To answer this question we need to know what we mean by 'explain' and how it differs from 'explain away.'
2. A second point, equal in obviousness unto the first, is that a decent explanation cannot issue in the elimination of the explanandum, that which is to be explained. You cannot explain beliefs and desires by saying that there are no beliefs and desires. A successful explanation cannot be eliminativist. It cannot 'explain away' the explanandum.
3. Summing up (1) and (2): the very project of explanation presupposes the existence of the explanandum, and success in explanation cannot result in the elimination of the explanandum.
4. Daniel Dennett points out that there can be no explanation without a certain 'leaving out': "Leaving something out is not a feature of failed explanations, but of successful explanations." (Consciousness Explained, p. 454.) Thus if I explain lightning as an atmospheric electrical discharge, I leave out the appearing of the lightning to lay bare its reality. That lightning appears in such-and-such a way is irrelevant: I want to know what it is in reality, what it is in nature apart from any observer. The scientist aims to get beyond the phenomenology to the underlying reality.
5. It follows that if consciousness is to be explained, it must be reduced to, or identified with, something else that is observer-independent. Dennett puts this by saying that "Only a theory that explained conscious events in terms of unconscious events could explain consciousness at all." (454) For example, if your explanation of pain in terms of C-fibers and Delta A-fibers still contains the unreduced term 'pain,' then no satisfactory explanation has been achieved. There cannot be a "magic moment" in the explanation when a "miracle occurs" and unconscious events become conscious. (455)
6. Now if a successful explanation must explain conscious events in terms of unconscious events, then I hope I will be forgiven for concluding that consciousness CANNOT be explained. For, as I made clear in #2 above, a successful explanation cannot issue in the elimination of that which is to be explained. In the case of the lightning, there is a reduction but not an elimination: lightning is reduced to its observer-independent reality as electrical discharge. Now suppose you try the same operation with the sensory qualia experienced when one observes lightning: the FLASH, the JAGGED LINE in the sky, followed by the CLAP of thunder, etc. You try to separate the subjective appearance from the observer-independent reality. But then you notice something: reality and appearance of a sensory quale coincide. Esse est percipi. The being of the quale is identical to its appearing. This is what Searle means when he speaks of the "first person ontology" of mental data.
7. It follows from #6 that if one were to explain the conscious event in terms of unconscious events as Dennett recommends, the explanation would fail: it would violate the strictures laid down in #2 above. The upshot would be an elimination of the datum to be explained rather than an explanation of it. To reiterate the obvious, a successful explanation cannot consign the explanandum to oblivion. It must explain it, not explain it away.
8. I conclude that consciousness cannot be explained, given Dennett's demand that a successful explanation of consciousness must be in terms of unconscious events. What he wants is a reduction to the physical. He wants that because he is convinced that only the physical exists. But in the case of consciousness, such a reduction must needs be an elimination.
9. To my claim that consciousness cannot be explained, Dennett has a response: "But why should consciousness be the only thing that cannot be explained? Solids and liquids and gases can be explained in terms of things that are not solids, and liquids, and gases. . . . The illusion that consciousness is the exception comes about, I suspect, because of a failure to understand this general feature of successful explanation." (455)
Dennett's reasoning here is astonishingly weak because blatantly question-begging. He is arguing:
A. It is a general feature of all successful explanations that F items be explained in terms of non-F items
B. Conscious items can be explained
Ergo
C. Conscious items can be explained in terms of nonconscious items.
(B) cannot be asserted given what I said in #6 and #7. I run the argument in reverse, arguing from the negation of (C) to the negation of (B): conscious items such as pains are irreducible.
10. Recall from #4 that Dennett said that successful explanations must leave something out. But in the case of a conscious item like a pain, what is left out when we explain it is precisely what we needed to explain! For what is left out is precisely the sensory quale, the felt pain, the Feiglian "raw feel,' the Nagelian "what it is like."
11. Amazingly, on p. 455 he retracts what he said on the previous page about successful explanations having to leave something out. He now writes:
Thinking, mistakenly, that the explanation leaves something out, we think to save what otherwise would be lost by putting it back into the observer as a quale -- or some other "intrinsically" wonderful property. The psyche becomes the protective skirt under which all those beloved kittens can hide. There may be motives for thinking that consciousness cannot be explained, but, I hope I have shown, there are good reasons for thinking it can. (455)
Do you see how Dennett is contradicting himself? On p. 454 he states that a successful explanation must leave something out, which seems plausible enough. Then he half-realizes that this spells trouble for his explanation of consciousness -- since what is left out when we explain consciousness in unconscious terms is precisely the explanandum, consciousness itself! So he backpedals and implies that nothing has been left out, and suggests that someone who affirms the irreducibility of qualia is like a lady who hides her 'kwalia kitties' under her skirt where no mean neuroscientist dare stick his nose.
The whole passage is a tissue of confusion wrapped in a rhetorical trick. And that is the way his big book ends: on a contradictory note.
12. To sum up. A successful explanation cannot eliminate the explanandum. That is nonnegotiable. So if we agree with Dennett that a successful explanation must leave something out, namely, our epistemic access to what is to be explained, then we ought to conclude that consciousness cannot be explained.
Bill,
Thanks for a brilliant post!
(I should preface my remarks by confessing that I have no formal philosophical training, so my remarks are admittedly non-rigorous.)
You maintain that Dennett's work fails because he explains away the explanandum rather than merely reducing it. Is it possible to compress that even further? For example, could Dennett's work be considered an explanatory category mistake. In other words, is trying to explain a first-person experience in third-person terms like trying to explain numbers in terms of colors?
If I am understanding my own argument right, that would seem to make consciousness a basic explanatory category that therefore cannot be explained in terms of something else. Do you buy that?
Posted by: Marshall | Wednesday, April 22, 2009 at 09:10 PM
Can 'truth' be explained in terms of non-truth? To assert that it can is rather a contradiction in terms, isn't it?
'Truth' is itself, it doesn't reduce to anything else. Likewise, it seems to be the case, with consciousness: it is itself, it doesn't reduce to anything else.
Or, as C S Lewis explained: "All explanations come to an end." To explain a thing is to see/understand it in terms of something else (generally, something more "basic" or epistemologically "prior"); but all explanations must, soon or late, reach the thing which is itself and cannot be further explained.
Posted by: Ilíon | Thursday, April 23, 2009 at 02:20 AM
Hi Bill,
Consciousness is a very tricky issue, indeed. It seems we are left with denying consciousness (materialism) denying matter and keeping consciousness (idealism) keeping them both and having an uncomfortable gap between them (dualism) or positing the existence of some stuff that can do both things (aspect dualism). None of these is remotely satisfactory as far as I can tell.
But I think maybe we err in thinking of consciousness as a thing. Maybe we should be thinking of consciousness more as a verb. Note that consciousness is always transitive. You are always conscious of something. It makes no sense at all to say "I'm conscious but not conscious of anything." I'm not sure how much that helps.
By the way, since we are on the topic of mental states, I'd like to reiterate that I think you are wrong when you say that pain is not an intentional state. A doctor asking a patient if he has any pains is looking at pain as a sign of disorder in the body, a sign of harm. It is not merely self-referencing.
When I'm writing something and I'm stuck, I use the following trick on myself, "Sure, I don't have the answer, but if I did would it look like?" Maybe you should begin by explaining what an explaination of consciousnes would have to look like.
Posted by: Spencer | Thursday, April 23, 2009 at 08:54 AM
To whom (or what) do the terms "Feiglian" and "Nagelian" refer to?
Posted by: Brian | Thursday, April 23, 2009 at 12:13 PM
Brian,
Herbert Feigl and Thomas Nagel. See http://www.umass.edu/philosophy/PDF/Aune/feigl.pdf. And http://philosophy.fas.nyu.edu/object/thomasnagel.
Posted by: Bill Vallicella | Thursday, April 23, 2009 at 01:32 PM
Marshall,
Thanks for the kind words. My point is that consciousness is irreducible. If you try to reduce it to something else, you will simply eliminate it. The reduction collapses into an elimination. So I am not saying that Dennett eliminates it when he should be reducing it, but that the very project of attempting a reductive account of consciousness is doomed from the start since it leads inevitably to the elimination of consciousness. It shows a fundamental lack of understanding of what consciousness is. It makes sense to identify lightning with an electrical discharge. That of course leaves out the appearing of the lightning to a subject. This is not a problem. But if you then try to give a reductive materialist account of the appearing of the lightning to a subject, you will end up leaving out the very thing that needs accounting, namely, the appearing.
What you say above is basically right. Consciousness cannot be explained in terms of anything unconscious.
Posted by: Bill Vallicella | Thursday, April 23, 2009 at 02:00 PM
Spencer writes, >>Note that consciousness is always transitive. You are always conscious of something. It makes no sense at all to say "I'm conscious but not conscious of anything." >I think you are wrong when you say that pain is not an intentional state. A doctor asking a patient if he has any pains is looking at pain as a sign of disorder in the body, a sign of harm. It is not merely self-referencing.<<
You just don't understand the issue. I've gone over this many times.
Posted by: Bill Vallicella | Thursday, April 23, 2009 at 02:05 PM
Illion,
Truth does indeed seem basic and irreducible. If you try to deny it ('There is no truth!') you presuppose it ('It is true that there is no truth.')
Posted by: Bill Vallicella | Thursday, April 23, 2009 at 02:08 PM
You cannot explain beliefs and desires by saying that there are no beliefs and desires. A successful explanation cannot be eliminativist.
Is this right? In one sense yes, of course. Yet shouldn't we distinguish between meaning and reference here? The eliminativist (of which I am not) is basically saying that the words we use are unhelpful and ultimately don't properly refer. It's akin to how many scientists think of something like ki (chi). If someone says, "explain meridian lines," it seems fair for a scientist to reply, "there are none."
Posted by: Clark Goble | Thursday, April 23, 2009 at 08:18 PM
Spencer,
No dualist worth their salt believes that consciousness is a 'thing'. Some (like me) believe that it's a substance, but in the philosophical sense of substance, the definition of which does not in any way imply materiality. When dualists talk about consciousness, they aren't talking about some thin, shade-like piece of quasi-matter, the 'ghost' of Ryle's 'ghost in the machine'. They are talking about interiority, subjectivity, experience itself, which, intrinsically and by definition, cannot be seen, weighed, or measured, even granted the tools and theory of an ideal physical science. When materialists assume that consciousness, if it is to exist, must be some quasi-material thing, they beg the question by assuming that all substances must be physical substances.
I'm afraid the idea that consciousness is just a verb makes no sense to me. It sounds like a very strange category mistake. Truth, for example, is irreducible; as Illion and Bill pointed out, to reduce truth to something else is to presuppose the irreducible truth of the reduction. We would be have explained nothing if we just sat back and said that truth is a noun; precisely what's at issue is the nature of the reality to which the noun refers. Similarly for consciousness. While we might use verbs to describe consciousness, what's at issue isn't so much the terms we use, but the reality to which those terms refer. You might say that the term refers to no reality, like Clark Goble's meridian line example. But to do so would lead to an elimination, not an explanation, of consciousness.
Posted by: Brodie Bortignon | Thursday, April 23, 2009 at 08:43 PM