There are a few basic arguments which just don't follow to me. Maybe you're relying on some intuition I don't share?
> Thus, the argument is as follows.
>
> 1 A being could be physically identical to me but could not be conscious.
>
> 2 Two beings that are physically identical must have all physical properties in common
>
> Therefore consciousness is not a physical property.
As far as I can tell all you can conclude from this argument is that consciousness is not _necessarily_ a physical property. But as long as we're talking about thought experiments in 1 surely this can't rule out consciousness as a physical property. After all, this is a purely logical argument, devoid of any facts about the world.
> If consciousness were a reductively explainable physical property then knowing all of the facts about the brain would make it possible to know what it's like to see red, despite being colour blind. However, this is clearly impossible.
It's not obvious to me that this is impossible. It may well be, but you don't prove it.
But, maybe more importantly, it's obvious to me that if consciousness were a reductively explainable physical property then knowing all the facts about the brain would make it possible to know whether that brain was conscious, or seeing red. But I don't see why it should necessarily give knowledge of what an experience feels like.
> disembodied heat is impossible
Yet I'm sure I can imagine disembodied heat. I know that's a paradox because I have a microscopic model of heat, but if I didn't it would feel very possible.
.....
It is seeming to me like some of this confusion might stem from using the word consciousness to mean two different things: the property of having internal subjective experiences, and the property of being able to make decisions.
It seems to me like lots of these arguments are much easier to accept in the latter (basically epiphenomenological, if that's the right word) case. Though even there I have an itch telling me something is missing from real understanding.
Basically at the end of the day I see a choice between postulating new fundamental laws and postulating the existence of a hard-to-derive consequence of existing laws. And it's very unclear to me which way Occam's razor should swing.
P.S.: Right at the end you claim
> Non physicalist theories are testable and make predictions
but the links here are to two different theories where consciousness is reduced to purely physical phenomena (electromagnetic fields are as physical as any particle, and quantum physics is still physics). This confused me.
Well I have some strong sense that it’s wrong in many ways
First and foremost the claim that “all knowledge in the world cannot teach someone what’s it like to see red without explicitly seeing red” seems quite strange. People have quite complex features through which they perceive the world around them (e.g. eyes, ears, etc), so it’s fairly easy to accept from reductionalist perspective that if one hasn’t perceived a colour through seeing it, one cannot have an experience of “what it’s like to see red” if the key method to know that is through actually perceiving waves coming into one’s eyeballs and whole complex physical process that transfers information into neurons, etc
Consider the following example:
If you have some kind of AI, and you want it to know some features that are informationally orthogonal to (e.g. cannot be derived from) the train set of an AI, then it’s sort of obvious that you cannot expect an AI to extrapolate those features. If your brain is a machine that has some perceptors, and you are saying “well it’s impossible for this machine to have a state that requires explicit and specific features to be perceived yet without actually introducing those features to machine’s perceptors”, then for me it seems like begging the question.
I assume that experience of seeing red is causally explained by actually seeing waves of red spectrum going to perceptors in the eye, and causally cause some signals to be transferred through neurons to the brain, etc, etc.
So for me it seems at least plausible that one can stimulate the exact same neurons in the same way to make someone “know what it’s like to see red”
I would argue that every single argument in here that is based on my capabilities of imagination is invalid, since my imagination is completely irrelevant to the question at hand. Imagination is a concept that is derived from the inside, and therefore attempting to make arguments about the properties of the system itself by using one of its own tools subjected to its internal set of rules is circular reasoning.
"This shows consciousness is not purely physical, as we can’t imagine a carbon copy of H20 that isn’t water." - What the fuck, who are you to tell me what I can imagine? I'm sorry, I really wish I was able to find kinder words for this. First, we are comparing three atoms to the most complicated and most complex (for once it is correct to use both at the same time) mechanism of the known universe. Second, I can totally imagine a carbon copy of H2O that isn't water because my imagination is not subjected to phsyical limitations, which is a necessary condition derived from your assertion that consciousness, and therefore imagination, is not physical, which is exactly why I consider this circular reasoning. Third, that is a zero, not an O. As you can tell, I am steaming mad at this point.
"If consciousness were a reductively explainable physical property then knowing all of the facts about the brain would make it possible to know what it’s like to see red, despite being color blind. However, this is clearly impossible." - Why? Why? Whyyyyyyyyyyyy? What the fuck is even happening? What is clear about this? How arrogant can a conscious being possibly be to consider this statement obvious? I repeat, the brain is the most complicated and most complex mechanism of the known universe, and the tools that we have developed thanks to it have barely reached the point of fully simulating singular neurons. I'm sorry, but we know diddly squat about the brain.
"Knowing all the facts about the brain" is a state so far beyond anything imaginable (lol) that it is complete nonsense to use this hypothetical state to make any argument. I am getting flashbacks to Eliezer Yudkowsky's moral dilemma of "would you prefer that one person be horribly tortured for fifty years without hope or rest, or that 3^^^3 people get dust specks in their eyes?". Yes, your scope insensitivity is this bad, I genuinely believe this. 86 billion neurons is very very very far away from 3^^^3, but given the current state of affairs, it seems reasonable to me that the amount of resources it takes to model a system of 86 billion neurons in a way that makes it solvable from the outside is closer to 3^^^3 than it is to 1.
As an aside, because I am putting myself at risk of commiting a logical fallacy of at least equal egregiousness, even the assumption that "knowing all the facts about the brain" is a possibility at all seems bonkers to me, given what we know about math thanks to the incompleteness theorems. The existence of solutions to an equation (and therefore its accessibility from inside the system) does not require those equations to be computable and therefore accessible from the outside, in fact the vast majority of problems that are guaranteed to have solutions are also guaranteed to be unsolvable. If quintic equations are complex, I'm sorry, I mean complicated enough to give us headaches in this regard, how bad do you think 86 billion neurons are going to be?
I am glad that you wrote this piece and I enjoyed becoming furiously angry with it though because I thoroughly enjoy this subject. So, thank you very much.
"However, we can imagine a physical carbon copy of you that lacks consciousness. This shows consciousness is not purely physical, as we can’t imagine a carbon copy of H20 that isn’t water."
I strongly disagree with this mode of reasoning. Just because you can imagine it does not mean it can exist. Just because you can imagine a consciousness-less copy of yourself does not mean that consciousness is separate from the physical world; it just means that you don't understand it well enough.
I can imagine a carbon copy of a diesel engine that still works in a world where the autoignition point of diesel is 1000°C, but that does not mean that it would be possible in the real world — it just means that I don't understand diesel engines.
If I'm bad enough at physics, I *can* imagine a carbon copy of H20 that is not water.
Premise 1 has not been established. Chalmers says that the reason we have to accept it comes down to intuition (presumably because it is quite difficult to "prove") but I don't immediately share the intuition, and neither do many philosophers.
Have you debated James Fodor? If I'm not mistaken he is a physicalist who prefers functionalist explanations. It's plausible to me that in your water = H2O example, we experience some properties of water like wetness and what water "feels like" but we aren't conscious of the particles that constitute water - and similarly, our inability to give a reductionist explanation because we only experience mental states and have extremely limited knowledge of how the brain works could not be the basis for an inference that reductionist accounts don't work
Wow, we actually agree for once! Panpsychism is so much more plausible than reductive physicalism.
There are a few basic arguments which just don't follow to me. Maybe you're relying on some intuition I don't share?
> Thus, the argument is as follows.
>
> 1 A being could be physically identical to me but could not be conscious.
>
> 2 Two beings that are physically identical must have all physical properties in common
>
> Therefore consciousness is not a physical property.
As far as I can tell all you can conclude from this argument is that consciousness is not _necessarily_ a physical property. But as long as we're talking about thought experiments in 1 surely this can't rule out consciousness as a physical property. After all, this is a purely logical argument, devoid of any facts about the world.
> If consciousness were a reductively explainable physical property then knowing all of the facts about the brain would make it possible to know what it's like to see red, despite being colour blind. However, this is clearly impossible.
It's not obvious to me that this is impossible. It may well be, but you don't prove it.
But, maybe more importantly, it's obvious to me that if consciousness were a reductively explainable physical property then knowing all the facts about the brain would make it possible to know whether that brain was conscious, or seeing red. But I don't see why it should necessarily give knowledge of what an experience feels like.
> disembodied heat is impossible
Yet I'm sure I can imagine disembodied heat. I know that's a paradox because I have a microscopic model of heat, but if I didn't it would feel very possible.
.....
It is seeming to me like some of this confusion might stem from using the word consciousness to mean two different things: the property of having internal subjective experiences, and the property of being able to make decisions.
It seems to me like lots of these arguments are much easier to accept in the latter (basically epiphenomenological, if that's the right word) case. Though even there I have an itch telling me something is missing from real understanding.
Basically at the end of the day I see a choice between postulating new fundamental laws and postulating the existence of a hard-to-derive consequence of existing laws. And it's very unclear to me which way Occam's razor should swing.
P.S.: Right at the end you claim
> Non physicalist theories are testable and make predictions
but the links here are to two different theories where consciousness is reduced to purely physical phenomena (electromagnetic fields are as physical as any particle, and quantum physics is still physics). This confused me.
Well I have some strong sense that it’s wrong in many ways
First and foremost the claim that “all knowledge in the world cannot teach someone what’s it like to see red without explicitly seeing red” seems quite strange. People have quite complex features through which they perceive the world around them (e.g. eyes, ears, etc), so it’s fairly easy to accept from reductionalist perspective that if one hasn’t perceived a colour through seeing it, one cannot have an experience of “what it’s like to see red” if the key method to know that is through actually perceiving waves coming into one’s eyeballs and whole complex physical process that transfers information into neurons, etc
Consider the following example:
If you have some kind of AI, and you want it to know some features that are informationally orthogonal to (e.g. cannot be derived from) the train set of an AI, then it’s sort of obvious that you cannot expect an AI to extrapolate those features. If your brain is a machine that has some perceptors, and you are saying “well it’s impossible for this machine to have a state that requires explicit and specific features to be perceived yet without actually introducing those features to machine’s perceptors”, then for me it seems like begging the question.
I assume that experience of seeing red is causally explained by actually seeing waves of red spectrum going to perceptors in the eye, and causally cause some signals to be transferred through neurons to the brain, etc, etc.
So for me it seems at least plausible that one can stimulate the exact same neurons in the same way to make someone “know what it’s like to see red”
Correct me if I’m wrong, please.
I would argue that every single argument in here that is based on my capabilities of imagination is invalid, since my imagination is completely irrelevant to the question at hand. Imagination is a concept that is derived from the inside, and therefore attempting to make arguments about the properties of the system itself by using one of its own tools subjected to its internal set of rules is circular reasoning.
"This shows consciousness is not purely physical, as we can’t imagine a carbon copy of H20 that isn’t water." - What the fuck, who are you to tell me what I can imagine? I'm sorry, I really wish I was able to find kinder words for this. First, we are comparing three atoms to the most complicated and most complex (for once it is correct to use both at the same time) mechanism of the known universe. Second, I can totally imagine a carbon copy of H2O that isn't water because my imagination is not subjected to phsyical limitations, which is a necessary condition derived from your assertion that consciousness, and therefore imagination, is not physical, which is exactly why I consider this circular reasoning. Third, that is a zero, not an O. As you can tell, I am steaming mad at this point.
"If consciousness were a reductively explainable physical property then knowing all of the facts about the brain would make it possible to know what it’s like to see red, despite being color blind. However, this is clearly impossible." - Why? Why? Whyyyyyyyyyyyy? What the fuck is even happening? What is clear about this? How arrogant can a conscious being possibly be to consider this statement obvious? I repeat, the brain is the most complicated and most complex mechanism of the known universe, and the tools that we have developed thanks to it have barely reached the point of fully simulating singular neurons. I'm sorry, but we know diddly squat about the brain.
"Knowing all the facts about the brain" is a state so far beyond anything imaginable (lol) that it is complete nonsense to use this hypothetical state to make any argument. I am getting flashbacks to Eliezer Yudkowsky's moral dilemma of "would you prefer that one person be horribly tortured for fifty years without hope or rest, or that 3^^^3 people get dust specks in their eyes?". Yes, your scope insensitivity is this bad, I genuinely believe this. 86 billion neurons is very very very far away from 3^^^3, but given the current state of affairs, it seems reasonable to me that the amount of resources it takes to model a system of 86 billion neurons in a way that makes it solvable from the outside is closer to 3^^^3 than it is to 1.
As an aside, because I am putting myself at risk of commiting a logical fallacy of at least equal egregiousness, even the assumption that "knowing all the facts about the brain" is a possibility at all seems bonkers to me, given what we know about math thanks to the incompleteness theorems. The existence of solutions to an equation (and therefore its accessibility from inside the system) does not require those equations to be computable and therefore accessible from the outside, in fact the vast majority of problems that are guaranteed to have solutions are also guaranteed to be unsolvable. If quintic equations are complex, I'm sorry, I mean complicated enough to give us headaches in this regard, how bad do you think 86 billion neurons are going to be?
I am glad that you wrote this piece and I enjoyed becoming furiously angry with it though because I thoroughly enjoy this subject. So, thank you very much.
"However, we can imagine a physical carbon copy of you that lacks consciousness. This shows consciousness is not purely physical, as we can’t imagine a carbon copy of H20 that isn’t water."
I strongly disagree with this mode of reasoning. Just because you can imagine it does not mean it can exist. Just because you can imagine a consciousness-less copy of yourself does not mean that consciousness is separate from the physical world; it just means that you don't understand it well enough.
I can imagine a carbon copy of a diesel engine that still works in a world where the autoignition point of diesel is 1000°C, but that does not mean that it would be possible in the real world — it just means that I don't understand diesel engines.
If I'm bad enough at physics, I *can* imagine a carbon copy of H20 that is not water.
Premise 1 has not been established. Chalmers says that the reason we have to accept it comes down to intuition (presumably because it is quite difficult to "prove") but I don't immediately share the intuition, and neither do many philosophers.
Have you debated James Fodor? If I'm not mistaken he is a physicalist who prefers functionalist explanations. It's plausible to me that in your water = H2O example, we experience some properties of water like wetness and what water "feels like" but we aren't conscious of the particles that constitute water - and similarly, our inability to give a reductionist explanation because we only experience mental states and have extremely limited knowledge of how the brain works could not be the basis for an inference that reductionist accounts don't work