Modern Philosophy - Qualia and Intentionality

[Jub]

This is a question spawned by a comment on a youtube video where a person stated that he was going to unsubscribe because the channel was pushing a materialist agenda. I haven't really stumbled across the idea that people aside from religious types were accepting a world view that acknowledges things outside of the grasp of physics before; but then again I tend not to hang out with philosophers very much. I engaged with this commenter and he brought up the ideas of qualia and intentionality being things that physics cannot test. I disagree and think that with greater understanding of the human brain and AGI we will one day be able to dirrectly see how another person perceives an event.

What I'm wondering is what do people here thing of the ideas of untestable bits of philosophy like qualia and intentionality. Do they hold any weight, or are they merely thought experiments by and for those who doubt the power of science?

[Ziggy Stardust]

The problem is that precise definitions of "qualia" and "intentionality" can be difficult to meaningfully describe. And most thought experiences that supposedly "prove" that these are non-materialist (like the ridiculous inverted spectrum argument) rely on flawed assumptions.

For example, perception of color. We have stringent physical criteria with which to classify colors and a strong understand of exactly how the mechanics and neurology of vision work to interpret light in such a way that can be predictably and accurately matched to these physical criteria. Heck, the fact that color blindness not only exists (or, for that matter, other perceptual oddities like color-related optical illusions or synesthesia), but we can reliably trace its development to specific genetic and physical defects, shows us how deeply preserved our perception of color really is, and how it is a direct function of measurable phenomena. But that won't stop any armchair philosopher from waving their hands and saying that we still can't "prove" that "my" red is the same exact thing as "your" red.

The problem is the inability to draw a meaningful description of what exactly that proof would require. We already have such a deep understanding of the way our perception of color operates that we can reliably trick people with optical illusions like this, and it is trivially easy to run behavioral experiments where people's subjective interpretation of color is shown to match up with those of others based on measurable quantities like wavelength, frequency, energy, etc. Just because something is subjective does not mean it isn't an inherently material phenomenon.

A similar argument can be presented for intentionality, though admittedly our scientific understanding of conceptual representations is still relatively sparse. The problem is that nobody is able to come up with consistent and testable criteria of what actually constitutes a representation or understanding in the way that intentionality requires. It isn't even entirely clear that objects even exist in our brains as a singular representational unit or that consciousness is a distinct state in the way that intentionality requires.

[Jub]

<snip>

Thanks, that was exactly the kind of reply I was looking for.

I was thinking of directly testing things, like in the future we could try doing a brain upload and see if that person's mind creates an output on he monitor that matches another person's output on a calibrated screen after both wereexposed to the same wave length of light. I didn't even think of all the ways that currently exist that allow us to say that, for example, this color is red and have others agree with us and to know why most people who can't agree with us see things differently.

[Ziggy Stardust]

You may also be interested in reading this thread from a few days ago, which relates to current research in modeling speech input. Specifically, researchers have begun to "reverse engineer" patterns of brain activity to recreate the original acoustic signal.

[Simon_Jester]

I think what it comes down to is that wherever there is an information-theoretic component to a system, a philosopher can, if they so choose, represent it as a non-materialistic component.

Because you can make a good case that the universe consists on one level of matter and on another level of the information that describes how this matter is organized. Information is definitely not matter, and has properties that can't be described in material terms.

Things like qualia, intentions, and consciousness are abstract concepts that exist in the realm of information, as well as being things with a material basis in our brains. Therefore, understanding the material basis on which our instances of these things run does not automatically confer upon us the ability to understand them fully. Just as understanding the workings of a computer chip in perfect detail would not necessarily allow us to understand the full nature and purpose of the software running on that chip. Even if it let us read the data in the device's memory, we might still fail to understand the behavior of the program at the user-interface level.

[Please note that the views I have just expressed are not the same as vitalism or even mind-body dualism.]

[Jub]

I think what it comes down to is that wherever there is an information-theoretic component to a system, a philosopher can, if they so choose, represent it as a non-materialistic component.

Because you can make a good case that the universe consists on one level of matter and on another level of the information that describes how this matter is organized. Information is definitely not matter, and has properties that can't be described in material terms.

Things like qualia, intentions, and consciousness are abstract concepts that exist in the realm of information, as well as being things with a material basis in our brains. Therefore, understanding the material basis on which our instances of these things run does not automatically confer upon us the ability to understand them fully. Just as understanding the workings of a computer chip in perfect detail would not necessarily allow us to understand the full nature and purpose of the software running on that chip. Even if it let us read the data in the device's memory, we might still fail to understand the behavior of the program at the user-interface level.

[Please note that the views I have just expressed are not the same as vitalism or even mind-body dualism.]

I'm not so sure that I can get behind classing information as something non-physical, at least in the sense that some people will make the claim. Basically my thought is that all information exists in a form that relies upon physical structures and thus has the potential to be analyzed and understood, or at least noted as existing in the case of information trapped within a black hole and unreachable, by sufficiently advanced sciences. Nothing, not even a thought never shared, can exist outside of physics - and thus the physical realm - because the mind that created the information is a purely physical construct which itself can be analyzed.

This is also why I don't believe we have free will any more than a human level AI would have free will. A purely physical device, if given the same initial inputs (including things such as quantum states and other such micro scale conditions, to create a perfectly repeatable scenario) will always output the same thing. Our minds, being driven by the same physics as everything else in the universe, must react in the same way. Thus for any given input there is only one possible outcome; our thoughts, analysis of the situation, and eventual action are all simple by products of the initial state and the added stimuli.

One might try to counter this by bringing up quantum randomness, but a random outcome is not the same as a choice; also I don't believe anything can be truly random. The appearance of randomness is simple an illusion based on incomplete information about an event.

You may also be interested in reading this thread from a few days ago, which relates to current research in modeling speech input. Specifically, researchers have begun to "reverse engineer" patterns of brain activity to recreate the original acoustic signal.

Thanks for linking to that, I missed that one somehow.

[Simon_Jester]

I'm not so sure that I can get behind classing information as something non-physical, at least in the sense that some people will make the claim.

I can't speak for a nebulous "some people" making an undefined claim, but...

Basically my thought is that all information exists in a form that relies upon physical structures and thus has the potential to be analyzed and understood... or at least noted as existing in the case of information trapped within a black hole and unreachable, by sufficiently advanced sciences. Nothing, not even a thought never shared, can exist outside of physics - and thus the physical realm - because the mind that created the information is a purely physical construct which itself can be analyzed.

Put this way. If I have two copies of the same novel, in different formats, are they the same novel? Physically they do not resemble each other. A paperback book and a PDF file stored on a flash drive are totally unlike each other in their physical composition, their nature, the means by which they can be interacted with, and so on.

And yet, we recognize them as having a quality of sameness that transcends their material composition.

It's not that the information isn't subject to physical analysis. It's not that it can't be read or understood. It's not vitalist, that's why I came right out and said it wasn't in my last post.

It's that this thing we call "information," and which has its own rigorously definable laws ('information theory') is not a kind of matter, nor is it any field of force. It does not follow the laws of 'atoms and void' that characterize our understanding of physics; it follows other laws that reflect a different subset of the broader realms of abstract mathematical possibilities. The rules get screwier if you start talking about 'information' in a sense broader than that of formal information theory, such as the abstract concepts like "sharpness" and "courage" and "the number five" that a philosopher might reference.

But in any case, information is a thing that is, when you get right down to it, not matter. Any given piece of information may be reduced to and instantiated some configuration of matter... but by the same token, any given instance of matter may be reduced to and represented in terms of information.

So philosophers will continue to think about how abstract concepts and information behave, supplementary to mechanistic analysis of how matter behaves.

This is also why I don't believe we have free will any more than a human level AI would have free will... Our minds, being driven by the same physics as everything else in the universe, must react in the same way.

I adopt the converse position. The AI, and me, both have free will. Because if "free will" somehow requires the ability to ignore the laws of physics operating on my brain, then "free will" is a meaningless term.

Except "free will" is not a meaningless concept; it is a term we use to mean the ability to make autonomous decisions. Which I have. You may argue that I am a puppet in the hands of the laws of physics- but the laws of physics do not themselves have any volition of their own. How can I be a puppet in the absence of a puppeteer?

I mean, by your argument, even if we otherwise had free will, it wouldn't be real free will as long as stepping off a cliff resulted in inevitable falling, because then the law of gravity is predictably constraining our actions.

But in what sense of "free will" does free will entail being free from gravity, or any other physical law? Nobody ever seriously considered that that was part of free will until neurochemistry came along. Only now do we get people claiming that because the laws of physics can analyze a brain, that brain and the software running on it is in some sense not an autonomous agent.

One might try to counter this by bringing up quantum randomness, but a random outcome is not the same as a choice; also I don't believe anything can be truly random. The appearance of randomness is simple an illusion based on incomplete information about an event.

Only a babbling cretin would bring up quantum randomness as evidence in favor of free will.

However, your disbelief in 'true randomness' in quantum mechanics is noted and patronizingly smiled at. The theory that quantum randomness is an illusion based on incomplete information is called a "hidden variable theory." It smacks forcefully into Bell's theorem. Some physicists are still trying (out of sheer intellectual high spirits) to find ways around that, but the results smell like epicycles to me.

Basically, it's a bad idea to make predictions about the laws of physics based on intuitive philosophical preferences.

[Ziggy Stardust]

And yet, we recognize them as having a quality of sameness that transcends their material composition.

What? We recognize them as having a quality of sameness because the words printed in those books are exactly the same. This is the problem I have with anti-materialist philosophy; it always feels like it is using semantic tricks rather than actual rigorous thought. Nobody would call two different editions of the same novel the same exact BOOK, because they are distinct physical objects. But there's no conflict there with the fact that those two different books have the same CONTENT. That content is still a material composition; the words physically exist on the page, and are the same in both copies. And those words have specific physical representations in our mind (as is demonstrated in the link in my previous post) that are themselves the result of a highly elaborate pattern recognition process that we call language acquisition, in which our brain maps discrete physical phenomena (sound waves) onto a representation of meaning. Further, these same words cannot be understood by people who don't speak the language in question, or will be less understood by people less proficient at this language, and that these perceptual deficits can be reliably correlated with specific physical phenomena in the brain.

These books contain the same material information. The ink used to print the pages makes up the same symbols, and those symbols are ones that have been culturally "agreed" upon to stand in for acoustic signals that exhibit specific and measurable effects on our brain. So strongly, in fact, that these effects can be reverse engineered to recreate the original signal. How is this information non-material? In every meaningful way we can interact with it, there is a direct and measurable physical phenomena involved. Even the abstract semantic representations we assign to words are just the result of physical interactions (both chemical and electrical) in our brain;

But in any case, information is a thing that is, when you get right down to it, not matter. Any given piece of information may be reduced to and instantiated some configuration of matter... but by the same token, any given instance of matter may be reduced to and represented in terms of information.

The problem is, I alluded to above, the lack of a precise and consistent definition of what "information" is. It is something that can be arbitrarily redrawn every time an example of its material impact is demonstrated, so that it can always remain theoretically abstract because it is a more or less meaningless concept.

If in every meaningful way we can interact with "information" we can connect it with discrete and measurable physical phenomena, by what criteria can we then claim it is still non-material without waving our hands and redefining what the "information" itself really is?

[Jub]

Put this way. If I have two copies of the same novel, in different formats, are they the same novel? Physically they do not resemble each other. A paperback book and a PDF file stored on a flash drive are totally unlike each other in their physical composition, their nature, the means by which they can be interacted with, and so on.

And yet, we recognize them as having a quality of sameness that transcends their material composition.

It's not that the information isn't subject to physical analysis. It's not that it can't be read or understood. It's not vitalist, that's why I came right out and said it wasn't in my last post.

It's that this thing we call "information," and which has its own rigorously definable laws ('information theory') is not a kind of matter, nor is it any field of force. It does not follow the laws of 'atoms and void' that characterize our understanding of physics; it follows other laws that reflect a different subset of the broader realms of abstract mathematical possibilities. The rules get screwier if you start talking about 'information' in a sense broader than that of formal information theory, such as the abstract concepts like "sharpness" and "courage" and "the number five" that a philosopher might reference.

But in any case, information is a thing that is, when you get right down to it, not matter. Any given piece of information may be reduced to and instantiated some configuration of matter... but by the same token, any given instance of matter may be reduced to and represented in terms of information.

So philosophers will continue to think about how abstract concepts and information behave, supplementary to mechanistic analysis of how matter behaves.

Ziggy put it better than I could myself, but I'll give this a go as well.

First define exactly what you mean by information. Make it something that we can have a meaningful discussion on, because metaphors involving similar books that are distinct objects aren't making a case. Second, I'd like to know if you think that information can be created by a non-physical process? I ask this because you seem to think that information exists outside of the physical sphere and if it does, it must be able to be created without input from the physical sphere and it must continue to exist after the universe has reached peak entropy as a physical reaction cannot destroy what is by definition a non-physical object.

I adopt the converse position. The AI, and me, both have free will. Because if "free will" somehow requires the ability to ignore the laws of physics operating on my brain, then "free will" is a meaningless term.

Free will is a meaningless term, it is meaningless to have a term for something that cannot exist.

Except "free will" is not a meaningless concept; it is a term we use to mean the ability to make autonomous decisions. Which I have. You may argue that I am a puppet in the hands of the laws of physics- but the laws of physics do not themselves have any volition of their own. How can I be a puppet in the absence of a puppeteer?

Does the universe have a puppeteer? Do we need to know that some guiding hand put the stars in the sky, or caused the planets to form? Do we need so badly to believe in some greater purpose in order to find satisfaction with our lives?

I mean, by your argument, even if we otherwise had free will, it wouldn't be real free will as long as stepping off a cliff resulted in inevitable falling, because then the law of gravity is predictably constraining our actions.

But in what sense of "free will" does free will entail being free from gravity, or any other physical law? Nobody ever seriously considered that that was part of free will until neurochemistry came along. Only now do we get people claiming that because the laws of physics can analyze a brain, that brain and the software running on it is in some sense not an autonomous agent.

If I as a creator set a robot on closed course and, via means of a random number generator, give that robot the ability to choose left and right at a junction does it have free will? No, of course it doesn't and no matter how much sophistication we program into the robot the choices will always be predetermined by a combination or pre-programmed responses and external stimuli. We are nothing more than biological machines programmed by DNA and by the outputs of other machines that have come before us and gathered information which they have passed on. I, just as we all do, have no freedom to change my response to stimuli.

Is our will free if via drugs, brain damage, disease, or chemical imbalance our choices are altered? Is our will free if so much of our biological function is run by processes that most people are unable to consciously control? How is your response to a question or a punch any different than your response to having your knee knocked with a reflex hammer?

However, your disbelief in 'true randomness' in quantum mechanics is noted and patronizingly smiled at. The theory that quantum randomness is an illusion based on incomplete information is called a "hidden variable theory." It smacks forcefully into Bell's theorem. Some physicists are still trying (out of sheer intellectual high spirits) to find ways around that, but the results smell like epicycles to me.

Basically, it's a bad idea to make predictions about the laws of physics based on intuitive philosophical preferences.

Your arguing for quantum particles being random sounds an awful lot like a Christian arguing for irreducible complexity. Just because we can't yet, and may never have the means to, perfectly predict quantum behavior doesn't make it random. Nor does my throwing a die or picking a number between x and y really result in a random number. Just because the outcome isn't known to the present parties doesn't mean the outcome is unknowable.

[Channel72]

The problem is the inability to draw a meaningful description of what exactly that proof would require. We already have such a deep understanding of the way our perception of color operates that we can reliably trick people with optical illusions like this, and it is trivially easy to run behavioral experiments where people's subjective interpretation of color is shown to match up with those of others based on measurable quantities like wavelength, frequency, energy, etc. Just because something is subjective does not mean it isn't an inherently material phenomenon.

Yeah - forget that - just ask any Fortune 500 marketing department. They rely on common human reactions to color all the time to sell products, and apparently it works well enough that "my red" is similar to enough to "your red" on average, that both of us want to buy iPads and drink Coke for some reason.

[Starglider]

The root cause of most of this philosobabble is how awful (inaccurate and confabulated) human introspection is. A rigorous neurological description of sensory perception isn't enough to shut them up, because they will just go on about the more abstract derrived perceptions in an endless loop of "but why does it feel like that?". A rigorous neurological trace of the reason why they say these things, from the reflective confabulation through the groping for bad analogies to the actual utterance, will definitively prove it to be nonsense. They still won't shut up, but at least then we could concisely dismiss with 'here is an exact constructive description of the mechanism that makes you spout silly philosobabble, good day'.

[Starglider]

P.S. Simon, you are generally correct but note that intelligence is a feature of information processing systems, and to reason correctly about it you must think about information transforming functions implemented as causal physical processes. Information and even semantics certainly independently of means to process it; consider the active SETI signals which are designed to be understood by an alien intelligence, using only physical and mathematical constants as shared knowledge. Minds are not static information though; given sufficient technology we could take a static copy but it is only considered a mind when combined with a mechanism capable of executing/insantiating the intelligence.

Quantum 'randomness' is a red herring, however the physics turn out; sadly a red herring much beloved by hack philosophers and new-age idiots, including ones who really should know better (e.g. Roger Penrose).

[Simon_Jester]

Ziggy put it better than I could myself, but I'll give this a go as well.

First define exactly what you mean by information. Make it something that we can have a meaningful discussion on, because metaphors involving similar books that are distinct objects aren't making a case. Second, I'd like to know if you think that information can be created by a non-physical process? I ask this because you seem to think that information exists outside of the physical sphere and if it does, it must be able to be created without input from the physical sphere and it must continue to exist after the universe has reached peak entropy as a physical reaction cannot destroy what is by definition a non-physical object.

In this case, talking about books is not a metaphor, because books store information in a measurable information-theoretic sense.. It is most commonly measured in terms of entropy, but that's just a unit of measure for information, not a definition. Whether information can exist in the absence of physical matter? Well, we wouldn't be able to see it if it did, so there's no way to answer the question.

I don't see why this poses a major problem.

I mean, matter cannot exist without having some defined position, velocity, and various quantum-mechanical parameters- but we don't say that "velocity" is a material thing. you cannot fill a teacup with velocity. Nor is it meaningful to talk about whether velocity exists in the absence of matter.

So the relationship between information and matter is quite simple and easy to grasp, just as the relationship between matter and space is. General relativity can be summed up in informal terms as "matter tells space how to curve, space tells matter how to move." But space is not itself a form of matter; there is no substance called "spacium," no space-force that causes space-particles to appear. Space is not matter, but interacts with matter, and remains meaningful as a concept even if we do not assume that matter is present.

Likewise, information is not matter, but describes matter, and remains meaningful as a concept even if we make no assumptions about the exact configuration of matter that represents the information.

[/color]

I adopt the converse position. The AI, and me, both have free will. Because if "free will" somehow requires the ability to ignore the laws of physics operating on my brain, then "free will" is a meaningless term.

Free will is a meaningless term, it is meaningless to have a term for something that cannot exist.

No.

What you're doing is, you're defining "free will" to mean "ability to ignore the laws of physics," then asserting that it doesn't exist because you can't ignore the laws of physics. This is circular logic: "I define X to be meaningless, therefore X is meaningless."

Thing is, that was never the normal definition of "free will." If you asked an ancient person whether "free will" implied the ability to fly by flapping one's arms, they would have looked at you like an idiot.

This notion that "free will" is not truly "free" unless it allows us to transcend physical laws and principles is new. It is a strawman created by people seeking to prove that "free will" does not exist.

Except "free will" is not a meaningless concept; it is a term we use to mean the ability to make autonomous decisions. Which I have. You may argue that I am a puppet in the hands of the laws of physics- but the laws of physics do not themselves have any volition of their own. How can I be a puppet in the absence of a puppeteer?

Does the universe have a puppeteer? Do we need to know that some guiding hand put the stars in the sky, or caused the planets to form? Do we need so badly to believe in some greater purpose in order to find satisfaction with our lives?

The point is that it is meaningless to talk about a thing being a puppet unless there is some source of decisions, some algorithm that governs its behavior.

Except that if we want to talk about the thing that is "I," our selves, those selves are not bodies, they are themselves the algorithms. I am not a big pile of meat sitting in a chair; I am the consciousness and thinking processes that make up "Simon"

So you're reduced to saying that "I" am a puppet governed by the decisions of an algorithm that is, well... me. I'm holding my own strings.

I mean, by your argument, even if we otherwise had free will, it wouldn't be real free will as long as stepping off a cliff resulted in inevitable falling, because then the law of gravity is predictably constraining our actions.

But in what sense of "free will" does free will entail being free from gravity, or any other physical law? Nobody ever seriously considered that that was part of free will until neurochemistry came along. Only now do we get people claiming that because the laws of physics can analyze a brain, that brain and the software running on it is in some sense not an autonomous agent.

If I as a creator set a robot on closed course and, via means of a random number generator, give that robot the ability to choose left and right at a junction does it have free will?

A random number generator is not a decision-making process.

Intelligence, consciousness, will, all these things are properties of decision-making processes, and of decision-making processes that refer back to themselves, that are conscious of themselves, that can potentially modify themselves by learning.

The analogy to a random number generator or a coin toss is therefore a false one.

No, of course it doesn't and no matter how much sophistication we program into the robot the choices will always be predetermined by a combination or pre-programmed responses and external stimuli. We are nothing more than biological machines programmed by DNA and by the outputs of other machines that have come before us and gathered information which they have passed on. I, just as we all do, have no freedom to change my response to stimuli.

I think the problem here is that you're not materialist enough.

See, traditional mind-body dualism claims that there is matter, and there is mind, they are qualitatively different.

You claim that there is the idea of matter, and there is the idea of mind, and they are qualitatively different... But then you claim that axiomatically, only matter can exist, therefore mind does not exist in reality.

Me, I claim that there is matter, and there is the information that describes matter and which is represented by various configurations of matter. I am an informational construct, not a physical object. I am the software that happens to run on a certain meat-computer. If there weren't physical laws governing the actions of the meat-computer, then I couldn't run on it at all, so there would be no "me."

My intellectual freedom comes from the fact that I am able to perform computations and use decision-making algorithms, because that is what thinking and deciding are. The processes by which I do these things may or may not be deterministic; I don't care. It's still me making the decisions, even if the decisions themselves are predictable when analyzed in terms of particles bouncing around in space.

Is our will free if via drugs, brain damage, disease, or chemical imbalance our choices are altered? Is our will free if so much of our biological function is run by processes that most people are unable to consciously control? How is your response to a question or a punch any different than your response to having your knee knocked with a reflex hammer?

My response to the punch or the question are routed through the decision-making algorithms that make up me.

My body's response to the reflex hammer is not routed through me. I am never consulted.

Therefore there is a difference.

...Basically, it's a bad idea to make predictions about the laws of physics based on intuitive philosophical preferences.

Your arguing for quantum particles being random sounds an awful lot like a Christian arguing for irreducible complexity. Just because we can't yet, and may never have the means to, perfectly predict quantum behavior doesn't make it random. Nor does my throwing a die or picking a number between x and y really result in a random number. Just because the outcome isn't known to the present parties doesn't mean the outcome is unknowable.

Bell's Theorem is an actual testable physical result, derived from actual testable physical laws. It's been tested, and is supported by experimental data.

This isn't a subject of philosophical speculation alone, it is a subject of physical study. And since it describes the behavior of matter, it's, well... material.

So now you are the one in the position of railing against "materialistic" science for making claims that reject your intuitions about how the world works.

The only class of 'hidden variable theory'* that can survive in the face of Bell's Theorem is the 'nonlocal' ones- the ones in which information propagates faster than the speed of light. At which point causality comes into question.

And while you may not want to believe in randomness, do you really prefer the idea of abandoning causality itself for the sake of getting randomness out of the picture?

*(i.e. the kind of belief you just expressed, only presented by actual physicists who understand quantum mechanics)

[Simon_Jester]

I had the post to Jub mostly drafted, so I just dashed it off; everything else is in here.

P.S. Simon, you are generally correct but note that intelligence is a feature of information processing systems, and to reason correctly about it you must think about information transforming functions implemented as causal physical processes.

I already do; I just didn't make it explicit enough.

Minds are not static information though; given sufficient technology we could take a static copy but it is only considered a mind when combined with a mechanism capable of executing/insantiating the intelligence.

Sure. But the basic point, which I think we agree on, is that mind (and philosophical concepts related to mind) refer to the realm of information, not to the realm of matter. They're made of bits*, and they're not made of particles.
_____________________

*Additional constructs other than the 'bit' may or may not be needed to fully describe the mind, but if they are, I don't know what to call them. Point being, they're units of information, not units of physical matter.

And yet, we recognize them as having a quality of sameness that transcends their material composition.

What? We recognize them as having a quality of sameness because the words printed in those books are exactly the same. This is the problem I have with anti-materialist philosophy; it always feels like it is using semantic tricks rather than actual rigorous thought. Nobody would call two different editions of the same novel the same exact BOOK, because they are distinct physical objects. But there's no conflict there with the fact that those two different books have the same CONTENT.

Excuse me. I had no intent of conflating 'the book' with 'the contents of the book;' I simply used 'the book' as shorthand.

My point, though, is that there are a lot of different formats in which the same 'contents of the book' can be represented. And we still call all these different formats instances of the same "book," in some sense. A copy of War and Peace is still a copy of War and Peace, even if it's chiseled on stone tablets instead of being stored in flash memory.

So when we point to a copy of War and Peace, we are talking about two different things at the same time- the information component and the material component. The material component is, well, whatever that particular copy of War and Peace is physically made of. Its particles are arranged in some way. The information component is the text of War and Peace, which can be represented in all sorts of different ways by different arrangements of matter, depending on our choice of format and medium.

In the material sense, a copy of War and Peace on stone tablets is totally different than a copy of War and Peace on a flash drive, because one is a pile of rocks and one is a memory chip.

In the information sense, the two copies of War and Peace are entirely the same because both contain the same information.

So no, the content of the book is not, in my opinion, a material composition. Material objects made of different substances are materially different. But copies of War and Peace made by precisely aligning carved void spaces on a slab of rock can contain the same content as copies made by precisely aligning electronic states in a slab of transistors. They are materially very different, but not different at all in information terms.

Now, our mind has its own response to the words in War and Peace. Like the copy of War and Peace itself, this information is stored through elaborate arrangements of atoms in the brain. But that doesn't mean that the information IS the elaborate arrangements of atoms in the brain. The storage medium is not the same as the thing being stored; this is the critical concept behind the idea of the Turing machine.

All exchanges of information we can possibly detect are mediated through interactions of matter, because matter is the only thing that can be directly observed. That does not mean information IS matter.

The problem is, I alluded to above, the lack of a precise and consistent definition of what "information" is. It is something that can be arbitrarily redrawn every time an example of its material impact is demonstrated, so that it can always remain theoretically abstract because it is a more or less meaningless concept.

If in every meaningful way we can interact with "information" we can connect it with discrete and measurable physical phenomena, by what criteria can we then claim it is still non-material without waving our hands and redefining what the "information" itself really is?

Most forms of information can, in principle, be measured in bits and subjected to analysis via information theory. Information theory contains numerous rigorous means by which such information can be quantified, manipulated, modeled, and analyzed.

There are some high-level abstractions (like "sharpness") which may resist being measured in bits, and which are arguably an entirely different thing that is neither matter nor information. However, in my opinion, they share enough of the basic properties of information* that I class them as information.

Granted, abstract concepts are information that is harder to understand than the routine stuff covered by information theory as we know it. This is much as the physical behavior of a plasma is harder to understand than the physical behavior of an electrically neutral gas.

*Off the top of my head, the basic physical properties of information are:
1) Being descriptive of various arrangements of other things, usually matter but occasionally space, or even other information.
2) Being represented by (various forms of) precisely arranged matter.
3) Being able to continue existing in complete form even when part of the medium storing them is destroyed, so long as a backup copy exists. Destroying one copy of a computer program doesn't in any sense damage the program, so long as a second complete copy exists. Whereas you cannot destroy part of a wall without damaging the wall; the existence of a second wall in another place is irrelevant.

[Jub]

Simon, I'm a bit under the weather and more than a little groggy at the moment, so it might be a day or two before I reply. Just figured I'd let you know that you can expect a reply at some point.

[Starglider]

They're made of bits*, and they're not made of particles. *Additional constructs other than the 'bit' may or may not be needed to fully describe the mind, but if they are, I don't know what to call them. Point being, they're units of information, not units of physical matter.

The causal analogue of bits is binary logic gates. All digital information processing systems can be described as a wiring pattern for some number of NAND (or NOR) gates (in abstract, we assume they are synchronously clocked). Analogue systems are similarly composable from basic functions, although it's a little more complex. A static copy of a mind is analogous to the gate wiring pattern (in some suitable binary encoding), but to understand that as a description of a mind (much less actually run it) you need a notion of causality. Of course you can also encode any information processing system as a tape for a universal Turing machine, or numerous other Turing-equivalent processing mechanisms, but logic gates are the simplest (in causal terms) and closest to physical reality. All digital processing elements that meet a functional specification are equivalent and in an abstract sense 'the same', just as different physical instantiations of a piece of information are 'the same' piece of information, but dynamic equivalence is different from static equivalence. It's not just the fact that pattern is temporal; the causal relationship is what makes processing elements useful and is what ultimately underlies a lot of higher level phisolophical constructs such as intentionality.

I'm sure you know all this in principle, but leaving out the key role of causality and causal dependency is a surprisingly common failing in reasoning about minds. Daniel Dennet went into this in detail back in the 1980s, proposing lots of thought experiments that show how nonsensical things get in extreme cases if you don't take this view of what minds are, but judging by the perennial popularity of the 'teleportation is murder' argument it still hasn't percolated into the mainstream. This stuff only seems obvious to me now because I've spent years trying to write AI reflection systems, and when grounding reflective patterns in propositional logic there is pretty much only one representation that actually makes sense.

Generally though I agree with everything you have posted.

[Simon_Jester]

OK, fair enough. I should have been able to sit down and remember this, it is as you say theoretically known to me, but thanks.

Basically, yeah, it makes sense that you represent causality and information processing in terms of, say, abstracted NAND or NOR gates. And, to amplify this to everyone else...

The point here is that both data and thinking-processes are something distinct from material substances. You can represent information in an unlimited number of different material forms, but information is not itself a material- not an element, not a particle, not a field. Likewise you can represent a thinking-process in terms of various arrangements of matter- but one arrangement is as good as any other, except for practical concerns like "will this computation get done any time this century?"

In theory, you could take all the information content of the universe and represent it with a sufficiently large grid of rocks. If you had enough rocks.

So it makes very much sense to talk about information and logic processes as having an existence of their own. Any given computing machine (like your brain) that implements those logic processes to understand that information is of course governed by the laws of physics. But that's the point, because the laws of physics are part of the process that (in hardware terms) makes the computer capable of actually processing the information.

Thus, the laws of physics are not a separate thing from "I" that imposes oppressive external constraints on "I." Rather, the laws of physics are what make it possible for a big pile of carbon and water to be "I."

Now, plenty of immature philosophical approaches exist that ignore all this.

[Channel72]

All digital processing elements that meet a functional specification are equivalent and in an abstract sense 'the same', just as different physical instantiations of a piece of information are 'the same' piece of information, but dynamic equivalence is different from static equivalence. It's not just the fact that pattern is temporal; the causal relationship is what makes processing elements useful and is what ultimately underlies a lot of higher level phisolophical constructs such as intentionality.

That seems arbitrary. Two comparable pieces of information are only "the same" (or different) based on some arbitrary equality comparison function. Just because x == y doesn't mean &x == &y. In the first case, the equality function may indeed indicate that x is the practical or functional equivalent of y, but that doesn't mean x and y have the same location in space-time or the same memory address - they may be existentially different - and that difference may actually matter in practice. It would certainly matter, at least, to any function which wanted to count the instances of a certain piece of information which you're claiming are all identical. Just ask Amazon how many "instances" of Stephen King's latest novel they've sold.

[Simon_Jester]

I don't see how anything he said contradicts anything you said.

[Channel72]

It seem he's saying that, philosophically, "different physical instantiations of a piece of information are 'the same' piece of information" - but that's only true according to some arbitrary equality comparison function. Since the cliche "transporter kills you" argument has inevitably already been brought up (and it seems Starglider disapproves of it), I usually argue that it does kill you, in the sense that it eliminates an instance of you. It "deallocates"* your instance, and instantiates a new instance of you. If we use a comparison function that compares your molecular structure before and after transport, that comparison function will find both of you equal. But if we use a comparison function that compares your 4-D path (world line) through space-time, you and your clone are decidely different objects.

* a process for which "death" is a valid semantic interpretation if we're talking about a living organism

[Simon_Jester]

Well, the catch is that the comparison function we use to say "this copy of War and Peace is the same as that copy" is important, and I would argue the correct one to use. Because it's the same comparison function that lets us even identify what does and does not constitute "a copy of War and Peace" in the first place.

Without a comparison function that extracts the information content from the material content, and analyzes structure as opposed to format... well, put this way. If you don't use such a comparison function, a hardback copy of War and Peace has little in common with an electronic copy of War and Peace. And rather more in common with a hardback copy of Classic Myths for Your Children or whatever.

But by that argument, there is literally no such thing as a distinct 'novel' called 'War and Peace.' There are only purely incidental arrangements of squiggles on a page, or of bits in a flash drive. The entire concept that "the novel War and Peace" is a real thing hinges on our ability to analyze its textual content, and compare it to the textual content of other things.