I feel that you do not understand the real problem at hand. I am now going
to
try my best to explain it to you and hopefully you'll then understand.
You wrote
> Your argument from the beginning seem predicated on your thinking that
> a digital system is well-defined whereas an analogue system is 'random'
Now read carefully what follows. I have been trying to put the idea across
that
a neuron cannot process information in an analogue manner USING graded
responses. This is because of the randomness produced during each graded
response cannot be replicated when needed. I am tired of giving examples for
how
to make a neuron analogue (see earlier postings) - can you think of one? No
one
has yet given a suitable candidate(s). Remember, the states of the
continuously variable
quantity that you choose should be easily reproducible - for example, if you
choose the total number of ions taking part in a graded response, which in
turn
produces a 'specific state' in the neuron, then do you honestly believe that
the exact
same number of ions can take part in a future graded response to produce the
same 'specific state' when called for? [See my last posting on the workings
of an
analogue system, the snipped parts].
The important thing to remember is that a neuron is unbelievably dynamic
with
respect to ion concentration, membrane permeability, protein synthesis,
protein
transport etc. So much so, that one state at a single point in time may
never be
reproduced in its remaining lifetime! In conclusion, how can information
processing in the conscious human brain rely on a system that most likely,
produces unique conditions that exist only once, momentarily in time?
Going back to your quote, it is blatently obvious that you have no idea
about
the suggestions that I have been making, if this is my fault then excuse me.
When I am talking about a digital system and that the input/output mechanism
is
well-defined, I am talking about AP generation. Anybody knows that an AP is
an
all-or-nothing response - it either happens or it doesn't - to quote once
more 'there
is no messing around'. In this sense, a neuron has the means to create a
well-defined communication network, so that the AP are generated when needed
for
cross-talk and information processing. I think we need to get away from the
idea
that a single neuron can process information on its own, instead we should
now
begin to focus on neuronal networks within the brain, whether they are small
or
large. At this level, the well-defined level with respect to AP, a coherency
is
observed that can never be seen in an analogue neuron. And yes, AP
generation is
in binary form, but what is the main difference between this binary state
and
that found in computers? Time, of course - the frequency coding concept, the
way these ON and OFF events are able to be produced in trains of action
potentials with differing time intervals between each AP. And remember, the
information contained within these 'constructs' could equally be found, not
only
in the AP itself, but also in the time between AP, i.e. both ON and OFF
events.
You can now see that the amount of information that can be contained within
a
system like this is almost limitless, but unlike an analogue neuron, easily
reproducible when needed. The problem we as scientists face, is the
inability to intrepret
these signals - indeed, we are still unable to track these constructs in the
smallest neuronal network in situ, let alone be able to understand them.
There is never any discrepancy, either the
neuron produces the AP or it doesn't, and therein lies the key to how this
system works, a neuron does not generate an AP but waits until it's forced
to or
has to (how this is achieved is a big mystery - I don't mean the
biophysics -
but the actual reason behind its generation).
Fortunately, evidence supporting information processing in this manner
comes from the use of drugs. In depression, or other neuro
conditions, drugs act to suppress or disrupt AP generation by many different
molecular
mechanisms. Susan Greenfield gives a vivid explanation in her latest book,
about
how these drugs suppress neuronal communication, and thus influence the
emotional and/or conscious state of the individual. Albeit, these drugs may
work
on the molecular scale, but again, the argument I present suggests that no
coherent pattern of graded responses could be reproduced when needed, and in
fact, the introduction of exogenous drugs that disrupt the molecular ebb and
flow would only increase the apparent randomness of the analogue neuron.
The original argument way way back, was that the brain processed information
only in an analogue manner. This was then redefined to include binary
processing
on a larger scale, so that an analogue/binary system was developed. However,
by
using my argument, I suggest that the brain was never a true analogue system
in
the first place because no one has yet given a suitable candidate for the
physical
variable required for an analogue neuron. Instead, if you think about it,
the
variable that I have introduced is TIME, so in that case, my proposed
continously variable physical quantity was the fourth dimension. Indeed,
using this
concept one could say the brain processes information in a
psuedoanalogue-binary
manner (is time physical? -hence pseudo). Whether this gives rise to the
conscious human brain is always in
question, nevertheless it removes the unimaginable complexity faced with
when
dealing with analogue neurons.
T.L.S.
For those who believe,
no explanation is necessary.
For those who do not,
none will suffice.
Richard Norman <rsnorman at mediaone.net> wrote in message
news:Ecdr5.11798$_e4.468005 at typhoon.mw.mediaone.net...
>> "Theophilus Samuels" <theophilus.samuels at btinternet.com> wrote in message
> news:8ojiam$3cq$1 at plutonium.btinternet.com...> >
> >.....<snip a lot of stuff> The key thing to
> > remember here, is that the digitial input/output mechanism is a
> well-defined
> > operation - it either happens or it dosen't, there's no messing around
[I
> > will come back to this again later>
> > .....<snip another lot of stuff > Now here
> > lies the difference between the randomness of an analogue system based
on
> > graded responses (this was the original argument) and .... <snip a lot
> more>
>> Your argument from the beginning seems predicated on your thinking that
> a digital system is well-defined whereas an analog system is "random".
You
> have included this type of statement now in several of your posts.
>> This is non only non-sensical, it departs from everything we have learned
> about the universe in the history of science. There is underlying reason
> why
> a graded or analog system cannot be determinate. The mathematics of
> partial differential equations well describes a tremendous variety of
> phenomena that are continuously variable in space and in time. In my
> mind, digital systems with probabilistic state transitions are far more
> "random" than the clockwork machinery of the Newtonian universe.
>>>