On 24 Aug 2014, at 00:50, meekerdb wrote:
On 8/23/2014 9:09 AM, David Nyman wrote:
On 23 August 2014 05:02, meekerdb <[email protected]> wrote:
What we observe in practice are physical devices of various kinds
(indeed, in principle, indefinitely many kinds) that we accept FAPP
as adequately instantiating particular classes of computation
within certain fairly stringent limits.
What we observe, aside from observing our own thoughts and maybe
even then, is always theory laden. Partly we see the world through
a theory of objects that evolution has provided us, sort of naive
physics, but a theory nonetheless as optical illusions demonstrate.
And the relevance of this remark to my point is what, precisely?
Just a cautionary remark that observing a physical device already
involves assumptions that it is an experience that can in principle
be shared and agreed on by other observers - i.e. that it assumes
some physics.
To put it another way, we are prepared to interpret the normal
physical behaviour of such devices *as if* it instantiated the
mathematical notion of computation.
Right, because it usually (modulo a dropped bit or so) does.
Obviously. And your point is...?
These considerations should make it clear that any description of
the normal behaviour of a physical device as computation can only
be in a sense that is, ultimately, metaphorical.
I think you are too hung up on ontology. You denigrate everything
that's not in terms of the ur-stuff of (some unknown) true ontology
as metaphor or fiction. Why not accept that knowledge, including
knowledge of ontology, is always provisional and uncertain and it's
best to think of it as a model summarizing our best idea - but not
necessarily the one TRUE idea.
Maybe you are getting a little too hung up on what you imagine me
to be hung up on. If that is the case, it might make you somewhat
unreceptive at the outset to what you assume to be my line of
argument. I'm not trying to grind any axe in particular but only to
articulate what I suspect are sometimes unrecognised assumptions as
clearly as I can and then examine the consequences. Of course I may
well be wrong on any point and so my aim is to encourage discussion
from which I might learn.
In this particular case what I'm driving at isn't that either
matter or computation need be considered as some kind of mystical
"ur-stuff" (whatever that would be). What I'm questioning is
whether it is really coherent to attribute *first-person*
consciousness to "computation" against the background of any theory
that is committed to a physically primitive level of explanation.
You're saying it may be incoherent to reduce consciousness to
computation, if computation is reducible to physics? Why would that
be incoherent? Must 'reduction' necessarily be reduction to the
bottom to be coherent? Or are you assuming consciousness can't be
reducible to physics therefore it can't be reducible to some
intermediate (computation) that risks be reducible to physics?
I know you cross swords with Bruno over the meaning of primitive in
this context, but I don't see why this has to be problematic.
Primitive simply means the level of explanation to which it is
*assumed* every other level can be reduced.
I think we only "cross swords" when Bruno implies that any reduction
of consciousness to physical processes is the same as asserting that
some (undefined) physics is primitive. The way theoretical physics
has developed over the last century I wouldn't be surprised if
someone developed a fundamental theory of physics based on homotopy
theory or knots or numbers. But it wouldn't affect the reduction of
chemistry to physics and it might not affect the reduction of
computation to physics. I take "physics" to mean the most
fundamental science of what we can empirically agree on - whatever
it's ontology is. I think Bruno only differs in that he thinks we
can agree on things that are not empirical, like arithmetic.
The point is not that we can know any particular theory of this
sort to be TRUE, but only that we should rigorously pursue its
consequences *as if it were*.
My point then is that we should start by treating a theory of
physical primitivism "as if true". If so, it is only consistent to
suppose that any phenomenon under consideration in terms of that
theory must be assumed to be adequately and fully accountable (at
least in principle) at its lowest level of physical reduction. You
persistently demur from this line of argument, but I think you miss
my point, which is entirely harmless in every case except (I
contend) that of the "1p part" of consciousness. It is entirely
possible to understand a physically-instantiated computation (and
hence, on CTM, an associated state of consciousness) to be the same
physical process regardless of the "level of reduction" at which it
is considered. After all, any such "level" is, in the end, merely a
term of art associated with the theory in question. But what I'm
questioning is whether it is coherent to (tacitly) treat the 1p
part as merely such a term of art.
?? Does "term of art" imply a third person description?
My sense is that you equivocate on this, because if we only
consider the 3p part (as in your analogy of the Mars Rover) the
point (i.e. 3p-reducibility) is indeed harmless. But the 1p part
resists 1p-reduction. It stubbornly is what it is. Hence my
question essentially is about the kind of theory required to make
sense of associating an irreducible 1p part with a reducible 3p
part. AFAICS such a question cannot even be posed coherently in
terms of physical primitivism.
Suppose that you have a theory which reliably tells you which 1p
qualia go with which 3p brain processes (the engineering level
solution to the mind-body problem). Does that "make sense of
associating an irreducible 1p part with a reducible 3p part."? Or
are you saying that such an association can't be a brute fact?
Indeed you have suggested that it is unreasonable to ask for this.
What inclines me to Bruno's ideas (assuming CTM of course) is that
this particular question may be better posed in terms of a theory
that takes computation, not physics, as its primitive.
Yes, I agree that it might be. But even if it provides a better
answer to the relation of qualia to brains, that doesn't make it a
better TOE if it can't solve the "white rabbit" problem. It will
just have traded one problem for another.
But that's what is interesting. UDA indeed does only that: it reduces
a part of the mind-body problem into a precise body believe problem in
arithmetic.
Then AUDA confirms, spectacularly I would like to say, that the
enterprise makes sense. We *can* have already conversations with
surprisingly smart machine, with rich cognitive abilities, like PA and
ZF are. It is "standard" classical mathematical logic.
Bruno
I think the remainder of your remarks equivocate on precisely this
3p-1p distinction, so I won't comment on them specifically. If I've
read you wrongly on this I'd be grateful for clarification.
No, I think that's a fair reading. As you note, I don't think it's
possible to have a 1p-reduction of the 1p part (if I understand what
you mean by 1p-reduction). I think the engineering level solution
may be the best that can be done.
Brent
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it,
send an email to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
http://iridia.ulb.ac.be/~marchal/
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
