Dennett's "Consciousness Explained": Chpt 2

post by PhilGoetz · 2010-01-10T23:38:55.449Z · LW · GW · Legacy · 25 comments

Contents

  Chapter 2: Explaining Consciousness
    1. Should consciousness be demystified?
    2. The mystery of consciousness
    3. The attractions of mind stuff
    4. Why dualism is forlorn
    5. The challenge
    Summary
None
25 comments

Chapter 2: Explaining Consciousness

This chapter is all about dualism and why it's bad.  I find this chapter incoherent, because Dennett never defines dualism carefully, and confuses it with the theistic views historically associated with dualism.

[Long stretches of my own ideas will be in brackets, like this.]

1. Should consciousness be demystified?

D begins with a defense against those who don't want an explanation of consciousness.  I'm not even going to read this section.

2. The mystery of consciousness

"What could be more obvious or certain to each of us than that he or she is a conscious subject of experience, an enjoyer of perceptions and sensations, a sufferer of pain, an entertainer of ideas, and a conscious deliberator? ... How can living physical bodies in the physical world produce such phenomena?"

3. The attractions of mind stuff

D says Searle describes functionalism by saying that if a computer program reproduced the entire functional structure of a human wine taster's cognitive system, functionalism says that it would reproduce all the mental properties, including enjoyment.  This is followed by a [long, pointless, popular] discussion of whether volcanos, hurricanes, etc., have souls.

The "mind stuff" position D sets up in order to knock down is not John Searle's "mind stuff" position, nor Roger Penrose's, but something spiritual and non-physical.  He says that the mind-stuff view is that "the conscious mind... cannot just be the brain... because nothing in the brain could appreciate wine."  [D should pursue this further, because the summary given of the "mind stuff" view is too incoherent to attack.]

4. Why dualism is forlorn

D says that the "mind stuff" view is that the brain is "composed not of ordinary matter", and is dualism.  Materialists, OTOH, say we can account for all mental phenomena using the same principles that explain other things.  Dualism is bad; materialism is good.

D says that if dualism were true, then the mind would need to act on our body, so we could move our arms.  But to move, a physical thing needs energy; and how can non-matter produce energy?  Dualism is thus dead, QED.

[IMHO this muddies the waters terribly, in a way I've seen them muddied many times before.  The dichotomy between dualism and materialism is a false dichotomy.  There are no "materialists" who believe only in a kinematic world.  (Ironically, Descartes, the classic dualist, was just such a materialist when it came to the physical world; he refused to believe in gravity for that reason.)  The physics that we believe in is full of other forces such as gravity, electricity, and magnetism.  Energy can exist in all these forms, as well as in matter.  D is one of the many who reject "dualism", yet accept the everyday mysterious forces; and presumably approve of people who accepted them long before there was any material explanation of them.

If we want to be able to dismiss people for positing new "stuff", the way we currently do by calling them dualists, then we need a way to distinguish an acceptable new fundamental force or type of being, such as gravity, from an unacceptable one.  Obeying the law of conservation of energy is one such rule; if Descartes wants to have a non-material soul, it must exchange energy with ordinary matter in predictable ways.  To put it another way, rejecting "dualism" can make sense if we define dualism as the belief in things that don't obey conservation laws.

Once we've done that, though, we find that all the old enemies we hoped to dismiss as dualists can sneak back in by claiming to observe the conservation laws.  Even magic systems in fantasy worlds often obey energy conservation laws.

Even this conservative position is problematic.  EY believes in many worlds.  Many worlds seems, at least to many people like me who consider it a respectable position without understanding it, to require a stupendous, continual violation of conservation laws.  But we don't usually therefore call EY a "bad dualist" and dismiss many-worlds.

My intuition is that we are willing to consider even very crazy-sounding new proposed extensions to physics, if we believe they are made with a sincere desire to understand the universe.  Historically, a "dualist" is usually someone, such as Descartes, who is trying to come up with an excuse for not trying to understand the universe.  The term "dualist" is an accusation against a person's intent masquerading as an objection to their physics.]

D also refutes dualism by saying that mind stuff can't both elude physical measurement, and control the body; as anything that escapes our instruments of detection can't possibly interact with the body to control it.  [The problem with this argument is that, 200 years ago, if someone had told you that the brain used electrical impulses, you could have used the exact same argument to "prove" that that was impossible.]

D considers this angle on the next page, without realizing that it devastates his last several pages of argument:  "Perhaps some basic enlargement of the ontology of the physical sciences is called for in order to account for the phenomena of consciousness."

He also gives us what I take to be the defining characteristic of "bad" dualism: "The few dualists to avow their views openly have all candidly and comfortably announced that they have no theory whatever of how the mind works -- something, they insist, that is quite beyond human ken."  [This shows us that what D dismisses as "dualism" has nothing to do with making a dualistic distinction between matter and non-matter; and everything to do with the intentions of the theorist.  What D objects to as "dualism" is actually the God argument that has been historically associated with dualism:  Stopping further inquiry by positing the existence of something that a) explains, and b) cannot be explained.]

5. The challenge

D lays down rules for himself to follow: To explain consciousness, with only existing science, while acknowledging his own conscious experience.

Summary

[This chapter will confuse more than inform.  Its only purpose is to refute dualism; but D hasn't taken a close look at what he means when he uses that word, so he mingles together its meaning and its associations and historical contingiencies indiscriminately.]

25 comments

Comments sorted by top scores.

comment by Jack · 2010-01-11T02:02:34.387Z · LW(p) · GW(p)

What we need is a way to distinguish an acceptable new fundamental force or type of being, from an unacceptable one.

Phil meet science; science, this is Phil.

Replies from: PhilGoetz
comment by PhilGoetz · 2010-01-11T04:22:01.974Z · LW(p) · GW(p)

Your point being?

comment by Paul Crowley (ciphergoth) · 2010-01-11T17:30:05.864Z · LW(p) · GW(p)

I'm thinking about a top-level post on heterophenomenology. I'd like to hear from people who don't believe in p-zombies, but don't think that heterophenomenology is enough to set out the problem of consciousness, on why you don't see a contradiction between those two positions. Thanks!

Replies from: Wei_Dai
comment by Wei Dai (Wei_Dai) · 2010-01-12T03:53:36.845Z · LW(p) · GW(p)

Ok, I don't believe in p-zombies, in the standard sense of there being a logically possible world which is physically identical to this one, but where the inhabitants are not conscious. But I do believe that someone (perhaps a super-intelligent being) could possibly emulate my outward behavior perfectly, while having very different conscious experiences on the inside (i.e., by lying). I don't think you can distinguish between these two cases without reference to what I'm really experiencing, as opposed to just what I say about what I experience.

Replies from: komponisto, ciphergoth
comment by komponisto · 2010-01-12T04:18:26.673Z · LW(p) · GW(p)

This is a traditional objection to the "behaviorism" of philosophers such as Carnap. I recall arguing in an undergraduate term paper that this was a misunderstanding of behaviorism: there is no reason that "behavior" should not encompass e.g. the behavior of neurons, which are in principle just as publicly observable as a subject's verbal behavior. So the question is whether any being could have a brain observably identical to yours and yet have different experiences.

Replies from: Wei_Dai
comment by Wei Dai (Wei_Dai) · 2010-01-12T05:46:58.843Z · LW(p) · GW(p)

Ok, let me try this again. I want a way to map between the internal and external views of a mind. That is, given what I know about what I'm experiencing, what can I deduce about the physical structure of my brain? And given a physical description of a mind, what can I know about what it is experiencing? Perhaps this is already considered a legitimate part of the problem of consciousness according to heterophenomenology, or "behaviorism" (are they the same thing?), but if so I think it's at least not a part of the problem that those approaches tend to emphasize. In any case, I'd appreciate it if ciphergoth could address this topic a bit in his post.

(Why am I so interested in this part of the problem? Mainly because I need the solution in order for UDT1 to be usable by human beings.)

comment by Paul Crowley (ciphergoth) · 2010-01-12T08:42:45.343Z · LW(p) · GW(p)

But a theory that people are deliberately lying about their internal experience isn't really going to fly. Even kids ask "when you see red, do you see the same colour I do?". No-one prompted them to lie.

comment by Zack_M_Davis · 2010-01-10T23:47:22.737Z · LW(p) · GW(p)

many worlds requires a stupendous, ongoing, exponential violation of conservation laws.

My domain knowledge is weak, I'm still pretty sure it doesn't work that way. "Many worlds" is just a cute name; there aren't really new worlds appearing from nowhere. Rather, reality is fundamentally made out of this wavefunction-thingy, and from our perspective it looks like "worlds" "branching."

(EDIT: quotation truncated for relevance)

Replies from: Alicorn, Tyrrell_McAllister
comment by Alicorn · 2010-01-10T23:55:02.500Z · LW(p) · GW(p)

from our perspective it looks like "worlds" "branching."

Nothing has ever looked like that to me.

Replies from: Zack_M_Davis
comment by Zack_M_Davis · 2010-01-11T06:11:21.238Z · LW(p) · GW(p)

Try squinting.

(A beat.)

Right, I phrased that poorly. From our perspective it looks like objective probability, and it's pedagogically useful to speak of the different possible outcomes happening in different "worlds."

comment by Tyrrell_McAllister · 2010-01-11T00:35:34.860Z · LW(p) · GW(p)

Conservation laws are mathematical constraints. If Many Worlds violated any conservation law, it would be possible to show this mathematically. Phil, do you believe that this has been done? If not, why do you think that MW violates conservation laws?

Perhaps you think that MW says that, at each "branching", a distinct universe, made out of distinct material stuff, is created out of nothing. That would be a violation of conservation, but that's not what MW says.

Replies from: PhilGoetz, PhilGoetz
comment by PhilGoetz · 2010-01-11T01:28:30.547Z · LW(p) · GW(p)

Perhaps you think that MW says that, at each "branching", a distinct universe, made out of distinct material stuff, is created out of nothing. That would be a violation of conservation, but that's not what MW says.

That is what the Wikipedia entry on many-worlds describes it as saying, as I read it.

It further says:

Common objections and misconceptions:

  • Conservation of energy is grossly violated if at every instant near-infinite amounts of new matter are generated to create the new universes.

MWI response: Conservation of energy is not violated since the energy of each branch has to be weighted by its probability, according to the standard formula for the conservation of energy in quantum theory. This results in the total energy of the multiverse being conserved.

Please remember, my larger point is not to argue for or against many worlds. My subsidiary point is that many worlds is considered an acceptable view to hold, and therefore not considered dualism; even though it appears - to most people, even if not to MW experts - to fail the only sensible test I was able to come up with to distinguish good dualism from bad dualism.

  • It does not matter whether experts currently agree that MWI actually does or does not violate the conservation of energy or not. It only matters that many people who would say, if asked, that MW violates conservation of energy, would nonetheless not call it dualism.
  • My conclusion was not that many worlds is wrong; my conclusion was that the concept of "dualism" is bankrupt.

More generally, not to Tyrrell specifically, but to the multitudes aggressively down-voting this post: Relax! I didn't say anything bad about EY! Just follow the chain of reasoning, people; stop hyperfocusing on phrases that upset you, and getting upset because something you like was used in the same context as something you don't like.

Replies from: Tyrrell_McAllister, cousin_it, Cyan, orthonormal
comment by Tyrrell_McAllister · 2010-01-11T04:36:37.736Z · LW(p) · GW(p)

More generally, not to Tyrrell specifically, but to the multitudes aggressively down-voting this post: Relax! I didn't say anything bad about EY! Just follow the chain of reasoning, people; stop hyperfocusing on phrases that upset you, and getting upset because something you like was used in the same context as something you don't like.

I don't think you're being downvoted just for criticizing MW or EY. Some of Mitchell Porter's posts criticizing the MW orthodoxy here have been highly upvoted.

Replies from: Eliezer_Yudkowsky
comment by Eliezer Yudkowsky (Eliezer_Yudkowsky) · 2010-01-11T08:39:52.758Z · LW(p) · GW(p)

You're not being downvoted for criticizing the dominant paradigm. You're being downvoted for gross misunderstanding of the dominant paradigm followed by unwillingness to accept correction.

Also, I personally simply do not want to see a chapter-by-chapter review of Consciousness Explained on LW, and may veto this even if it's not net-downvoted - keeping in mind that while any reader can upvote, only commenters have downvotes to use, and so the fact that this has gotten a number of downvotes is still quite alarming even if some upvotes canceled them out.

Replies from: ciphergoth
comment by Paul Crowley (ciphergoth) · 2010-01-11T09:48:51.825Z · LW(p) · GW(p)

Yes, I'm also sorry to see a second part of this. CS is a good book but not worthy of this sort of crawl-over. And if it really were worthy of it, I would want to read it from someone who had finished the book before they started posting.

comment by cousin_it · 2010-01-11T11:20:34.352Z · LW(p) · GW(p)

MWI response: Conservation of energy is not violated since the energy of each branch has to be weighted by its probability, according to the standard formula for the conservation of energy in quantum theory. This results in the total energy of the multiverse being conserved.

This sounds extremely weird to me. Who cares about the total energy of the multiverse? When worlds split, can one world gain energy from another? I thought conservation laws applied to each single worldline as it developed, as it's the only way we single-world beings could notice and make use of them. So neither the "objection" nor the "MWI response" actually make sense. Or am I misreading the quote?

comment by Cyan · 2010-01-11T02:18:15.048Z · LW(p) · GW(p)

More generally, not to Tyrrell specifically, but to the multitudes aggressively down-voting this post: Relax! I didn't say anything bad about EY! Just follow the chain of reasoning, people; stop hyperfocusing on phrases that upset you, and getting upset because something you like was used in the same context as something you don't like.

Voted up just for this.

comment by orthonormal · 2010-01-11T01:47:29.140Z · LW(p) · GW(p)

It does not matter whether experts currently agree that MWI actually does or does not violate the conservation of energy or not. It only matters that many people who would say, if asked, that MW violates conservation of energy, would nonetheless not call it dualism.

If this is your point, please rephrase the controverted paragraph in the post. I think that as currently written, it distracts from your main point by opening up a different argument.

Replies from: PhilGoetz
comment by PhilGoetz · 2010-01-11T04:32:38.270Z · LW(p) · GW(p)

Okay. Rephrased.

Replies from: orthonormal
comment by orthonormal · 2010-01-11T05:45:23.205Z · LW(p) · GW(p)

Thanks; I find it much improved.

I think the general pattern is that it's counterproductive to baldly state† multiple independent controversial claims in the same place— one winds up in a muddle of all the various arguments. Bracketing off all but one claim at a time, via disclaimers (like the one you added) or other forms of rhetorical modesty, is IMO the most productive way to go.

† (or, of course, to write so that a reader might think you are baldly stating several such claims)

comment by PhilGoetz · 2010-01-11T00:48:07.715Z · LW(p) · GW(p)

Then I don't know what MW says. I'd be surprised if you can formulate many-worlds in a way so that there are as many worlds in the past as in the future, and still have entropy increase.

Replies from: Tyrrell_McAllister
comment by Tyrrell_McAllister · 2010-01-11T00:54:30.817Z · LW(p) · GW(p)

I'd be surprised if you can formulate many-worlds in a way so that there are as many worlds in the past as in the future.

The number of worlds isn't necessarily conserved. But the things called "worlds" in MW aren't made out of distinct pieces of material stuff. (ETA: That is, you don't have one collection of atoms constituting one world while a disjoint collection of atoms constitutes another world.)

Replies from: PhilGoetz
comment by PhilGoetz · 2010-01-11T04:24:43.643Z · LW(p) · GW(p)

I don't think I'm going to understand many-worlds today.

Replies from: Kaj_Sotala
comment by Kaj_Sotala · 2010-01-11T20:17:49.261Z · LW(p) · GW(p)

Good and Real explains it pretty well. (Just resumed reading it today, and realized I liked the chapters on physics more than I thought I would.)

comment by Joanna Morningstar (Jonathan_Lee) · 2010-01-11T14:54:08.749Z · LW(p) · GW(p)

I haven't read Consciousness Explained in a while; checking with a summary here confirmed my recollections.

Section 3 isn't asserting a consistent definition of mind-stuff; it's noting the heuristics and feelings that lead to an assertion of mind-stuff existing as separate from existing known phenomena. It's Dennet's version of this.

From section 4:

D also refutes dualism by saying that mind stuff can't both elude physical measurement, and control the body; as anything that escapes our instruments of detection can't possibly interact with the body to control it.

Mind-stuff must interact with a physical system (your nervous system), and thus in principle can be detected by purely physical measuring devices. It is in principle detectable.

Furthermore if ontology needs extending, as in

but anything that can move a physical thing is itself a physical thing (although perhaps a strange and heretofore unstudied kind of physical thing).

[p35, emphasis added.]

then this mind-stuff must explain more than it introduces; regressing conciousness to mind-stuff doesn't help in and of itself.

I also note that you don't suggest a third possibility; merely accuse Dennet of muddying the waters; what is your alternative where the mind is not reducible to the material in your skull yet is still within the ken of science?

Edit: blockquote fail