"John Wilkins"  wrote in message
news:c7rcpi$kgo$1{at}darwin.ediacara.org...
> Guy Hoelzer  wrote:
>
> > Hi John,
> > in article c7hmdk$hu9$1{at}darwin.ediacara.org, John
Wilkins at
> > john_SPAM{at}wilkins.id.au wrote on 5/7/04 9:07 PM:
> >
> > > Guy Hoelzer  wrote:
>
[snippage]
.....
> Allow me to interrupt for a second. It is a common
criticism of
> "reductionism" that they must insist that we refer *only*
to the
> components of an entity to explain its behavior. This is
both false, as
> you note, and unnecessary. A *full* description of a
system must take
> account of both the componential properties *and also* the
initial and
> boundary conditions. So cognition must involve also an
account of the
> past history of the neurones, the inputs they have
received and are
> receiving (both sensory and cultural), and of course the
evolutionary
> heritage of the body that incorporates them.
>
> > clear that cognition is not even manifested in part at
the level of a single
> > neuron, that the properties of water (e.g., surface
tension) are not even
> > manifested in part by a single water molecule, and that
pressure is not even
> > manifested in part by single molecules.   All of these
phenomena emerge
>
> As to water, the microstructure of water is eminently
computable, and it
> generates surface tension effects as well as the other
factors of
> solutes and the like. It's an exciting (and beautifully
reductionistic)
> field.
>
> > through the interaction of parts at reduced scales.  It
is utterly
> > irrelevant that you may have a reductionistically-based
sense of
> > understanding as to how the parts might interact, and be
able to explain the
> > phenomena that emerge in collections of interacting
units.  This is why I
> > have always shied away from unpredictability as part of
the definition of
> > either emergence or complexity.  Of course these things
are predictable, in
> > principle, if you know enough about the qualities of the
interacting units
> > and the environmental context in which they are to
interact.  This is why
> > computational modeling is an effective way to study
emergence in complex
> > systems.
>
> Then your emergence = my reduction.
[snip]

I also agree that in talking about "complexity,"
"emergence," and even reductionism, one should shy-away from
ideas of unpredictability as a defining factor and
concentrate on "in-principle."

One little note to your examples of neurons and the
micro-structure of water though. While it is easy to say
that the "boundary conditions" in reductionism need to be
taken account of, I'd just like to emphasize that this is
not at all so easy in the real world--certainly not when
there are very many factors, and/or where a number of them
are not even imagined.

Taking the water example, all the reductionist theory in the
world on molecular water doesn't help when your water has a
unknown trace of sodium metal dropped in it--not without a
whole lot of added theory one had not anticipated relating
to other possible elements, compounds, substances, (or a
zillion other things,) and unknown conditions present that
are a factor.

Similarly for the case of neurons, one has to at least
provide for all the possible different variations in
_function_ under a variety of differing
conditions/circumstances...or again, the presence of a
zillion other substances which may, or _may not_ be present
at a
given time, viz. biochemicals such as known or unidentified
neurotransmitters.

My only point is neither the Holistic nor the Reductionism
approach are always simple--and both need to be used in
theoretical research, with a careful determination of the
_all_ the important factors that affect the process(s) under
consideration. Also, _function_ under different
circumstances, is extremely important. As always though, the
devil is in the details.            ...tonyC

> >
> > Cheers,
> >
> > Guy
>
>
> -- 
> Dr John S. Wilkins, www.wilkins.id.au
> "I never meet anyone who is not perplexed what to do with
their
>  children" --Charles Darwin to Syms Covington, February
22, 1857
>
---
þ RIMEGate(tm)/RGXPost V1.14 at BBSWORLD * Info{at}bbsworld.com

---
 * RIMEGate(tm)V10.2áÿ* RelayNet(tm) NNTP Gateway * MoonDog BBS
 * RgateImp.MoonDog.BBS at 5/13/04 1:39:48 PM