REPLY Forming a System Dynamics Institute (SD6566)

[SDMAIL Bill Harris]

Posted by  Bill Harris <bill_harris at facilitatedsystems.com>

"SDMAIL Kim Warren" <Kim at strategydynamics.com> writes:
>> In his Keynote address at the Society conference, Professor Forrester
>> made a persuasive argument that the progress of SD's influence might be
>> accelerated if it were not hampered by its status in relation to most
>> Universities' established departments [the establishment of Engineering
>> schools over a century ago may indicate a precedent]. It seems unlikely
>> that this situation will change any time soon, so a search for an
>> alternative might be fruitful. [I hope I have not mis-stated Jay's point
>> here].


Kim,

Allow me to express a different view, if you will.  I see much in
current business that seems to be an acceleration of the building of
intellectual silos: Six Sigma, BSC, etc. (I only name those two because
you did; I suspect we can all find others).  That acceleration seems
part of a boom phenomenon, driven perhaps somewhat by increased
specialization amongst people dealing from the mentality (policy) of
"growth is good" or "you grow or you die."  I think we all know what can
happen in things that (try to) grow forever.

Six Sigma is an interesting case, if only because I read an interesting
article in Fortune about it last summer
(http://www.facilitatedsystems.com/weblog/2006/07/growth-and-winds-of-change.html).
In particular, that article suggests that companies which have adopted
Six Sigma have fared less well than those which have not.  Building SD
up into a mainstream approach at the top of the hill risks putting SD in
that position and risks studies showing that organizations which apply
SD do less well than those which don't.

With all due respect to all that Jay and many others here (and
elsewhere) have accomplished, SD is "simply" feedback control theory.
Admittedly, we have to deal with strongly nonlinear phenomena, while
many of our engineering colleagues can get away with linearizing their
models or their systems.  Admittedly, we've established some skill in
creating practical models from rather complex situations and using them
to make important and impressive changes (as well as processes to do
that repeatably).  Still, it's control theory -- the stuff that Black,
Nyquist, Nichols, and others worked on in the first half of the last
century, except with little of the analytical side they developed thanks
to their focus on linear systems.

  Lest you think I'm picking on SD, I see Six Sigma as largely Deming's
  work of simple statistics with a dose of OD and other ideas mixed in.
  I see BSC as a good idea to put potentially useful measures in front
  of managers on a regular basis.  I think both may have suffered from
  what Fred Kofman has described as an ontological inversion: making the
  concept (Six Sigma, BSC, etc., and their certifications) more
  important than the issue (skillfully improving processes, working well
  with people, managing an organization).  That way, I think, lies
  danger, as it does with most (all?) reification efforts.

If we keep adding new disciplines (and I think we will, absent some
major transition), then, if we don't figure out how to incorporate SD
into our thinking processes (along with all the rest), we run the risk
of fewer and fewer getting better and better at something no one
eventually cares about.  When I was a practicing engineer, I wasn't a
control systems designer (even though there are such things); I was an
analog circuit designer, a digital circuit designer, a feedback
engineer, a test and measurement engineer, ....  Many of the things I
did wouldn't have been possible if I had only been one of those.

If we're to improve the quality of thinking, discourse, and progress
against some tough problems society faces, I posit that we need thinkers
who can blend together multiple approaches as needed to address the
problems at hand.  Yes, we'll need people with deep expertise in control
theory (SD) for certain problems, we'll need a broader range of those
with reasonable expertise in SD and other fields for a broader range of
problems, and we'll need a still broader range of those with some
appreciation for what we're doing but who lack the skill to do it.  But
what we need fundamentally is critical thinkers with a broad array of
tools in our toolboxes, not critical thinkers each armed with one 10mm
spanner.  (One challenge is that it is indeed more challenging to apply
SD than to apply a number of other tools.)

If I thought that we were in danger of losing the recipe for learning
how to do technically expert SD, I might agree that a special institute
could help.  As it is, I don't see that, and I think that we'll have
more of an impact if we focus more on making problems go away and less
on making SD grow big.  I do think it's appropriate to have people doing
research in and getting PhDs in SD; I think it's just important not to
make SD more important than it is.

I realize I may have glossed over important issues, and I welcome dialog
in developing a deeper understanding of these ideas that Kim has shared
and I have commented on.  In particular, Kim, I'm not stating that
you're wrong, but I am saying that I have a view that seems to differ.

Thoughts?

Bill
Posted by  Bill Harris <bill_harris at facilitatedsystems.com>
posting date  Mon, 03 Sep 2007 20:23:43 -0700