Weightings in neural networks. You are thinking in terms of brains containing representations, but neural nets are not representational. — Banno

Why do you say that? Any pattern could symbolize something. And not all symbols necessarily appear like symbols to everyone. So I don't see any basis for the claim that "neural nets are not representational". To me, it looks far more likely that they actually are, as any ordered structure can be said to be representational. That's fundamentally what "ordered" is, a representation of the intent which orders, in other words, meaning.

This isn't addressing the issue. Your eyes and my eyes are stimulated by the same light, reflected by the same external world source. Yet I see red and you see green. If "red" and "green" refer to some hidden state in the external world cause then what does it mean for me to "see red" and you to "see green" in this situation? The "red" and "green" are referring to some quality of our experiences. — Michael

The sensation of colour is best described by reference to the activities of the "cones" and the "rods" of the eyes, rather than reference to some external source. But even this does not account for the huge role of brain activity. Since we dream in colour, it may actually be brain activity which causes the sensation of colour. The role of REM in relation to dream imaging is not well understood.

Some object is red1 if it causes most humans to see red2 — Michael

Herein is the problem. There's no reason at all to consider the existence of red2. We respond to red1. We reach for the word 'red', we imagine other things which are red...

What there's no evidence at all for is some entity called red2 which we 'see'. The process of 'seeing' does not, in any way, consult an internal colour chart. There's no entity matching your description of red2.

So the supposed 'external cause' of the sensation is not a cause at all, and cannot even be truthfully said to be a necessary condition. Therefore it is not at all irrational to be skeptical of the reality of a proposed 'external world'.

There's no reason at all to consider the existence of red2. — Isaac

When I see the dress as white and gold and you see the dress as black and blue, what do the words "white", "gold", "black", and "blue" refer to?

They don't refer to some hidden state. They refer to immediately apparent features of our experiences. The words "white" and "gold" refer to features present in my experience that aren't present in your experience, and the words "black" and "blue" refer to features present in your experience that aren't present in my experience.

When I see the dress as white and gold and you see the dress as black and blue, what do the words "white", "gold", "black", and "blue" refer to? — Michael

The properties of hidden states.

They don't refer to some hidden state. — Michael

Why not?

the words "black" and "blue" refer to features present in your experience that aren't present in my experience. — Michael

Then how did we learn to use the words? If they describe private experiences, how is it I ever learnt their use. How do we even know that what I call 'black' today is the same thing I called 'black' yesterday?

Why not? — Isaac

Because I can see the difference between red and blue. It's immediately apparent. Therefore red and blue aren't hidden states.

Then how did we learn to use the words? If they describe private experiences, how is it ever learnt their use. — Isaac

Because we're shown a bunch of things that share the same colour-appearance and told that this colour is to be called "red". We then come to associate the word "red" with this appearance.

A better question is this: if colour is a hidden state then how can we learn to use colour words?

How do we even know that what I call 'black' today is the same thing I called 'black' yesterday? — Isaac

Because we have a memory and can remember how things appeared in the past and how they appear now, and which words we used then and which words we use now to refer to that appearance? I don't really understand your question.

The properties of hidden states. — Isaac

So when I see a white and gold dress and you see a black and blue dress we're seeing different properties of hidden states?

The eye has a lens and photo-receptors. The ear has cilia that send an electrical facsimile of vibrations. Obviously a representation is being delivered to the brain. We just don't know how the brain creates a holistic experience out of that.

I can see the difference between red and blue. It's immediately apparent. Therefore red and blue aren't hidden states. — Michael

Hidden states cause effects in us. Them being hidden refers to their being outside of the Markov blanket, not to their having no effect. So you distinguishing red from blue is easy. Red has a different effect on you to blue.

we're shown a bunch of things that share the same colour-appearance — Michael

How is it that things all share the same colour experience if they've no intrinsic properties relating to colour? How is it that I can point to a post box, a rose, a bus, a traffic light...to teach my son how to use the word 'red' if all of those things have no intrinsic property called 'red'? Is it just luck that he has the same response to each such that he can see the similarity I'm getting at?

if colour is a hidden state then how can we learn to use colour words? — Michael

Because hidden states have fairly consistent and similar effects on us.

we have a memory and can remember how things appeared in the past and how they appear now — Michael

How could you possibly know that? Why would you even think that? I mean, I'm not going to rehash Wittgenstein's private language argument, you probably know it, but without some external reference you'd have no default reason to even think your 'black' of today was the same experience as your 'black' of yesterday.

So when I see a white and gold dress and you see a black and blue dress we're seeing different hidden states? — Michael

No. Same hidden states. I don't understand why you're having so much trouble with the idea of a hidden state having a different effect on different people or in different contexts. Did you not understand the example I gave earlier of the constellation Orion? The stars of Orion are arranged in the form of a man with a bow. The fact that they could also be seen as several other things doesn't mean they're not in the form of a man with a bow, it's just that they're also in other forms.

They're definitely not in the form of next week's winning lottery numbers. There's something intrinsic about them which constrains what pattern we can see. A man with a bow is one such option.

In the case of the dress, its colour is either white and gold or black and blue. It's not pink and green. There's something intrinsic about the hidden states which constrains our range of normal responses. That something is its colour.

We just don't know how the brain creates a holistic experience out of that. — Tate

Well. We know quite a lot about how the brain creates a holistic experience out of that. Not a complete picture. I'm not sure what your point is.

No. Same hidden states. I don't understand why you're having so much trouble with the idea of a hidden state having a different effect on different people or in different contexts. — Isaac

Hidden state X causes me to see red and you to see blue. What does "red" and "blue" refer to here? It doesn't refer to hidden state X, otherwise we would both be seeing red and both be seeing blue.

Well. We know quite a lot about how the brain creates a holistic experience out of that. Not a complete picture. I'm not sure what your point is. — Isaac

Actually we have theories for how experience is created by the brain.

The schematic of the nervous system gives us ample reason to believe that a spider's brain, for instance, is receiving a representation of its environment. For some reason you're wanting to ignore that obvious fact and say that hidden states and the representation are the same thing.

Hidden state X causes me to see red and you to see blue. What does "red" and "blue" refer to? — Michael

The colour of hidden state X.

It doesn't refer to hidden state X, otherwise we would both be seeing red or both be seeing blue (or both be seeing some other colour). — Michael

Why? Why must hidden state X have a colour such that it causes the same response in you as it does in me? In this case, its colour is 'red or blue'. We might, if we were to keep finding such hidden states, come up with a new word for such a colour, but since such states are exceedingly rare, we haven't the need. Most of the time the same state will cause the same (or similar) response in normally sighted people.

The schematic of the nervous system gives us ample reason to believe that a spider's brain is receiving a representation of its environment. — Tate

Does it? In what way?

The colour of hidden state X. — Isaac

If both "red" and "blue" refer to hidden state X then red and blue are the same colour, and the person who sees red and the person who sees blue are seeing the same thing.

This is evidently not the case. Your account of colour makes no sense.

The schematic of the nervous system gives us ample reason to believe that a spider's brain is receiving a representation of its environment.
— Tate

Does it? In what way? — Isaac

The lenses, the photo-receptors, the optic nerves delivering electrical signals.

If both "red" and "blue" refer to hidden state X then red and blue are the same colour — Michael

Not at all. If I think the person in the doorway is called Jim and you think he's called Jack, we're still both referring to the same person. That doesn't make Jim the same name as Jack. It makes one of us wrong, or it makes the person in the doorway possessed of two names, or it makes the name of the person in the doorway ambiguous, or unresolved.

Likewise with a hidden state which causes you to reach for the word 'blue' and me to reach for the word 'red' doesn't make red and blue the same thing, it makes one of us wrong, or the colour of the hidden state ambiguous, or unresolved.

The lenses, the photo-receptors, the optic nerves delivering electrical signals. — Tate

You're just naming parts of the optic system. Your claim was that they give us reason to believe that a spider's brain is receiving a representation of its environment.

If the same hidden state causes you to see the colour on the left and me to see the colour on the right then we are seeing different colours. You're seeing red and I'm seeing blue. The word "red" refers to the colour that you see and that I don't see (the colour on the left), and the word "blue" refers to the colour that I see and that you don't see (the colour on the right).

If you want the words "red" and "blue" to also refer to the hidden state that causes us to see what we do then we have two different meanings of "red" and "blue" as I mentioned above.

Red₁ is the colour on the left, blue₁ is the colour on the right, and red₁ and blue₁ are different colours.

Red₂ is the hidden state that causes you to see the colour on the left, blue₂ is the hidden state that causes me to see the colour on the right, and red₂ and blue₂ are the same colour.

This makes us susceptible to equivocation and a confusing metaphysics, as exemplified by your comments.

The lenses, the photo-receptors, the optic nerves delivering electrical signals.
— Tate

You're just naming parts of the optic system. Your claim was that they give us reason to believe that a spider's brain is receiving a representation of its environment. — Isaac

The electrical signals it receives from the optic nerve are a representation. What did you think that data was?

If the same hidden state causes you to see the colour on the left and me to see the colour on the right then we are seeing different colours. — Michael

An impossible situation from the outset. Hidden states cannot cause us to see colours. There's no mechanism by which that can happen. Seeing a colour is a process which starts with the property of a hidden state and ends with a series of responses (one of which might be to reach for a colour word).

You're imagining that 'seeing a colour' is some internal process, but you've given no reason why you'd imagine such a thing.

An impossible situation from the outset. Hidden states cannot cause us to see colours. There's no mechanism by which that can happen. Seeing a colour is a process which starts with the property of a hidden state and ends with a series of responses (one of which might be to reach for a colour word).

You're imagining that 'seeing a colour' is some internal process, but you've given no reason why you'd imagine such a thing. — Isaac

:up:

The electrical signals it receives from the optic nerve are a representation. — Tate

Uh huh. But we don't 'see' the electrical signals. We see the external world object. 'Seeing' involves those electrical signals, they're part of the process of seeing. They're not what we see.

Hidden states cannot cause us to see colours. — Isaac

What? Here are your words:

I'm saying that 'green' is a property of a hidden state which cause most humans in most situations to respond in the way we describe as 'seeing green' — Isaac

What? Here are your words: — Michael

Yeah. Look at the words. I was quite careful.

respond in the way we describe as 'seeing green' — Isaac

So? What's the difference between saying "x causes us to see green" and saying "x causes us to respond in the way described as 'seeing green'"?

It's the exact same thing, you're just wording it in a needlessly convoluted way.

It's like saying "that's the person named 'Jack'" instead of just saying "that's Jack."

The electrical signals it receives from the optic nerve are a representation.
— Tate

But we don't 'see' the electrical signals."
— Isaac

Not without an ampmeter, no. Nevertheless those signals are representative. Are you denying that? Because you would be grossly out of step with biology if you do.

What's the difference between saying "x causes us to see green" and saying "x causes us to respond in the way described as 'seeing green'" — Michael

One has green as a property of some mental representation, the other as a property of the hidden state.

The former is without warrant.

those signals are representative. Are you denying that? — Tate

No. I'm denying that they're what we 'see'. They're part of 'seeing', they're not what we see.