At 10:25 04/02/04 -0200, Eric Cavalcanti wrote:

EC: Entering the discussion here...

----- Original Message -----
From: "Pete Carlton" <[EMAIL PROTECTED]>

> PT: But even this goes way out in front of what we can possibly know.  You
> say we have no idea what these feelings are like to experience--but why
> should we assume we even are entitled to ask this question?

EC: And why assume that we are not? I prefer not to have
my right to ask waived. If I cannot ask it, then I should
understand why not, and that would be an answer in a

BM: Yes. Somehow it would be an answer at a higher level, or even a meta-level.
Now later I will perhaps try to explain that consciousness will be linked
(among other things) with some collapse between a level and a meta-level, so
it's good to keep in mind those levels, and it is good to expect that if
some question lacks an answer, we will perhaps be able to justify that very
fact, by a level shifting indeed.

> To borrow a bit from Wittgenstein -- imagine you have completely > translated these aliens' language, and they tell you that each of them > has a box with something inside it. Although they talk a lot in rather > vague terms about what's in their box, they insist you can't really > know what is inside it. Now what is the logical conclusion here: > a) There may or may not be something in the box. > b) There's definitely something in the box, and I have absolutely no > idea what it is. > > What on earth could possibly make someone conclude (b) here? It's not > logical at all. Yet this is what people conclude when they bend over > backwards talking about "qualia" and how ineffable they are.

And you seem to conclude a (c) hypothesis: "there is
definitely nothing inside the box". I stay with (a). And
to try to find out if there is something there or not, we
need to talk about it, and "qualia" is the word for the
hypothetical contents of the box.

OK. Although I assume you are using the word "content" with the meaning of "meaning", qualia being an interpretation of a state of the "brain" by the owner of the brain, grosso modo. Saying that "qualia are ineffable" is very interesting, and Carlton's mention of Wittgenstein is quite relevant (imo...), and indeed, shifting from the level of what consciousness can be I propose to try to capture it by the possible self-referential discourse, and ineffability is the key point for, at first, a very large sense of consciousness (including all Protogoras Virtue, conscience, sensation, self-identity, cleverness, honesty,...) So here is a first proponent for an "axiom" of those "virtues":

AX 1 It is ineffable or uncommunicable

I call it "Wittgenstein" principle (in my Brussel's thesis), and we
can abbreviate it by 'existential) formula (equation)

AX 1 Not Communicable x,

and even translate it in modal logic language as -[]x.

Note that Wittgenstein is in trouble here, because when he says
in the tractatus "we cannot talk on those things" he is really
talking about, and we are entitled to ask him what does he really
talk about. In the same way we will need to find "models", or
mathematical structures to classify our valid forms of reasoning
with those sentences (but then that's all what logic is about).

A trivial solution for a honest machine (or being, I am not
using comp here) would be F (your favorite falsity, like 1 = 0, ...)
But Wittgenstein was hardly giving a moral statement like

"you will not communicate the false" ( -[]F )

nor can we suppose Wittgenstein was saying "you will not be wrong",
or "you will not be mad", etc...

So F is perhaps a little bit too trivial, but note that this solution
will remains stable for a very vast amount of semantics for the box.

But Godel and the logicians found many other solutions...
Actually for self-referentially correct (SRC) machine, the Wittgenstein principle
is itself a solution of the Wittgenstein principle:

-[] (-[]F)

Actually, a SRC machine can even prove that, that is, that if she would be
honest (consistent,  ...) then she would not been able to prove that she is

-[]F -> -[] (-[]F)

That is just Godel's second incompleteness theorem. It gives an arithmetical model for most of what we can be derived, informally or not, from the ineffability assumption.

> > So, in addition to the empirical data, there is this extra bit of > > information, neither contained in the data nor able to be derived from > > it using the laws of physics: what it actually feels like to be the > > one experiencing the subjective sensation. If someone can think of a > > better way to describe it than "extra bit of information" or can come > > up with a way to formalise it, I would be happy to hear about it. > > A better way to describe what, exactly? "What it actually feels like"? > But why do you first commit yourself to the view that this question > makes any sense?

Suppose a blind man did understand all the chain of events
that lead from the light reaching the eyes to the retina, then
to the brain and finally to the qualia of red. Would he see the
red? Why not?

I can imagine that I have my vision scrambled in such a way
that red is exchanged with blue. Red is the color that I associate
with an apple, and blue to the sky. I can imagine that the sky
appeared to have the color of an apple and an apple appeared
to have the sky's color. But if that happened, from then on, I
could change the names of the colors in such a way that I still
called the apple 'red' and the sky 'blue'. If there is no such a
thing as 'qualia', then nothing really happened. But I could tell
that things are different. In what sense could they be different?

Good question.

> > I suppose there will still be some who insist that if you know all
> > about the physiology etc. behind the alien response to gamma rays,
> > then you know all there is to know. I think this response is analogous
> > to the "shut up and calculate" attitude to the interpretation of
> > quantum mechanics.
> Yes, I am one of these people.  You say "if you know all about", and
> you must be taken seriously here:  you would really have to know >all<
> about it.  But if you did, you would be able to entirely trace the
> causal pathway from the receipt of the gamma rays, to whatever internal
> responses go on inside the alien's body, to the subsequent report of "I
> feel that pleasant, odd-multiple feeling".  Let's say you had that
> entire explanation written out.  And "subjective experience" doesn't
> appear anywhere on this list.  So what reason on earth do you have to
> assert that it exists?

When the alien says "I feel that pleasant feeling", he is just
saying that he knows that chain of events is happening
in his body right now. Suppose you are watching him with
equipments that let you know that same thing. Could you
also say "I feel that pleasant feeling" too? Why not, if there
is nothing beyond the chain of events? What could make the
alien's knowledge different from yours? One obvious answer
is that "he is the organism where those events are happening".
But this means that each organism is entitled to "feel
something" about himself , an experience that is inaccesible to
others, no matter how comprehensive their knowledge is. But
that is something that you are claiming that does not exist.
How can we explain this without something as "qualia"?

...or without making explicit, like you did, the difference between
a pure 3-person description of a chain of events ("made of" a number
quantifiable happenings) and the way an organism is able to
get higher order and unquantifiable information or knowledge
about itself: that is, some first person apprehension.

But now what could be that "unquantifiable" knowledge?
Well I want to stay modest but it's here that my thesis makes
a "tour de force". Actually, the "tour de force" has been made
by Thaetetus when trying to define science and knowledge
to Socrate.  Now the crazy things is that the godel's incompleteness
phenomenon introduces so many nuance between the notion of
provable, true, consistent, correct, that it makes it possible to
translate quasi-literally the (many) Thaetetus attempt to define
knowledge. In particular, if you define knowable(p) by
provable(p) and true(p), you get a knower which has no name,
no description, and cannot know he could be a machine, or any
object having a pure 3-person description, showing there is a
necessary sort of gap between 1 and 3 pov for honest machine.
Than this is related to Brouwer's theory of soul and consciousness,
and intuitionnistic logic. Eventually it is another thaetetic variant
of provability which provide the logic of the "probability 1" physical
propositions, from which a quantum logic can be extracted, and
from which a similarity measure for consistent
computational extensions can be mathematically apprehended.

Sorry for making so many questions. I don't intend to be
pedantic, but I really don't know the answers.

> Of course subjective experience exists in a way -- but it's just a way
> of talking about things.  It isn't a "primitive".  When I see red, I
> have a subjective experience of red, sure -- but all this means is just
> that my brain has responded to a certain stimulus in the way it
> normally does.

And maybe it is not all that this means...

I would say it most certainly means something else. It is more the fact that
the brain has responded in some way which means for you that
you are seeing red. Would Pete Carlton say "I love that person but
actually it just means my (and btw what does mean "my" here) brain
has been put in such configuration"?

I will try to explain more of the theory of "consciousness" ... I am at last
rereading my brussel's thesis, which is the same as the lille thesis,
except everything is explained in details (In the lille thesis I wrote just
6 pages on the theory of consciousness (this is justified by the fact that
it is essentially just logical variant of the "well-known" logic of self-reference G and G*),
but in the brussels's thesis those 6 pages makes 400 pages ...mmmmh...

I hope you have all buy Boolos 93 now!



Reply via email to