Explain consciousness to the layman.

[Piggy]

So when you say "brain state", do you really mean "the interpretation of conventional symbols in a pre-established language"?

No.

I mean the actual physical state of a real human brain... or some other brain that's able to use external objects as symbols.

Please stop trying to interpret every statement I make about the real physical world as if I were trying to make a statement about something abstract or symbolic.

The question before us has to do with the behavior of physical objects in real time and space.

Unless we look at them, we can't answer questions about them.

 

[Piggy]

You're misunderstanding what I'm doing. My complaint isn't about removing brain states from the picture--it is about the way you chose to define "physical addition". I won't get into it in this reply.

I'm aware of that quibble, and it's irrelevant.

However, we need some way of expressing the fact that the machine physically channels 6 or fewer marbles, iirc, into a group at the bottom, and yet it can be used to make a brain think of the aggregation of much larger groups.

Somehow we need to establish that fact and explore its implications.

However you want to phrase that, it's probably fine with me, but we need to be able to get that much established anyway.

 

[Piggy]

My analysis in the long post was based on playing with the machine instead of watching the video. But it was not based on interpreting those symbols at all.

No, it's not obvious to me at all. In fact, quite the opposite. This is a simple puzzle--the physical machine would be a fairly decent puzzle, just handed to someone, without any rules. The style of analysis that I suggested in my long post has nothing to do with what is painted there. You can put green squares on there if you like, but I would like to one up you and suggest removing all of them. It's still possible to figure out what the machine does.

Only the descriptions that match what the machine actually does are possible. And a description of this machine's function is a description of its meaning.

Well, let me give arbitrary labels to these things just for discussion. Label the columns, from the right to the left, $, #, @, !, and *.

You can get a marble in column # by adding a single marble to column #. You can also get a marble in column # by adding two marbles to column $. So one marble input into column # is the same as two marbles in column $. All of this is simply how the machine behaves.

And that is where the two-ness of the number in column # is.

There is also a corresponding two-ness of the number in column @; that would be the fact that one marble shoved into column @ causes the machine to get into the same state as two marbles shoved into column #. And since column # has a two-ness equivalent to the marbles you can shove into column $ to get it, then column @ has a mega-twoness about it--a twoness of a twoness. That is, if you shove two marbles into column $, and you shove two more marbles into column $, you get the machine in the same state as shoving one marble in column @.

Note that nothing in this explanation appeals to the interpretation of pre-established symbols.

The machine behaves this way regardless of how you guess that it does. The only issue is figuring out how it works. But "how it works" is simply how that machine works--so it's all in the machine.

Your claim that it's impossible to figure out what this machine does is simply wrong.

Well, it's not quite what I said.

What I said was that you could figure it out if you decided to try to figure out every possible use for the machine, real or symbolic, but you'd have no more reason to think that would be the right choice than any other option.

So that's better summarized as "it's impossible to figure out what the machine must be used for".

And I hate to point this out, but... of course the machine behaves that way. Of course those relationships are there.

If they weren't, it wouldn't work the way it's supposed to for folks who know what the symbols mean.

The problem is, these relationships could be coincidental, or related to some other purpose than performing symbolic addition.

Why consider those relationships to be meaningful, and not the relationships among the lengths of the channels (as in a pipe organ)?

Why imagine that it's supposed to add anything at all?

In short, there is no object which can only be put to exactly one use, or made to represent exactly one thing.

And if you wanted to substitute this thing for a part in another apparatus, then you'd better hope that the other apparatus doesn't need to group large groups of marbles at one time, because this machine won't be able to do that, regardless of the mathematical results of one of the symbolic uses it could be put to.

So if you want to talk about this thing as an object, and ask what work it can do as an object in a physical system, you're going to have to strike "aggregation of large numbers of marbles" off the list.

For precisely the same reason, if you want to swap a simulator machine for a brain, you're going to have to ignore the symbolic values you attach to whatever the machine is doing, and look at its real matter-and-energy outputs exclusively.

Since they won't match the brain, any real work that the brain is doing at that point in the chain will not be performed. In other words, whatever the brain is doing won't really happen anymore.

Since conscious awareness is generated in the body by the brain at that span in the physical chain, it would not occur in such an apparatus.

 

[Piggy]

I'm getting the feeling, from your previous post, that you're referring to the labels above the column when you mean symbolic.

But the physical machine also has symbols in it. A rocker is either leaning towards the right trapping a marble, or leaning towards the left with no marbles in the vicinity. Those are symbols, and they are also what I would imagine you would call physical. And they are part of the machine.

They are always "part of the machine". In other words, in the system which includes the machine alone, the position of a rocker is part of the machine.

But they are not symbols in that system. (In some other system that includes a user, they could symbolize a number of things, or nothing.)

Symbol systems do not exist without a brain to imagine the association between the symbol and its referent. Nothing is a symbol if no one decides to imagine an association with it.

This goes right back to all logical computations being overlayed onto physical ones... the physical changes are real, and the logical ones are imaginary.

You know, there's a board game discovered in ancient Ur, and nobody knows how to play it, and nobody will ever know how it was played unless we find instructions, because there are many logical possible symbolic overlays.

This is true for any system, including the marble machine.

Yes, the relationships described by the Ur game will be inherent in the parts if one day we find the rules written somewhere.

But they won't be our only option til we find the rules.

If I put that marble in column #, it represents two marbles being shoved in column $. It does this because there's one machine state that can be achieved by either adding one marble to column # or two marbles to column $. So this state is a symbol.

Wrong.

It can do that, but you cannot say "it does this".

That relationship is there, but there are many other relationships one might list, and none of them can be symbolic of anything -- that is, they can't represent an abstraction or another physical object or an imaginary object -- without somebody somewhere deciding to imagine the association.

This is just the fact of the matter.

For instance, the relationship between the weight of a quarter and the weight of a dime is real. But it turns out not be symbolic of their relative monitary value. Ditto the size.

For specie, though, you can go by weight.

That fact is not discernable from the coins themselves without any outside information.

 

[Leumas]

Ftfy.

What is SRIP?

 

[westprog]

What is SRIP?

Self-referential information processing. If you don't agree that the only possible explanation for consciousness is SRIP, you necessarily believe in magic beans. Sorry, it's either or.

 

[PixyMisa]

Self-referential information processing. If you don't agree that the only possible explanation for consciousness is SRIP, you necessarily believe in magic beans. Sorry, it's either or.

I don't know if that's true, but there's certainly a strong correlation.

 

[yy2bggggs]

What I said was that you could figure it out if you decided to try to figure out every possible use for the machine, real or symbolic, but you'd have no more reason to think that would be the right choice than any other option.

Okay, so I agree that this machine can be used for multiple purposes. A farmer might use it to keep track of eggs in the henhouse. The builder might have intended it to balance his checkbook. And I have already stated that this sort of thing is not a kind of thing you can guess at. So if this is what you mean, I agree.

But I think your use of the word "right" is questionable. There's no right way to use the machine. There is only a way that particular people intend to use it. Even if it was built to balance the builder's checkbook, he can sell it to the farmer to keep track of eggs in the henhouse. Neither of these is the "right" way, and neither is how the machine "must" be used.

It's just how particular people do use it. But again, sure. It's impossible to figure out what they're using it for by only looking at the marbles.

The problem is, these relationships could be coincidental, or related to some other purpose than performing symbolic addition.

Unless we're just using the machine wrong, whatever use the machine may have should at least map to symbolic addition. But if we're using the machine wrong, then you're talking about a different machine.

Why consider those relationships to be meaningful, and not the relationships among the lengths of the channels (as in a pipe organ)?

Example? Show me what sort of invariants would be involved.

Why imagine that it's supposed to add anything at all?

What it is "supposed to do" is irrelevant--see the henhouse/checkbook example above. The machine does what it does, and you can derive meaning from how its states relates to its inputs. Beyond that you don't know; but up to that, you do know.

In short, there is no object which can only be put to exactly one use, or made to represent exactly one thing.

Sure, but this applies to the brain as well. The brain is a very nutritious source of protein nutrients. It could be used to derive meaning from the world, based on its physical structure. Or, it could be used as a food source.

So if you want to talk about this thing as an object, and ask what work it can do as an object in a physical system, you're going to have to strike "aggregation of large numbers of marbles" off the list.

Obviously. The machine only has 64 states. But you can glue onto the left of it to build a machine that uses more states.

For precisely the same reason, if you want to swap a simulator machine for a brain, you're going to have to ignore the symbolic values you attach to whatever the machine is doing, and look at its real matter-and-energy outputs exclusively.

The same reason as what? I don't understand how it follows that you can use this machine another way, that when used this way as part of a bigger machine, the latter cannot produce consciousness.

The symbol machine is being argued as the thing that can produce consciousness. And presumably you are arguing against this. Arguing that the symbol machine isn't used as one, or is used a different way, is a non-sequitur.

PixyMisa might build a giant marble android--a Pixydroid. Part of the Pixydroid's mechanics will involve this marble machine, in this way, to produce states--maybe the machine's states will represent how red or green a particular pixel is. That the machine can be burned and used as a fuel source is about as meaningful as saying that a brain could be used as a food source. Well, it can, but the fact that you can eat a brain doesn't well mean that your structures don't contribute to consciousness, does it?

Since they won't match the brain, any real work that the brain is doing at that point in the chain will not be performed. In other words, whatever the brain is doing won't really happen anymore. Since conscious awareness is generated in the body by the brain at that span in the physical chain, it would not occur in such an apparatus.

Yeah, you stated your premise already.

 

[yy2bggggs]

Symbol systems do not exist without a brain to imagine the association between the symbol and its referent. Nothing is a symbol if no one decides to imagine an association with it.

But you're begging the question. There is an association between that state of the machine and a referent--a particular set of inputs.

Furthermore, on the flip side, we do not actually in practice imagine the association between the symbol and its referent when performing an addition. In fact, some people are perfectly able to perform an addition and don't even know the association between the symbol and its referent. They just know how to apply a set of rules.

ETA: Fun related video to this part:

This goes right back to all logical computations being overlayed onto physical ones... the physical changes are real, and the logical ones are imaginary.

We learn logic by interacting with physical entities.

You know, there's a board game discovered in ancient Ur, and nobody knows how to play it, and nobody will ever know how it was played unless we find instructions, because there are many logical possible symbolic overlays.

Sounds like you have the entities but not the transformations.

If I put that marble in column #, it represents two marbles being shoved in column $. It does this because there's one machine state that can be achieved by either adding one marble to column # or two marbles to column $. So this state is a symbol.

Wrong.

It can do that, but you cannot say "it does this".

You're talking a different language, perhaps because you're so obsessed with brain states.

I've already told you. I'm not trying to figure out what that builder had in mind. I'm trying to figure out what the states of that machine imply about the inputs. So when I say that the marble in # represents two marbles being added to the column in $, I don't care if nobody on the planet has ever conceived of the machine working that way, including the builder. Because I'm not after brain states--you are. All I'm after is what that marble being in # means.

And it means the same thing whether one marble gets put into #, or two into $. Therefore, I say, because it behaves that way, regardless of anyone's brain state but my own serving only as the guy figuring out what the meaning of that marble being in # is, that that marble in # represents two marbles in $.

So if you say this:

That relationship is there,

Then you agree with me.

For instance, the relationship between the weight of a quarter

...has nothing to do with the marble machine.

 

[Modified]

They might well be electronic, but they would be electronic devices that are capable of replacing actual, physical neurons. That means they have to have very specific physical properties. This is sufficiently difficult to do that in spite of trying for many years, surgeons and scientists have been unable to do it. It isn't possible to just plug any arbitrary component into the nervous system. It isn't valid to claim that a network of computers or processors would be equivalent to a network of neurons, artificial or not. And calling something a "neural network" doesn't make it one.

Yes, but we were assuming that you can plug in a single neuron. The original point stands. If you can plug in and connect many of these to make a brain, then you can plug in a computer that emulates the entire system in the same way.

 

[westprog]

Then please answer my question: how do hearts and oceans perform algorithms and calculations ?

If you give me a definition of what you think an algorithm consists of, or of what calculations are, then I'll tell you exactly how.

I should say that I don't think this. I think that algorithms and calculations are human-specific concepts, and that it makes no sense to describe an algorithm being executed except in reference to a person wanting it to execute. However, the people who think that there's some kind of objective reality to algorithms in the real world, absent human interaction, have as yet not given a definition that would satisfactorily exclude any physical process whatsoever.

 

[westprog]

Yes, but we were assuming that you can plug in a single neuron. The original point stands. If you can plug in and connect many of these to make a brain, then you can plug in a computer that emulates the entire system in the same way.

And such a computer would be made of artificial neurons. A computer not made of artificial neurons wouldn't work. We know this because we don't have artificial neurons yet. We are unable to replace damaged nerve tissue. If we could do it we would.

 

[Modified]

And such a computer would be made of artificial neurons. A computer not made of artificial neurons wouldn't work. We know this because we don't have artificial neurons yet. We are unable to replace damaged nerve tissue. If we could do it we would.

No, the computer would not be made of artificial neurons, but simply emulated artificial neurons. If artificial neurons are implemented by digital electronics or by a computer, then there is nothing a network of such neurons can do that a single processor system with the same inputs and outputs can't also do. Given arbitrarily large storage, no computer has more functionality than any other, and no network of computers has more functionality than any single computer. Computable is computable.

 

[punshhh]

I'm getting the feeling, from your previous post, that you're referring to the labels above the column when you mean symbolic.

But the physical machine also has symbols in it. A rocker is either leaning towards the right trapping a marble, or leaning towards the left with no marbles in the vicinity. Those are symbols, and they are also what I would imagine you would call physical. And they are part of the machine.

If I put that marble in column #, it represents two marbles being shoved in column $. It does this because there's one machine state that can be achieved by either adding one marble to column # or two marbles to column $. So this state is a symbol.

But it is also physical. I can point to the exact chamber where the marble will be trapped to represent this symbol. If there were no such physical correlate, you could not say that this marble machine simulated addition.

This is what I've been saying. It has nothing to do with the Arabic decimal numerals painted onto the columns.

It is the presence of a brain(your brain) which enables you to see it in this way and so can I and piggy.

But what if it were a squid, it would have no subjective appreciation of these distinctions at all. It would be able to watch what was going on though and would be aware of what was happening to the marbles physically as we are.

It is due to the presence of a subjective understanding in our human brains which can interpret the activity of the marble machine as a form of processor. Without this subjective involvement it is entirely devoid of meaning and is just some marbles rolling around on a board, as observed by the squid.

It would be possible to program a robot to detect the marbles and determine the same relation of the marbles to the board that you and I can understand and respond/mimic the response of a human by performing a calculation. But the robot would be nothing more than a processor itself. Entirely unconscious, unaware like the squid of any subjective interpretation of marbles. In fact entirely unaware of anything. It is nothing more than an aggregation of cogs, wheels, on off switches and semi conductors.

If it had not been specifically programmed by a human brain to respond to the marble machine, it would not understand/interpret it if it stared at it for years.

 

[yy2bggggs]

It is the presence of a brain(your brain) which enables you to see it in this way and so can I and piggy.

But what if it were a squid, it would have no subjective appreciation of these distinctions at all. It would be able to watch what was going on though and would be aware of what was happening to the marbles physically as we are.

I've already said that I'm using my brain to look at it. But I have a problem with your approach.

If I have a strong emotional response to a really good movie, and start crying over it, then I would be feeling and appreciating the meaning of the movie in a very strong and visceral way.

But what if your squid were watching me? I doubt your squid would appreciate that I am experiencing the meaning of the movie. So I question whether the squid should have his place in the jury box.

 

[PixyMisa]

It is due to the presence of a subjective understanding in our human brains which can interpret the activity of the marble machine as a form of processor. Without this subjective involvement it is entirely devoid of meaning and is just some marbles rolling around on a board, as observed by the squid.

De Chelonian Mobile.

It would be possible to program a robot to detect the marbles and determine the same relation of the marbles to the board that you and I can understand and respond/mimic the response of a human by performing a calculation. But the robot would be nothing more than a processor itself.

And so are you.

 

[Piggy]

Okay, so I agree that this machine can be used for multiple purposes. A farmer might use it to keep track of eggs in the henhouse. The builder might have intended it to balance his checkbook. And I have already stated that this sort of thing is not a kind of thing you can guess at. So if this is what you mean, I agree.

You know those folks I mentioned who have one-side-blindness... the ones who see only half the world but can't tell that they're missing the other half... who'll trip over a footstool and swear that it wasn't because they couldn't see it?

Your way of thinking on this issue is precisely like that.

You find it impossible, not just to accept a view of the marble machine as an object, but even to contemplate such a view.

How do you know that the particular relationships in the machine's construction which you're focusing on are significant to its use, rather than coincidental? And how do you know that other relationships of the machine (say, its height, width, and depth, or the length of the channels) aren't used as indicators to perform further mathematical functions?

Perhaps if the marbles come through a rectangular machine, you're supposed to multiply the result by 10, and if they come through a square machine, you're supposed to multiply by 7.

Or perhaps the purpose of the machine is only to route marbles along those channels and through those baffles, and the particular mathematical relationships are coincidental, or are there because this part needs to fit into another part which has those relationships for reasons which are significant to it instead, or because the designer was a mathematician and an artist who found the relationships elegant.

Perhaps this machine works with another machine that uses the positions of the marbles to do something other than adding numbers. Perhaps it's the color of the marble that matters to the other machine, and the physical set-up of the marble sluice is to route different marbles to different positions to have their colors detected by the proper device.

Perhaps this device's only purpose is to limit the number of marbles moving through a certain point at a certain time to fewer than 7 at a time.

You have no way of knowing.

Of course, if you're exploring all the possibilities, and you find one like "it could be used as an adding machine", and you decide that there's very little chance that a human being would set it up like that accidentally, and you decide that there's very little else a person might want it to do, then you might think "It's got to be an adding machine."

But notice all the information you had to drag in from elsewhere. And notice that this information has to do with how human brains tend to act.

But the machine itself, if it were self-aware and if it knew everything that could be known about its body and nothing else -- in other words, if we're looking at the circle with just the machine in it, the machine as an isolated system -- the apparatus itself has no way to guess all these things, and so every possible use of the machine is just a good a guess about its body's purpose and function as any other.

(Odds are, though, that the machine won't come up with any ideas about its body which begin with "Suppose there were some other being using my bodily functions symbolically....")

So your world, in which our only choices are folks using the machine to add stuff, is a partial world, and you're not only ignoring the rest of the world, you're refusing even to see it.

Yet again -- and for precisely the reasons I've explained -- you find yourself right back smack-dab in the circle of the information system, which includes the machine and information in (which is to say, a physical state of) at least one human brain.

All I'm asking you to do is to take your eyes off that system and look at the other circle for a moment, the one with just the machine in it.

But you refuse to look there, even momentarily.

As long as you listen to what I'm saying about the system that includes just the machine, but you keep your eyes glued to the "information system" circle that also has a human brain in it, and you try to interpret what I'm saying about the machine alone into that picture, what I'm saying isn't going to add up.

That's why what I'm saying seems confusing or wrong to you.

 

[Piggy]

Unless we're just using the machine wrong, whatever use the machine may have should at least map to symbolic addition. But if we're using the machine wrong, then you're talking about a different machine.

Amazing!

Your definition of the "machine" you're looking at... apparently doesn't include the object you're looking at.

To you, the "machine" is an abstraction based on a single possible symbolic use of the object which has been described to you.

Wow.

But the thing is, if you're just looking at an object in isolation, there can be no "wrong" use, because there can be no intended "right" use. (Remember, the intended "right" use is off in a brain somewhere.)

There are only "possible" and "impossible" uses. (E.g. the marble machine is not a space shuttle.)

This object, as an object, has no "right" use.

ETA: If you don't believe me, paint over the numbers with colors instead, and give it to a 7 year old kid along with some marbles, and you'll see all sorts of real and imaginative uses that it could be put to.

 

[Piggy]

The machine only has 64 states.

Wanna bet?

It has a helluva lot more than that.

Try using colored marbles and see how many it has.

Stop ignoring the bone pile and see how many it has.

 

[Piggy]

I don't understand how it follows that you can use this machine another way, that when used this way as part of a bigger machine, the latter cannot produce consciousness.

The symbol machine is being argued as the thing that can produce consciousness. And presumably you are arguing against this. Arguing that the symbol machine isn't used as one, or is used a different way, is a non-sequitur.

It's not a non-sequitur at all if you take the entire argument into consideration.

First of all, it's not being argued that any "symbol machine" can create consciousness, but rather that physical objects can (and do).

The assertion that physical objects can make consciousness occur is scientifically sound.

The assertion that a "symbol machine" when used as such can do such a thing is not even scientifically coherent.

The brain itself, as a physical object, cannot be a symbol machine like the marble sluice, because when we look at the operation of the brain, we see only physical activity... and we don't find any 3rd party in an unconscious brain that could decide that anything in the brain was symbolic of anything else and thereby (how?) "produce consciousness".

If the use of the brain as a symbol machine somehow produces consciousness, then by all means, explain this to me:

1. If the unconscious brain is a symbol machine -- remember, if this process creates consciousness, we have to begin with a brain that's not conscious yet -- what are the physical symbols (if the symbols themselves also have no material or energetic properties, then we're really in fairy land!), what do they represent, how (and by whom) are the associations between the symbols and their referents created, how are those associations remembered by this entity, how does this entity then interpret the symbols later, and what does it do once it interprets them?

2. How does this process lead to conscious experience of the self and the world?

PixyMisa might build a giant marble android--a Pixydroid. Part of the Pixydroid's mechanics will involve this marble machine, in this way, to produce states--maybe the machine's states will represent how red or green a particular pixel is. That the machine can be burned and used as a fuel source is about as meaningful as saying that a brain could be used as a food source. Well, it can, but the fact that you can eat a brain doesn't well mean that your structures don't contribute to consciousness, does it?

The use of the brain as a food source is just as real as the brain's activity when it generates consciousness.

The two do not exclude one another.

And neither purpose is the "right" purpose for the brain or the "wrong" purpose for the brain.

And although the marble machine could be some sort of body part in a conscious machine, if we want that machine to be conscious, we've got to give it a brain that does whatever ours does to be conscious.

ETA: Note that your droid there isn't conscious yet. Consciousness is not required in order for a system to respond to color.