I have followed this discussion with most interest. However, even though
I feel very comfortable with what System Dynamics is (having also read
various definitions in the literature over time), I have the feel that
the problem of achieving a widely shared and accepted definition has to
do with its requirements. Yes, like a model, a definition, in itself,
must have a purpose.
Why are we defining? To be able to understand ourselves what the concept
is about? To re-organise and bring rigour to our current understanding
and make this explicit? To distinguish the concept from similar ones? To
transmit our personal understanding of the concept to someone else? To
sell the concept to a potential buyer, making it look helpful to a
particular problem, and within a particular context? To discredit the
concept if we do not believe it, dislike it, or see it as a competitor
to our alternative concepts? To *discover* what the concept *really* is
and make *the truth* accessible to any human being willing to learn
about the concept? (and does it make sense?)
Different needs will naturally drive different definitions. In this
discussion, I have read some that focused on the need to solve a
certain category of problems. These definitions look attractive to one
living such problems, but lack rigour, and restrict the scope of the
potential audience. Other definitions, instead, searched rigour and a
generic scope: they look attractive to the thorough thinker, the
scientist, or the thorough modeller, and have the potential to fit
within a wide range of applications. However, they also require
abstraction and, thus, a kind of intellectual thinking with which most
of the "Client" audience can hardly cope. Finally, other definitions
sought support in the concept of "modelling", borrowing robustness form
it; these lack the ability of clearly differentiating System Dynamics
from other modelling techniques.
So, like a model, does it make sense to propose a definition without a
purpose? If the definition of System Dynamics can serve various
purposes, then the purpose is, in itself, one dimension of the
definition, which we often forget to mention -- I felt better after
reading Eric Wolstenholmes definition.
I particularly liked the following statement:
"Things today are the things of yesterday plus any changes. The changes
are the result of the things of yesterday. Now extend this to
tomorrow."
From:
bonnell@kodak.com
This raises the issue that System Dynamics is one way, among many, of
thinking about and linking the past, the present, and the possible
future. Most models do this, and so does our mind everyday. So, System
Dynamics becomes as particular way of thinking and perceiving the world;
one that brings a specific rigour to this thinking, and one that may
help us feeling more comfortable about how we perceive the world, and in
particular the future. As we learn the discipline, we adjust our
framework of thinking to it, and if we feel better with we adopt it as
one of various ways of understanding reality.
Then I was struck by the eternal need of selling, and the consequent
imperative that a definition that does not sell does not survive. Whilst
true, I wondered whether the concept would survive even without really
helping people, but simply by selling. I wondered whether System
Dynamicists would become the "Leornardo DaVinci"s in the
modelling/management arena, able to sell beautiful models that managers
would gladly buy because they look nice, speak their language, say what
they like, and, more dangerously, help them to reinforce their
long-cemented ideas and impose these over the others. Personally, I
feel, quite strongly, that the answer should be no. A major role of
System Dynamics is precisely to induce a change in mental models towards
increased consistency. System Dynamics has the great virtue of forcing
people to think (often how wrong they might be), as opposed to relax the
need to think deeper. This makes the role of System Dynamics a most
challenging one in the arena of real practical management. Good
definitions of System Dynamics must support the mission.
Alexandre Rodrigues
From: Alex Rodrigues <
Alex.Rodrigues@PA-Consulting.com>