On Consciousness

[Fudbucker]

Well I've given him a few days, but I guess Piggy dropped our conversation thread. I shall have to rein in my disappointment.


The feedback loop thing goes way waaay back in this thread. To summarize: "consciousness" is a crappy term that means something slightly different to absolutely everybody, yet everbody will swear up and down that they know it when they see it. Pixy's SRIP was a counterargument: ridiculous as it is, you cannot prove it wrong because it's just as unfalsifiable as everyone else's definition.

(by the way, a purely subjective definition means it's pretty damned useless as a concept to study scientifically, and should honestly be thrown out entirely until someone can find something falsifiable to demonstrate about it)

Regarding your change, I'd object to it even if we weren't talking about the color thing. Calling all subjective experiences "subjective experience" implies they're a unified whole, which we've learned enough to say is completely false.


I think the idea behind shuttit's post was that, if the networks could be exchanged wholesale, if you could take the world's finest tweezers and swap the red cones for green (or something infinitely more complex downstream), you would see red for green, and vice versa. I agree, but I don't feel that detracts from the point I think most others are making; that "red" and "green" are entirely arbitrary distinctions that mean nothing more or less than "this population" versus "that population." If the two were swapped and you didn't already have a facebook of associations linking them to their respective (and now out-of-) contexts, you'd never know the difference.


Unconsciousness comes with a small period of retrograde amnesia. Aided by anesthesia you go down quick and stay there, so the last thing you remember is a bit before the gas really hits you, with no periods of light sleep/dozing awakeness to give you a sense of time.

Yes, but unconsciousness eventually arrives. if consciousness = SRIP, we would never be unconscious. SRIP seems to me a necessary condition, but doesn't tell the whole story. Maybe an SRIP of sufficient complexity...?

 

[Piggy]

Yes, but unconsciousness eventually arrives. if consciousness = SRIP, we would never be unconscious. SRIP seems to me a necessary condition, but doesn't tell the whole story. Maybe an SRIP of sufficient complexity...?

SRIP is all over the brain, and is not a useful criterion for discerning consciousness.

As has been explained upthread, the signatures of conscious awareness are certain activities in the brain stem, a trio of deep brain waves, and synchronized oscillations across disperse brain real estate.

When all that's going on, you're experiencing "qualia" or phenomenology or consciousness (the terms all mean the same thing).

When it's not, you're not.

When you fall asleep, the deep brain waves weaken then lose coherence. When you go under general anesthesia they lose coherence without weakening. When you wake up or come to from anesthesia, the waves cohere then gain strength.

 

[Beelzebuddy]

Last I heard, it was you who hadn't answered my questions, and I had observed that nothing in your cited article contradicted the physicalist position.

In fact, I was wondering when you were going to get around to posting anything that actually supported your position or responded in a meaningful way to the questions you've been dodging.

Odd interpretation of events on your part.

Ah, you must have missed it then. Here you go.

Admittedly it's not the answer you want, but I don't think that's altogether my problem.

Yes, but unconsciousness eventually arrives. if consciousness = SRIP, we would never be unconscious. SRIP seems to me a necessary condition, but doesn't tell the whole story. Maybe an SRIP of sufficient complexity...?

There can't be multiple layers of consciousness? You can't be more or less conscious of something, only conscious versus un-?

It's just as easy to throw up a screen of special pleading for SRIP as it is for other definitions of consciousness. They're all equally terrible. That is the point of SRIP, as I understand it.

 

[Fudbucker]

There can't be multiple layers of consciousness? You can't be more or less conscious of something, only conscious versus un-?

The only other layer I know of is the subconscious, which, by definition is an unconscious part of the mind.

Or do you mean like strata layers, where you rise up out of consciousness slowly? But that would just be a single consciousness, going from partial to full consciousness. SRIP entails not layers of one single consciousness, but multiple consciousnesses going on all the time as every instance of SRIP is consciousness.

It's just as easy to throw up a screen of special pleading for SRIP as it is for other definitions of consciousness.

How is it special pleading? If SRIP can't account for unconscious brain activity (both in general and regarding involuntary systems our brain regulates), it fails as a sufficient theory of consciousness. This would only be special pleading if the criticism only applied to SRIP. I consider any theory that can't account for unconscious brain activity to be a failure. And I'm not sold on insect consciousness. Is anyone seriously even studying it? Has there been a paper published in a peer reviewed journal about it?

They're all equally terrible. That is the point of SRIP, as I understand it.

OK, Pixy is advancing a terrible theory. He seems oddly committed to it, in spite of its terribleness.

 

[Beelzebuddy]

The only other layer I know of is the subconscious, which, by definition is an unconscious part of the mind.

Or do you mean like strata layers, where you rise up out of consciousness slowly? But that would just be a single consciousness, going from partial to full consciousness. SRIP entails not layers of one single consciousness, but multiple consciousnesses going on all the time as every instance of SRIP is consciousness.

No, really. The idea is that what we call "consciousness" is the aggregate of millions of smaller consciousnesses. That there's no true "subconscious" or "unconscious," only "less conscious," and with consciousnesses that don't have access to episodic memory.

For why that last bit's important, think back to your anesthesia experience. Although you don't remember time passing, time did pass and you were conscious for some of it, you just don't remember.

I consider any theory that can't account for unconscious brain activity to be a failure.

I consider any theory of consciousness that regards it as anything more than post-hoc rationalization for episodic memory to be a failure, so we're even there.

OK, Pixy is advancing a terrible theory. He seems oddly committed to it, in spite of its terribleness.

Just like EVERY theory of consciousness.

 

[PixyMisa]

A point against Pixy is that we're not actually conscious of all these SRIP's going on.

And I'm not conscious of what you are thinking. That's not a problem, just reality.

 

[PixyMisa]

How is it special pleading? If SRIP can't account for unconscious brain activity (both in general and regarding involuntary systems our brain regulates), it fails as a sufficient theory of consciousness.

Unconscious brain activity is either not self-referential, or it is self-referential and it's a separate process to which you only have limited access.

And I'm not sold on insect consciousness. Is anyone seriously even studying it? Has there been a paper published in a peer reviewed journal about it?

Cf. Sphex wasps.

OK, Pixy is advancing a terrible theory.

No, actually, I'm pointing out that this is the definition in common use, when you eliminate all the hand-waving. See Descartes, for example.

 

[Prometheus]

...To say that "red" or "redness" is somehow an inherent quality of light -- of any type of light at all -- is to make a fundamental mistake that will lead to errors in conclusions.

It's the other way around, actually: (Information about) Light (of a particular range of wavelengths) is an inherent quality of "red".

The most common problem which arises on threads like these is that folks will say, well, as long as a brain -- organic or machine -- distinguishes among various wavelengths of light, then we've got a brain which "sees color"

This is incorrect, however.....

I agree. When one is seeing red s/he is not merely distinguishing among wavelengths of light. The "red" quale also includes information about the subject's past experiences, and this metadata is being accessed when the brain is producing "red".

You can build a machine, for instance, that distinguishes among various types of objects striking it, for example, and responding differently to them, but you have not thereby created a machine which feels pain.

The pain of being struck by a bullet, hit by a hammer, and injected with a needle are all different. (Similarly, a throbbing toothache feels different from a piercing stab wound or a sunburn.) But designing a machine which responds differently to these does not mean you've designed a machine that feels pain, because pain is not "information" contained in the hammer or bullet or needle, so it is not "information" which can be received from them.

Right. The pain includes information about being struck or whatever, AND information about the victim himself (current state of wellbeing, similar past experiences, etc. This is quite messy for a couple of reasons: 1) the brain is an analog device so, among other differences to digital computers, this means that the information it processes is constantly being altered or degraded; 2) Since the brain is a product of evolution rather than design, it doesn't need to be perfect, just good enough to get by, so instead of a "color equals wavelength" one-to-one relationship which is highly accurate, the brain associates bits of data via one-to-many and many-to-one relationships. By strengthening and weakening neural pathways according to use, its even storing meta-meta-information about the usefulness of the meta-information. A lot less accurate, but lot's of data can be stored with a very high compression ratio.

This is important. Crucial, in fact.

By the same token, there is no information about color in photons / light waves. Therefore there is no "information" about color to be received from light.

Color is something produced by the brain, specifically by the processes in the brain which are dedicated to producing conscious experience.

No brain processes upstream from those particular processes have anything to do with color, and neither does light.
<snip>

The non-hi-lighted portion is consistent with my hypothesis, above. My idea also provides a plausible explanation for why it's useful for the brain to produce a conscious experience at all. That conscious experience is a form of information processing, just not one like we're used to seeing in digital computers. The massively parallel circuitry of the brain, combined with its extremely non-linear data structures enable it to access a vast amount of data extremely quickly, but at some expense of accuracy.

The highlighted portion is where you're wrong. Those upstream processes do have something to do with color--they're a subset of all the data that makes up a color.

This is crucial in that it means that, while a digital computer can, in theory, be made to mimic or simulate an analog one, in order to emulate the brain it would take an amount of processing power and memory orders of magnitude more than would be needed to merely render sensory input.

 

[steenkh]

I think the idea behind shuttit's post was that, if the networks could be exchanged wholesale, if you could take the world's finest tweezers and swap the red cones for green (or something infinitely more complex downstream), you would see red for green, and vice versa. I agree, but I don't feel that detracts from the point I think most others are making; that "red" and "green" are entirely arbitrary distinctions that mean nothing more or less than "this population" versus "that population." If the two were swapped and you didn't already have a facebook of associations linking them to their respective (and now out-of-) contexts, you'd never know the difference.

I think I agree. Exactly what the colours look like for our inner eye is of no consequence. What matters is the connections to physical colours, our experience with colours, such as the ability to distinguish ripe from unripe fruits, or stop when we see a red traffic light. The sky is blue because that is what we know the sky should be, and it has always been blue, even if your would think that I saw the sky as green if you were able to see inside my head.

 

[kirstenholme]

I don't have the knowledge to fully participate in this conversation but I'd like to throw a fresh bunch of circumstances into your discussion if I may--epilepsy sufferers experience several different types of consciousness. When they have tonic clonic seizures, there is no way for them to retain information about what happened because they were entirely unconscious and their brains did not take in any data.

Another type of seizure allows them to repeat automatic behaviours without the decision-making process--they would, during an absence seizure, continue to walk but would not choose the direction they were headed in. These events are sometimes remembered and sometimes not. There are sets of automatic behaviours that are unique to the focus of the epilepsy--a half-decent neurologist can narrow in on where the seizure focus is by asking a set of exhaustive questions about habitual actions answered by the patient's family members and friends.

On another front, temporal lobe epilepsy has a powerful impact on mood, leaving seizure sufferers with extreme swings that may last ten minutes or a few hours, only to end when the epileptic activity stops. Furthermore, temporal lobe epilepsy patients switch senses sometimes--a taste is experienced as a smell; a sight becomes a sound, colours can be felt on the skin and so forth.

These changes can happen when only portions of the brain are experiencing increased electrical activity. A quarter of the brain may have isolated activity while the rest of the brain experiences nothing unusual.

Then there are hallucinations, which are common in some epilepsy patients. They happen during some seizures yet are often remembered in tiny detail and in full colour. Those hallucinations often follow the belief sets of the person in question. Take Karen Armstrong as an example: as a nun, her hallucinations included angels and other distinctly Christian visions, which all disappeared when she was eventually diagnosed and treated for TLE (after which she laid down her habit and began looking for a new belief system.)

I think there is a lot to gain from looking at consciousness as it is experienced by those with temporary or chronic neurological conditions stemming from the brain. I'm waiting to get my hands on some Michael Shermer books. Apparently this is the sort of thing he writes about.

 

[Belz...]

Unconscious brain activity is either not self-referential, or it is self-referential and it's a separate process to which you only have limited access.

Wait. In this context, what's "you" ? Isn't that a bit circular ?

 

[PixyMisa]

Wait. In this context, what's "you" ?

One of the conscious processes running in your brain.

Isn't that a bit circular ?

Why?

 

[Belz...]

One of the conscious processes running in your brain.


Why?

You define consciousness as SRIP. Someone asked how SRIP can happen in the brain without your conscious knowledge, challenging that SRIP is sufficient for consciousness. You answer that unconscious processes could be SRIP that isn't accessible to you.

First, that's a tacit admission that SRIP is not sufficient. Second, unconscious processes are not accessible to "you" by definition, so that's not very helpful. That doesn't address the question, in my opinion.

It may not be circular, but it sounds unfalsifiable, so I'd like you to be clearer.

 

[PixyMisa]

You define consciousness as SRIP. Someone asked how SRIP can happen in the brain without your conscious knowledge, challenging that SRIP is sufficient for consciousness. You answer that unconscious processes could be SRIP that isn't accessible to you.

First, that's a tacit admission that SRIP is not sufficient.

No, not at all. The point is that a so-called "unconscious" process could, in reality, be separate but fully conscious in its own right.

Second, unconscious processes are not accessible to "you" by definition, so that's not very helpful. That doesn't address the question, in my opinion.

It may not be circular, but it sounds unfalsifiable, so I'd like you to be clearer.

There are multiple conscious processes running in your brain. You are one of them, or, to put it another way, one of them is you. You don't have special access to the others, any more than you have special access to my consciousness. The others feed data to the one that is you, but then, so do I.

As has been noted, split-brain patients present us with striking examples of this sort of thing. Cut that connection and in many respects we end up with two distinct minds - sometimes strikingly distinct, and yet unaware of one another's very existence. And these consciousnesses can actually direct physical behaviours.

 

[Mr. Scott]

How about you simply respond to the question I put to you?

"Which quale do you think this person would be experiencing?"

How about that?

If you answer that question, then we have something we can talk about.

ETA: If you want to know why I sometimes lose patience, it's the question-dodging that's a large contributor.

The question contains assumptions I don't find valid, so I've been addressing those assumptions. I'll try to respond in a way that doesn't come across as a dodge.

BTW, you seem to keep suggesting the idea of "tiger" is subconscious.

I surmise that, if the brain can produce 6 million color qualia, a seemingly infinite number of sound qualia, likewise for smell, taste, touch, emotion, that the brain is most likely able to synthesize new qualia as a result of input from its environment, including combining more than one sense, e.g. 6 million colors that are combinations of just three primary colors.

So, to answer your question, this person would be experiencing a tiger quale which the brain had synthesized from instinct and environmental experience.

I hope that suffices.

You know, the guy who thought his mother was an imposer is a better example. Let's play with that one, shall we? Would it not be valid to say this person who sees and hears a woman who, despite actually being his mother, experiences the stranger quale rather than the mother quale? Remember, the formal definition of qualia says nothing about the five senses. Only the subjective experience.

 

[Mr. Scott]

Piggy, a question for you:

Where do you find the confidence in, or evidence for, your assertion that a computer, ordinary in design but extraordinary in size and speed, could not ever be conscious?

 

[Guybrush Threepwood]

No, not at all. The point is that a so-called "unconscious" process could, in reality, be separate but fully conscious in its own right.


There are multiple conscious processes running in your brain. You are one of them, or, to put it another way, one of them is you. You don't have special access to the others, any more than you have special access to my consciousness. The others feed data to the one that is you, but then, so do I.

Does your next paragraph not suggest that it's more like multiple conscious processes acting together are you, and if you stop them acting together you can end up with two yous in the same brain. Or alternatively through brain damage remove a number of processes but still end up with a conscious you, albeit one who is missing some normal parts of consciousness?

As has been noted, split-brain patients present us with striking examples of this sort of thing. Cut that connection and in many respects we end up with two distinct minds - sometimes strikingly distinct, and yet unaware of one another's very existence. And these consciousnesses can actually direct physical behaviours.

 

[PixyMisa]

Does your next paragraph not suggest that it's more like multiple conscious processes acting together are you, and if you stop them acting together you can end up with two yous in the same brain. Or alternatively through brain damage remove a number of processes but still end up with a conscious you, albeit one who is missing some normal parts of consciousness?

Yeah, even though the model is conceptually simple, the reality of it is pretty complicated. A lot of the things that we assume are an essential part of human consciousness - the emotional response to a photo of mom, for example - are actually handled in a modular manner and integrated into the whole, and can fail independently. Oliver Sacks' books are full of that sort of thing; Sacks himself suffers from prosopagnosia (he can't recognise faces).

Where and how (and why) we draw boundaries when we want to discuss what makes you you is not a simple matter.

 

[Perpetual Student]

The only reason we speak of "red light" or any color light is because of what that light (sometimes) does inside our skulls.

Yes

If the topic at hand is that very behavior inside our skulls, then it is important for us to understand that "red" is (sometimes) our brains' response to light, and that this response is in no way implicit in the physical properties of the light itself.

If the wavelength of the light is within the 620–750 nm range we see red; if it is a little shorter we see orange. Our response depends on the properties of the light.

ETA: If you'd like to explain the physics of how red can be a property of light, or of how it can be determined which brains' particular response is the inherent quality and why all others aren't, or how red could be predicted from the neural activity or the wavelength by an observer with a brain that doesn't respond the way ours do, then be my guest.

I already explained the physics above. Asking for light in the 620–750 nm range to be predicted by neural activity is like asking for light in the 620–750 nm range to be predicted by the letters r,e and d. The brains activity produces responses that are representative for that range of light, just as it produces responses that represent a particular face. These are internal representations of some real external thing. Because the internal representations can be reduced to certain neural activities, it does not make the external reality any less real.

 

[Piggy]

Yes

If the wavelength of the light is within the 620–750 nm range we see red; if it is a little shorter we see orange. Our response depends on the properties of the light.

I already explained the physics above. Asking for light in the 620–750 nm range to be predicted by neural activity is like asking for light in the 620–750 nm range to be predicted by the letters r,e and d. The brains activity produces responses that are representative for that range of light, just as it produces responses that represent a particular face. These are internal representations of some real external thing. Because the internal representations can be reduced to certain neural activities, it does not make the external reality any less real.

I don't disagree with any of that.

My only point relative to this, and it's a crucial point in understanding consciousness, is that red is not a representation of red.

Red is (sometimes) a response to -- or representation of -- a particular kind of light. But since red cannot possibly be an inherent property of any kind of photon or wave out there in the world, red is unique to brain behavior, and specifically to the subset of brain behavior which we call consciousness or phenomenology, or sometimes "qualia".

With regard to potential machine consciousness, what this means is that merely building a machine which responds differentially to various wavelengths of light, or "processes information about" various wavelengths of light, is not sufficient to allow us to say we've built a machine which is conscious of light or which "sees color".

To say that is it, is what I call informational opportunism, and it's an error in reasoning.

We can demonstrate this clearly in the lab, for example, using binary visual confusion experiments, in which two different sets of visual cues are sent into the brain simultaneously.

The brain real-estate responsible for non-conscious behavior can handle both neural responses at the same time, and we can observe this neural behavior. But the electro-mechanical brain processes which are tasked with producing conscious experience cannot. They can only produce / perform a conscious response to (i.e. awareness of) one at a time.

The result is that our "Observer A" has the experience of an image which "flips" back and forth between the two, as if a pair of images were being displayed intermittently.

This fact, and the blindsight experiments, effectively puts the kibosh on informational opportunism.

Consciousness is not just any old sort of "information processing" -- it is a specific brain function which, like all bodily functions, demands its own dedicated physical mechanism to accomplish. Of course, in the brain, there's a lot of crossover, with many bits of real-estate participating in multiple functions, e.g. perception and imagination.

But that's really the only point I was trying to make regarding this issue in particular, and given what you've posted, I don't see any reason why you should disagree with that, at least.