Re: A name I've never heard mention of...

["Tim Gwinn" <***>]

Per Bak was known for his idea of "self-organized criticality" (SOC), which
he felt was an important part of 'complex systems'. He also has a book on
the topic, which is available at Amazon.com.

I personally found the notion uninteresting in regard to Rosennean complex
systems. But perhaps I am misinformed on some point(s).

A math-free 1999 paper of his, "Self-Organization of Complex Systems", is
located here:
http://xxx.lanl.gov/abs/cond-mat/9906077

(A list of archived papers is at:
http://xxx.lanl.gov/find/cond-mat/1/au:+Bak_P/0/1/0/all/0/1 )

He sums it up on p. 2: "We assert that punctuated equilibrium dynamics is
the essential dynamical process for everything that evolves and becomes
complex, with a specific behavior that is strongly contingent on its
history."

The classic example of "punctuated equilibrium dynamics" is a sandpile where
grains continue to be added to it. Eventually the stable sandpile form will
reach a point of criticality and then avalanche (the "punctuation"). The
sandpile retains its "organized" state. The criticality points follow a
regular pattern or "power law".

Curiously, he asserts that at these points of criticality:
"Only at the critical state, does the compromise between order and surprise
exist that can qualify as truly complex behavior. There are very large
correlations, so the individual degrees of freedom cannot be isolated. The
infinity of degrees of freedom interacting with one another cannot be
reduced to a few. This irreducibility is what makes critical systems
complex."

This appears to follow the von Neumann attitude to complexity: if you can
just get "enough" Newtonian pieces put together in the right way, then it
will cross a 'threshold' into complexity. I do not know if he (or related
authors) ever describe in detail (mathematical physics) how this threshold
is supposedly crossed, or what physics occurs that specifically endows these
points of criticality with "infinite degrees of freedom" unique from any
other points of evolution of the system. Without that, this seems little
more than discovery of an interesting statistical law.

I think the underlying mentality is that systems that are thermodynamically
not at equilibrium are somehow inherently unstable or merely metastable.
Therefore, any non-equilibrium "organized" system is always on the verge of
collapse. What keeps it from collapse is that the system uses energy to keep
itself stable at these points of criticality. ("It is intuitively clear that
complex systems must be situated at this delicately balanced edge between
order and disorder in a self-organized critical (SOC) state." p. 3)

But all this rests on the notion that thermodynamically closed systems
constitute *the* standard for all physics. I agree with Rosen that
attempting to understand inherently open systems - like organisms - in a
context of closed-system thermodynamics is ill-conceived. (see EL ch. 12 &
16)

Regards,
Tim


> -----Original Message-----
> From: ROSEN Forum [mailto:*** Behalf Of Judith
> Rosen
> Sent: Sunday, July 27, 2003 9:30 AM
> To: ***
> Subject: A name I've never heard mention of...
>
>
> Does anyone on the list know who Per Bak was? He's touted as a "founder of
> complexity theory" but I've never heard my father mention him, and his
> "ground-breaking" paper was supposedly only published in 1987,
> which is long
> after my father was writing about Complexity. What is the opinion of list
> members of this fellow's contribution and how does it fit with Rosennean
> Complexity?
>
> Thanks!
> Judith
> Website address: http://www.rosen-enterprises.com/