John Wolfenden asked about system dynamics and general systems theory.
Hes right that there is a lack of clarity (to say the least) about how
they relate and how they differ. The initial confusion comes from their
common use of the dreaded word "system." Confusion is compounded by
their common interest in looking at the world more or less
"holistically." But system dynamics and general systems theory have very
different aims and they come from very different origins.
System dynamics, we should remember, started out as "industrial
dynamics." Then we had "R&D dynamics," "urban dynamics" and "world
dynamics," and finally just "any old system" dynamics. The field
emerged from servomechanisms engineering, focusing on the dynamics
endogenously generated by efforts to control dynamic systems. The focus
has always been on solving management problems in dynamic systems, from
inventories to corporate growth to ecosystem management to problems of
governing (note that word) almost anything. Practitioners have often
tried to invent other names for what we study -- the most recent is
McKinsey & Companys coinage of the term "business dynamics" (practically
a reissue of Forresters original).
General systems theory started out as an effort to propose testable
principles applicable in general to "the whole range of systems," from atoms
and molecules to solar systems (and, by the way, human systems). The aim
was theory building, not problem solving, although one could argue that
the purpose of good theory is to get ready to solve problems. The
origins of GST were the writings of von Bertalanffy (biologist) and a
group of scientists centered originally at the University of Chicago.
The great reference is James G. Millers "Living Systems," published in
1978, after much of the systems theory movement had played itself out.
The discussions in GST focused on the implications of concepts like
"open" and "closed" systems, "equifinality," stability, the structure of
generic homeostatic mechanisms, and so on. The discussions in system
dynamics focused on the (stock-and-flow/feedback) structure underyling
dynamic behavior, management policies intended to improve problematic
dynamics, naturally ocurring structures that compensate or compound
imposed solutions, and so on. The purpose and the resulting discussions are
very different.
Looking back, some smart people have made some serious mistakes trying to
put the two fields (lets say theyre fields) into common context.
People have said feedback systems are closed systems, and since human
systems are open systems then feedback must not apply to human systems.
That false conclusion comes from a confusion of "closed" in the GST sense
(no exhange of anything with the environment) with "closed loop" in the
feedback sense and "closed (causal) boundary" in the system dynamicists
endogenous point of view. Almost no system dynamics model is a "closed
system" since the little clouds on the ends of the flows show that stuff
is exchanged with the systems invironment. But to see internally
generated dynamics, perhaps in response to imposed management policies,
we intentionally draw a closed causal boundary around the system under
investigation and let the internal dynamics vividly appear.
People have fought over whether the word "feedback" should apply to
natural systems, like, say, a predator/prey system. Some in GST wanted to
reserve the word for information that is deliberately "fed back" to
control a system. System dynamicists have ignored the controversy and
have tended to use the notion of a feedback loop to stand for all manner
of causal loops existing or thought to exist in all manner of systems.
The biggest current confusion between the aims of GST and system dynamics
seems to me to be our current efforts to create "archtypical" insightful
structures. We should recall that there have always been hints of such
archtypical interests in the field -- Forresters Urban Dynamics tried to
capture the structure and dynamics of ANY city in a limited area, not a
particular city. In his acerbic book on systems analysis, Dan Berlinsky
ridiculed that aim in both Forresters work and GST. But our archetypes,
and our efforts to understand generic structures appearing in a number of
dynamic systems, do make us sound like GST. We have to be careful about
that.
The difference, I think, is that there is a grounded, problem-solving
focus to good system dynamics work. There are clients (imagined or real),
an audience, a problem to solve or a set of policies to explore, a focus
on implementation [in the best work that focus exists from start to
finish], and a drive to capture "what is really going on" in terms
completely acceptable to those who know the most about the system. And
good system dynamics work must go way beyond identifying system archetypes
in a given problem area, for those are but a few of the commonly occurring
patterns and they need to be focused tightly on the particulars of a
given problem.
These thoughts are much like what Wolfenden is finding in current GST
system dynamics literature. Personally, I think the distinctions are most
easily understood from the origins of GST and system dynamics (you might
guess that bias from Feedback Thought). In their origins they have very
little in common. What we see today in common goals seems to come from
the sometimes corssing meanderings of different intellectual paths of
the two fields. It does not help me a bit that GST and system dynamics
have the same word in their names.
[As Bob Eberlein noted in his hosts note, theres a lot of this in
Feedback Thought in Social Science and Systems Theory, and a lot of good
stuff in the special issue of the System Dynamics Review, 10(2-3) (1994).
Please note that the book is "Feedback Thought in ...," not "Feedback
Though." I think several versions of the system dynamics bibliography
went out without the final T, and even the esteemed and knowledgeable
host of this list left it out.]
...GPR
-----------------------------------------------------------------------------
George P. Richardson
G.P.Richardson@Albany.edu
Rockefeller College of Public Affairs and Policy Phone: 518-442-3859
University at Albany - SUNY, Albany, NY 12222 Fax: 518-442-3398
-----------------------------------------------------------------------------