[From Erling Jorgensen (2005.01.20 2200 EST)]
(This post just launches off from the discussion of
MOL, & is more about how control systems are related.)
Bruce Gregory (2005.0120.2022)
The individual _is_ one system, at least in the model
I propose. I suppose I could consider a model in which
autonomous control systems are at war, but I'd like to
try a simpler version to start.
In my opinion, the task is tremendously simplified by
considering a hierarchical arrangement of "autonomous
control systems" (i.e., ECU's). To partition the tasks
of control this way seems much easier than bringing
on-line all-at-once one massive control system that
does everything.
I don't mean to overstate the ease, here, because
obviously we are having difficulty modeling complex
layers of controlled perceptions. But the control
task for any given layer is greatly simplified if it
can delegate the implementing details of control to
lower levels controlling their own (semi-)"autonomous"
perceptions. I also think that builds in a great deal
of robustness in terms of available degrees of freedom.
It is also my guess that this is the way evolution
stumbled on the amazing complexity we see in almost
any organism you'd want to pick. More complex ways of
controlling were built (blindly, or so we postulate
with E-coli reorganization) out of existing forms of
control. Why reinvent something when you can just
feed off the results?
I suspect that we have yet to fully appreciate the
implications of this way of linking control loops.
Once any kind of stable, negative feedback control
arises, extraordinarily complex chains of control
can emerge, stabilizing new types of collective
variables with new types of emergent properties.
Maybe this is a stretch, but I think the following
(admittedly diverse) situations exemplify this
process --
1) Isn't this the mechanism behind symbiosis between
different life forms?
2) Isn't something like this what we see with the
Krebs cycle of the body's metabolism?
3) And isn't this the feature that Jay Forrester &
the Systems Dynamics folks have built upon, when they
undertake to simulate comlex macro-phenomena, despite
not knowing many of the underlying details?
The point is, if you get a stable result from some
form of negative feedback, you can _use_ the result,
without having to know anything about the details of
where it came from. Whole ecosystems are constructed
out of that dynamic, aren't they?
I admit, that is not quite the same as constructing
a precise "hierarchy" of control, complete with setting
reference standards for those other levels of control.
And I do see by your final remarks that your interests
lie there --
No, I'm simply trying to build a model consisting
of hierarchically organized ECU's.
But couldn't a hierarchy of tightly integrated control
systems be the _special case_, of this same dynamic of
utilizing the (unknown) fruits of other control systems,
without having to care how the stabilized results of
those systems arose?
To me, this seems the key for understanding complexity,
not as "design", but as "emergence".
Does anyone else see these dynamics this way? Or am I
off in my own little eddy here?
Anyway, thanks for the opportunity to splash around in
the water...
All the best,
Erling
NOTICE: This e-mail communication (including any attachments) is CONFIDENTIAL and the materials contained herein are PRIVILEGED and intended only for disclosure to or use by the person(s) listed above. If you are neither the intended recipient(s), nor a person responsible for the delivery of this communication to the intended recipient(s), you are hereby notified that any retention, dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify me immediately by using the "reply" feature or by calling me at the number listed above, and then immediately delete this message and all attachments from your computer. Thank you.
<<<<RCMH>>>>