This argument and Jackson’s Knowledge Argument are fairly similar. The most relevant line of response to the Knowledge Argument to this one is the phenomenal concepts strategy. I’m skeptical that it works, but it’s worth thinking through the details.
The Knowledge Argument straightforwardly begs the question by stipulating that Mary knows all the physical facts but doesn't know the "what-it's-likeness" of consciousness. The physicalist obviously disputes that is possible... otherwise they wouldn't be a physicalist. The Knowledge Argument does nothing to advance the debate between the physicalist and nonphysicalist, and its popularity is a mark against the supposed clarity and rigor of analytic philosophy.
Given that only Dennett went that way and many other physicalists (including Jackson, who changed his mind) nevertheless wrote responses, your ‘obvious’ here is not obviously apt.
What independent evidence does Frank Jackson give that justifies the claim that Mary learns something new when seeing red for the first time? He only says something along the lines of "It just seems obvious Mary learns something new." This is begging the question.
Compare: Mary knows all the physical facts, so obviously she learns nothing new - she already knew everything about what it's like to see red. Obviously a nonphysicalist isn't going to be moved by this argument - because I'm just asserting the negation of the nonphysicalist's thesis - that Mary doesn't know what it's like from all the physical facts.
It's easy to see the representation of the purple square when we look at your brain, just as we see a representation of a purple square in a jpeg of a purple square on a hard drive.
Likewise, if you open the hard drive you will also not see a purple square just a certain state of the matter.
And yet there is a tangible sense in which one hard drive contains purple square while other does not. Certain state of the matter encode a specific images which can be decoded and shown.
That's a fair response. I suppose after reading Nick Humphrey's "Sentience" I'm really tempted by illusionist views, as absurd and offensive as they first sounded to me. Maybe the purple square you are imagining doesn't really exist and your brain only believes it does.
>It's easy to see the representation of the purple square when we look at your brain
Point of order, it's not actually easy at all. We do not currently have the capability of identifying the neurons that encode a perception of anything. I can understand having faith that we will someday, but still, we can't do that currently.
Other than that I agree with Philip, when my monitor creates a purple square that purple square is made of red, blue, and green light. What is the purple square in my mind's eye made of?
We may not be able to identify the _specific_ neurons that can do it, but that seems unimportant, because we can build an algorithm that can reconstruct the contents of your visual field based on fMRI representations of neural activity.
Yeah the purple square is made out of RGB light, and there isn't an entire network of interconnected protocols, rendering engines, parallelized computations in a GPU directed by a multiprocessor CPU, operating systems, and distributed databases and servers making that purple square possible. You just look and see whether there's purple light there, just like you look at the brain and don't see purple neurons so you know you've done your due diligence in understanding the problem and what you're asking.
I don't dispute the protocols, engines, hardware, or software, I just note that the process ends in red, blue, and green light. I can observe that light, I can measure that light, that light explains why I see a purple square: because there is a purple square made of light out there that I'm observing. But what am I observing when I imagine a purple square? Where is it?
The purple square isn't the RGB lights. You yourself give another candidate criterion for identity of the purple square - the photons that are emitted each pulse from the RGB lights. Another candidate is the GPU operations that cause the monitor RGB lights to produce those photons. Another candidate is the image file. Another candidate is the packets that were involved in downloading the image file. Asking which one of these is the "real" purple square doesn't make sense. We can cause interruptions anywhere along this complex path that will redirect our focus when we ask "Where is the purple square?"
What happens with consciousness is that the information processing involved gets encapsulated into something simpler for consumers to interpret, a "quale," that abstracts all the technical and specialized information into something easy for a metacognitive system that has to manage many incompatible subsystems to use. One such visual "quale" would include metadata about its position among other visual qualia - used to construct e.g. a "visual field." But the neurons don't have to be in a 2 or 3d array to represent their own spatial location, this is something that gets calculated during the encapsulation process and is afforded to the "visual field manager," which knows how to aggregate these individual qualia into even more meaningful abstractions for the conscious subject (itself other producers and consumers of neuronal data in the brain) - producing edge detection, color interpolation depending on shadows, etc.
So "where" is the purple square? It's distributed across a variety of interneuronal connections that don't need to use their own spatial location to represent the spatial features of the square, and that don't need to use mental paint to represent the purple. Just like there isn't a "real" purple square that gets scanned by the computer and beamed into your eyes - it's something ephemeral that constantly changes its form from electric currents to packets to photons to what you naively call "the" purple square.
And that turns into an experience how? Is my GPU having an experience every time I play a game? Does a calculator have qualia? If I should take it on faith that my experiences are somehow produced by neuronal activity, then it seems like I should believe the same is true of GUP activity.
It doesn't turn into an experience, it is the experience. Experience is what you're afforded by the information processing in your brain that accumulates smaller details into a simplified gift for the "conscious subject" - there's that <neuronal construct I can easily identify again> when looking at e.g. a red apple. Just like in object oriented programming when you define a "Car" class with wheel, weight, and door class variables, but only expose a single method, e.g. Car.drive() for external objects to interact with the "Car" class. Everything in consciousness also (roughly) follows this paradigm - your experience is yet another abstraction you can reason about because your brain affords various introspective mechanisms about your experience like where the red apple is located with respect to other objects, but you don't have an introspective mechanism to e.g. figure out why or how you're experiencing in that specific way or experiencing at all, otherwise we could just do neuroscience by introspecting.
Your GPU doesn't have experience for the same reason there aren't two chess AIs playing chess in your neuronal connections - there are effectively no similarities between what evolution constructed human brains to do and what we constructed GPUs to do. There is no "chess AI" abstraction in your brain and no "experience" or "qualia" abstraction in your GPU. Also, if you're a dualist, you might as well believe that GPUs have qualia, since apparently no functional decomposition will ever falsify or confirm whether another thing is conscious. My theory has an easy explanation for why GPUs aren't conscious - yours doesn't.
Inside the hard drive you will find a set of electrons disposed to arrange other atoms into a visual representation of a purple square. But once the image is generated such that we can perceive it, the representation is obviously composed entirely of atoms. In the case of the brain, when you perceive the representation of the square, the representation itself is obviously not composed of atoms. If you doubt this, stipulate that the picture of the purple square you're imagining is larger than your head.
This is a circular argument. The physicalist does not agree that the representation is "obviously not composed of atoms."
It makes no difference how big you imagine the square to be, because the representation of a square bigger than your head isn't the same thing as an actual square larger than your head. This would be like saying that I can't have a picture of the Empire State Building on my computer because my computer is smaller than the Empire State Building.
You can’t have a picture on your computer that’s bigger than your computer screen. You can’t have a representation of a purple square in your brain that’s bigger than your brain. This is supposed to motivate the idea that it’s ridiculous to imagine tiny images of purple squares inside your brain somewhere, that your brain somehow perceives.
As I’ve said in another comment, the physicalist can easily proceed by saying there is no actual purple-square-image, only the experience of seeing one.
> This is supposed to motivate the idea that it’s ridiculous to imagine tiny images of purple squares inside your brain somewhere, that your brain somehow perceives.
Nothing is needed to motivate this idea. Everyone agrees it's ridiculous, and it's not what physicalists actually believe. The physical representation of the purple square is not the same size as the square itself is imagined to be. This is true regardless of what theory of consciousness is true, since, on dualism, the representation of the purple square has zero size.
Yes (well, except maybe the tiny part), in the same way there really exists a physical representation of a purple square in a computer that is processing an image of one. Even dualists should admit this, since our brain is certainly doing *something* to create our mental images even if there's some non-physical component as well. But the representation is a pattern of neural activity, not an actual purple square.
Wouldn't the physicalist just reject Premise 1 by having the ontology that the purple square isn't a thing? The purple square itself isn't part of reality, reality comprises only the set of atoms that constitute your imagining of the purple square.
I think they would say that there is a thing (quale) that is "what it feels like to imagine (or imagine seeing) a purple square", and that thing exists and is composed of atoms in the brain. But the actual representation of the purple square that you perceive doesn't exist at all.
First, I think the purple square example is confusing because purple squares are abstract objects.
The physicalist view is that the only datum we have is you having the experience of perceiving a particular purple-square-image. You might ask where the *experience* is, and they'll tell you it's in the brain, which is already their view. But the experience of seeing a square-image and the square-image itself aren't the same thing, and on the physicalist view only the former exists.
That doesn't really help matters, because then you're left with having to explain what the "experience" is made out of, and it doesn't make any more sense for the experience to be mad of atoms then the purple square itself.
This seems like the right response for a physicalist. They aren't eliminating consciousness, but just denying that its *objects* are real objects. (Compare: We can be realists about fiction books, while insisting that their represented *contents* are wholly fictional, not located anywhere, etc.)
That argument would also apply to non-fiction books; their content is also not located anywhere. Does that mean their content is equally fictional? (If yes, I can disregard your written comment)
How on Earth does the argument that fictional books don't refer to real objects apply to non-fiction books as well? That's the entire distinction between fiction and non-fiction.
Because the "contents" of both fiction and non-fiction books are concepts and thoughts, encoded using paper, ink, and a system of letters. The argument seems to be that the concepts encoded in a fiction book don't represent anything real, so the purple square in your mind also doesn't represent anything real so therefore we don't have to explain it. This is a complete side step: regardless of whether the purple square corresponds to a "real" purple square it still equally requires an explanation because it exists in the mind, even if nowhere else. In the same way a fictional story does exist as concepts in your mind, and just because the concepts are fictional doesn't mean that your experience of those concepts doesn't exist! The concepts in non-fiction books are the same experientially as the concepts in fiction books, there's nothing that makes them special because they correspond with real objects.
To summarize, saying that the contents of fictional books are not "real" (meaning, do not correspond with physical reality) doesn't git rid of the problem because if the concepts aren't "real" they are "real" in the sense that I am experiencing the concepts, regardless of whether they fall under "fiction" or "non-fiction".
Here is my master argument for why computers aren't physical. My computer shows a shopping cart when I go to amazon.com, and the physicalist is committed to thinking the shopping cart has a physical location. But where is the shopping cart? It's clearly not at my house, nor at Amazon's headquarters, nor inside of my CPU, because it wouldn't fit! Therefore the shopping cart and my computer are nonphysical.
If I draw a picture, what the picture *is* is not the photons that reflect off the ink and into my eyes, the picture *is* the ink particles. If I see a gorilla, the gorilla is not photons, the gorilla is the gorilla.
All you've stated is that the picture of the shopping cart is the picture of the shopping cart. I'm not disputing that. I'm disputing why it would be in the screen rather than some place else, like on the photons the screen emits, or in the GPU operations, or in the alternating current. Unless what you meant by your original comment was "the image of your shopping cart is the screen," but then I can replace my screen with another and still have the image of the shopping cart.
The picture of the shopping cart is liquid crystal and polarizers oriented in a particular way for photons to pass through. If you replaced the screen with a different one, you’d have a different picture depicting the same thing.
Similarly, if I draw a picture, I am arranging atoms in a particular way such that it constitutes a picture. The picture isn’t in my brain, or in the pen, or in the light, the picture comprises the picture-atoms, which are the atoms in the ink. I can copy the picture, but then it would be a different picture.
I remember David Chalmers on the 80,000 hours podcast making the claim that it's very difficult to be a physicalist but not an illusionist about consciousness. The argument you make in the post is somewhat convincing against non-illusionist physicalism, but it doesn't seem to make much of a dent in illusionism, which is the strongest form of the physicalist argument.
Agree! Keith Frankish (arch-illusionist) makes the same point, that it is very hard to make sense of claims that some ‘raw feel’ is literally identical to some physical thing.
Premise 1 is false, not Premise 3. The problem with this argument is that it conflates the mental image of a purple square with a literal, real-life purple square. But these are clearly not the same thing. So it's false that if physicalism is true, the purple square exists in the physical world. The purple square doesn't exist anywhere, regardless of whether physicalism is true or not.
Consider the converse argument in favor of physicalism:
1. The purple square exists in the mind.
2. If physicalism is false, anything that exists in the mind is non-physical.
3. But a purple square cannot be non-physical. (After all, a square exist in space, which is physical, by definition, and color is also a physical property related to the light being reflected and emitted by an object.)
4. Therefore, physicalism is true.
Obviously, this is a terrible argument for physicalism, but it's based on the same reasoning that leads to the phenomenal object argument against physicalism.
The purple square doesn't have to be anywhere. That's what is misleading in the argument. For jbstabce, a computer can informationally represent Super Mario without having SuperMario or an image of SuperMario in its memory.
Presumably you agree memories, beliefs, judgements, computation — the kinds that computers can support — can be purely physical. My laptop stores images of cool space photos in its memory, but if I broke it open I wouldn't be suprised to find only wires and circuits and no cool space photos.
But I agree that qualia are different! Qualia are essentially private, ineffable, unitary, non-composite, etc. And neurons obviously don't have those properties. The physicalist can insist that qualia and certain physical states/processes are literally identical, but this is absurd! It's meaningful to say some physical states/processes can give rise to reports of the experience of seeing purple, but it's really confusing to say they are *literally* the experience of seeing purple. The emperor has no qualia.
I think this isn't a case of “when we opened the brain we expected to see the color purple, instead we just saw neurons”. I think it's a case of “if qualia are physical, but also private, ineffable, etc., I have *no idea what I am supposed to see* when I scan/dissect someone's brain, because I can't make sense of this position”.
One response is dualism, but I think it's not the best response. I think the better response is to deny qualia. We say things about having private, ineffable, etc. experiences; but we are wrong. I don't think this leaves unexplained explananda, and posits one less thing in the world (qualia), so I think we should favour this view (‘strong illusionism’) on priors.
You and I feel strongly that we do have qualia, but I don't see how I can rule out illusionism in a non question-begging way. “There is something it is essentially like to see purple, dammit!” might not be unlike “I'm free in the libertarian sense because I know I am, dammit”.
“If physicalism is true, then when I imagine a purple square, the purple square is part of the physical world.”
This premise is dubious and largely semantic. A pattern of neural activation that represents a purple square need not be a purple square. A computer can represent a chessboard with a series of 12 bitboards or a 8x8 array containing a character in each cell. Neither of these things is a chess board, but they represent chess boards well enough for computers to crush human opponents.
I think you're assuming that the purple square has to be visible or like immediately obviously a purple square to be considered physically there, but it could be encoded in any number of ways. The problem is that then you have to define what counts as an "encoding" but I resolve this by just weighting the anthropic probability by how complicated it is to decode, similar to how Solomonoff induction weighs possible universes by how complicated their physical laws are.
I dont think a physicalist would accept that there is an actually existing purple square that comes into existence when you imagine a purple square, just a state that represents a purple square (with representation being some kind of causal/physical relationship). The view of imagination you are assuming seems analagous to a kind of sense data theory of perception where perception involves the existence of concretely existing mental entities. But I think all physicalists would/should deny that picture of perception and by extension deny that picture of imagination.
You might say that is denying consciousness. But then youre just using the word 'consciousness' to refer to a particular conception of consciousness that physicalists would (probably?) deny. And of course it's a tautology that physicalists deny a conception of consciousness that physicalists deny.
What is stopping one from embedding representational systems in a physical substrate? There exists systems that embed physical interactions into higher order systems (aka writing "purple square") in such a way that it is interpretted by other physical processes as that representation. The purple square you imagine is representation that triggers the same mental states as physical interaction with a purple square.
I think you need to say something more dramatic, that the idea of the purple square is essential to any sufficient representation of the physical world. Perceived equivalent lengths of projections onto a plane and electromagnetic perception at particular wave lengths is necessary for any higher order interaction with the world. Therefore these principles are inherent to the level of physical complexity, pre-existing any learning by a physical system.
Although presumably unintentional, this is simply another version of the most elementary strawman attack on physicalism.
Physicalism does not require that everything be made out of atoms or other matter. Or that phenomena be physical "stuff". It only requires that phenomena arise in some way or another from physics.
My rudimentary defense of physicalism in this case centers on the purple square as a projection based on information received from prior experience, which is composed more from electricity and energy than of atoms. Atoms (neuronal pathways, synapses et al) set the stage for the energy to form the projections, but that energy piece is critical. Love and emotions involve those coupled with hormones and reflexes from other parts of the body via the nervous system.
While I don't want to lean on brains-as-computers too much, I would also point to the fact that every image under the sun can be produced/reproduced from a series of transistors and silicon arranged and configured into different components. Following from that, I don't understand why a brain (along with nervous system, gut, etc.) can't do everything we attribute to consciousness. Perhaps there's simply a corollary to God of the Gaps we're dealing with, particularly with neuroscience still being a toddler in terms of age.
Everytime a computer generates an image, what the image actually *is*, is a series of pixels which are themselves composed of physical matter. The question in this case isn't how the brain might come to produce the square, but what the square *is*, or rather, of what it is composed. The square can't be made of energy because energy isn't a physical thing.
There is a state of 0's and 1's in the computer's memory that stores the information needed to reproduce the purple square on the screen. Nevertheless if we reduced the screen to its parts like the example of opening up a brain, we don't find purple squares either, just transistors, capacitors, diodes and so on. There is no reason yet discovered which says that the brain cannot operate in an analogous manner. In this sense, the "screen" is what we call the mind's eye and the information used to create the image is stored in our memory (obviously not the same as 0's and 1's).
If there's a special set of atoms inside my brain that constitute the images I project, how can I see the images if my eyes are outside my head and light isn't reflecting off of them? If I enlarge the brain-image inside my head, are more atoms required to generate the image? What happens when I think about something that doesn't have a 2D visual representation?
I don't see why the neurons and such inside the brain need to be special for each inage. Take the computer screen again. Do we need special parts (transistors, diodes, etc) to display different images on the screen? No, we don't. They may be specific to displays and the display circuitry, but not to each image. As such, we would probably find parts of the brain associated with mental imagery (and images generated via light from the outside world) and neurons specific to that task, but it seems unlikely we would find a set of neurons associated with something specific like purple squares.
When I perceive a computer screen projecting a purple-square-image, I am seeing a particular arrangement of atoms. That's what the image *is*. If the arrangement of the atoms changed, the image would be different, or the screen would cease to be a screen entirely. When you say the mind's eye is like a computer screen, it seems like you're saying that the brain analogously manipulates some set of atoms into the form of a purple-square-image, such that your brain can perceive it. But that seems weird? Like what's the mass of this image?
More precisely, if state of the transistors representing 1's and 0's in a computer's memory changes, then the image will change...as long as the change is compatible with the rest of the system. For example as long the 1's and 0's in memory conform to some standard that the various computer code for the different hardware (CPU, GPU, etc) isrecognized as valid, then an image is formed. It's not enough to say that changing the atoms will get a new image (for example, what atoms are we talking about?), but it's a change of the system as a whole. This is what I am arguing is analogous for the brain. So suppose I have learned what the color purple looks like and I know what a square looks like. Therefore, because the different parts of my brain that deal with memory and visual imagery have something to work with, my mind's eye can now show a purple square. I am not saying this is how it works, since I am not a brain or neuro-scientist, but rather this is to show that we don't need dualism to explain thoughts. The computer example helps because humans know how they work. For what it's worth, what is the weight of a purple square on a computer screen?
I should note that the conclusion here is that we cannot find purple squares by reducing a system to its parts. Rather it is an emergent property of the whole system given a certain state of it's parts.
There is ambiguous reference that occurs when people speak about a computer image. If the image file is corrupted it would be silly to "fix it" by altering the values of the RGB lights displaying the corrupted region. If your bit adder is overflowing each 255 value to be 0, it's silly to download another copy of the image file to "fix it." Failing to notice the problem is much more complex than you can formulate superficially meaningful questioms about it "Where is the image?" is the source of much of the confusion surrounding consciousness.
Funny that Huemer thinks that failure to locate the square should lead you to abandon sense data views in favour of direct realism. Perhaps his craziest view, more crazy considering he’s a substance dualist!
This argument and Jackson’s Knowledge Argument are fairly similar. The most relevant line of response to the Knowledge Argument to this one is the phenomenal concepts strategy. I’m skeptical that it works, but it’s worth thinking through the details.
https://philpapers.org/browse/phenomenal-concepts/
The Knowledge Argument straightforwardly begs the question by stipulating that Mary knows all the physical facts but doesn't know the "what-it's-likeness" of consciousness. The physicalist obviously disputes that is possible... otherwise they wouldn't be a physicalist. The Knowledge Argument does nothing to advance the debate between the physicalist and nonphysicalist, and its popularity is a mark against the supposed clarity and rigor of analytic philosophy.
Given that only Dennett went that way and many other physicalists (including Jackson, who changed his mind) nevertheless wrote responses, your ‘obvious’ here is not obviously apt.
What independent evidence does Frank Jackson give that justifies the claim that Mary learns something new when seeing red for the first time? He only says something along the lines of "It just seems obvious Mary learns something new." This is begging the question.
Compare: Mary knows all the physical facts, so obviously she learns nothing new - she already knew everything about what it's like to see red. Obviously a nonphysicalist isn't going to be moved by this argument - because I'm just asserting the negation of the nonphysicalist's thesis - that Mary doesn't know what it's like from all the physical facts.
It comes down to intuitions. So does everything else.
It's easy to see the representation of the purple square when we look at your brain, just as we see a representation of a purple square in a jpeg of a purple square on a hard drive.
What we see is certain states of the brain. But we never see a purple square!
Likewise, if you open the hard drive you will also not see a purple square just a certain state of the matter.
And yet there is a tangible sense in which one hard drive contains purple square while other does not. Certain state of the matter encode a specific images which can be decoded and shown.
That's a fair response. I suppose after reading Nick Humphrey's "Sentience" I'm really tempted by illusionist views, as absurd and offensive as they first sounded to me. Maybe the purple square you are imagining doesn't really exist and your brain only believes it does.
>It's easy to see the representation of the purple square when we look at your brain
Point of order, it's not actually easy at all. We do not currently have the capability of identifying the neurons that encode a perception of anything. I can understand having faith that we will someday, but still, we can't do that currently.
Other than that I agree with Philip, when my monitor creates a purple square that purple square is made of red, blue, and green light. What is the purple square in my mind's eye made of?
We may not be able to identify the _specific_ neurons that can do it, but that seems unimportant, because we can build an algorithm that can reconstruct the contents of your visual field based on fMRI representations of neural activity.
> What is the purple square in my mind's eye made of?
This is an interesting question that we will hopefully be able to answer one day. But first lets take a step back.
> when my monitor creates a purple square that purple square is made of red, blue, and green light.
How can *purple* square be made from non-purple light? Where does the *purpleness* coming from then?
Or going one level deeper.
Where does this "color" thing even comming from? How can non-colourful quarks produce colorful objects?
Yeah the purple square is made out of RGB light, and there isn't an entire network of interconnected protocols, rendering engines, parallelized computations in a GPU directed by a multiprocessor CPU, operating systems, and distributed databases and servers making that purple square possible. You just look and see whether there's purple light there, just like you look at the brain and don't see purple neurons so you know you've done your due diligence in understanding the problem and what you're asking.
I don't dispute the protocols, engines, hardware, or software, I just note that the process ends in red, blue, and green light. I can observe that light, I can measure that light, that light explains why I see a purple square: because there is a purple square made of light out there that I'm observing. But what am I observing when I imagine a purple square? Where is it?
The purple square isn't the RGB lights. You yourself give another candidate criterion for identity of the purple square - the photons that are emitted each pulse from the RGB lights. Another candidate is the GPU operations that cause the monitor RGB lights to produce those photons. Another candidate is the image file. Another candidate is the packets that were involved in downloading the image file. Asking which one of these is the "real" purple square doesn't make sense. We can cause interruptions anywhere along this complex path that will redirect our focus when we ask "Where is the purple square?"
What happens with consciousness is that the information processing involved gets encapsulated into something simpler for consumers to interpret, a "quale," that abstracts all the technical and specialized information into something easy for a metacognitive system that has to manage many incompatible subsystems to use. One such visual "quale" would include metadata about its position among other visual qualia - used to construct e.g. a "visual field." But the neurons don't have to be in a 2 or 3d array to represent their own spatial location, this is something that gets calculated during the encapsulation process and is afforded to the "visual field manager," which knows how to aggregate these individual qualia into even more meaningful abstractions for the conscious subject (itself other producers and consumers of neuronal data in the brain) - producing edge detection, color interpolation depending on shadows, etc.
So "where" is the purple square? It's distributed across a variety of interneuronal connections that don't need to use their own spatial location to represent the spatial features of the square, and that don't need to use mental paint to represent the purple. Just like there isn't a "real" purple square that gets scanned by the computer and beamed into your eyes - it's something ephemeral that constantly changes its form from electric currents to packets to photons to what you naively call "the" purple square.
And that turns into an experience how? Is my GPU having an experience every time I play a game? Does a calculator have qualia? If I should take it on faith that my experiences are somehow produced by neuronal activity, then it seems like I should believe the same is true of GUP activity.
It doesn't turn into an experience, it is the experience. Experience is what you're afforded by the information processing in your brain that accumulates smaller details into a simplified gift for the "conscious subject" - there's that <neuronal construct I can easily identify again> when looking at e.g. a red apple. Just like in object oriented programming when you define a "Car" class with wheel, weight, and door class variables, but only expose a single method, e.g. Car.drive() for external objects to interact with the "Car" class. Everything in consciousness also (roughly) follows this paradigm - your experience is yet another abstraction you can reason about because your brain affords various introspective mechanisms about your experience like where the red apple is located with respect to other objects, but you don't have an introspective mechanism to e.g. figure out why or how you're experiencing in that specific way or experiencing at all, otherwise we could just do neuroscience by introspecting.
Your GPU doesn't have experience for the same reason there aren't two chess AIs playing chess in your neuronal connections - there are effectively no similarities between what evolution constructed human brains to do and what we constructed GPUs to do. There is no "chess AI" abstraction in your brain and no "experience" or "qualia" abstraction in your GPU. Also, if you're a dualist, you might as well believe that GPUs have qualia, since apparently no functional decomposition will ever falsify or confirm whether another thing is conscious. My theory has an easy explanation for why GPUs aren't conscious - yours doesn't.
Inside the hard drive you will find a set of electrons disposed to arrange other atoms into a visual representation of a purple square. But once the image is generated such that we can perceive it, the representation is obviously composed entirely of atoms. In the case of the brain, when you perceive the representation of the square, the representation itself is obviously not composed of atoms. If you doubt this, stipulate that the picture of the purple square you're imagining is larger than your head.
This is a circular argument. The physicalist does not agree that the representation is "obviously not composed of atoms."
It makes no difference how big you imagine the square to be, because the representation of a square bigger than your head isn't the same thing as an actual square larger than your head. This would be like saying that I can't have a picture of the Empire State Building on my computer because my computer is smaller than the Empire State Building.
You can’t have a picture on your computer that’s bigger than your computer screen. You can’t have a representation of a purple square in your brain that’s bigger than your brain. This is supposed to motivate the idea that it’s ridiculous to imagine tiny images of purple squares inside your brain somewhere, that your brain somehow perceives.
As I’ve said in another comment, the physicalist can easily proceed by saying there is no actual purple-square-image, only the experience of seeing one.
> This is supposed to motivate the idea that it’s ridiculous to imagine tiny images of purple squares inside your brain somewhere, that your brain somehow perceives.
Nothing is needed to motivate this idea. Everyone agrees it's ridiculous, and it's not what physicalists actually believe. The physical representation of the purple square is not the same size as the square itself is imagined to be. This is true regardless of what theory of consciousness is true, since, on dualism, the representation of the purple square has zero size.
Are you saying there exists a tiny physical representation of a purple square?
Yes (well, except maybe the tiny part), in the same way there really exists a physical representation of a purple square in a computer that is processing an image of one. Even dualists should admit this, since our brain is certainly doing *something* to create our mental images even if there's some non-physical component as well. But the representation is a pattern of neural activity, not an actual purple square.
Wouldn't the physicalist just reject Premise 1 by having the ontology that the purple square isn't a thing? The purple square itself isn't part of reality, reality comprises only the set of atoms that constitute your imagining of the purple square.
But then they’re eliminating consciousness. Chairs exist despite just being collections of atoms.
I think they would say that there is a thing (quale) that is "what it feels like to imagine (or imagine seeing) a purple square", and that thing exists and is composed of atoms in the brain. But the actual representation of the purple square that you perceive doesn't exist at all.
Right but then they’re eliminating the data! I have a mental image of a purple square and they deny it exists!
First, I think the purple square example is confusing because purple squares are abstract objects.
The physicalist view is that the only datum we have is you having the experience of perceiving a particular purple-square-image. You might ask where the *experience* is, and they'll tell you it's in the brain, which is already their view. But the experience of seeing a square-image and the square-image itself aren't the same thing, and on the physicalist view only the former exists.
That doesn't really help matters, because then you're left with having to explain what the "experience" is made out of, and it doesn't make any more sense for the experience to be mad of atoms then the purple square itself.
"Experience is made of atoms" is the (slight) majority view among philosophers and likely the (strong) majority view among educated laypeople.
This seems like the right response for a physicalist. They aren't eliminating consciousness, but just denying that its *objects* are real objects. (Compare: We can be realists about fiction books, while insisting that their represented *contents* are wholly fictional, not located anywhere, etc.)
That argument would also apply to non-fiction books; their content is also not located anywhere. Does that mean their content is equally fictional? (If yes, I can disregard your written comment)
How on Earth does the argument that fictional books don't refer to real objects apply to non-fiction books as well? That's the entire distinction between fiction and non-fiction.
Because the "contents" of both fiction and non-fiction books are concepts and thoughts, encoded using paper, ink, and a system of letters. The argument seems to be that the concepts encoded in a fiction book don't represent anything real, so the purple square in your mind also doesn't represent anything real so therefore we don't have to explain it. This is a complete side step: regardless of whether the purple square corresponds to a "real" purple square it still equally requires an explanation because it exists in the mind, even if nowhere else. In the same way a fictional story does exist as concepts in your mind, and just because the concepts are fictional doesn't mean that your experience of those concepts doesn't exist! The concepts in non-fiction books are the same experientially as the concepts in fiction books, there's nothing that makes them special because they correspond with real objects.
To summarize, saying that the contents of fictional books are not "real" (meaning, do not correspond with physical reality) doesn't git rid of the problem because if the concepts aren't "real" they are "real" in the sense that I am experiencing the concepts, regardless of whether they fall under "fiction" or "non-fiction".
Here is my master argument for why computers aren't physical. My computer shows a shopping cart when I go to amazon.com, and the physicalist is committed to thinking the shopping cart has a physical location. But where is the shopping cart? It's clearly not at my house, nor at Amazon's headquarters, nor inside of my CPU, because it wouldn't fit! Therefore the shopping cart and my computer are nonphysical.
The image of the shopping cart is in your screen.
Why not on the photons the screen emits? What's special about the screen?
If I draw a picture, what the picture *is* is not the photons that reflect off the ink and into my eyes, the picture *is* the ink particles. If I see a gorilla, the gorilla is not photons, the gorilla is the gorilla.
All you've stated is that the picture of the shopping cart is the picture of the shopping cart. I'm not disputing that. I'm disputing why it would be in the screen rather than some place else, like on the photons the screen emits, or in the GPU operations, or in the alternating current. Unless what you meant by your original comment was "the image of your shopping cart is the screen," but then I can replace my screen with another and still have the image of the shopping cart.
The picture of the shopping cart is liquid crystal and polarizers oriented in a particular way for photons to pass through. If you replaced the screen with a different one, you’d have a different picture depicting the same thing.
Similarly, if I draw a picture, I am arranging atoms in a particular way such that it constitutes a picture. The picture isn’t in my brain, or in the pen, or in the light, the picture comprises the picture-atoms, which are the atoms in the ink. I can copy the picture, but then it would be a different picture.
What fact am I getting wrong if I identity the picture with the picture-photons?
I remember David Chalmers on the 80,000 hours podcast making the claim that it's very difficult to be a physicalist but not an illusionist about consciousness. The argument you make in the post is somewhat convincing against non-illusionist physicalism, but it doesn't seem to make much of a dent in illusionism, which is the strongest form of the physicalist argument.
Agree! Keith Frankish (arch-illusionist) makes the same point, that it is very hard to make sense of claims that some ‘raw feel’ is literally identical to some physical thing.
Premise 1 is false, not Premise 3. The problem with this argument is that it conflates the mental image of a purple square with a literal, real-life purple square. But these are clearly not the same thing. So it's false that if physicalism is true, the purple square exists in the physical world. The purple square doesn't exist anywhere, regardless of whether physicalism is true or not.
Consider the converse argument in favor of physicalism:
1. The purple square exists in the mind.
2. If physicalism is false, anything that exists in the mind is non-physical.
3. But a purple square cannot be non-physical. (After all, a square exist in space, which is physical, by definition, and color is also a physical property related to the light being reflected and emitted by an object.)
4. Therefore, physicalism is true.
Obviously, this is a terrible argument for physicalism, but it's based on the same reasoning that leads to the phenomenal object argument against physicalism.
The purple square doesn't have to be anywhere. That's what is misleading in the argument. For jbstabce, a computer can informationally represent Super Mario without having SuperMario or an image of SuperMario in its memory.
Presumably you agree memories, beliefs, judgements, computation — the kinds that computers can support — can be purely physical. My laptop stores images of cool space photos in its memory, but if I broke it open I wouldn't be suprised to find only wires and circuits and no cool space photos.
But I agree that qualia are different! Qualia are essentially private, ineffable, unitary, non-composite, etc. And neurons obviously don't have those properties. The physicalist can insist that qualia and certain physical states/processes are literally identical, but this is absurd! It's meaningful to say some physical states/processes can give rise to reports of the experience of seeing purple, but it's really confusing to say they are *literally* the experience of seeing purple. The emperor has no qualia.
I think this isn't a case of “when we opened the brain we expected to see the color purple, instead we just saw neurons”. I think it's a case of “if qualia are physical, but also private, ineffable, etc., I have *no idea what I am supposed to see* when I scan/dissect someone's brain, because I can't make sense of this position”.
One response is dualism, but I think it's not the best response. I think the better response is to deny qualia. We say things about having private, ineffable, etc. experiences; but we are wrong. I don't think this leaves unexplained explananda, and posits one less thing in the world (qualia), so I think we should favour this view (‘strong illusionism’) on priors.
You and I feel strongly that we do have qualia, but I don't see how I can rule out illusionism in a non question-begging way. “There is something it is essentially like to see purple, dammit!” might not be unlike “I'm free in the libertarian sense because I know I am, dammit”.
“If physicalism is true, then when I imagine a purple square, the purple square is part of the physical world.”
This premise is dubious and largely semantic. A pattern of neural activation that represents a purple square need not be a purple square. A computer can represent a chessboard with a series of 12 bitboards or a 8x8 array containing a character in each cell. Neither of these things is a chess board, but they represent chess boards well enough for computers to crush human opponents.
I think you're assuming that the purple square has to be visible or like immediately obviously a purple square to be considered physically there, but it could be encoded in any number of ways. The problem is that then you have to define what counts as an "encoding" but I resolve this by just weighting the anthropic probability by how complicated it is to decode, similar to how Solomonoff induction weighs possible universes by how complicated their physical laws are.
I dont think a physicalist would accept that there is an actually existing purple square that comes into existence when you imagine a purple square, just a state that represents a purple square (with representation being some kind of causal/physical relationship). The view of imagination you are assuming seems analagous to a kind of sense data theory of perception where perception involves the existence of concretely existing mental entities. But I think all physicalists would/should deny that picture of perception and by extension deny that picture of imagination.
You might say that is denying consciousness. But then youre just using the word 'consciousness' to refer to a particular conception of consciousness that physicalists would (probably?) deny. And of course it's a tautology that physicalists deny a conception of consciousness that physicalists deny.
What is stopping one from embedding representational systems in a physical substrate? There exists systems that embed physical interactions into higher order systems (aka writing "purple square") in such a way that it is interpretted by other physical processes as that representation. The purple square you imagine is representation that triggers the same mental states as physical interaction with a purple square.
I think you need to say something more dramatic, that the idea of the purple square is essential to any sufficient representation of the physical world. Perceived equivalent lengths of projections onto a plane and electromagnetic perception at particular wave lengths is necessary for any higher order interaction with the world. Therefore these principles are inherent to the level of physical complexity, pre-existing any learning by a physical system.
Although presumably unintentional, this is simply another version of the most elementary strawman attack on physicalism.
Physicalism does not require that everything be made out of atoms or other matter. Or that phenomena be physical "stuff". It only requires that phenomena arise in some way or another from physics.
Thoughts aren't made up of atoms, but perhaps atoms are made up of thoughts.
My rudimentary defense of physicalism in this case centers on the purple square as a projection based on information received from prior experience, which is composed more from electricity and energy than of atoms. Atoms (neuronal pathways, synapses et al) set the stage for the energy to form the projections, but that energy piece is critical. Love and emotions involve those coupled with hormones and reflexes from other parts of the body via the nervous system.
While I don't want to lean on brains-as-computers too much, I would also point to the fact that every image under the sun can be produced/reproduced from a series of transistors and silicon arranged and configured into different components. Following from that, I don't understand why a brain (along with nervous system, gut, etc.) can't do everything we attribute to consciousness. Perhaps there's simply a corollary to God of the Gaps we're dealing with, particularly with neuroscience still being a toddler in terms of age.
Everytime a computer generates an image, what the image actually *is*, is a series of pixels which are themselves composed of physical matter. The question in this case isn't how the brain might come to produce the square, but what the square *is*, or rather, of what it is composed. The square can't be made of energy because energy isn't a physical thing.
There is a state of 0's and 1's in the computer's memory that stores the information needed to reproduce the purple square on the screen. Nevertheless if we reduced the screen to its parts like the example of opening up a brain, we don't find purple squares either, just transistors, capacitors, diodes and so on. There is no reason yet discovered which says that the brain cannot operate in an analogous manner. In this sense, the "screen" is what we call the mind's eye and the information used to create the image is stored in our memory (obviously not the same as 0's and 1's).
If there's a special set of atoms inside my brain that constitute the images I project, how can I see the images if my eyes are outside my head and light isn't reflecting off of them? If I enlarge the brain-image inside my head, are more atoms required to generate the image? What happens when I think about something that doesn't have a 2D visual representation?
I don't see why the neurons and such inside the brain need to be special for each inage. Take the computer screen again. Do we need special parts (transistors, diodes, etc) to display different images on the screen? No, we don't. They may be specific to displays and the display circuitry, but not to each image. As such, we would probably find parts of the brain associated with mental imagery (and images generated via light from the outside world) and neurons specific to that task, but it seems unlikely we would find a set of neurons associated with something specific like purple squares.
When I perceive a computer screen projecting a purple-square-image, I am seeing a particular arrangement of atoms. That's what the image *is*. If the arrangement of the atoms changed, the image would be different, or the screen would cease to be a screen entirely. When you say the mind's eye is like a computer screen, it seems like you're saying that the brain analogously manipulates some set of atoms into the form of a purple-square-image, such that your brain can perceive it. But that seems weird? Like what's the mass of this image?
More precisely, if state of the transistors representing 1's and 0's in a computer's memory changes, then the image will change...as long as the change is compatible with the rest of the system. For example as long the 1's and 0's in memory conform to some standard that the various computer code for the different hardware (CPU, GPU, etc) isrecognized as valid, then an image is formed. It's not enough to say that changing the atoms will get a new image (for example, what atoms are we talking about?), but it's a change of the system as a whole. This is what I am arguing is analogous for the brain. So suppose I have learned what the color purple looks like and I know what a square looks like. Therefore, because the different parts of my brain that deal with memory and visual imagery have something to work with, my mind's eye can now show a purple square. I am not saying this is how it works, since I am not a brain or neuro-scientist, but rather this is to show that we don't need dualism to explain thoughts. The computer example helps because humans know how they work. For what it's worth, what is the weight of a purple square on a computer screen?
I should note that the conclusion here is that we cannot find purple squares by reducing a system to its parts. Rather it is an emergent property of the whole system given a certain state of it's parts.
There is ambiguous reference that occurs when people speak about a computer image. If the image file is corrupted it would be silly to "fix it" by altering the values of the RGB lights displaying the corrupted region. If your bit adder is overflowing each 255 value to be 0, it's silly to download another copy of the image file to "fix it." Failing to notice the problem is much more complex than you can formulate superficially meaningful questioms about it "Where is the image?" is the source of much of the confusion surrounding consciousness.
The image of a purple square on a computer screen is not a computer file, it’s a set and particular orientation of particles. There’s no problem.
Funny that Huemer thinks that failure to locate the square should lead you to abandon sense data views in favour of direct realism. Perhaps his craziest view, more crazy considering he’s a substance dualist!