Re: Latest psychophysics/QM brain article

bjflanagan <wordsmyth1@xxxxxxxxxxxxx> writes:

>Khavkine:
>> You keep citing references which conjecture a direct relationship
>> between the functioning of the brain and quantum mechanics, to be
>> specific, a much more direct connection than, say, between quantum
>> mechanics and the flow of water in a stream. To me, these conjectures
>> remain as extraordinary as the first time I heard them. But I have yet
>> to see, in any of them, the "extraordinary proof" that would support
>> them.

>Flanagan
>No, it is neither more nor less direct. The working thesis remains: All
>material things consist of quantum fields. There is nothing
>extraordinary about this notion whatever. Thus, e.g., Hawking and
>Ellis: "the view of physics that is most generally accepted at the
>moment is that one can divide the discussion of the universe into two
>parts. First, there is the question of the local laws satisfied by the
>various physical fields. These are usually expressed in the form of
>differential equations."

>Hawking, SW, and Ellis, GFR. The large-scale structure of space-time,
>Cambridge, 1973.

Your response seems to indicate that as far as you know quantum
mechanics has no more to do with brains than it has to do with the
flow of water in a stream.

>Khavkine:
>> The only evidence that I see presented is that linear algebra is used
>> both in quantum mechanics and in NN models of the brain. Fine, this
>> establishes correlation (linear algebra used in both places), but says
>> nothing about causation (it's used in NNs because it's used in QM). In
>> fact, I addressed the low probability of such causation in a previous
>> post. There is insufficient evidence to make *any* conjectures.

>Flanagan
>You make no comment re: the precise correspondence between the
>mathematics of color superposition and the mathematics of the QM
>superposition.

You had earlier said:

>Flanagan
>I presented for visual inspection the facts of color (vector)
>superposition, side-by-side with an illustration of the concomitant
>wave (vector) superposition. These are facts which are public,
>reproducible, well known, and which (no doubt it's a coincidence)
>coincide mathematically.

Your agument appears to be (and please correct me if I'm
wrong):

1. Colours (that is, the experience of seeing colours) can be added
2. Wavefunctions, being vectors, can also be added
therefore
3. The experience of seeing colours might be a wavefunction.

You also make several references to electromagnetism and
its role within the nervous system, and appear to suggest
that the experience of seeing colours might be identified
with electromagnetic fields, which also have an additive
property.

Now, I should first point out that the addition of electromagnetic
fields and the addition of quantum states are not the same thing,
and you would need to clarify (for me at least) which of the two
you mean to identify with the experience of seeing colour.

Secondly, in some of your earlier posts, you emphasised
the importance of projective geometry, which is relevant
for the addition of quantum states. You hinted at the
idea that this was related to the fact that the set of
colours (at constant brightness) occupy a convex subset
of the real projective plane. There are two points to be
made here: One is that projective geometry is not relevant,
in the same way, for the addition of electromagnetic fields.
The other is that the set of colours occupies merely
a convex subset of the real projective plane, and is
simply connected, and hence has a different toplogy to
any projective space. The set of colours cannot, then,
be identified with, for example, the set of quantum
states of a system, since the latter has a different
topology to the former.

Thirdly, I have said to you many times that when colours
are added together, it is not the experience of seeing
one colour which is added to the experience of seeing
another colour. One ink is added to another, or one
beam of light is mixed with another, and it is that
addition (of lights and inks) which constitutes the
additive property of colours. Hence I can say that
adding red light to blue light produces purple
light, but the addition is not an addition of
experiences, but rather an addition of light.

I feel that this last point is most important, and, from
what I understand, it completely undermines your entire
argument. However, you have never once tried to address
it, despite my having brought it to your attention
many times. If you have a good answer to it, I would
like you to enlighten me, because what you suggest
would be very interesting if it were in fact true.

However, from my point of view, the likelihood that
it is true is very small, and this is not a reflection
of my mere closed-mindedness, but is a reflection of the
fact that there appears (to me at least) to be a gaping
hole in your argument, namely that you rely upon the
hypothesis that experiences can be added together
when it is in fact lights and inks which are added.

And if you once again refuse to address the question,
and continue to assert that the experience of seeing
colours should be identified with quantum states, or
electromagnetism, because they can both be added together,
then what inference would you expect me to draw from that?

You have also, on more than one occasion, in this
newsgroup, responded to requests, that you clearly explain
what you are talking about, by terminating the conversation,
claiming that you're too busy or offended to continue
dealing with such closed-minded entities. This is
hardly a way to appear persuasive or coherent.

>The simplest and most economical
>explanation for this predictive, quantitative relation consists in
>positing an identity relation between the "two" sorts of vectors, along
>the lines sketched out by Chalmers -- and also Feigl, Lockwood and me:

>http://wordassociation1.net/FieldWork.html

The addition which is reflected in the linear algebra used in
artificial neural networks corresponds to the addition of
membrane potentials coming in from various dendrites at
the soma of a neuron in the brain. Hence, the addition there
is the addition of an electromagnetic entity - electric
potential, and is mediated by ions of various types going
through various channels in the membrane. This is modelled
fairly well by the Hodgkin-Huxley equations, without
any reference to quantum mechanics. That is, Igor was
correct in his assessment that the linear algebra of
quantum mechanics was not the same as the linear
algebra of biological neural networks.

Indeed, the vector space in quantum mechanics is complex,
while membrane potentials are always real.

In your paper, "Are Perceptual Fields Quantum Fields?",
you say:

I suspect we will continue to encounter a measure of not
unintelligent resistance from those who, perhaps not wishing
to trouble themselves with the admitted difficulties of quantum
theory, may reply that, even if we do have a remarkable
correspondence between the mathematics of neural networks and
QFT (as would seem evident), we have no need to pursue matters
at the quantum level, because we can say all that needs to be
said at the neural level. We are not much moved by such replies,
which would seem to have more to do with intellectual inertia
than with logic or reason, let alone the spirit of scientific
inquiry.

Please rest assured that I've troubled myself with the difficulties
of quantum theory (and of neural networks), and that, having done
so, the putative "correspondence between the mathematics of neural
networks and QFT" is neither remarkable nor even evident. The
fact that linear algebra is used in both is entirely unremarkable,
since linear algebra is used in almost every field of human endeavour
which is subject to mathematical representation.

I should also point out to you that, while in quantum mechanics and
quantum field theories, the evolution of the system is entirely
linear, in the sense that the time-evolution of |a>+|b> is given
by the time evolution of |a> plus the time evolution of |b>, this
is not the case for artificial neural networks, in which a non-linear
function such as tanh or 1/(1+exp) is applied to the dynamical
variables in between successive applications of the linear
transformation. It is also not true for biological neural networks,
in which linearity of the addition of subthreshold membrane
potential fluctuations holds only up to a specific threshold,
at which the neuron "fires", or "spikes", and then resets to
a default membrane potential. In short, the evolution of
quantum systems is linear (in the wavefunction), while the
evolution of neural systems is nonlinear (in the variables
which characterise the state of the neural network).

Your response to the correct assertion that behaviour can be
understood at the neural level (i.e. we don't need to introduce a
quantum mechanical analogue of the Hodgkin-Huxley equations) appears
to consist entirely of name-calling, along with a declaration that
you yourself are not impressed.

However, if you have calculated the scale of quantum corrections
to Hodgkin-Huxley (which one would expect in advance to be exceedingly
small), then that would be interesting in its own right. If, for
some reason which nobody has yet anticipated (unless Roger
Penrose is correct), the addition of quantum corrections would
lead to significantly different effects at the level of the
behaviour of the organism, then scientists would take an
interest in the relation between quantum mechanics and behaviour.
Would I be right if I guessed that you haven't calculated any quantum
corrections, and that you have no intention of ever doing so?

>> > Khavkine:
>Is there anything that does not "embody quantum field processes" then?

>Flanagan
>All material things -- what about that don't you understand?

He was just asking for clarification. If you are saying something
coherent you should be happy when given a chance to clear
up misconceptions. The relevance of Igor's question is that,
if you really have no reason for claiming that quantum
mechanics is relevant for neural processes which doesn't
also apply to wheelbarrows, then the implication is that,
as far as you know, quantum mechanics has no more to
do with neural networks than it has to do with wheelbarrows.

That is, if you want to say that we should consider
quantum mechanics to be more relevant for neural
networks than it is for wheelbarrows, then you have
to say what relation quantum mechanics has to
neural networks which it doesn't also have to wheelbarrows.

Merely the fact that neurons are physical objects
does indeed establish a connection between them
and quantum mechanics, but, you understand, wheelbarrows
are also physical objects, and, as such, have as
much relation to quantum mechanics as neurons do,
if neurons are related to quantum mechanics merely
in virtue of their being physical objects, and
for no other reason.

>>From what you have written before, I would expect
you to respond to this by telling me that I'm too
closed-minded. However, whether I am closed-minded
or not, this is a genuine question, and without
an answer to this question, nobody has any chance
of understanding what you are talking about. It's
not even a complicated question. If you don't provide an
answer, or if you respond with insults or claims of your own
greatness, then it would appear to be the case that
you don't have any interest in clearly communicating
what you are saying, which raises the question of
why you began talking in the first place.

>Khavkine:
>> > The fact that ANNs successfully employ matrix transformations in many
>> > real-world applications is only a piece of corroborating evidence,
>> > albeit a large, multifarious and persuasive piece.
>>
>> Most likely a coincidence -- not at all persuasive to me.

>Flanagan
>Given the density of the matter at hand, that's not terribly
>surprising. I suggest you let the issue rest and busy yourself with
>other things -- billiards, perhaps. You may notice that the balls'
>motions can be described by vector mathematics, but no doubt that is a
>coincidence as well. Au revoir.

It's not the density of the matter at hand which is the problem,
but rather the apparently poorly motivated leaps of logic (x
is additive, y is additive, therefore y is x), the reliance
on quotations as evidence, the apparent confusion of
addition of electromagnetic waves and the addition of
quantum states, and the lack of any answer to
any question of substance.

For example, I've pointed out to you before that the real
projective plane is not simply connected while the set of
colours is, but you completely ignored it. I even went to some
trouble to make it clear exactly what it means for something
to be simply connected, and why it is that the set of colours
is simply connected. Was my explanation unclear, or did you
have an answer which would have fully explained how a thing
with one topology can be "identified" with a thing with
a completely different topology? If you did have an answer,
why did you keep it to yourself?

It seems that in the end, when genuine questions are
put to you which require any technical answer, you
ignore them or resort to quotations, or say "I'm
very busy doing important hard work, so go away
physicist. This is all too complicated for stupid
little you." What would your opinion be of somebody
who behaved like that?

R.

.

Relevant Pages

  • Re: The TOTAL disproof of Relativity and current QM
    ... > also completely disproves both Relativity and current Quantum Mechanics. ... In general, this is an equation for the *relativistic* mass, ... This is the *definition* of the Compton wavelength! ...
    (sci.physics)
  • Re: process physics (was Re: Aether is the empty space in which the Universe sits)
    ... >The important point he makes is that absolute motion is not incom- ... >different way of understanding the origin of space, the quantum etc. ... >This is far from complete as even Cahill readilly admits. ... I think understanding quantum mechanics in terms ...
    (sci.physics.relativity)
  • qm2 highlights 2001-2005
    ... undergrad QM, bell, entanglement, photon experiments ... of existing quantum mechanics theory seeking to resolve ... wolfram/fredkin, physics as information flow ...
    (sci.physics)
  • Re: process physics (was Re: Aether is the empty space in which the Universe sits)
    ... >The important point he makes is that absolute motion is not incom- ... >different way of understanding the origin of space, the quantum etc. ... >This is far from complete as even Cahill readilly admits. ... I think understanding quantum mechanics in terms ...
    (sci.physics)
  • Re: Quantum physics study - Where to begin?
    ... >>I am very interested in studying quantum physics but i don't really know ... > People will say, and for reasonable cause, that quantum mechanics is ... > Hamilton and Lagrangian), linear algebra, calculus and partial ...
    (sci.physics)