"The Next Step" CD from High Performance Sys LO12749

Phil (podence@hps-inc.com)
Tue, 4 Mar 97 06:44:38 -0500

Replying to LO12733 --

>From Peter Marks' reactions to Dennis M's comments on "Systems Thinking:
The Next Step."

>I think the real *initial* lesson that needs to be learned - particularly
>by a team - is that to be worthy of trust, a model's *structure* must be
>open to arguing over, modifying, and testing. The really successful
>'conversions' I have seen occurred only when the results of successive
>attempts to model an objector's trial counter-hypotheses were seen to not
>accomplish the desired ends.

I quite agree with Peter. However, what many in the SD/ST field have found
after many years is that it is just plain hard to get teams to the point
where they are able and motivated to go through this incredibly valuable
process. Most folks I have encountered who see the power of Systems
Thinking are looking for ways to introduce it to others in their
organizations, so that they can achieve the kinds of "successful
'conversions" to which Peter refers. That was the motivation for The Next
Step.

>Unless I overlooked some critical option, this kind of packaging simply
>propagates the illusion that system thinking is about *running* a
>(presumed correct) model with different parameter sets. I would have
>thought that the now-decades-ago reaction to "Urban Dynamics" and "Limits
>of Growth" would have laid this one to rest. In this sense a literal
>flight simulator *is* better than Next Step because its underlying model
>is well-accepted.

It's worth making a distinction between interacting with a flight
simulator or Learning Environment and running a model in ballistic mode.
Enabling users to make periodic decisions based on model conditions
effectively puts them in the middle of various feedback loops. Rather than
the relationships in those loops being "presumed," they're created by the
users policies. For example you might discover and embed in a model the
assertion that training gets cut back when money gets tight. A ballistic
model would let one explore the implications. However, users interacting
with a model and finding *themselves* cutting training every time money
gets tight will actually "feel" the policy and its implications.

Peter's point still holds, model builders always learn more than those not
involved in the model. But well-crafted Learning Environments can help
those others to recreate the key insights achieved by the modelers.

Phil Odence
High Performance Systems
45 Lyme Road, Suite 300
Hanover, NH 03755
voice- 603 643 9636 x107, fx- 603 643 9502, web- http://www.hps-inc.com

-- 

Phil <podence@hps-inc.com>

Learning-org -- An Internet Dialog on Learning Organizations For info: <rkarash@karash.com> -or- <http://world.std.com/~lo/>