I upvoted your comment/house because I think it can be looted for valuables, but not because I think it’s sturdy enough to live in.
A lot of these so-called materialisms are actually dualisms, but they are property dualism rather than substance dualism: the mind is the brain, but it has properties like “being in a certain state of consciousness”, which are distinct from, yet somehow correlated with, properties like “being made of atoms arranged in a certain way”.
This is not true. The reductionist claim is that the arrangement of the atoms is entirely sufficient to produce consciousness, and not that there is consciousness and then the atoms. Until you shake this style of thought, you will never be able to see single-level-of-reality reductionism as anything more than a mutated form of dualism, which is not what it is.
But abstracted causal models are the whole of natural-scientific ontology at the present time, and materialists try to believe that that is the fundamental nature of reality,
No! Of course, if a more accurate map of reality is developed, the reductionists will say that “this is the closest we have to knowing the true base level of reality.” Only strawman-level reductionists will say “this is the most accurate map we have? Okay, that’s base reality.” It could be that the laws of physics do fit in 500 bits, or it could be that they’re just like onion layers and for whatever reason there is no bottom layer or no one ever finds it. But it is not the case that reductionism is the claim that the extent to which we have figured out how our subjectivitity is delusional, that that is The True Reality. But it’s far better than just plucking from the naive intuitions. We know where they came from, after all, and it wasn’t from a deep experimental study into reality.
and the aspect of reality which we experience more or less directly in subjectivity, is some sort of alien overlay.
Also not true! Why, if there was a direct one-to-one correspondence between subjective experience and reality, there would never be any surprising facts, and there would be no need to distinguish the map and the territory. In fact, I confess I have no idea what such a world would look like. What would it be like to be the universe? It is a wrong question, certainly. The subjective delusions arise from experiencing reality imperfectly, or else, once again, we would have already known about atoms and gluons and whatever-mathematics-are-really-down-there.
But I will assert emphatically that the crude reductionisms we have available to us now are radically at odds with the facts of subjective experience, and so therefore they are wrong.
And I will assert twice as emphatically that the reductionism we have available to us now, while incomplete (and knowing that it is so), is not at odds with subjective experience (they add up to normality, after all) and do more to explain the facts of subjective experience than any dualism, substance or otherwise.
You have said in the past that the computational theory of mind implies dualism. When I first saw this, I was outraged and indignant and did not wish to read any further. Later I discovered that you make much more sense than this initial impression led me to think, so I read more of your work, and yet I never found an argument that supported this claim. Do show me, if you’ve got one.
I will show, however, that even if the computational theory of mind is wrong (as implying dualism would necessarily force it to be), this does not matter for transhumanist realism. For even if you could not copy the brain on a computer, obviously the brain exists, so there is some way of creating them. It can and will be understood so that new brains can be made, even if its substrate isn’t “computations”. (I admit, though, that I have no idea what else it might be doing, that isn’t computable).
Also, curse you for getting me to write in your style of incredibly long comments!
Edit: This comment was upvoted three seconds after I posted it. I don’t know how or why.
There were and are materialisms which explicitly talk about multiple levels of reality. Someone who believes that the brain is made of atoms but that consciousness is “strongly emergent” is still a materialist—at least compared to someone else who believes in a separate soul-substance.
But yes, mostly I am saying that a lot of materialism involves stealth dualism—the materialists are property dualists and don’t realize it.
One place you can see this, is when people talk about consciousness as “how it feels to be an X”, where X is something material (or computational). For example, X may be a certain arrangement of atoms in space. And how it feels to be X is… some detailed specific conjunction of sensations, thoughts, intentions, and so on, that adds up to a single complex experience.
Obviously we could make a 3D plot of where all those atoms are, and zoom around it and into it, view it from different angles, and we’ll still see nothing but a constellation of atoms in space. You won’t “see the experience from the inside” no matter how many such views you try.
“Single level of reality” implies that there is nothing more to those atoms than what can be seen in such a view. Yet the experience is supposed to be there, somewhere. I conclude that a conventional materialist theory of consciousness involves positing that the brain has properties (the “feels” or “qualia” that make up a conscious experience) in addition to the properties already stipulated by physics.
But abstracted causal models are the whole of natural-scientific ontology at the present time, and materialists try to believe that that is the fundamental nature of reality,
No! Of course, if a more accurate map of reality is developed, the reductionists will say that “this is the closest we have to knowing the true base level of reality.” Only strawman-level reductionists will say “this is the most accurate map we have? Okay, that’s base reality.”
You’ve missed my real point. Yes, a materialist is happy to say that their currently favored model is probably not the whole story. I’m saying that all the available models will suffer from the same deficit.
Consider the argument I just gave, about how the “feels” are nowhere to be seen in the atom plot, yet they are supposed to exist, yet only atoms are supposed to exist. This is a contradiction that will not be affected by adding new atoms or rearranging the old ones. All models of the world as atoms in interaction are “abstracted causal models”, the result of a centuries-long effort to understand the world without talking about so-called secondary properties, which have to be reintroduced once you want to explain consciousness itself. And it’s at that point that these subjective properties form an “overlay”—they have to be added to the physical base.
and the aspect of reality which we experience more or less directly in subjectivity, is some sort of alien overlay.
Also not true! Why, if there was a direct one-to-one correspondence between subjective experience and reality, there would never be any surprising facts, and there would be no need to distinguish the map and the territory.
There’s supposed to be a 1-to-1 correspondence between subjective experience and the physical reality of the part of the brain responsible for being the experience—not a 1-to-1 correspondence between subjective experience and the physical world external to the brain.
I hope it’s now clear that I’m not accusing materialists of identifying their models with reality at that level. It’s the identification of experiences themselves with physical parts of the brain where the problem lies, given the physical ontology we have. Obviously, if physics already posited the existence of entities that could be straightforwardly identified with elementary qualia, the situation would be rather different.
the reductionism we have available to us now, while incomplete [...] is not at odds with subjective experience (they add up to normality, after all)
Adding up to normality is a slogan and a (doomed) aspiration here. I believe 2+2=5, I know it sounds strange, but it’s OK because it adds up to normality! Except that normality is 4, not 5. Or in this case, “normality”, i.e. reality, is that experiences exist. Even if we were to take a virtual trip through an atom-plot of a brain, and we arrived somewhere and you pointed at a specific cluster of atoms and said, “There’s part of an experience! That cluster of atoms is one pixel of a visual sensation of red”, I’m still not going to see the redness (or even see the “seeing of redness”) no matter what angle I choose to view that cluster of atoms. If the redness is there, it is there in addition to all the properties that feature in the physics.
You have said in the past that the computational theory of mind implies dualism. [...] I never found an argument that supported this claim. Do show me, if you’ve got one.
Maybe it implies trialism. We end up with three levels here: the level of atoms (i.e. the fundamental physical level), the computational state machine which describes cognition and consciousness, and the experiences which we are supposed to be explaining.
A computational theory of consciousness says that a given state of consciousness just is a particular state in a particular state machine. The argument for dualism here is similar to the argument for dualism I gave for the arrangement of atoms, except that now we’re not dealing with just one arrangement of atoms, we’re dealing with an enormous equivalence class of such arrangements—all those arrangements which instantiate the relevant state machine. Pick any instance, any individual member of that equivalence class, and the previous argument applies: you won’t “see the experience from the inside”, no matter how you examine the physical configuration. The existence of the experience somewhere “in” the configuration implies extra properties beyond the basic physical ones like position and momentum.
Systematically associating conscious states with computational states will allow you to have a systematic property dualism, but it will still be property dualism.
Okay, I understand your position much better. Here’s why it is wrong:
Your argument about arranging atoms to make consciousness
also applies to arranging atoms to make apples.
You can look at this arrangement of atoms all you want, but you still won’t “see” the appleness unless you’re some sort of lifeform that has mechanisms that recognize apples easily, like humans.
Presumably consciousness is a lot more complicated than apples, and worse yet is how it isn’t a relatively durable object that humans can experience with all of their senses (indeed, none of the classical ones). So it intuitively feels like it’s different, but that doesn’t make it so.
I will see some aspects of the apple but not others. I will see its shape, because you can make shapes by arranging atoms in space, but I won’t see its color. Then there are attributes like the fact that it grew on a tree, which I will be able to “see” if the atom-plot extends that far in space and time.
Before we go any further, I would like to know if this “counterargument by apple” is something you thought up by yourself, or if you got it from somewhere. I have an interest in knowing how these defensive memes spread.
ETA: I will try to write a little more in the way of rebuttal. But first, I will allow myself one complaint, that I have made before: arguments like this should not even be necessary. It should be obvious that, e.g., if you had a universe consisting of an arrangement of particles in space whose only properties are their relative positions, that nothing in that universe has a color. The property of being colored just does not exist there. And so, if you want to maintain that conscious mental states exist in such a universe, and that they include the experience of color, you are going to have to introduce color as an additional property somehow—a property that exists somewhere inside the assemblages of
particles that are supposed to be the experiences.
So what of the attempt to rebut this with “appleness”, as a reductio ad absurdum? Well, we can start by distinguishing between the apple that exists in the external world, the experience of the apple, and the concept of an apple. Before atomism, before neuroscience, human beings are supposedly naive realists who think that what they experience is the thing itself—though if they are grown up just a little, they will already be positing that reality is a little different to their experience, just by supposing that entities continue to exist even when they are out of sight.
But let’s suppose that we have come to believe that the world of experience is somehow just “in our minds” or “in our brains”, and that it is an imperfect image or representation of an external world. This distinction has been understood for centuries. It is presupposed by the further distinction between primary and secondary properties that has been methodologically important for the development of physics: we will develop theories of space, time, shape, and motion, but we won’t worry about color, taste, or smell, because those qualities are in the perceiver only, not in the external world.
So here I sit, I see an apple, and it looks red. The physicist tells me that the apple in the external world is not red in that way. It is a colorless object made of colorless particles, but they have the property of reflecting light at a certain wavelength, and when that arrives in my eye it stimulates my brain to construct the experience of redness with
which I am familiar. All right; it may be disorienting to the former naive realist to suppose that the external world doesn’t contain color, that it’s just an arrangement of atoms possessing the property of location but no property of coloredness. But the scientific realist just has to get used to the idea that everything they are seeing is in their head, including the colors.
But wait! Now it’s the era of neuroscience and molecular biology and cognitive science. The inside of your head is now also supposed to be made of colorless atoms. So it now seems like there’s no place left in the universe where you can find an object that is actually colored. Outside your head and inside your head, there is nothing but colorless particles arranged in space. And yet there are the colors, right in front of you. The apple looks as objectively red as it ever did.
Historically, property dualism and strong emergence has been a common response to this situation, among people who thought clearly enough to see the difficulty. For example, see Bertrand Russell writing about two types of space, physical space and subjective space. Physical space is where the atoms are located, subjective space is where the colors and the experienced objects are located.
So why don’t functionalists and other contemporary materialists openly avow property dualism? I think a lot of them just habitually associate experiences and mental activity with “brain states” and “computation”, and don’t actually notice that they are lining up two different things. The attitudes of instinctive programmers towards computers probably also contribute somehow. People get used to attributing semantic states and numerous other properties to what goes on in a
computer, and forget, or never even learn, that those attributed properties are not intrinsic properties of the physical computer, no more than the shapes of letters on a page are intrinsically connected to the sounds and the meanings that they represent. The meanings that are associated with those shapes are a product of culture and of the mental intentionality of the person actively interpreting those shapes as symbols. This also applies to just about everything that goes on in a computer. A computer is a universal state machine capable of temporarily instantiating specific state machines which can causally model just about anything. But the computer doesn’t literally contain what it is causally modeling, just as emails don’t literally contain the meanings that people extract from them.
Another confusion that occurs is treating basic sensory properties like categories. There is no reason to believe in a fundamental property of “appleness”. If I identify an object I experience as an apple, it is because it possesses a conjunction of other properties, like shape, color, perhaps taste, perhaps physical context, which lead me to deduce that this thing in front of me is one of those edible objects, grown on a plant, that I have encountered before. But consider the
properties on the basis of which that identification is made. Sometimes it is argued that, for example, “red” or “redness” is also just a category, and so if you can show that the brain is a computer which computationally classifies optical stimuli according to wavelength, you have accounted for the existence of colors. It may also be added that different cultures have different color words, whose scope is not the same, so there is no reason to believe in colors above and
beyond cognitive and cultural constructs, and wavelengths of light.
But what color categories classify are specific instances of specific shades of color. We can group and regroup the spectrum of shades differently, but in the end the instances of color have an existence independent of, and prior to, the words and categories we use to designate them. And that is the level at which the existence of color refutes any claim to the ontological completeness of a physics of colorless particles. You can organize the motions of particles so
that they form state machines undergoing conditional changes of state that can be termed “classification of stimuli”. But you do not thereby magically bring into being the existence of color itself.
Ironically, in a sense, such magic is precisely what a functionalist theory of consciousness (and of the existence of conscious persons) claims: that just the existence of the appropriate state machine is enough to guarantee the existence of the associated experience or the associated person. Since the ontological ingredients of these experiences can be lacking in the computational substrate, the implication is that they come into being when the state machine does, in a type of lawful property dualism where the fundamental laws of psychophysical parallelism refer to computational properties on the physical side.
Now of course, people who believe in mind uploading would viscerally reject the idea that they are saying that nonmaterial qualia or even nonmaterial souls would materialize when their emulation started running on the computers of the post-singularity future. That’s supposed to be a dumb idea reserved perhaps for Hollywood, and writers and an audience whose minds are still half-choked with spiritual delusions about the nature of personhood, and for whom computers and technology are just props for a new type of magic. CGI can show a misty soul congealing around the microprocessors, ghosts of the departed can show up in virtual reality, Neo can have his “matrix vision” even when
he’s unplugged and in the real world…
My thesis is that people who believe in standard materialist theories of mind, and who would pride themselves on knowing enough to reject that sort of hokum, are doing exactly the same thing on a higher level. These aren’t childish delusions because they are based on a lot of genuine knowledge. It is actually the case that you can put a chip in someone’s brain and it will restore certain simple neurological functions. It does appear that large tracts of the nervous system truly can be understood as a type of physical computer. But that’s because we are describing unconscious
activities, activities that take place “out of sight”—more precisely, out of awareness—so problems like “where is the color” don’t even arise. “Consciousness” or “experience” is the problem, because it is the repository for all the types of Being that we experience, but which are not present in the ontology of the natural sciences.
It should be obvious that, e.g., if you had a universe consisting of an arrangement of particles in space whose only properties are their relative positions, that nothing in that universe has a color.
I assume that by “color”, you mean the subjective experience of colour, not the fact that an object reflects or emits certain kinds of light. Because “reflecting and emitting certain kinds of light” can be explain in terms of “arrangement of particles”, in our universe.
And so, if you want to maintain that conscious mental states exist in such a universe, and that they include the experience of color, you are going to have to introduce color as an additional property somehow.
I bet you don’t actually think like that. If it is obvious to you that an “arrangement of particles” universe cannot have subjective experience of colour in it, that’s because in the first place, it is obvious to you that it can’t have subjective experience period.
I do not have the energy to properly respond to your comment. It is simply too long. Instead, at least for now, I will just respond to this:
Before we go any further, I would like to know if this “counterargument by apple” is something you thought up by yourself, or if you got it from somewhere.
I came up with it myself. It’s a good question, because that is not true of most of the arguments I wield.
There were and are materialisms which explicitly talk about multiple levels of reality. Someone who believes that the brain is made of atoms but that consciousness is “strongly emergent” is still a materialist—at least compared to someone else who believes in a separate soul-substance
The problem with “strong emergence” is that it can be used to “explain” anything and is thus worthless.
If evolutionary biology could explain a toaster oven, not just a tree, it would be worthless. There’s a lot more to evolutionary theory than pointing at Nature and saying, “Now purpose is allowed,” or “Evolution did it!” The strength of a theory is not what it allows, but what it prohibits; if you can invent an equally persuasive explanation for any outcome, you have zero knowledge.
I like quantum mind, but despite the unity of superpositions matching the apparent unity of subjective experience, does it really give us much? I think the answer is no, at least until we have a better understanding of the physics of (quantum) computation, a better theory of computation in light of that, and a highly advanced computationalism/monadology in light of that. And even then Leibniz’ solution to the mind-body problem was literally Goddidit. (Which is an intriguing and coherent theory that explains all the evidence, but you’d think there’d be something better. Also Leibniz’ God causally influences monads, which aren’t supposed to be influence-able, so his metaphysic seems sort of broken, even if you can fix that bug with a neat trick or two maybe.) Quantum mind might help us do uploads, but it still wouldn’t have the answer to the mind-body problems, we still wouldn’t know if the uploads were conscious. Or is apparently matching a phenomenological property with a physical property (unity of experience/superposition) somehow a big philosophical step in the right direction?
I upvoted your comment/house because I think it can be looted for valuables, but not because I think it’s sturdy enough to live in.
This is not true. The reductionist claim is that the arrangement of the atoms is entirely sufficient to produce consciousness, and not that there is consciousness and then the atoms. Until you shake this style of thought, you will never be able to see single-level-of-reality reductionism as anything more than a mutated form of dualism, which is not what it is.
No! Of course, if a more accurate map of reality is developed, the reductionists will say that “this is the closest we have to knowing the true base level of reality.” Only strawman-level reductionists will say “this is the most accurate map we have? Okay, that’s base reality.” It could be that the laws of physics do fit in 500 bits, or it could be that they’re just like onion layers and for whatever reason there is no bottom layer or no one ever finds it. But it is not the case that reductionism is the claim that the extent to which we have figured out how our subjectivitity is delusional, that that is The True Reality. But it’s far better than just plucking from the naive intuitions. We know where they came from, after all, and it wasn’t from a deep experimental study into reality.
Also not true! Why, if there was a direct one-to-one correspondence between subjective experience and reality, there would never be any surprising facts, and there would be no need to distinguish the map and the territory. In fact, I confess I have no idea what such a world would look like. What would it be like to be the universe? It is a wrong question, certainly. The subjective delusions arise from experiencing reality imperfectly, or else, once again, we would have already known about atoms and gluons and whatever-mathematics-are-really-down-there.
And I will assert twice as emphatically that the reductionism we have available to us now, while incomplete (and knowing that it is so), is not at odds with subjective experience (they add up to normality, after all) and do more to explain the facts of subjective experience than any dualism, substance or otherwise.
You have said in the past that the computational theory of mind implies dualism. When I first saw this, I was outraged and indignant and did not wish to read any further. Later I discovered that you make much more sense than this initial impression led me to think, so I read more of your work, and yet I never found an argument that supported this claim. Do show me, if you’ve got one.
I will show, however, that even if the computational theory of mind is wrong (as implying dualism would necessarily force it to be), this does not matter for transhumanist realism. For even if you could not copy the brain on a computer, obviously the brain exists, so there is some way of creating them. It can and will be understood so that new brains can be made, even if its substrate isn’t “computations”. (I admit, though, that I have no idea what else it might be doing, that isn’t computable).
Also, curse you for getting me to write in your style of incredibly long comments!
Edit: This comment was upvoted three seconds after I posted it. I don’t know how or why.
There were and are materialisms which explicitly talk about multiple levels of reality. Someone who believes that the brain is made of atoms but that consciousness is “strongly emergent” is still a materialist—at least compared to someone else who believes in a separate soul-substance.
But yes, mostly I am saying that a lot of materialism involves stealth dualism—the materialists are property dualists and don’t realize it.
One place you can see this, is when people talk about consciousness as “how it feels to be an X”, where X is something material (or computational). For example, X may be a certain arrangement of atoms in space. And how it feels to be X is… some detailed specific conjunction of sensations, thoughts, intentions, and so on, that adds up to a single complex experience.
Obviously we could make a 3D plot of where all those atoms are, and zoom around it and into it, view it from different angles, and we’ll still see nothing but a constellation of atoms in space. You won’t “see the experience from the inside” no matter how many such views you try.
“Single level of reality” implies that there is nothing more to those atoms than what can be seen in such a view. Yet the experience is supposed to be there, somewhere. I conclude that a conventional materialist theory of consciousness involves positing that the brain has properties (the “feels” or “qualia” that make up a conscious experience) in addition to the properties already stipulated by physics.
You’ve missed my real point. Yes, a materialist is happy to say that their currently favored model is probably not the whole story. I’m saying that all the available models will suffer from the same deficit.
Consider the argument I just gave, about how the “feels” are nowhere to be seen in the atom plot, yet they are supposed to exist, yet only atoms are supposed to exist. This is a contradiction that will not be affected by adding new atoms or rearranging the old ones. All models of the world as atoms in interaction are “abstracted causal models”, the result of a centuries-long effort to understand the world without talking about so-called secondary properties, which have to be reintroduced once you want to explain consciousness itself. And it’s at that point that these subjective properties form an “overlay”—they have to be added to the physical base.
There’s supposed to be a 1-to-1 correspondence between subjective experience and the physical reality of the part of the brain responsible for being the experience—not a 1-to-1 correspondence between subjective experience and the physical world external to the brain.
I hope it’s now clear that I’m not accusing materialists of identifying their models with reality at that level. It’s the identification of experiences themselves with physical parts of the brain where the problem lies, given the physical ontology we have. Obviously, if physics already posited the existence of entities that could be straightforwardly identified with elementary qualia, the situation would be rather different.
Adding up to normality is a slogan and a (doomed) aspiration here. I believe 2+2=5, I know it sounds strange, but it’s OK because it adds up to normality! Except that normality is 4, not 5. Or in this case, “normality”, i.e. reality, is that experiences exist. Even if we were to take a virtual trip through an atom-plot of a brain, and we arrived somewhere and you pointed at a specific cluster of atoms and said, “There’s part of an experience! That cluster of atoms is one pixel of a visual sensation of red”, I’m still not going to see the redness (or even see the “seeing of redness”) no matter what angle I choose to view that cluster of atoms. If the redness is there, it is there in addition to all the properties that feature in the physics.
Maybe it implies trialism. We end up with three levels here: the level of atoms (i.e. the fundamental physical level), the computational state machine which describes cognition and consciousness, and the experiences which we are supposed to be explaining.
A computational theory of consciousness says that a given state of consciousness just is a particular state in a particular state machine. The argument for dualism here is similar to the argument for dualism I gave for the arrangement of atoms, except that now we’re not dealing with just one arrangement of atoms, we’re dealing with an enormous equivalence class of such arrangements—all those arrangements which instantiate the relevant state machine. Pick any instance, any individual member of that equivalence class, and the previous argument applies: you won’t “see the experience from the inside”, no matter how you examine the physical configuration. The existence of the experience somewhere “in” the configuration implies extra properties beyond the basic physical ones like position and momentum.
Systematically associating conscious states with computational states will allow you to have a systematic property dualism, but it will still be property dualism.
Okay, I understand your position much better. Here’s why it is wrong:
You can look at this arrangement of atoms all you want, but you still won’t “see” the appleness unless you’re some sort of lifeform that has mechanisms that recognize apples easily, like humans.
Presumably consciousness is a lot more complicated than apples, and worse yet is how it isn’t a relatively durable object that humans can experience with all of their senses (indeed, none of the classical ones). So it intuitively feels like it’s different, but that doesn’t make it so.
I will see some aspects of the apple but not others. I will see its shape, because you can make shapes by arranging atoms in space, but I won’t see its color. Then there are attributes like the fact that it grew on a tree, which I will be able to “see” if the atom-plot extends that far in space and time.
Before we go any further, I would like to know if this “counterargument by apple” is something you thought up by yourself, or if you got it from somewhere. I have an interest in knowing how these defensive memes spread.
ETA: I will try to write a little more in the way of rebuttal. But first, I will allow myself one complaint, that I have made before: arguments like this should not even be necessary. It should be obvious that, e.g., if you had a universe consisting of an arrangement of particles in space whose only properties are their relative positions, that nothing in that universe has a color. The property of being colored just does not exist there. And so, if you want to maintain that conscious mental states exist in such a universe, and that they include the experience of color, you are going to have to introduce color as an additional property somehow—a property that exists somewhere inside the assemblages of particles that are supposed to be the experiences.
So what of the attempt to rebut this with “appleness”, as a reductio ad absurdum? Well, we can start by distinguishing between the apple that exists in the external world, the experience of the apple, and the concept of an apple. Before atomism, before neuroscience, human beings are supposedly naive realists who think that what they experience is the thing itself—though if they are grown up just a little, they will already be positing that reality is a little different to their experience, just by supposing that entities continue to exist even when they are out of sight.
But let’s suppose that we have come to believe that the world of experience is somehow just “in our minds” or “in our brains”, and that it is an imperfect image or representation of an external world. This distinction has been understood for centuries. It is presupposed by the further distinction between primary and secondary properties that has been methodologically important for the development of physics: we will develop theories of space, time, shape, and motion, but we won’t worry about color, taste, or smell, because those qualities are in the perceiver only, not in the external world.
So here I sit, I see an apple, and it looks red. The physicist tells me that the apple in the external world is not red in that way. It is a colorless object made of colorless particles, but they have the property of reflecting light at a certain wavelength, and when that arrives in my eye it stimulates my brain to construct the experience of redness with which I am familiar. All right; it may be disorienting to the former naive realist to suppose that the external world doesn’t contain color, that it’s just an arrangement of atoms possessing the property of location but no property of coloredness. But the scientific realist just has to get used to the idea that everything they are seeing is in their head, including the colors.
But wait! Now it’s the era of neuroscience and molecular biology and cognitive science. The inside of your head is now also supposed to be made of colorless atoms. So it now seems like there’s no place left in the universe where you can find an object that is actually colored. Outside your head and inside your head, there is nothing but colorless particles arranged in space. And yet there are the colors, right in front of you. The apple looks as objectively red as it ever did.
Historically, property dualism and strong emergence has been a common response to this situation, among people who thought clearly enough to see the difficulty. For example, see Bertrand Russell writing about two types of space, physical space and subjective space. Physical space is where the atoms are located, subjective space is where the colors and the experienced objects are located.
So why don’t functionalists and other contemporary materialists openly avow property dualism? I think a lot of them just habitually associate experiences and mental activity with “brain states” and “computation”, and don’t actually notice that they are lining up two different things. The attitudes of instinctive programmers towards computers probably also contribute somehow. People get used to attributing semantic states and numerous other properties to what goes on in a computer, and forget, or never even learn, that those attributed properties are not intrinsic properties of the physical computer, no more than the shapes of letters on a page are intrinsically connected to the sounds and the meanings that they represent. The meanings that are associated with those shapes are a product of culture and of the mental intentionality of the person actively interpreting those shapes as symbols. This also applies to just about everything that goes on in a computer. A computer is a universal state machine capable of temporarily instantiating specific state machines which can causally model just about anything. But the computer doesn’t literally contain what it is causally modeling, just as emails don’t literally contain the meanings that people extract from them.
Another confusion that occurs is treating basic sensory properties like categories. There is no reason to believe in a fundamental property of “appleness”. If I identify an object I experience as an apple, it is because it possesses a conjunction of other properties, like shape, color, perhaps taste, perhaps physical context, which lead me to deduce that this thing in front of me is one of those edible objects, grown on a plant, that I have encountered before. But consider the properties on the basis of which that identification is made. Sometimes it is argued that, for example, “red” or “redness” is also just a category, and so if you can show that the brain is a computer which computationally classifies optical stimuli according to wavelength, you have accounted for the existence of colors. It may also be added that different cultures have different color words, whose scope is not the same, so there is no reason to believe in colors above and beyond cognitive and cultural constructs, and wavelengths of light.
But what color categories classify are specific instances of specific shades of color. We can group and regroup the spectrum of shades differently, but in the end the instances of color have an existence independent of, and prior to, the words and categories we use to designate them. And that is the level at which the existence of color refutes any claim to the ontological completeness of a physics of colorless particles. You can organize the motions of particles so that they form state machines undergoing conditional changes of state that can be termed “classification of stimuli”. But you do not thereby magically bring into being the existence of color itself.
Ironically, in a sense, such magic is precisely what a functionalist theory of consciousness (and of the existence of conscious persons) claims: that just the existence of the appropriate state machine is enough to guarantee the existence of the associated experience or the associated person. Since the ontological ingredients of these experiences can be lacking in the computational substrate, the implication is that they come into being when the state machine does, in a type of lawful property dualism where the fundamental laws of psychophysical parallelism refer to computational properties on the physical side.
Now of course, people who believe in mind uploading would viscerally reject the idea that they are saying that nonmaterial qualia or even nonmaterial souls would materialize when their emulation started running on the computers of the post-singularity future. That’s supposed to be a dumb idea reserved perhaps for Hollywood, and writers and an audience whose minds are still half-choked with spiritual delusions about the nature of personhood, and for whom computers and technology are just props for a new type of magic. CGI can show a misty soul congealing around the microprocessors, ghosts of the departed can show up in virtual reality, Neo can have his “matrix vision” even when he’s unplugged and in the real world…
My thesis is that people who believe in standard materialist theories of mind, and who would pride themselves on knowing enough to reject that sort of hokum, are doing exactly the same thing on a higher level. These aren’t childish delusions because they are based on a lot of genuine knowledge. It is actually the case that you can put a chip in someone’s brain and it will restore certain simple neurological functions. It does appear that large tracts of the nervous system truly can be understood as a type of physical computer. But that’s because we are describing unconscious activities, activities that take place “out of sight”—more precisely, out of awareness—so problems like “where is the color” don’t even arise. “Consciousness” or “experience” is the problem, because it is the repository for all the types of Being that we experience, but which are not present in the ontology of the natural sciences.
I assume that by “color”, you mean the subjective experience of colour, not the fact that an object reflects or emits certain kinds of light. Because “reflecting and emitting certain kinds of light” can be explain in terms of “arrangement of particles”, in our universe.
I bet you don’t actually think like that. If it is obvious to you that an “arrangement of particles” universe cannot have subjective experience of colour in it, that’s because in the first place, it is obvious to you that it can’t have subjective experience period.
I do not have the energy to properly respond to your comment. It is simply too long. Instead, at least for now, I will just respond to this:
I came up with it myself. It’s a good question, because that is not true of most of the arguments I wield.
The problem with “strong emergence” is that it can be used to “explain” anything and is thus worthless.
Eliezer Yudkowsky http://lesswrong.com/lw/kr/an_alien_god
Probably someone saw your comment/house analogy and found it very clever and upvoted before reading on.
All cleverness credit goes to steven0461.
I like quantum mind, but despite the unity of superpositions matching the apparent unity of subjective experience, does it really give us much? I think the answer is no, at least until we have a better understanding of the physics of (quantum) computation, a better theory of computation in light of that, and a highly advanced computationalism/monadology in light of that. And even then Leibniz’ solution to the mind-body problem was literally Goddidit. (Which is an intriguing and coherent theory that explains all the evidence, but you’d think there’d be something better. Also Leibniz’ God causally influences monads, which aren’t supposed to be influence-able, so his metaphysic seems sort of broken, even if you can fix that bug with a neat trick or two maybe.) Quantum mind might help us do uploads, but it still wouldn’t have the answer to the mind-body problems, we still wouldn’t know if the uploads were conscious. Or is apparently matching a phenomenological property with a physical property (unity of experience/superposition) somehow a big philosophical step in the right direction?