Has consciousness been fully explained?

[Ichneumonwasp]

Well an action that is not an action of something (i.e. that physically exists) would not be a change in the real world. Similarly, an action that does not have locality cannot be a change in the real world. My understanding is that you are claiming a simulation is "an action" that does not have locality and is not an action of a thing (but rather of another action).

Then I am still not being clear, so please bear with me.

Once again, the only reason why I even mentioned non-locality was to draw the parallel with Descartes -- to explain away, hopefully, his error.

Any simulation in a computer occurs within the computer. It is actions themselves, as actions, that don't have a locality precisely. So, say you see a horse running. Where exactly is the 'running'? The horse is right there, well now way over there, but that's the horse. Where is the 'running'? It doesn't have a locality in the sense the horse does because it isn't a thing. It is an action. It's action only makes sense at it occurs in relation to the horse, but it is not some sort of separate thing that can have a location or not have a location. It is a relation of parts. Location is not a word that one should ascribe to it. We can speak of the location of the horse with impunity, though.

The same is true of anything that a computer does - it occurs within the confines of the computer, but it isn't proper to speak of its location in the same way that it isn't really proper to speak of the location of 'running'. It's the horse (and the computer) that has a location, not the action (which is a relation of the parts that have a location).

That is the only sense in which a simulation can be said not to have a location. It isn't that it doesn't happen anywhere, but that location isn't really a proper label for it.

 

[Malerin]

Summation is the operation of combining a sequence of numbers using addition; the result is their sum or total.
http://en.wikipedia.org/wiki/Summation

OK, so what you are saying is that summation, even in a general sense, is arithmetic in base ten?

The computational definition of summation is adding numbers. The neurological drfinition of summation is adding inputs and outputs to neurons. Just because it happens to share the same name doesn't mean it has anything to do with computation.

That's a form of computation and neurons are not doing that. Isn't the term you're using actually "spatial summation"?

Are you saying that spatial summation (it's actual spatial and temporal, but that's another matter entirely) is not a form of summation? What are we really supposed to call it?

Call it a form of summation then. That doesn't mean the neuron is computing.

Definition of COMPUTE
transitive verb
: to determine especially by mathematical means; also : to determine or calculate by means of a computer
http://www.merriam-webster.com/dictionary/compute

Is a neuron doing anything remotely like that?

No one argues that neurons do addition in the way that humans talk about doing it (so, I'm really not sure why you would try to imply that).

When you say that neurons are computing, that's exactly what you're saying!

What neurons do is take inputs, add them together (sometimes subtracting), to arrive at a final input that is coded temporally. The behavior of neurons follows rules that are based in physics and biology and these rules are not completely chaotic like rocks falling.

A falling pachinko ball is not chaotic. It's behavior is goverened by physics. Is it computing it's way down?

They are quite controlled and produce a limited number of outputs. Why is that not a form of summation, not a form of calculation?

Because nothing is being calculated!

Definition of CALCULATE
transitive verb
1a : to determine by mathematical processes <calculate the rate of acceleration>

Where are you getting your definitions?

Isn't the idea of calculation that inputs are summed following a set of rules to produce an output? Isn't that the essence of summation?

No. The essence of summation is adding numbers. NUMBERS!

Because it would not be a simulation of real consciousness. Do you go unconscious when no one's observing you?

Wait a second. Is there real consciousness and not real consciousness? Could there not be simply different forms of it.

Each particular instance could be a form. I have no idea what your conscious experience is actually like. However, you would agree that no type of consciousness dicovered so far is observer dependent, correct?

I don't believe for a second that a simulation ceases to be a simulation when no one is looking, so this is a moot point, but your point that it is a problem for RD and Pixy simply falls flat.

If no one is there to interpret the pixels on the screen, then that's all you have: pixels on a screen. It takes an observer to go to the next step and assign meaning to what's on the screen. This goes back to the falling abacus: just because the beads are moving doesn't mean its computing anything.

The simulation is either conscious or not unless you can decide on what constitutes real and not-real consciousness. There could potentially be any number of types of consciousness that do not follow a human pattern.

Potentially. There is potentially a tea cup around Jupiter, right?

See above. It would be qualiatatevly different than real consciousness because real consciousness is not observer dependent. I'm not sure observer-dependent consciousness is even a coherent concept.

If it isn't a coherent concept then why did you bring it up.

I'm not sure if it's coherent or not.

Simulations continue to occur whether anyone is looking at them, so now I am totally unsure what your point was.

Definition of SIMULATION
1: the act or process of simulating
2: a sham object : counterfeit
3a : the imitative representation of the functioning of one system or process by means of the functioning of another <a computer simulation of an industrial process>

How do you get imitative representation without an observer?

I'm suggesting if it comes down between a recognized authority like Searle and a bunch of anonymous forum posters, the smart money is on the authority. Perhaps someone here has published something as influential as the Chinese Room? Anyone?

Sure and you'd generally be right if you were just going to decide on the person and not look at the argument. Are you telling me that you don't understand the argument or that you don't want to work through the argument? I didn't just say Searle was wrong. I gave you a reason why he was wrong.

You're reasoning is wrong, your definitions are wrong, and Searle's reasoning is right.

He is simply wrong. You don't have to believe me -- look at what neurons do. They summate. He's wrong. There is simply no reason to appeal to authority in this sort of situation, so I don't understand why you would want to go that way.

Because he's right?

Again, I tell you: summation is the operation of combining a sequence of numbers using addition; the result is their sum or total.

Can you at least see that inputs coming into a neuron are just that? Each EPSP in most CNS synapses is 1/30. They summate at the axon hillock to 1 (threshold) so that they neuron fires. We can call them numbers, or EPSPs or whatever we want.

We can call them numbers because we understand the concept of what a number is. There are no numbers involved in a neuron adding inputs and outputs. It's not calculating anything, by any stretch of the definition I've provided.

The point is that something summates to create a total. That total can then do something. It doesn't matter if you recognize it as a calculation or addition or anything for it to do what it does.

I'm not denying there are more or less outputs after it's done "summating". I'm asking how you're going from adding inputs and outputs to "determining, especially by mathematical means". There's quite the missing gap there.

Adding and subtracting numbers. Isn't that what computation is all about? Do you think numberless computation is possible?

What is a number?

A mathematical object.

I certainly think that computation without numbers is possible.

And now we're back to Westprog's point: everything becomes computation. Why stop at neurons? Rockslides and asteroid impacts and supernova are computation. It becomes a meaningless concept.

Computers do it all the time. They use electricity that amounts to the same thing as a number. Neurons use ion channels and synapses to do the same thing.

Computers are a collection of a lot of tiny switches. Computation only occurs when we assign meaning and values to the results of those switches. Do you think a chess computer is actually playing chess?

And how can electrcity "amount to the same thing" as a mathematical object?

Except we have multiple definitions going on and you're playing fast and loose with them. You don't seem so gung-ho about definitions as you were earlier in the thread.

How am I any less gung-ho about definitions than I was earlier?

Because now it's become a "semantic cesspool".

As I have said all along, I don't care what you want to call the process that neurons do. But what they do is summate inputs. It isn't complicated.

Not at all. Except when you go from spatial summation ----> computation. Then it becomes complicated. And wrong.

If you want to restrict your definition of computation to what humans do with numbers, as I have said all along, go right ahead.
But neurons are going to keep summating their inputs no matter what anyone wants to call it.

And rockslides will continue to ADD rocks to the ground. It's not computation.

 

[brokefoot]

Has consciousness been fully explained?

No.

 

[Meed]

Ah, OK, then I have not been clear, which is my fault and I apologize. As I tried to explain above, I used the non-locality only as a reference to the way that Descartes discussed the problem. I am not arguing in any way that there is no relation to the real world. I am not aware of any action that ultimately does not depend on a physical reality. Please excuse my clumsiness if you ever got that idea.

Okay, I may have misunderstood in that case. Not a problem.

What I was trying to argue previously is that I don't see what the problem is in having an action result in further actions, just as we use abstract concepts to drive other abstractions. In other words, why could not the action of a simulated particle be able to join with the actions of another simulated particle to drive a simulation that could produce a conscious simulated person. Of course it is all tied to an underlying physical being -- a computer. I don't see how we could interact with anything not tied to a physical being.

It is an interesting question and not something I can say I know the answer to for sure. But I definitely don't think it's clear that this could result in a "conscious simulated person" as opposed to a "simulated conscious person", for the same reason I would not expect a simulation to produce a "wet simulated puddle" as opposed to a "simulated wet puddle". Because the simulation at its implementation (i.e. 'real') level is not wet. Rather, a subset of it and/or its behavior is isomorphic to a wet puddle. This seems to show that a simulator exhibiting an isomorphic mapping to something does not actually produce its qualities. I mean, we agree that a simulation of an orange is not an orange and does not produce an orange. Therefore I think it's very reasonable to argue that a simulation of a conscious person does not produce either a person or consciousness.

If I understand you correctly, you would say that that is true of a simulated "thing", but not a simulated "action" -- and conscious is an action. I don't really agree either that consciousness can be defined merely as an action or that a simulated action (unlike a simulated thing) can be said to be the same as the action it is simulating. I don't think your argument is outlandish though.

I just don't see why it is the case that there cannot be another level of abstraction between the physical being (computer) and the action we are discussing (consciousness) -- that other level of 'abstraction' being the simulation that makes up the computer world. I'm not sure that I accept without reservation the physicalist argument (though my sympathies still lie there) that action must only occur with physical beings alone. Yes, they must be connected with those physical beings, but what is the reason that the actions initially produced by the physical being cannot result in further action? That is what a simulation does.

Well, I think "abstraction" must, by definition, be something that only exists in the mind of a cognizer. Abstractions are immensely useful, in fact we probably can't think about or perceive the world at all without abstraction. And they can be useful to consider real even though they aren't (outside our brains at least).

I haven't claimed that actions can't result in further actions. I've just claimed that those further actions would still be actions of "things". If action is change in the physical world (my preferred definition), then of course changes in the physical world can lead to other changes in the physical world.

 

[Meed]

Any simulation in a computer occurs within the computer. It is actions themselves, as actions, that don't have a locality precisely. So, say you see a horse running. Where exactly is the 'running'? The horse is right there, well now way over there, but that's the horse. Where is the 'running'? It doesn't have a locality in the sense the horse does because it isn't a thing. It is an action. It's action only makes sense at it occurs in relation to the horse, but it is not some sort of separate thing that can have a location or not have a location. It is a relation of parts. Location is not a word that one should ascribe to it. We can speak of the location of the horse with impunity, though.

Ah. Yeah, I misunderstood what you meant by locality, then. I thought you had conjured up some sort of anti-realist sorcery. I don't think I disagree with most of your above explanation. However, if we asked "where is the horse running?" an answer involving location would be reasonable. Same with "from where did the horse run?", "where did the horse run to?" and "what route did the horse run?" Since an action always applies to a thing, we can probably speak of location in regard to any action, so long as we pose it in terms of the thing it applies to rather than the action itself.

 

[westprog]

Nope.

Already explained how mathemacal concepts can be assigned a physical reality.

Try again.

It's quite simple, really. All physical interactions can be mathematically described. All of them. If neurons are doing calculations, then so are melting ice cubes, so are asteroids, so is the interior of the sun.

Computation is either non-existent in the real world, or it is universal. I expect that this will be characterised as "Westprog can't tell the difference between computation and cheese" but I don't anticipate any kind of refutation.

 

[Democracy Simulator]

Are you serious?

Yes, i'm just interested in testing your definition of consciousness.

If we take an example of consciousness thought that is expressed by the words,

'Plane travel is faster than travelling by boat'

Where is the self reference in this thought?

 

[westprog]

No, I'm sorry, but that is just not the case. Computation is not just two things ending up together, two rocks falling together, so 1+1=2. Computation is a process that requires some set of rules that are followed. One of the reasons that we tend only to think of it as observer dependent is that most of the examples we encounter in the world do require an observer to interpret what she sees as computation -- like the example of rocks falling or using an abacus (or a computer for that matter). But that is simply not the case with neurons. They actually perform a function according to a set of rules decided on by natural selection -- they summate inputs to arrive at an output. I'm not at all sure why this is a controversial issue. It's something that just *is*. There isn't any way to argue against the reality of what neurons do.

That you consider what they do to be adding two numbers is purely the meaning that you assign to it. It's not doing mathematics any more than the Missouri is doing mathematics when it joins the Mississippi.

If there were a physical definition of computation, such that we could describe when it happens in the physical world, then we'd see such a description in the physics textbooks. There isn't, and we don't. Computation is not a physical concept.

 

[westprog]

The inference an unbiased person is supposed to make upon thinking of such a simulation, the inference that you have repeatedly shown that you are unable to make, is that if the Planck scale really represents some kind of a fundamental limit of fidelity in our own universe then oh, maybe there is no way to prove we are not in a simulation to begin with.

Stick your head in the sand all you want Mr. Ostrich, it doesn't change that fact.

I'm well aware of the Matrix hypothesis. I also know the limitations of "maybe there is no way to prove we are not" reasoning.

 

[westprog]

Nope.

Already explained how mathemacal concepts can be assigned a physical reality.

Try again.

Mathematical concepts don't have a physical reality. However, mathematics can be used to describe the physical world.

There seems to be considerable confusion about this, and has been for a long time. It's not surprising that physicists have to be excluded from the discussion.

 

[westprog]

Go on then. I'm keen to learn.

What is the definition of 'Self' in SRIP?

Whatever the definition of "self" in a C++ program, it's very different for a person. A person defines his self by the parts of the world that provide sensory input. That's why most people regard their hands and feet as being a part of themselves, are ambivalent about their nails and hair, and don't consider their houses, cars or clothes as being integral.

Consider a robot controlled by a computer from a remote location - with real time feedback from its limbs. However, the computer and the program controlling the robot has no information about its own state. Then consider a computer program running a simulation program, with no external connections, but with self-referencing so that the program monitors everything about itself and the computer it's running on.

Which is closest to how a human being actually works?

 

[Ichneumonwasp]

Well, I think "abstraction" must, by definition, be something that only exists in the mind of a cognizer. Abstractions are immensely useful, in fact we probably can't think about or perceive the world at all without abstraction. And they can be useful to consider real even though they aren't (outside our brains at least).

I haven't claimed that actions can't result in further actions. I've just claimed that those further actions would still be actions of "things". If action is change in the physical world (my preferred definition), then of course changes in the physical world can lead to other changes in the physical world.

Well, isn't the whole 'simulation' thing a bit of a polite fiction anyway? I am no computer whiz -- I built a couple of them and took one programming course in Pascal many, many years ago -- but isn't programming just a top-down way of getting the electrons in the machine to go where we want them to?

There is still a physical process going on with physical inputs and outputs. The big difference I see between how the brain and how a computer does it is that the brain does it from the bottom up. The flow of information in a brain is largely decided by genetics and early development, though there are obvious changes through learning.

We still talk about programs as though they are a thing themselves, but really they are just a set of instructions that tell lower levels how the electrons are supposed to flow through 'gates'. Computers and simulations are a top-down type of design as opposed to what brains do. But I still don't see why they couldn't do the same thing as a brain, that is if we got better at our top-down designs.

That's where I think Searle might have over-reached. He took a fairly valid criticism of one type of solution and generalized it to all programming. I don't think he's right, but I certainly can't prove it.

 

[Dancing David]

And what do we interpret with?



If we can't do anything with a mind, how can a "theory" arise?

So why not call it 'the sum of brain processes' rather than a non-existant 'mind'?

That mind is as real as the orange on the screen.

 

[Ichneumonwasp]

That you consider what they do to be adding two numbers is purely the meaning that you assign to it. It's not doing mathematics any more than the Missouri is doing mathematics when it joins the Mississippi.

If there were a physical definition of computation, such that we could describe when it happens in the physical world, then we'd see such a description in the physics textbooks. There isn't, and we don't. Computation is not a physical concept.

Once again, if that is the only way the word computation is used, then that is fine; it is simpy the wrong word to apply in this situation. But it still does not negate what neurons do -- which is summate. They take information, add it to get a new output and do something with it. No observer needs to define any of this as information (or as anything) for the neuron to do what it does.

Unless we really do want to wallow in a semantic mess, because computers at their most basic level don't compute either then. They take inputs following a set of rules to produce a new output. That is what neurons do -- the constraining rules being very important to the process.

I'm not sure how many times I have to say that I'm not claiming that neurons "do sums". If the words 'computation' and 'summation' are so tightly constrained in their meaning, then I humbly suggest that we have all been using the wrong words to talk about computers as well.

 

[Dancing David]

Go on then. I'm keen to learn.

What is the definition of 'Self' in SRIP?

I am not sure because there is no 'self', there is a body in humans and its attendant proecsses. But as far as recreating a self in terms of SRIP, I haven't s clue. I think that SRIP can be made to model the processes in humans and to replicate them once they are understood.

The self is a construct of memory more than anything else, our pattern of 'self' is inconsistent and fluid.

 

[Dancing David]

Yes, i'm just interested in testing your definition of consciousness.

If we take an example of consciousness thought that is expressed by the words,

'Plane travel is faster than travelling by boat'

Where is the self reference in this thought?

I am not defending SRIP, as i am not strong enough in programming to really appreciate it, I do feel that intuitively it makes sense to me, especially given the way neural structures develop, but that is just a feeling.

Now part of the medical definition of 'conscious' and at full capacity requires that the individual show 'orientation' to the three construct or 'person, place and time'. Other parts of the mental status exam also involve 'general knowledge' of past events and current events.

So there is the test of how the body of the individual is commonly labeled, where that body is and what day, week, year and seasonal rhythms it is currently in, and how much of the general history around that body can be recalled.

 

[Dancing David]

Has consensus ever been reached on anything in the philo forum?

The closest is that theories of dualism are incoherent.

 

[Dancing David]

I already did. Summation is a form of computation. If neurons don't summate, what do they do? If adding inputs to create a new output is not summation, then what is the definition of summation?

One theory is they act more like probable gates, detrmining the open status through association with other neurons. It could be modeled in a summation matrix, that does not mean it is a summation process.

 

[Dancing David]

I agree Searle's Chinese room argument was fallacious in the jump from "the man in the room does not understand Chinese" to "therefore a computer with the same capabilities couldn't understand Chinese" (paraphrasing what I think his argument was).

Where I'd agree with him is that functionalist definitions of mental states are not adequate, except from a purely pragmatic standpoint.

Um when the functional model has one trillion units and 7,000 connections for each unit?

It is Searle's model that fails, not functional models. (They could but the Chinese room is a very simple toy model.)

 

[Ichneumonwasp]

One theory is they act more like probable gates, detrmining the open status through association with other neurons. It could be modeled in a summation matrix, that does not mean it is a summation process.

That is certainly what they do at a system level. Is that not similar to what a computer does? That is my ultimate point. At their most basic level do computers compute according to this definition? It's electron flow through gates, just as with a single neuron we have EPSPs and IPSPs summating at the axon hillock and, in systems, differing levels of association through synaptic plasticity.