On 09 Mar 2011, at 12:28, 1Z wrote:
On Mar 9, 7:24 am, Bruno Marchal <marc...@ulb.ac.be> wrote:
On 08 Mar 2011, at 15:54, 1Z wrote:
On Mar 8, 1:43 pm, Bruno Marchal <marc...@ulb.ac.be> wrote:
On 07 Mar 2011, at 21:48, David Nyman wrote:
On 7 March 2011 15:56, Bruno Marchal <marc...@ulb.ac.be> wrote:
Reduction is not elimination
Ontological reduction does not necessarily entail epistemological
*elimination*, but it does entail ontological *elimination*.
Bruno, this is what I was trying to say some time ago to Peter.
"ontological reduction does not necessarily entail epistemological
*elimination*" is of course precisely the question that mustn't be
dodged or begged, which is what I'm convinced Peter is doing by
insisting dogmatically that "reduction is not elimination". The
is that a primitive-materialist micro-physical theory is
(if not explicitly) committed to the claim that everything that
is *just* some arrangement of ultimate material constituents.
literally *all there is*, ex hypothesi. Despite the fact (and, a
fortiori, *because* of the fact) that this is not what any of
observers, actually finds to be the case, we can nonetheless
deny or ignore this "inconvenient truth". But if we do not so
we can perhaps see that here we have the materialist Hard
perhaps its purest form: why should there be anything at all
ensemble of quarks? (or whatever this month's "ultimate
everything" is supposed to be). And why should any subset of an
ensemble of quarks be localised as "here" or "now"?
Adding "computation" to the materialist mix can't help, because
computation is also just an arrangement of quarks, or whatever,
talking about emergence, or logical levels etc, can achieve
because after any amount of this logical gyrating *it's still all
quarks*. Of course, funnily enough, we manage nonetheless to talk
about all these additional things, but then to claim that this
can be materially "identical" to the quarks "under some
is just to play circular and futile games with words. Plugging
conclusion into the premise can of course explain nothing, and
begs the critical question in the most egregious way.
The crucial difference in your theory, Bruno, to the extent that
understood it, is that it is explicitly both analytic AND
That is, it postulates specific arithmetical-computational
components" and their relations, AND it further specifies the
emergence of conscious first-person viewpoints, and their layers
composite contents, through an additional subtle filtering and
synthesis of the relational ensemble. Hence, through a kind of
duality of part and whole, it is able to avoid the monistic
and consequently isn't forced to deny, or sweep under the rug, the
categorical orthogonality of mind and body. In such a schema, the
entire domain of the "secondary qualities", including matter, time
space themselves, is localised and personalised at the
these analytic and synthetic principles.
I think that you have a clear understanding of that rather subtle
The problem with the materialist, once they use comp, and thus does
not "materialize the soul" is that they have to identify a state of
mind with a state of matter. Pain become literally neuronal
quark interaction. But pain is not neuronal firing, pain is a non
pleasant subjective reality lived by person, and to equate pain
neuronal firing leads the most honest materialist scientist to the
conclusion that pain somehow does not exist (eliminativism).
Another option is open to them: it is a brute fact that the neuronal
IS the pain, but physicalese descriptions of neuronal firing don't
capture that because
they are inadequate.
Except that it is not a brute fact that the neuronal firing is the
pain. That does not make sense. A neuronal firing is entirely
descriptibe in a thrid person way, but a pain is not at all.
If it is entirely describable, the identification fails. If the
is true as a brute fact, then it is the description that is failing.
it fails, we can't grasp the brute fact, but brute facts aren't
graspable...that is the point of the jargon term "brute fact". In the
world of mathematics, ideas such as brute facts and unreachable
noumena don't make sense, because mathematical objects are fully
describable. Brute facts belong to realism, where certain things "just
and there is no guarantee they will be comprehensible.
Arithmetical truth is definable by Zermelo Fraenkel (a Löbian
machine), but not by Peano Arithmetic (another simpler one). Most big
object in math are not fully graspable, and if we are machine any
notion of truth-about-us, is beyond our reach, yet a priori
mathematical, assuming comp.
"fully describable" is ambiguous. natural numbers are already not
fully describable. The least non constructive ordinal is not fully
describable, even if nameable in ZF + Church thesis, etc.
associate them, but you can't equate them. You can identify 3-heat
with molecular cinetic energy, but you cannot equate the sensation of
heat with neuronal firing in the same way. The problem is that *any*
physicalese or not, third person description of what could be the
will fail, because the pain quale is just not a third person
I dare say any physicale
se will fail, but then we can keep
the identity between the pain and the neural activity, and reject
the identity between the nerual firing and the physicalese description
OK. I guess you associate pain to the primitive matter. But that is
even more incoherent with respect to the comp hypothesis. Also, it
identify two mysteries, and of very different nature. That looks more
and more pseudo-religious to me.
Comp solves the problem by identifying the
pain with what appears to be existing non describable, by numbers,
attribute of numbers' relation.
eventually, some realize that if neuron plays a role in pain, they
only *associate* it to neuronal firing, and this leads to dualism,
which most materialist abhor. That is even truer for monist
materialist who are then force to accept a form of
Prima facie, comp, which is also a form of reductionism, might
to be lead to a similar problem,
It leads to a worse problem. The objection to identifying qualia
physical happening is that felt qualitiies are not identifiable with
The approach outlined above resolves that
with the idea that concrete physical events have a noumenal
hinterland which is not captured by physicalese descriptions.
Well, this is introducing magical thing in the picture. If such
noumenal things exist, they have to escape the comp description.
Is that so shocking? As I have said, the problems with
computationalism, as a theory of qualia, are the same are the same as
Computationalism is not per se a theory of qualia. Comp is the
assumption that qualia are preserved through functional substitution
at some level. It happens that when a machine introspects itself it
brings on a plateau a theory of qualia (Z1*, X1*, and perhaps S4Grz1).
And both UDA and AUDA makes it testable, because quanta are the
sharable part of those qualia, so that we can compare them to empiric
say "yes" to the doctor if my consciousness and qualia is related
noumenal hinterland of the matter in my physical brain. That noumenal
matter hinterland contradicts the idea that there is a level of
description of myself where matter and physical structure can be
replaced by arbitrary different matter and structure, once they
preserve the computational relations, which are arithmetical, by
Yep. Comp is a bad theory of qualia.
As I said, Comp is the assumption that qualia are preserved through
functional substitution at some level. A theory of qualia emerges from
the self-reference logic.
But then if you abandon comp at this stage, it means I have made my
point. I thought you were defending COMP + MAT(*)
If you have a theory of qualia using primitive matter, and coherent
with comp, then you should be able to use it to extract a flaw in the
(*) For the new people: MAT = weak materialism = the common doctrine
that primitive matter exists, or that matter exists at the basic
We don't know
how to write subroutines for phenomenality.
Assuming comp, it is enough to write the code of a universal machine.
For example RA (Robinson Arithmetic) has qualia, but RA lacks the
cognitive ability to understand the notion of qualia. If you give it
the induction axioms (getting PA), you get a Löbian machine, and it
has the full power to find its own theory of qualia.
That artificial people
do not have "real feelings" is a staple of sci fi.
However, in the realm of pure math, without "stuffy matter",
no such hinterland is available: neuronal firings have to be
identical with their physicalese (and hence mathematical)
The contrary happens. The physical stuff lost the possibility to be
entirely describe in mathematical terms. We need theological terms,
and the whole self-reference logics.
I don't see why
It is the point of the Universal Dovetailer Argument. See Sane04
If the quale isn't there, it isn't anywhere: it has no place to
It has the whole theological realm, which exists *epistemologically*
for any universal machine introspecting itself. And the self-
logic justifies entirely their non communicable feature, without
denying them and without trying to localize them in any way, like
numbers are not localized in any place. Localization is a higher
epistemological emerging notion, not a primitive one, in the comp
I don't think you can model qualia just as being incommunicable.
I never suggested I did that.
Qualia have many other properties like being immediately
apprehensible, physically sensible, having a logic with fields
semantics, (like colors), they are undoubtable, etc.
It *happens* that physics is a sort of first person plural quale,
making the comp theory (+ classical knowledge theory) testable.
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To post to this group, send email to firstname.lastname@example.org.
To unsubscribe from this group, send email to
For more options, visit this group at