Replying to LO23764 --
Since now two people have asked me what I meant by "dynamic models" I
guess it's time to reveal the "big secret" behind organizational learning:
Peter Senge was a student of Dr. Jay Forrester in the MIT system dynamics
program! Many of the seminal concepts underlying his conceptualization of
OL (though certainly not all of them) came from system dynamics. In my
humble opinion, anyone who tries to work with OL without at least a
rudimentary familiarity with system dynamics is missing the most powerful
aid to understanding (and improvement) going nowadays.
If you're unfamiliar with system dynamics modeling, I recommend you begin
with Jay Forrester's "Industrial Dynamics" which, although it was written
40 years ago it is the seminal book for the entire field and remains a
brilliant, clear introduction to the methods and promise of system
dynamics, crammed with practical and eye-opening examples of its
application.
I believe one of the first challenges in OL is to capture learning that
has already occurred in an organization, and most of it occurs without
organizational intent (or, at least, without purposeful organizational
intervention.) I used to go into organizations and, with "fresh eyes,"
just observe how work was processed (and talk with people, asking them to
describe how they did things and on what bases they made decisions, etc.)
I would diagram what I learned in a kind of flow diagram and present it
back to the very people I had observed and talked with -- and they would
be astounded!! (while, at the same time, agreeing that the diagram was how
they did things.)
One of the ongoing "arguments" I've been having with KM'ers on the LO list
is about this kind of learning: if the people who actually create a work
process are not aware of its dynamic structure, "where" in the
organization does this "knowledge" reside? Is it useful to refer to this
as "knowledge" at all?
My answers:
1. It resides nowhere except in the dynamic structure of the work process
itself.
2. Perhaps (if we agree to expand what we mean by "knowledge.")
A simple workflow diagram such as thing is one way to capture and preserve
and transmit the structure of work processes, of course -- but there's an
even better way: a system dynamics ("stock and flow") model. Not a
causal loop diagram, although they have some educational (transmitting)
uses, but a stock and flow (or level and rate) model that can simulate the
current critical "behavior" of the process or system of interest. So, by
"feedback structure" or "dynamic structure" I mean precisely the sequence
of levels and rates (and their defining equations) that are required to
"explain" (i.e., to simulate) the behavior of an organizational process
or system.
I'll mention just one system dynamics idea that's most relevant here:
What appears to be "evolution" in an organization's behavior may not be.
That is, the "same" system can generate quite different (or different
appearing) behavior in different circumstances or at different times.
For example, suppose you put money into an investment that compounds
interest. If you look at a graph of how the money grows in the first
several years, you'll see what seems to be almost a flat growth rate; a
couple of years later the growth will seem to be more rapid and nearly
linear; but a few years later the growth will look to be explosive!! This
can appear to be the result of evolution (fundamental changes in
structure) of the investment "system" -- but it is not. The system has
stayed exactly the same.
System dynamics models can differentiate between changes due to changed
structure and those that just occur in existing structure. I think that's
a very useful and important distinction, don't you?
Does this help at all? Please feel free to write again.
John W. Gunkler
jgunkler@sprintmail.com
--"John Gunkler" <jgunkler@sprintmail.com>
Learning-org -- Hosted by Rick Karash <rkarash@karash.com> Public Dialog on Learning Organizations -- <http://www.learning-org.com>
"Learning-org" and the format of our message identifiers (LO1234, etc.) are trademarks of Richard Karash.