>The writers of this article once again called in the trouble-shooting
>team from 'Complexity Analysts, Inc.' Citing Stuart Kauffman who
>spins his order-from-chaos and self-organization stories to explain
>this difficult phenomena. This is the same guy _Time_ called
>"suspiciously trendy" and "new age" and "at best inchoate but fruitful
>and, at worst inchoate and sterile" in their Dec. 28, 1992 article on
>spontaneous origin-of-life theories. Stuart Kauffman didn't get
>dis'ed in this article. Perhaps it finally dawned on the writing
>staff of _Time_ that if new complexity theories fail, they may be
>left without a theory.
I had read this article the day before John posted his review and was
also struck by the more positive treatment of the complexity crew.
As is probably apparent, I have found a lot to like in the complexity/
self-organization stuff, but there is also a lot of stuff not to like.
Unfortunately, a lot of complexity folks seem to be hanging their hats
on so-called self-organized criticality (SOC) as perhaps the general
principle of complexity that everyone has been looking for. This is
basically what Kauffman is talking about, although he doesn't mention
the key buzz words (SOC). The paradigm of SOC is Bak's sand piles,
which is mentioned (without Bak's name): "Rather, just as the tiniest
touch can cause a steeply angled sand pile to slide ...".
John Maynard Smith has been an outspoken critic of self-organization.
In February of this year, the Linnean Society hosted a debate between
Kauffman and Maynard Smith. Regarding Bak's sand piles Maynard Smith
commented "I just find the whole enterprise contemptible".
One of the highlights of the debate was the following exchange:
JMS: "My problem with Santa Fe, is that I can spend a whole week
there...and not hear a single fact"
SK: "Now that's a fact!"
Another was Kauffman's joke:
"What do you get when a deconstructionist joins the Mafia?
An offer you can't understand."
But the main problem with Per Bak's sand piles is that sand piles
apparently do not exhibit self-organized criticality. How could
such an embarrassing situation occur? There is a tie in here with
previous discussions on this relflector about the suitability of
models. It turns out that Bak decided to develop his ideas based
on a computer model of sand piles rather than by using real sand
piles. [This reminds me of the story (is it true?) of how Aristotle
developed his "theory" that women have fewer teeth than men without
bothering to look in his wifes mouth and count her teeth ;-)]. Recent
studies using real sand piles have shown Bak's model to be seriously
in error, no criticality. One author even suggests an experiment as
simple as slowly tilting a bowl of sugar is sufficient to refute
Bak's model!
This is not to say that self-organized criticality is not a real
phenomena. It may be. If it is, it is very unfortunate that the
principal illustration of the effect fails.
==
========================
Brian Harper |
Associate Professor | "It is not certain that all is uncertain,
Applied Mechanics | to the glory of skepticism" -- Pascal
Ohio State University |
========================