Bugmaster comments on The raw-experience dogma: Dissolving the “qualia” problem - Less Wrong

2 Post author: metaphysicist 16 September 2012 07:15PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (340)

You are viewing a single comment's thread. Show more comments above.

Comment author: Bugmaster 18 September 2012 04:27:32PM 0 points [-]

An althouh we don't have a theory of *how qualia depen on brain states -- which is, in fact, the Hard Problem

In my response, I was trying to say that "qualia" are brain states. I put the word "qualia" in quotes because, as far as I understand, this word implies something like, "a property or entity that all beings who see this particular shade of red share", but I explicitly deny that such a thing exists.

Everyone's brains are different, and not everyone experiences the same red, or does so in the same way. The fact that our experiences of "red" are similar enough to the point where we can discuss them is an artifact of our shared biology, as well as the fact that we were all brought up in the same environment.

Anyway, if "qualia" are brain states, then the question "how do qualia depend on brain states" is trivially answered.

Comment author: Peterdjones 18 September 2012 07:56:01PM *  1 point [-]

In my response, I was trying to say that "qualia" are brain states

My use of "depend" was not meant to exlude identity. I had in mind the supervenience principle, which is trivially fulfilled by identity.

"a property or entity that all beings who see this particular shade of red share"

I am not sure where you got that from. C I Lewis defined qualia as a "sort of universal", but I don't think there was any implication that everyone sees 600nm radiation identicallty. OTOH, ones personal qualia must recur to a good degree of accuracy or one would be able to make no sense of ones sensory input.

Anyway, if "qualia" are brain states, then the question "how do qualia depend on brain states" is trivially answered.

Interestingly, that is completely false. Knowing that a bat-on-LSD's qualia are identical to its brain states tells me nohting about what they are (which is to say what they seem like to the bat in question..which is to say what they are, since qualia are by definition seemings.[If you think there are two or three meanings of "are" going on there, you might be right]).

Comment author: Bugmaster 19 September 2012 01:04:33AM 0 points [-]

OTOH, ones personal qualia must recur to a good degree of accuracy or one would be able to make no sense of ones sensory input.

Agreed. I was just making sure that we aren't talking about some sort of Platonic-realm qualia, or mysterious quantum-entanglement qualia, etc. That's why I personally dislike the word "qualia"; it's too overloaded.

Knowing that a bat-on-LSD's qualia are identical to its brain states tells me nohting about what they are (which is to say what they seem like to the bat in question..

If I am correct, then you personally could never know exactly what another being experiences when it looks at the same red object that you're looking at. You may only emulate this knowledge approximately, by looking at how its brain states correlate with yours. Since another human's brain states are pretty similar to yours, your emulation will be fairly accurate. A bat's brain is quite different from yours, and thus your emulation will not be nearly as accurate.

However, this is not the same thing as saying, "bats don't experience the color red (*)". They just experience it differently from humans. I don't see this as a problem that needs solving, though I could be missing something.

(*) Assuming that bats have color receptors in their eyes; I forgot whether they do or not.

Comment author: Peterdjones 20 September 2012 02:20:30PM 1 point [-]

Agreed. I was just making sure that we aren't talking about some sort of Platonic-realm qualia,

I don't think anyone has raised that except you.

If I am correct, then you personally could never know exactly what another being experiences when it looks at the same red object that you're looking at.

Alhough, under may circumstances, I could know approximately.

However, this is not the same thing as saying, "bats don't experience the color red".

Bats have a sense that humans don't have, sonar, and if they have qualia, they presumably have some kind of radically unfamiliar-to-humans qualia to go with it. That is an issue of a different order to not knowing exactly what someone else's Red is like. And, again, it is not a problem solved by positing the identity of the the bat's brain state and its qualia. Identity theory does't explain qualia in the sense of explaining how variations in qualia relate to varations in brain state.

Comment author: Bugmaster 20 September 2012 06:42:00PM *  0 points [-]

Alhough, under may circumstances, I could know approximately.

Agreed.

Bats have a sense that humans don't have, sonar, and if they have qualia, they presumably have some kind of radically unfamiliar-to-humans qualia to go with it.

I wasn't talking about sonar, but about good old-fashioned color perception. A bat's brain is very different from a human's. Thus, while you can approximate another human's perception fairly well, your approximation of a bat's perception would be quite inexact.

Identity theory does't explain qualia in the sense of explaining how variations in qualia relate to varations in brain state.

I'm not sure I understand what you mean. If we could scan a bat's brain, and understand more or less how it worked (which, today, we can't do), then we could trace the changes in its states that would propagate throughout the bat when red photons hit its eyes. We could say, "aha, at this point, the bat will likely experience something vaguely similar to what we do, when red photons hit our eyes". And we could predict the changes in the bat's model of the world that will occur as the result. For example, if the bat is conditioned to fear the color red for some reason, we could say, "the bat will identify this area of its environment as dangerous, and will seek to avoid it", etc.

If the above is true, then what is there left to explain ?

Comment author: Peterdjones 20 September 2012 07:17:06PM *  1 point [-]

If the above is true, then what is there left to explain ?

Radically unfamiliar-to-humans qualia. You have picked an easy case, I have picked a difficult one. If we wan't to know what the world sonars like to a bat on LSD, identity theory doens't tell us.

Comment author: Bugmaster 20 September 2012 07:32:11PM 0 points [-]

You have picked an easy case, I have picked a difficult one. If we wan't to know what the world sonars like to a bat on LSD, identity theory doens't tell us.

Well, in point of fact, I've personally never done LSD, so I don't know what color perception is like for another human on LSD, either. I could make an educated guess, though.

In case of the bat sonar, the answer is even simpler, IMO: we lack the capacity to experience what the world sonars like to a bat, except in the vaguest terms. Again, I don't see this is a problem. Bats have sonars, we don't.

Note that this is very different from saying something like "we can't know whether bats experience anything at all through their sonar", or "even if we have scanned the bat's brain, we can't predict what changes it would undergo in response to a particular sonar signal", etc. All I'm saying is, "we cannot create a sufficiently accurate mapping between our brain states and the bat's, as far as sonaring is concerned".

Again, I'm not entirely sure I understand what additional things we need to explain w.r.t qualia.

Comment author: Peterdjones 20 September 2012 08:17:33PM 1 point [-]

In case of the bat sonar, the answer is even simpler, IMO: we lack the capacity to experience what the world sonars like to a bat, except in the vaguest terms. Again, I don't see this is a problem

I see that as a problem for the claim that mind-brain identity theory explains qualia. It does not enable us to undestand the bat's qualia, or to predict what they would be like. However, other explanations do lead to understanding and predicting.

Again, I'm not entirely sure I understand what additional things we need to explain w.r.t qualia.

Understanding and predicting.

Comment author: Bugmaster 20 September 2012 08:36:56PM 0 points [-]

I guess I'm not entirely sure what you mean by "understanding" and "predicting". As I said, if we could scan the bat's brain and figure out how all of its subsystems influence each other, we would know with a very high degree of certainty what happens to it when the bat receives a sonar signal. We could identify the changes in the bat's model of the world that would result from the sonar signal, and we could predict them ahead of time.

Thus, for example, we could say, "if the bat is in mid-flight, and hungry, and detects its sonar reflecting from a small object A of size B and shape C etc., then it would alter its model of the world to include a probable moth at the object's approximate location (*). It would then alter course to intercept the moth, by sending out signals to its wing muscles as follows: blah blah".

Are predictions of this sort insufficient ? If so, what additional predictions could be made by those other explanations you mentioned ?

(*) Disclaimer: I don't really know much about the hunting habits of real-life bats.

Comment author: Peterdjones 21 September 2012 12:02:03AM 1 point [-]

Are predictions of this sort insufficient ?

More irrelevant. None of them are actualy about qualia, about how things seem to experiencing subjects. You have Substituted an Easier Problem.

Comment author: gwern 21 September 2012 02:40:05AM 0 points [-]

Well, in point of fact, I've personally never done LSD, so I don't know what color perception is like for another human on LSD, either. I could make an educated guess, though.

Normally I'd assume that I know what you meant and move on, but since this involves LSD... You don't know what it's like? Or you do, but it's an educated guess? What?

Comment author: Bugmaster 21 September 2012 08:45:44AM 0 points [-]

I've never done LSD myself, but I've talked to people who did, and I've read similar accounts in books, online, etc. Thus, I can make a guess as to what LSD would feel like, assuming my brain is close to the average.

Comment author: bogus 18 September 2012 04:55:45PM -1 points [-]

Anyway, if "qualia" are brain states, then the question "how do qualia depend on brain states" is trivially answered.

It still makes sense to ask what these "brain states" actually are, physically. Since we seem to have direct experiential access to them as part of our subjective phenomenology, this suggests on Occamian grounds that they should not be as physically or ontologically complex as neurophysical brain states. The alternative would be for biological brains to be mysteriously endowed with ontologically basic properties (as if they had tiny XML tags attached to them) which makes no sense at all.

Comment author: TheOtherDave 18 September 2012 05:18:58PM 1 point [-]

It still makes sense to ask what these "brain states" actually are, physically

I would agree that it makes sense to ask what sorts of brain states are associated with what sorts of subjective experiences, and how changes in brain states cause and are caused by those experiences, and what sorts of physical structures are capable of entering into those states and what the mechanism is whereby they do so. Indeed, a lot of genuinely exciting work is being done in these areas by neurologists, neurobiologists, and similar specialists as we speak.

Comment author: bogus 18 September 2012 05:31:48PM 0 points [-]

Indeed, a lot of genuinely exciting work is being done in these areas by neurologists, neurobiologists, and similar specialists as we speak.

I agree, and I would add that a lot of interesting work has also been done by transcendental phenomenologists - the folks who study the subjective experience phenomenon from its, well, "subjective" side. The open question is whether these two strands of work will be able to meet in the middle and come up with a mutually consistent account.

Comment author: shminux 18 September 2012 05:48:33PM 0 points [-]

"transcendental phenomenology" is not a natural science but philosophy, so there is no middle to meet in.

Comment author: Peterdjones 18 September 2012 08:03:22PM 1 point [-]

Except that there is, since there are plenty of subjects which have been studied from both sides. The natures of space, time and causality for a start.

Comment author: shminux 18 September 2012 09:12:22PM 0 points [-]

The natures of space, time and causality for a start.

Having studied these subjects from the physics side, I find that there is little useful input into the matter from the philosophy types, except for some vague motivations.

Comment author: Peterdjones 20 September 2012 03:18:08PM 1 point [-]

You may not like the Middle, but it is there.

Comment author: shminux 20 September 2012 04:27:11PM -1 points [-]

Feel free to give an example.

Comment author: Peterdjones 20 September 2012 05:05:40PM 0 points [-]

The natures of space, time and causality for a start.

Comment author: TheOtherDave 18 September 2012 05:38:02PM 0 points [-]

You say "has been done"... is that to suggest that there is no active work currently being done in transcendental phenomenology?

Comment author: Mitchell_Porter 18 September 2012 10:37:24PM 2 points [-]

If I can jump in... It's useful to distinguish between phenomenology in general, as the study of consciousness from "within" consciousness; various schools of phenomenological thought, distinguished by their methods and conclusions; and then all those attempts to explain the relationship between consciousness and the material world. These days the word "phenomenology" is used quite frequently in the latter context, and often just to designate what it is that one is trying to "correlate" with the neurons.

It's part of the general pattern of usage whereby an "-ology" comes to designate its subject matter, so that "biology" means life and not the study of life - "we share the same biology" doesn't mean our biology classes are in agreement - "psychology" means mind and not the study of mind, and "sociology" means social processes and not the study of them. That's an odd little trend and I don't know what to make of it, but in any case, "phenomenology" is often used as a synonym for the phenomena of consciousness, rather than to refer to the study of those phenomena or to a genuine theory of subjectivity.

Thus people talk about "naturalizing phenomenology", but they don't mean taking a specific theory of subjective consciousness and embedding it within natural science, they just mean embedding consciousness within natural science. Consciousness is treated in a very imprecise way, compared to e.g. neuroscience. Such precision as exists is usually in the domain of philosophical definition of concepts. But you don't see people talking about methods for precise introspection or for precise description of a state of consciousness, or methods for precise arbitration of epistemological disputes about consciousness.

Phenomenology as a discipline includes such methodological issues. But this is a discipline which exists more as an unknown ideal and as an object of historical study. Today we have some analytic precision in the definition of phenomenological concepts, and total imprecision in all other aspects, and even a lack of awareness that precision might be possible or desirable in those other aspects.

Historically, phenomenology is identified with a particular movement within philosophy, one which attached especial significance to consciousness as a starting point of knowledge and as an object of study. It could be argued that this is another sign of intellectual underdevelopment, in the discipline of philosophy as a whole - that phenomenology is regarded as a school of thought, rather than as a specific branch of philosophy like epistemology or ethics. It's as if people spoke about "the biological school of scientific thought", to refer to an obscure movement of scientists who stood out because they thought "life" should be studied scientifically.

So to recap, there is a movement to "naturalize phenomenology" but really it means the movement to "naturalize consciousness", i.e. place consciousness within natural science. And anyone trying to do that implicitly has a personal theory of consciousness - they must have some concept of what it is. But not many of those people are self-consciously adherents to any of the theories of consciousness which historically are known as phenomenological. And of those who are, I think there would be considerably more enthusiasm for "existential phenomenology" than for "transcendental phenomenology".

This distinction goes back to the divide between Husserl and his student Heidegger. Husserl was a rationalist in an older, subjective sense and by temperament - he was interested in analytical thought and in the analytical study of analytical thought; the phenomenology of propositional thinking, for example. Heidegger was his best student, but he became obsessed with the phenomenology of "Being", which became a gateway for the study of angst, dread, the meaning of life, and a lot of other things that were a lot more popular and exciting than the intentional structure of the perception of an apple. The later Heidegger even thought that the best phenomenology is found in the poetic use of language, which makes some sense - such language evokes, it gets people to employ complex integrated systems of concepts which aren't so easy to specify in detail.

Meanwhile, Husserl's more rationalistc tendencies led towards transcendental phenomenology, which even among philosophers was widely regarded as misguided, the pursuit of a phantasmal "transcendental ego" that was (according to the criticism) an artefact produced by language or by religious metaphysics. Husserl literally fled Nazi Germany in order to continue his work (while Heidegger tried to accommodate himself to the sturm und drang of the regime) and died with only a few loyalists developing the last phase of his ideas. After the war, Heidegger was excoriated for his politics, but existential phenomenology remained culturally victorious.

If we come closer to the present and the age of cognitive science, there are now many people who are appreciative of Husserl's earlier analyses, but transcendental phenomenology is still mostly regarded as misguided and metaphysical. Existential phenomenology is also a somewhat exotic affiliation among scientists, but it does get some recognition among people who are into the importance of "embodiment" in cognitive science and consciousness studies. Husserl's phenomenology is so verbal and verbalizing, whereas existential phenomenology, in its attention to "raw existence", can lead (among other destinations) to a 1960s-style rediscovery of the senses, e.g. in Merleau-Ponty's phenomenology, and from there to the embodied cognition of Rodney Brooks et al.

So in the contemporary world, transcendental phenomenology is very obscure and mostly it's a subject of historical research. You could make the analogy between Husserl and Einstein, with transcendental phenomenology as Husserl's unified field theory. Einstein was regarded as a founder of modern physics but his later interests were regarded as misguided, and it's much the same with Husserl. But fifty years after Einstein's death, unified theories are a standard interest, it's just that they're quantum rather than classical. Similarly, it's likely that the spirit of transcendental phenomenology will be revived eventually.

Comment author: Bugmaster 18 September 2012 05:00:18PM 0 points [-]

Since we seem to have direct experiential access to them as part of our subjective phenomenology, this suggests on Occamian grounds that they should not be as physically or ontologically complex as neurophysical brain states.

How so ? I don't follow your reasoning, and I'm not sure what you mean by "neurophysical brain states" -- are there any other kinds ? Ultimately, every human brain is made of neurons...

Comment author: Peterdjones 18 September 2012 08:04:34PM 1 point [-]

I didn't understand that either.

Comment author: TheOtherDave 18 September 2012 05:13:41PM 0 points [-]

Ultimately, every human brain is made of neurons...

Not exclusively. There are glial cells, for example.

Comment author: Bugmaster 18 September 2012 05:43:08PM 1 point [-]

Good point. I should've said, "made of neurons or other physical substances" :-)