> no modeling session will bring success unless you are able to define a
>common problem to the group. In my opinion, this is where the first crucial
>mistake is made in modeling attempts.
I can point you to hundreds of modeling sessions that were successful, in
which the preparation did not include defining a common problem to the
group. The first crucial mistake is in assuming that you have to do that.
First, the semantic issue. A "problem" is someone's perception of
something. It is not a tangible thing in itself.
Second, the scope issue. Today, as is well-understood by people who read
Russ Ackoff's publications, organizations frequently have to deal with
what Ackoff labeled a "mess" back in the early 1970's. He defined a mess
to be "a system of interacting problems". Note that with this definition,
the word problem is not allowed to be used to represent the whole
situation. This is a very valuable distinction, because a problem is a
construct of one individual--not of a group. If you insist on specifying
the problem, you are asking others to accept your formulation.
Third, the utilitarian language. Rhoderick is certainly right in the
sense that you have to specify something. Here is what can be spelled out
ahead of time:
Within the broad scope, and the context that is defined specifically to
orient the group, and help decide who should be invited to take part,
people in the group will be free to specify "problems" that they perceive.
Ultimately, the entire mess can be structured, using Interactive
Since it's been done several hundred times, I feel no compunction about
saying that what I have just said is relevant.
-- John N. Warfield Johnwfield@aol.com