The subjective observer is a fictional character

[SpeedOfSound]

So this is where you guys have been hiding. Did anybody bring the beer? I've been reading some of the first few pages, but after I got to Surendra's third mention of spoonbending, I figured I hadn't missed much...or I'll get caught up later on.

Surendra, it's not that tough to understand the world view of others. It's one thing to understand it; it's another thing entirely to agree that the person knows what they're talking about.

As for "taking pot shots at rationality," I haven't seen that jamest or anybody else is doing that. However, let me be among the first to take pot shots as those self-proclaimed clear and rational thinkers who would have others believe they are more rational than thou. Simply because some rigid thinkers, in all of their wishful thinking, would like to have others believe they are rational doesn't make it so.

From what I've read of Graham's posts, I think that which you are trying to identify as "the subjective observer" is very real, and I'll be happy to start getting into that. But first, somebody pass me a beer and some of those chips and dip...

I suggest you read the last 9 pages of the thread.

[Surendra Darathy]

I suggest you read the last 9 pages of the thread.

Your blondes have arrived. It's just not the Swedish Cheerleaders. LaMont just wants to have fun. (Daddy don't preach.)

What James is engaging in (now with help from the peanut gallery) is a classic case of goalpost shifting. You'd think that with those abilities, he'd be able to bend a couple of spoons.

All of a sudden, we're no longer investigating whether the brain has anything to do with cognition, and are off trying to figure out the locus in the parietal lobe for prayer meetin's.

[Surendra Darathy]

So this is where you guys have been hiding.

Gittes: Mulvihill! What are you doing here?
Mulvihill: They shut my water off. What's it to you?
Gittes: How'd you find out about it? You don't drink it; you don't take a bath in it... They wrote you a letter. But then you have to be able to read.

[LaMont Cranston]

SOS, I'll do what I can to get caught up. I haven't seen your ass since we were talking about Jesus on RDF. I was in St. Louis a couple weeks ago, and it was so fucking cold I could not believe it, and, as I understand it, that place is warm compared to where you live. By the way, we brought some O'Doul's for you...

[LaMont Cranston]

Surendra, I've already explained to you that spoonbending is no big thing. You grab one end of the spoon with one hand and the other end with your other hand and apply pressure is any number of directions. I think there are even some formulas out there; Gee, I wish Mr. Wizard was still on TV.

Seriously though, (and we all know how seriously we take this stuff), I pounded you into submission on that other thread, and you took your woo, wibble and spoonbending and headed for the exits. To quote the great Joe Louis, "The Brown Bomber," "You can run, but you can't hide..."

[SpeedOfSound]

SOS, I'll do what I can to get caught up. I haven't seen your ass since we were talking about Jesus on RDF. I was in St. Louis a couple weeks ago, and it was so fucking cold I could not believe it, and, as I understand it, that place is warm compared to where you live. By the way, we brought some O'Doul's for you...

What was your user name at RDF?

Fuck O'Douls get me some Moussy. (Sibra Products, the importer of Moussy non-alcoholic beer.) What the hell happened to these guys?

[LaMont Cranston]

SOS, I was Dr. Robert Klass on RDF. So many people had a shit fit because I had a Dr. in front of my name. Besides, I always thought The Shadow was very cool, my kind of super-hero. Too bad that movie with Alec Baldwin playing LaMont Cranston was such a piece of shit. See you soon...

[SpeedOfSound]

All of a sudden, we're no longer investigating whether the brain has anything to do with cognition, and are off trying to figure out the locus in the parietal lobe for prayer meetin's.

If I got this right james is going to prove logically that the brain can't work but he doesn't want to talk about that sciency stuff like about how the brain works.

That about right?

[SpeedOfSound]

Off topic but where did we get the idea that we should teach the illiterate to read and write?

[SpeedOfSound]

how can the brain really know what any of its internal structures actually correlate to, externally to itself?

:pissed: Okay. I'm calm now.

The brain doesn't know like you think you know. You pretty much ARE the brain. So when you know things the brain is doing something and that something is not being a thing separate from the brain that knows things. You need to get the Dual out or you will continue to think like a sausage.

I would drop internal structures in favor of Neural Nets (NN) if I were you. The brain has many internal structures and a very specific way to wire them up that is encoded by your genes. This is present at birth if all went well. I'm beginning to suspect that it went better for me than you but that's not important right now.

Let's swap the sun for a ripe orange that your mommy becomes obsessed with showing to you shortly after birth and thereafter for about three years.

The visual system is very complicated and not what we all think it is. It has about 50 modules for tearing apart and image and creating and firing patterns of electrical spikes in NN's.

It is a self-teaching self-constructing system. The brain 'knows' that orange is orange because orange always enters the retina and then goes to the same spots in the brain that have always entertained orange info. It knows the circle that the orange looks like in the same way. It taught itself to see circles and orange and to put them together by seeing oranges many times.

If you showed an orange to a 7-month old fetus (just guessing on the timing here) there would be no qualia of orange but there would be something like a color blob floating around a weird edge thing that had little resemblance to a circle. But it is unlikely that there would be consciousness of any of that so what I said about how things look is just a way of explaining, not strictly real.

The infant can move it's eyes and learns to focus and then moving it's eyes crisscross over the orangy thingy will start to bring the color blob into the edges and circle shape modules will start to be carved in the brain.

There are many things that are getting learned here and all at once. Edge detection (mostly built in). Motion detection, Spatial fields and place, color, color blob attachment to the spatial field system, shape categories and detection, and more. A big part of More is the movements of the eyes and focusing of attention on parts of the visual field. The more this is done the more the color blobs get associated with the shapes and the location.

Billions of bits of information are carving a pathway for the orange experience in the infants brain.

The they are hooking up to other systems, not visual, in the brain such as the word orange and your mommie. If the bitch lets you eat the orange you will have olfactory and taste info as well.

For the rest of your life seeing an orange will light up these same clusters of neurons. If you imagine an orange many, but not all the visual, clusters will light up 'as if' you were seeing an orange.

There is massive traffic going backward from the places were things are imagined to the places where things are sensed.

SO the answer to your question is that you know an orange because you have seen many oranges and the paths that lead to your orange qualia have been carved, by oranges, in the systems that your senses are hooked up to.

Touching an orange with the tip of your toe in the dark is never going to be mistaken for seeing an orange with your eyeballs.

Also. The equation for the recognition or knowledge of an orange in your specific brain, if written out, would be larger than the galaxy even if you used a small sharp script. I'm guessing wildly but it would be really, really, fucking BIG!!!

[GrahamH]

Okay, let us - for the sake of argument - assume that the qualities usually associated with 'experience' are not actual. After all, I'm here to entertain such theories. Let us consider that 'yellow', for instance, doesn't have a physical correlate... and that 'physical correlations' are just associated with details of some external event. Likewise for each and every experiential quality that we can think of. So, the model we are considering (as I understand it) is one in which numerous localities within some region(s) of the brain correlate with sensed external qualities of the world. For example, there would be a specific physical correlate for the photonic energy (formerly known as 'yellowness') of the sun (likewise for all such external qualities that are reported to the brain via the sense receptors of the body). Now, at any one moment, the brain itself would have to take on the role of 'the individual' in putting together a meaningful 'picture of the world' to facilitate appropriate behaviour within that world. That is, there would have to be a singular 'assessment' of all relevant physical correlates, as a whole. Our reactions to the world are not just automatic - 'we' (commensurate here with the brain itself) often assess each scenario that we are confronted with prior to responding. So, we cannot escape the need for this oft-required singular review of the physical correlates as a whole. Now, with the details of the model in place, we can now assess whether there are any logical flaws within it.

The first problem I see, is that a considered response to a specific event within the world would amount to a considered response to brain states - those numerous localities of physical correlates associated with that event. That is, the brain would really be considering and responding to itself. The problem here, is that the brain would have to associate meaning with each physical correlate associated with any event. That is, the brain would have to know what each physical correlate meant in relation to any event. For instance, the brain would have to know that the physical correlate(s) associated with the photonic energy of the sun, was synonymous with the photonic energy of the sun. It might even assign tags to each correlate, such as 'yellow', to facilitate simpler processing of the information. And since we say that the sun is 'yellow', this would indeed be the case.
So, the question begs, how does the brain assess itself and know what any particular 'physical correlate' means?

SoS and SD have pointed out your error on this. You are thinking naive R1 and dualism when you are supposed to be looking at things as physical processes.

It should be obvious that by the model under discussion the brain isn't a dualist observer of brain states. Brain states are not perceived, they are perception. Brain states are not considered, they are cognition.

The brain doesn't "know what each physical correlate means". The meaning is inherent in the fact that the NN responds to the object or event. A NN that recognises a tree means a tree is perceived when the NN is active. It doesn't require an extra level of semantic processing to work out 'what it means'.

There are no 'tags' involved besides the NN themselves. The growth and reinforcement of connections in response to stimuli generates spontaneous new category 'tags' just by being a different NN responding. You might need to think this through a few times before the meaning sinks in.

Here, the problem is one of semantics - what philosophers of the mind have referred to as intentionality (also known as 'aboutness'). How, for instance, could the brain assess a physical correlate within itself, associated with some external event, and know that it meant the photonic energy associated with the sun ('yellow'). That is, how can the brain know what a localised physical structure/event, within itself, is about (aboutness)? I've thought of a simple way to illustrate the scope of this problem:

Try something simpler first:

Imagine a creature with a brain and eyes that sees an object it has never encountered before. It has no prior experience of that object. If the object has similarities with other things it has encountered before NN recognisers for those things may be activated by the object, but the response is muted and incomplete. This is a recognisable situation, something new. The creature doesn't know what it means. If contact is maintained, say by staring at the new object, a new NN recogniser will form. Activation of that NN means (I am) seeing object X. The ability to recognise object X allows the creature to learn things about object x, such as object x is trying to eat (me), or object x tastes good. These new facts about object X are of the same NN form as recognition of object x.

Because NNs recognise objects, situations, relations, events, there is no need for interpretation. The recognisers are 'interpreting' the scene by activating, or not. The activation of NN produces action. The premise is that in human brains the action can be as complex as vocalising 'I see an X hiding under that bush and he looks hungry'. Or maybe sense of joy. The 'consideration' is performed in the complex interactions of vast numbers of NNs. The 'considerations' are also patterns learned by 'experience', or generalised and recapitulated versions thereof.

Note that NNs are interconnected to increadible degrees, so a 'layer' of NNs performing sensory recognition may provide inputs to other 'layers' that recognise percepts as aspects of compound objects that involve many different percepts. The deeper this goes the more abstract things can get, We can have recognition of A about B about C about D ... We can have recognition of causal relationships by recognising A then B over and over. Remembering that A has just occurred primes for the possibility that B will occur. If it does the recognition is strengthened. This is prediction based on causal relations. This is a mechanism of inductive inference.

We can include recognition of facial expression and body language as indicators of 'state of mind'

I don't want to delve into this just now, I just want to indicate that we aren't limited to front-line perception.

Imagine that you are sat in a room and have no idea what is going on outside that room. However, I come into the room and present you with numerous lego structures. I've utilised different structures and colours-of-lego and each singular structure is synonymous with a specific detail of an event happening outside the room. To make things easier for you, I put these structures into groups - one representing sight; one representing sound; one representing smells (we won't bother with taste and touch information). Now, even in eternity, do you think that you could tell me what was going on outside? No, you couldn't, because you wouldn't have a clue what any particular structure was ABOUT, except generally (a sight; a sound; a smell). The problem is, then, that you would need to know what each structure was about in order to tell me what was happening outside. But if somebody doesn't tell you, then you're literally in the dark, forever. That is, a physical structure contains no meaning about anything, other than itself.

Re-read your paragraph and see if you can see your error.

Did you spot it? You assumed a 'conscious observer' inside a Cartesian Theatre.

Revise your scenario. There is no homunculus observing symbols. The 'symbols' function as recognisers and assemble themselves in response to patterns on the wall of the box. The recognisers don't 'know what a smell is' but some of them do what smell is - they recognise the presence of specific stimuli arising from aromatic molecules binding to receptor molecules in sensory neurons to recognise a particular aroma, a recipe of scents is thus analysed and might elicit the responses of a vocalisation 'I smell roses' and a memory <my grandma smelled of rose perfume>.

So, if we reconsider the brain assessing its own 'lego structures', we come to the same conclusion. That is, the brain assessing its own internal structures as a means to understanding what's going on outside, wouldn't have any clue (other than those structures were of sight; touch; taste; sound; smell) about what those structures actually meant.

You need to lock up your dualistic tendencies for the duration of this discussion James. Revise your thinking on what it means to 'mean something'.

This is a big rational problem for anyone harbouring theories similar to the one that Graham has presented - which is why it's been an issue for contemporary philosophers. So, you can't just whitewash it - you actually need to provide a rational solution to that problem. That is, how can the brain really know what any of its internal structures actually correlate to, externally to itself?

The brain doesn't need to know about its internal structures. Consciousness does not include explicit awareness of brain states as brain states and there is no inner observer sorting through unlabelled tags.

This has been a lengthy post, so I will end here and await responses. Further rational problems for such 'models' could be discussed, but I'll leave them for another day. There's enough to consider, here.

In summary, your 'objections' in this post are misapprehensions on your part, not flaws in the model at all.

[jamest]

SoS and SD have pointed out your error on this. You are thinking naive R1 and dualism when you are supposed to be looking at things as physical processes.

It should be obvious that by the model under discussion the brain isn't a dualist observer of brain states. Brain states are not perceived, they are perception. Brain states are not considered, they are cognition.

Graham, first of all you are contradicting yourself, as is obvious from something you said on p8 of this thread:

an individual brain does necessarily have a singular view, if it do any interpretation of 'data'.

You've said it yourself - and my post just amouts to there being a "necessary singular view of data to be interpreted".

Secondly, how can you accuse me of being a dualist when the model I was assessing was one of the brain itself processing numerous localities of brain states? That is, I was clearly considering a monistic model of reality. Your judgement about me discussing a dualist model is just bullshit whitewash. Do you even know what dualism is? It appears not, since dualistic philosophies regard a material brain AND an immaterial observer. Clearly, I wasn't discussing any thing of an immaterial nature in that post.

There are no 'tags' involved besides the NN themselves.

This must be incorrect, since clearly our language is replete with those tags. And since you say that the brain is responsible for our language, then it must be the brain that assigns tags to any particular NN. If, for instance, I say "I see yellowness", then that yellowness is a tag associated with a particular NN. Otherwise, I'd be saying something like "I see a NN with physical characteristics xyz".

Imagine a creature with a brain and eyes that sees an object it has never encountered before.

Actually, we should just imagine scenarios with human beings, so that we can relate to any responses that might come about from said scenarios.

Btw, your model only allows for brains seeing new NNs that they haven't encountered before, not 'objects'. Brains don't see anything external to themselves, Graham - they just process brain states therein.

It has no prior experience of that object.

Correction, it has no prior experience of that NN.

If contact is maintained, say by staring at the new object, a new NN recogniser will form.

Again, any "staring" would have to be focused upon the NN itself. Which, btw, enforces my point about the brain having to assess its own brain states, as a meaningful whole.

Activation of that NN means (I am) seeing object X.

Actually, it can mean nothing more than 'I' am seeing NN X.

The ability to recognise object X allows the creature to learn things about object x, such as object x is trying to eat (me), or object x tastes good. These new facts about object X are of the same NN form as recognition of object x.

This is significant:

If the brain is only aware of NNs, then correlating those NN's to external events/objects requires semantics about different 'things' (things different to the NN's themselves).
'Reality' for the brain, is itself - the states of itself. Clearly, for the brain to think that its own internal states refer to another reality (other than itself), requires the utilisation and understanding of an ontology that the brain simply has no direct encounter with. The problem now stands out like a sore thumb: that of the brain correlating events within itself, with events external to itself.

As I said, the brain can be aware of nothing, other than its own internal states. It's just an 'internal show' of relationships between NNs. This is what your theory reduces to, Graham. That is what known 'reality' amounts to, for the brain.

Do I really need to go on? Do you now understand the problem? Let me see if I can illustrate it better, with a revised edition of the lego scenario:

Here, you're aware of nothing other than events involving different structures of lego. This is the entirety of your reality. You realise, though, that there seems to be patterns that emerge from the lego events. You see for instance, that a red-square structure always immediately precedes the awareness of a yellow-rectangle structure. So, you conclude that red squares have a causal link to yellow rectangles.

... Now, what part of this process enables you to say something like: there's a lion outside that's trying to eat me? NOTHING, Graham - nothing at all facilitates that conclusion.
What would facilitate such a conclusion, then? Well, firstly, you'd have to assume the existence of a reality that is different to the one that is being encountered. And then, you'd have to assign external meaning to each and every internal NN.

There is no getting away from your predicament here: your model entails a brain that GIVES external meaning to its own internal states, upon the assumption(!) that there is a reality beyond itself.

My goodness, your model is in serious trouble.

[SpeedOfSound]

Secondly, how can you accuse me of being a dualist when the model I was assessing was one of the brain itself processing numerous localities of brain states? That is, I was clearly considering a monistic model of reality. Your judgement about me discussing a dualist model is just bullshit whitewash. Do you even know what dualism is? It appears not, since dualistic philosophies regard a material brain AND an immaterial observer. Clearly, I wasn't discussing any thing of an immaterial nature in that post.

Yes. You don't think you are a dualist but you have the deluded thinking of one. So you are.

[SpeedOfSound]

This must be incorrect, since clearly our language is replete with those tags. And since you say that the brain is responsible for our language, then it must be the brain that assigns tags to any particular NN. If, for instance, I say "I see yellowness", then that yellowness is a tag associated with a particular NN. Otherwise, I'd be saying something like "I see a NN with physical characteristics xyz".

No. It's more like a really complicated democracy. Everyone has a different strength of vote and they pass it on to thousands of acquaintances, and on, and on. At the end of the vote some big bureaucracy takes over and does something.

There have been a few of these tag neurons identified but it's not the norm.

[SpeedOfSound]

Btw, your model only allows for brains seeing new NNs that they haven't encountered before, not 'objects'. Brains don't see anything external to themselves, Graham - they just process brain states therein.
...
Correction, it has no prior experience of that NN.
...
Again, any "staring" would have to be focused upon the NN itself. Which, btw, enforces my point about the brain having to assess its own brain states, as a meaningful whole.
...
Actually, it can mean nothing more than 'I' am seeing NN X.

Bullshit. Graham is talking about an organism and they certainly do see objects.