an ordinary person and neglecting to imagine their experiences, rather than (impossibly) imagining the absence of any experience. (You can tell yourself “this person has no experiences” and then it will be true in your model that HasNoExperiences(ThisPerson) but there’s no necessary reason why a predicate called “HasNoExperiences” must track whether or not people have experiences.)
Here, I think, is how Chalmers might drive a wedge between the zombie example and the “water = H2O” example:
Imagine that we’re prescientific people familiar with a water-like substance by its everyday properties. Suppose we’re shown two theories of chemistry—the correct one under which water is H2O and another under which it’s “XYZ”—but as yet have no way of empirically distinguishing them. Then when we epistemically conceive of water being XYZ, we have a coherent picture in our minds of ‘that wet stuff we all know’ turning out to be XYZ. It isn’t water, but it’s still wet.
To epistemically but not metaphysically conceive of p-zombies would be to imagine a scenario where some physically normal people lack ‘that first-person experience thing we all know’ and yet turn out to be conscious after all. But whereas there’s a semantic gap between “wet stuff” and “real water” (such that only the latter is necessarily H2O), there doesn’t seem to be any semantic gap between “that first-person thing” and “real consciousness”. Consciousness just is that first-person thing.
Perhaps you can hear the sound of some hairs being split. I don’t think we have much difference of opinion, it’s just that the idea of “conceiving of something” is woolly and incapable of precision.
Thanks, I like the paper. I understand the core idea is that to imagine a zombie (in the relevant sense of imagine) you would have to do it first person—which you can’t do, because there is nothing first person to imagine. I find the argument for this persuasive.
And this is just what I have been thinking:
This is an interesting proposal, but we might ask why, if consciousness is not really distinct from the physical properties, is it so easy to imagine the physical properties without imagining consciousness? It’s not like we can imagine a microphysical duplicate of our world that’s lacking chairs. Once we’ve imagined the atoms-arranged-chairwise, that’s all it is to be a chair. It’s analytic. But there’s no such conceptual connection between neurons-instantiating-computations and consciousness, which arguably precludes identifying the two.
Only for people who haven’t properly internalized that they are brains. Just like people who haven’t internalized that heat is molecular motion could imagine a cold object with molecules vibrating just as fast as in a hot object.
Distinguish physical coldness from phenomenal coldness. We can imagine phenomenal coldness (i.e. the sensation) being caused by different physical states—and indeed I think this is metaphysically possible. But what’s the analogue of a zombie world in case of physical heat (as defined in terms of its functional role)? We can’t coherently imagine such a thing, because physical heat is a functional concept; anything with the same microphysical behaviour as an actual hot (cold) object would thereby be physically hot (cold). Phenomenal consciousness is not a functional concept, which makes all the difference here.
You are simply begging the question. For me philosophical zombies make exactly as much sense as cold objects that behave like hot objects in every way. I can even imagine someone accepting that molecular movement explains all observable heat phenomena, but still confused enough to ask where hot and cold come from, and whether it’s metaphysically possible for an object with a lot of molecular movement to be cold anyway. The only important difference between that sort of confusion and the whole philosophical zombie business in my eyes is that heat is a lot simpler so people are far, far less likely to be in that state of confusion.
This comment is unclear. I noted that out heat concepts are ambiguous, between what we can call physical heat (as defined by its causal-functional role) and phenomenal heat (the conscious sensations). Now you write:
Which concept of ‘hot’ and ‘cold’ are you imagining this person to be employing? If the phenomenal one, then they are (in my view) correct to see a further issue here: this is simply the consciousness debate all over again. If the physical-functional concept, then they are transparently incoherent.
Now, perhaps you are suggesting that you only have a physical-function conception of consciousness, and no essentially first-personal (phenomenal) concepts at all. In that case, we are talking past each other, because you do not have the concepts necessary to understand what I am talking about.
You are over-extending the analogy. The heat case has an analogous dichotomy (heat vs. molecular movement) to first and third person, but if you try to replace it with the very same dichotomy the analogy breaks. The people I imagine are thinking about heat as a property of the objects themselves, so non-phenomenal, but using words like functional or physical would imply accepting molecular movement as the thing itself, which they are not doing. They are talking about the same thing as physical heat, but conceptionalize it differently.
No, and I imagine you also have some degree of separation between the concepts of physical heat and molecular movement even though you know them to be the same, so you can e. g. make sense of cartoons with freeze rays fueled by “cold energy”. The fact that I understand “first-” and “third person consciousness” to be the same thing doesn’t mean I have no idea at all what people who (IMO confusedly) treat them as different things mean when they are talking about first person consciousness.
Yes and no. It’s a superficially open question what microphysical phenomena fills the macro-level functional role used to define physical heat (causing state changes, making mercury expand in the thermometer, or whatever criteria we use to identify ‘heat’ in the world). So they can have a (transparently) functional concept of heat without immediately recognizing what fills the role. But once they have all the microphysical facts—the Laplacean demon, say—it would clearly be incoherent for them to continue to see a micro-macrophysical “gap” the way that we (putatively) find a physical-phenomenal gap.
(knowledge that molecular movement is sufficient to explain observable macro-phenomena was assumed so the first half of the reply does not apply)
You and I would agree on that, but presumably they would disagree on being incoherent. And I see no important distinction between their claim to coherence and that of philosophical zombies, other than simplicity of the subject matter.
You can show that they’re incoherent by (i) explicating their macro-level functional conception of heat, and then (ii) showing how the micro functional facts entail the macro functional facts.
The challenge posed by the zombie argument is to get the physicalist to offer an analogous response. This requires either (i) explicating our concept of phenomenal consciousness in functional terms, or else (ii) showing how functional-physical facts can entail non-functional phenomenal facts (by which I mean, facts that are expressible using non-functional phenomenal concepts).
Do you think you can do one of these? If so, which one?
Okay, let’s imagine this. First, to explicate “macro functional facts”, we have the examples:
So, you try to show someone that jiggling around the molecules of mercury will cause the mercury to expand. How exactly would you do this? I’ll try to imagine it. You present them with some mercury. You lend them an instrument which lets them see the individual molecules of the mercury. Then you start jiggling the molecules directly by some means (demonic powers maybe), and the mercury expands. Or, alternatively, you apply what they recognize as heat to mercury, and you show them that the molecules are jiggling faster. So, in experience after experience, you show them that what they recognize as heat rises if and only if the molecules jiggle faster.
This is not mere observation of correlation, because you are manipulating the molecules and the mercury by one means or another rather than passively observing.
But what they can say to you is, “I accept that there seems to be some sort of very tight relationship between the jiggling and the heat, but this doesn’t mean that the jiggling is the heat. After all, we already know that there is a tight relationship between manipulations of the brain and conscious experiences, but that doesn’t disprove dualism.”
What could you say in response? Maybe: “if you jiggle the molecules, the molecules spread apart, i.e., the mercury expands.” They could reply, “you are assuming that the molecules are identical with the mercury. But all I see is nothing but a tight correlation between where the molecules are and where the mercury is—similar to the tight correlation between where the brain is and where the conscious mind finds itself, but that doesn’t disprove dualism.”
How do you force a reluctant person to accept the identification of certain macro facts with certain micro facts?
But of course, you don’t really have to, because when people see such strong correlations, their natural inclination is to stop seeing two things and start seeing one thing. They might even lose the ability to see two things—for example, when we look at the world with our two eyes, what we see is one image with depth, rather than two flat images (though we can see the individual images by closing one eye). So of course, someone who has experienced the correlation between a micro fact and macro fact will have no trouble merging them into one fact merely seen from two perspectives (micro versus macro).
In principle, the brain could be manipulated in all sorts of ways. Nobody would be willing to submit to arbitrary manipulations, but in principle it could be done, and someone who had undergone such manipulations might develop a strong identification with his physical brain.
They already agree on that, just like zombie postulators will (usually?) grant that a functional view will be sufficient to explain all outward signs of consciousness. Their postulated opinion that there is something more to the question is IMO only more transparently incoherent than the equivalent. If you were claiming that the functional view was insufficient to explain people writing about conscious experience that would mean not sharing the same incoherence.
For example, assume I stubbed by toe. From my first person perspective I feel pain. From a third person perspective a nerve signal is sent to the brain and causes various parts of the neural machinery to do things. If I look at what I call “pain” from my first person perspective I can discriminate various, but perhaps not all parts of the sensation. I can feel where it comes from, spatially, and that the part of my body it comes from is that toe. From a third person perspective this information must be encoded somewhere since the person can answer the corresponding questions, or simply point, and perhaps we can already tell form neuroimaging? From an evolutionary perspective it’s obvious why that information is present.
Back to first person, I strongly want it to stop. Also verifiable and explainable. I have difficulty averting my attention, find myself physically reacting in various ways unless I consciously stop it, I have pain related associations like the word “ouch” or the color red, and so on. Nothing I can observe first person except the base signal and baggage I can deduce to have a correlate third person stands out.
The signal itself seems uninteresting enough that I’m not sure if I would even notice if it was replaced with a different signal as long as all baggage was kept the same (and that didn’t imply my memories changed to match). I’m not even completely sure that I really perceive such a base signal and it’s not just the various types of baggage bleeding together. If such a base signal is there for me to perceive and is what made me write this it obviously also must also be part of the functional side. if it isn’t it doesn’t require any explanation.
Not so fast! That is possible, and that was EY’s point here:
And then he gave the later example of the flywheel, which we see as cooler than a set of metal atoms with the same velocity profile but which is not constrained to move in a circle:
Doesn’t touch the point of the analogy though. Add “disordered” or something wherever appropriate.
I think it does. Richard was making the point that your analogy blurs an important distinction between phenomenal heat and physical heat (thereby regressing to the original dilemma).
And it turns out this is important even in the LW perspective: the physical facts about the molecular motion are not enough to determine how hot you experience it to be (i.e. the phenomenal heat); it’s also a function of how much you know about the molecular motion.
If you met someone who said with a straight face “Of course I can imagine something that is physically identical to a chair, but lacks the fundamental chairness that chairs in our experience partake of… and is therefore merely a fake chair, although it will pass all our physical tests of being-a-chair nevertheless,” would you consider that claim sufficient evidence for the existence of a non-physical chairness?
Or would you consider other explanations for that claim more likely?
Would you change your mind if a lot of people started making that claim?
You misunderstand my position. I don’t think that people’s claims are evidence for anything.
When I invite people to imagine the zombie world, this is not because once they believe that they can do so, this belief (about their imaginative capabilities) is evidence for anything. Rather, it’s the fact that the zombie world is coherently conceivable that is the evidence, and engaging in the appropriate act of imagination is simply a psychological precondition for grasping this evidence.
That’s not to say that whenever you believe that you’ve coherently imagined X, you thereby have the fact that X is coherently conceivable amongst your evidence. For this may not be a fact at all.
(This probably won’t make sense to anyone who doesn’t know any epistemology. Basically I’m rejecting the dialectical or “neutral” view of evidence. Two participants in a debate may be unable to agree even about what the evidence is, because sometimes whether something qualifies as evidence or not will depend on which of the contending views is actually correct. Which is to reiterate that the disagreement between me and Lukeprog, say, is about epistemological principles, and not any empirical matter of fact.)
I agree that your belief that you’ve coherently imagined X does not imply that X is coherently conceivable.
I agree that, if it were a fact that the zombie world were coherently conceivable, that could be evidence of something.
I don’t understand your reasons for believing that the zombie world is coherently conceivable.
Are you assuming that in order for me to be able to justifiedly believe and reason from the premise that the zombie world is conceivable, I need to be able to give some independent justification for this belief? That way lies global skepticism.
I can tell you that the belief coheres well with my other beliefs, which is a necessary but not sufficient condition for my being justified in believing it. There’s no good reason to think that it’s false. (Though again, I don’t mean to suggest that this fact suffices to make it reasonable to believe.) Whether it’s reasonable to believe depends, in part, on facts that cannot be agreed upon within this dialectic: namely, whether there really is any contradiction in the idea.
At the moment, I’m asking you what your reasons are for believing that the zombie world is coherently conceivable; I will defer passing judgment on them until I’m confident that I understand them, as I try to avoid judging things I don’t understand.
So, no, I’m not making that assumption, though I’m not rejecting that assumption either.
Which of your other beliefs cohere better with a belief that the zombie world is coherently conceivable than with a belief that it isn’t?
If someone were to claim the following, would they be making the same point as you are making?
“The non-psychological fact that ‘SS0 + SS0 = SSSS0’ is a theorem of Peano arithmetic is evidence that 2 added to 2 indeed yields 4. A psychological precondition for grasping this evidence is to go through the process of mentally verifying the steps in a proof of ‘SS0 + SS0 = SSSS0’ within Peano arithmetic.
“This line of inquiry would provide evidence to the verifier that 2+2 = 4. However, properly speaking, the evidence would not be the psychological fact of the occurrence of this mental verification. Rather, the evidence is the logical fact that ‘SS0 + SS0 = SSSS0’ is a theorem of Peano arithmetic.”
Yes, when you make statements about how easy it is to imagine this thing or that thing, you do indeed seem to me to be presenting those statements as evidence of something.
If I’ve misunderstood that, then I’ll drop the subject here.
I claim to be wearing blue today.
It’s a restricted quantifier :-)
But these kinds of imagining are importantly dissimilar. Compare:
1) imagine the physical properties without imagining consciousness
2) imagine a microphysical duplicate of our world that’s lacking chairs
The key phrases are: “without imagining” and “that’s lacking”. It is one thing to imagine one thing without imagining another, and quite another to imagine one thing that’s lacking another. For example, I can imagine a ball without imagining its color (indeed, as experiments have shown, we can see a ball without seeing its color), but I may not be able to imagine a ball that’s lacking color.
This is no small distinction.
To bring (2) into line with (1) we would need to change it to this:
2a) imagine a microphysical duplicate of our world without imagining chairs
And this, I submit, is possible. In fact it is possible not only to imagine a physical duplicate of our world without imagining chairs, it is (in parallel to the ball example above) possible to see a duplicate of our world (namely the world itself) without seeing (i.e. perceiving, recognizing) chairs. It’s a regular occurrence that we fail to see (to recognize) what’s right in front of us in plain view. It is furthermore possible for a creature like Laplace’s Demon to imagine every particle in the universe and all their relations to each other without recognizing, in its own imagined picture, that a certain group of particles make up a chair, etc. The Demon can in other words fail to see the forest for the trees in its own imagined world.
Now, if instead of changing (2) to bring it into line with (1), we change (1) to bring it into line with (2), we get:
1a) imagine a microphysical duplicate of our world that’s lacking consciousness
Now, your reason for denying (2) was:
Converting this, we have the following proposition:
But this seems to be nothing other than the issue in question, namely, the issue of whether there is anything more to being a person than atoms-arranged-personwise. If you assume that there is, then you are assuming the possibility of philosophical zombies. In other words this particular piece in the argument for the possibility of philosophical zombies assumes the possibility of philosophical zombies.
Right, that’s the claim. I explain why I don’t think it’s question-begging here and here
How can can you perform that step unless you’ve first defined consciousness as something that’s other-than-physical?
If the “consciousness” to be imagined were something we could point to and measure, then it would be a physical property, and would thus be duplicated in our imagining. Conversely, if it is not something that we can point to and measure, then where does it exist, except in our imagination?
The logical error in the zombie argument comes from failing to realize that the mental models we build in our minds do not include a term for the mind that is building the model. When I think, “Richard is conscious”, I am describing a property of my map of the world, not a property of the world. “Conscious” is a label that I apply, to describe a collection of physical properties.
If I choose to then imagine that “Zombie Richard is not conscious”, then I am saying, “Zombie Richard has all the same properties, but is not conscious.” I can imagine this in a non-contradictory way, because “conscious” is just a label in my brain, which I can choose to apply or not apply.
All this is fine so far, until I try to apply the results of this model to the outside world, which contains no label “conscious” in the first place. The label “conscious” (like “sound” in the famous tree-forest-hearing question) is strictly something tacked on to the physical events to describe a common grouping.
In other words, my in-brain model is richer than the physical world—I can imagine things that do not correspond to the world, without contradiction in that more-expressive model.
For example, I can label Charlie Sheen as “brilliant” or “lunatic”, and ascribe these properties to the exact same behaviors. I can imagine a world in which he is a genius, and one in which he is an idiot, and yet, he remains exactly the same and does the same things. I can do this because it’s just my label—my opinion—that changes from one world to the other.
The zombie world is no different: in one world, you have the opinion that I’m conscious, and in the other, you have the opinion that I’m not. It’s your failure to notice that “conscious” is an opinion or judgment—specifically, your opinion or judgment—that makes it appear as though it is proving something more profound than the proposition that people can hold contradictory opinions about the same thing.
If you map the argument from your imagination to the real world, then you can imagine/opine that people are conscious or zombies, while the physical world remains the same. This isn’t contradictory, because it’s just an opinion, and you can change your opinion whenever you like.
The reason the zombie world doesn’t then work as an argument for non-materialism, is that it cheats by dropping out the part where the person doing the experiment is the one holding the opinion of consciousness. In your imagined world, you are implicitly holding the opinion, then when you switch to thinking about the real world, you’re ignoring the part that it’s still just you, holding an opinion about something.
And that is a question of cognitive science, is it not?
Ha, indeed, poorly worded on my part :-)
What was poor about it? The rest of your point is consistent with that wording. What would you put there instead so as to make your point more plausible?
Good question. It really needed to be stated in more objective terms (which will make the claim less plausible to you, but more logically relevant):
It’s a fact that a scenario containing a microphysical duplicate of our world but lacking chairs is incoherent. It’s not a fact that the zombie world is incoherent. (I know, we dispute this, but I’m just explaining my view here.)
With the talk of what’s easily imaginable, I invite the reader to occupy my dialectical perspective, and thus to grasp the (putative) fact under dispute; but I certainly don’t think that anything I’m saying here forces you to take my position seriously. (I agree, for example, that the psychological facts are not sufficient justification.)
Okay, but there was some evidence you were trying to draw on that you previously phrased as “it’s easy to imagine p-zombies...”—and presumably that evidence can be concisely stated, without having to learn your full dialectic perspective. Whether or not you think it’s “not a fact that the zombie world is incoherent”, there was something you thought was relevant, and that something was related (though not equivalent!) to the ease of imagining p-zombies. What was that?
(And FWIW, I do notice you are replying to many different people here and appreciate your engagement.)
World-models that are deficient at this aspect of world representation in ape brains.
That’s true. The difference between chairs and consciousness is that chair is a 3rd person concept, whereas consciousness is a 1st person concept. Imagining a world without consciousness is easy, because we never know if there are consciousnesses or not in the world—consciousness is not an empirical data, it’s something we speculate other have by analogy with ourselves.
Or in simpler terms: we can’t see how particular physics produces particular consciousness, even if we accept in general that physics produces consciousness. The conceivability of p-zombies doesn’t mean they are really possible, or that physicalism is false, but it does mean that our explanations are inadequate. Reductivism is not, as it stands, an explanation of consciousness, but only a proposal of the form an explanation would have.
My view on this question is similar to that of Eric Marcus (pdf).
When you think you’re imagining a p-zombie, all that’s happening is that you’re imagining an ordinary person and neglecting to imagine their experiences, rather than (impossibly) imagining the absence of any experience. (You can tell yourself “this person has no experiences” and then it will be true in your model that “HasNoExperiences(ThisPerson)” but there’s no necessary reason why an inert predicate called “HasNoExperiences” must track whether or not people have no experiences.)
Aleph basically has it right in his reply: ‘water’ is a special case because it’s a rigid designator, picking out the actual watery stuff in all counterfactual worlds (even when some other stuff, XYZ, is the watery stuff instead of our water).
Conceiving of the “twin earth” world (where the watery stuff isn’t H2O) is indeed informative, since if this really is a coherent scenario then there really is a metaphysically possible world where the watery stuff isn’t H2O. It happens that we shouldn’t call that stuff “water”, if it differs from the watery stuff in our world, but that’s mere semantics. The reality is that there is a possible world corresponding to the one we’re (coherently) conceiving of.
For more detail, see Misusing Kripke; Misdescribing Worlds, or my undergrad thesis on Modal Rationalism