Posted by Jack Harich <
register@thwink.org>
SDMAIL Jay Forrester wrote:
> On Jun 26, 2008, at 6:41 AM, SDMAIL Jack Harich wrote:
>
>> For example, unless I missed something, where are the goal suggestions from
the founder of the field,
>
> In my talk at the 2007 annual conference of the Society, I discussed the
present plight of system dynamics and my vision of how we could proceed over the
next 50 years. The paper is published in the Summer-Fall 2007 issue of the
System Dynamics Review.
> So far, I have seen almost no discussion or debate about the issues raised in
that paper. I hope that paper can serve as a basis for considering the future
of
the Society. Can we hear from members on it?
Jay,
I'm amazed there was not intense discussion at the conference when you gave the
talk. There were no questions during or after the talk? No one button holed you
in the halls and excitedly asked about your assessment and suggestions? After
the
conference was over, no one did the same?
I read ""System Dynamics - The Next Fifty Years"" when it came out in the SD
Review. Of all the articles in the 50 year anniversary special issue, it was THE
article to think about. I remember coming to the conclusion that this has either
caused a big buzz or is the result of one. But now we hear from you that is not
the case. There has been ""almost no discussion or debate about the issues raised
in that paper.""
Surely there was some discussion, because we are now embarked on trying to solve
the very problem your talk and paper covered. But there has not been enough.
Below are my comments about the issues you raise in the paper. Since this is a
goals setting thread, I will also attempt to extract the many goals mentioned.
*The Problem*
The problem you address is SD has failed to ""realize its potential. ... We are
now at the same state of advancement that engineering was when MIT first opened
its doors in 1865, or that medicine was in the late 1800s when the Johns Hopkins
School of Medicine was established. ... At present, with system dynamics on a
rather aimless plateau, the field seems to be catching its breath. ... Because
of
this stagnation, we see disappointment and frustration.""
Solving this problem needs to be expressed as a goal or goal set. This would be
the topmost goal of the problem solving project. That's the kind of goal Kim has
been looking for in this thread. Nowhere in the paper do I see a clear,
quantifiable top goal. Perhaps you had one in mind but didn't include it?
For example, my topmost goal suggestion was ""the ability to reliably solve
large,
pressing social problems."" The reliability level desired is easily quantified,
such as 95%. This goal does leave out cost and speed (the iron triangle of
function, cost and speed), so it is incomplete.
It's easy to see (from your words above and the results of the field) that the
proximate reason the field cannot yet reliably solve large pressing social
problems, as well as the easier ones of business management (really a social
problem type), is that the field is immature. Why it's immature we don't know.
How to solve the problem we don't know. So let's call this the System Dynamics
Immaturity Problem.
*Causes*
The paper hypothesizes several causes of the problem. This is important, because
if one doesn't resolve the underlying causes, then a problem cannot be solved
except for temporary quick fixes. The causes you present (I may have missed
some)
are:
Cause 1 - Low level of training - ""Most of the academic programs have stagnated
at the level of introductory courses taught to students who have no expectation
of developing expertise in the powerful professional field of nonlinear feedback
dynamics. We are turning out more and more people who are led to believe that
they have been taught system dynamics but who have only a superficial and
unworkable preview of the potential of the field. ...many people enter the field
without the training that would allow them to reach the full potential of system
dynamics.""
Cause 2 - Not enough innovation - ""There is little evidence of a strong reach
into new territory.""
(I'm not clear on this. What is new territory? Is it new ways to apply SD or
attempts to add to SD itself?)
Cause 3 - Corruption of the original idea - ""We see many people trying to 'dumb
down' system dynamics into 'systems thinking' and 'causal loop diagrams', which
lack the power that is inherent in system dynamics. ... [also to be avoided is]
using system dynamics for forecasting and placing emphasis on a model's ability
to exactly fit historical data.""
Cause 4 - Reasons that ""stand in the way of a new series of powerful books"" -
These are ""Lack of courage in the field to open oneself to severe debate and
criticism."" and ""The field has a very small number of people capable of such
publications."" and ""There is an assumption that expensive sponsorship must
precede an effort to address important issues."" and the dumbing down mentioned
in
cause 3.
Except for cause 2, these causes are shallow. They are intermediate rather than
root causes.
I see SD as one of many tools that make up the emerging field of social system
engineering (SSE). This is a more productive viewpoint, because the name of the
field incorporates the goal of the field: the ability to reliably, efficiently,
and quickly engineer social systems from scratch or solve existing social system
problems. The foundational tools would, I suspect, include a repeatable process
for solving SSE problems (like all mature engineering fields have), SD,
memetics,
reusable parts like those in electronics and architecture, and certain
principles
that would allow instant easy insight into complex social system behavior. These
principles would be similar in importance to the key principles of other fields,
like the three laws of motion, the concept of electricity, the molecular
composition of matter, the evolutionary algorithm, the Periodic Table, etc.
Thus to think and argue in terms of only SD is parochial. It is too limiting a
mindset. It constrains us to such a small pasture that we will never get a big
enough picture (and a comprehensive enough tool set) to solve the big hairy
audacious problems that are the true test of any emerging field.
The field of social system engineering is immature. I'd estimate that less than
10% of its major discoveries have been made. What these will be no one knows.
Given this immaturity, how can cause 1, a low level of training, possibly be a
root cause? How can a high level of training, such as at the phd level, train
anyone in how to reliably solve complex system social problems when how do that
is only 10% known?
Given this immaturity, cause 3, corruption of the original idea, cannot be a
root
cause because the original idea is incomplete. If it were not corrupted it would
still be immature.
Given this immaturity, consider cause 4. How likely is it we will see more books
like Urban Dynamics, World Dynamics, and The Limits to Growth? Those books were
a
combination of flukes, reliance on the brilliance of the founder of the field,
and picking the low hanging fruit. The last is probably the largest factor. That
more such books have not been created is not due to the four reasons listed.
It's
due to the fact that SSE is immature. In fact, it is proof SSE is immature.
Let's be honest. SSE is so immature that it has never solved a mega social
problem. The urban decay problem was never fully solved. It was only downgraded
from a crisis (in America) to a major, festering problem. We still have large
slums with high levels of crime, poverty, and alienation all over the world. We
still have occasional urban riots, even in first world countries such as France
in 2005 and 2007. The environmental sustainability problem was never solved. The
World2 and World3 models, as magnificent as the were, only identified the
problem.
Let's look at cause 2, not enough innovation. Assuming this means not enough
innovation of the field itself, then this is true. The discovery rate is too low
to bring the field to maturity. For a new promising technology that is still
immature, business calls this the ""not enough R&D"" cause.
Why not enough productive R&D? That's a good question.
*Solutions*
The paper presents various solution strategies as goals. These are:
Solution 1 - More influential books - ""a new series of powerful books to address
the important issues that dominant the newspaper headlines."" These would be
books
conveying SD analyses. There have been only three such books to date: ""Urban
Dynamics, World Dynamics, and The Limits to Growth. ... We need books addressed
to the public that are understandable, relevant, important, and dramatic.""
This is an impossible dream. As explained above, this cannot presently be done
because we have already picked the low hanging fruit and the field is so
immature
it cannot be reliably applied. This solution is an unrealistic short term goal.
Solution 2 - Better quality work in system dynamics - ""System dynamics is still
far from reaching the quality of work to which we should be aspiring. We need to
begin debating how to raise quality and scope in applications, published papers,
and especially in academic programs.""
If a field has only 10% of the foundational discoveries it needs to be mature,
how is raising the quality of the application or teaching of the 10% that does
exist going to result in success? It can't. This is like the general who urges
his ill-equipped troops on to win a battle. Their desperate charge fails,
because
they have only 10% of the arms and ammunition needed for the job. Better would
be
to stop the battle, go find what the troops need, give it to them, and then
return to the battle. The probable outcome would be significantly different.
Solutions 1 and 2 are symptomatic. More influential books and better quality SD
work would be symptoms that the basic capability of the field has dramatically
improved.
Solution 3 - Teaching SD in K to 12 - ""The goal is to have a cumulative program
from kindergarten through 12th grade that builds a systems background.""
This is appealing but flawed. SD is immature. The age at which it is learned
will
have little effect on that immaturity. I agree that thinking in terms of dynamic
structure is so difficult it's best learned at a young age. But this alone is
insufficient.
Have any people who were taught SD starting at a young age made breakthrough
contributions to the field? Have they turned out to be SD wunderkinds who can do
high quality analyses in record time? Probably not. This points to the fact that
it's not the age the tool is learned that is a critical factor.
Reading the Spring 2008 Creative Learning Center newsletter, I see progress has
been meager and temporary.
Solution 4 - SD centric management schools - ""A future school of corporate
design
would break the boundaries between disciplines.... We should think of management
education being more like that in engineering, with an undergraduate program
followed by a graduate program.""
But what will the students be taught? Again, this would fail because there is
simply not enough practical ability of SD to solve corporate problems, compared
to competing methods.
If social system engineering was mature and worked in the business world, there
would be a strong demand for trained practitioners. We would not have to create
such schools. They would spontaneously appear. Thus this is another symptomatic
solution.
Solution 5 - Reusable business structure patterns - ""Suppose we had 20 generic
structures that would cover more than 90 percent of the situations that a
manager
ever encounters. ... Each such generic structure would require a separate
textbook; each would be studied at least to the extend of a full semester
subject.""
A beautiful and productive vision. I like this. It directly addresses the
immaturity issue.
Solution 6 - A detailed plan - ""A first step would be to create a plausible
50-year detailed plan for the future of system dynamics.""
The success of all plans depends mostly on the strategies they are based upon.
If
the plan would implement solutions 1 to 5 it would fail, because except for
solution 5, these do not resolve the underlying cause of the problem. Even 5
appears insufficient, because there is no proof it would increase the maturity
of
the field by more than a small amount.
Are major business problems really that generic? If the 20 some structures were
created and taught, wouldn't competitive corporations digest them and then
quickly evolve to where other structures made the critical difference? If SD has
identified only one such structure in 50 years, how long is it going to take to
identify the rest? I love the idea Jay, but pragmatic skepticism prevails.
Solution 7 - Good leadership/promoters - ""The future that I have suggested
depends on there emerging a leadership group in the field that I have not yet
been able to identify. We need a group of full-time, enthusiastic, charismatic,
visionary, energetic promoters.""
This has some potential, but not in the promotional direction. At the risk of
repeating myself too many times, this solution relies on there being something
worthy of promotion. This is never the case for an immature product.
Rather the good leadership angle has potential. I don't think ""hundreds of
millions of dollars"" is necessary or even possible, in the short term. If the
leadership got behind a deep analysis and focused the Society's limited energies
on what that pointed to as worth doing, we would get results.
*Where's the model?*
This leads right into a main point I was planning to make: Where is the model
that would allow a sound look at the last 50 years and a successful approach to
the next 50 years?
In your ""Quality of Work in SD"" section, you make this insightful point:
""How often do you see a paper that shows all of the following
characteristics?
1. The paper starts with a clear description of the system shortcoming to be
improved.
2. It displays a compact model that shows how the difficulty is being caused.
3. It is based on a model that is completely endogenous with no external time
series to drive it.
4. It argues for the model being generic and descriptive of other members of a
class of systems to which the system at hand belongs.
5. It shows how the model behavior fits other members of the class as policies
followed by those other members are tested.
6. It arrives at recommended policies that the author is willing to defend.
7. It discusses how the recommended policies differ from past practice.
8. It examines why the proposed policies will be resisted.
9. It recognizes how to overcome antagonism and resistance to the proposed
policies.""
This is a great checklist.
But shouldn't we practice what we preach? Your paper attempts to solve a complex
social system problem where performance over time is the issue. So why have you
not followed the above checklist? Above all, where is the model in your paper?
The fact that there is none means we are doing just what too many others are
doing - we look at a difficult problem and come up with intuitively derived
causes and solutions.
Building this model is not easy, or you or others would have done it. The fact
that it is so difficult, and done so seldom, is one more proof that SD is still
too immature to have widespread applicability.
*Conclusions*
You said ""I hope that paper can serve as a basis for considering the future of
the Society.""
It's an excellent paper for getting the creative juices flowing. It raises the
call that we need to do something or SD will fade out of relevancy, with
probable
dire consequences for humanity.
But as the source of the topmost goals we need or the beginning of a strategy,
for me it falls short, because its central conclusions are all intuitively
derived. Similar papers could be written extolling entirely different goals and
strategies, and would sound just as plausible. That is, until one analytically
examined them.
But suppose we treated this paper as a first iteration. It has a lot of value.
What if we got behind it and listed its pros and cons, and then worked on those
cons?
Then your efforts over the last 50 years, and on this paper, would not have been
in vain.
SDMAIL ing. Augusto Carena - Dharma srl wrote:
> Why not trying to face the problem in a sort of ""SD style"", I mean building a
cooperative model (or models, or even antagonist models) of SD devlopment,
trying
to formally identify and understand the underlying dynamics?
>
> While I understand that is not too orthodox doing it without a clear statement
of goals (very ehterogeneous, as far as I can see), and I'm aware of the risk of
starting a Babel Tower project, it seems to me that the process itself, and the
methodological questions it would raise, could be useful to generate hints,
verify feasibility, and clarify the different positions through a common and
shared (I hope) language.
You must be a mind reader.
This is exactly what I hint at in my reply to Jay
Forrester.
By the way, this has already been suggested a few times. But the stone has not
yet begun to roll and collect moss.
I have long assumed that the strategy group would see this as one of their main
options.
SDMAIL Kim Warren wrote:
> Posted by ""Kim Warren"" <
Kim@strategydynamics.com>
>
> This is indeed helpful Jack. It is very close to the process we attempted to
initiate in April. It is of course complexified by the diffuse and distributed
nature of our community, but the principles are still sound [even though we
don't
have the opportunity to get the hundreds of Society members and thousands of
other stakeholders in one room for 8 hours!].
>
> Unfortunately, we have not yet moved beyond the brainstorming stage. I would
have moved us on long ago, had we received anything like the number and scope of
suggestions that might seem to represent passable coverage of the community's
interests. If we kill brainstorming now, we would be processing a badly
incomplete and partial set of prospective objectives [by 'partial' I mean that
whole domains of SD work are under-represented]. The risk is we would end up
with
a set of priorities, and a strategy for achieving them that no-one buys into.
> I do intend, as soon as I can carve out a few hours, to get a report back out
to everyone on where the contributions so far have got us to. I have them all,
and in a somewhat organised form. Please bear with me a short while.
>
> Kim Warren
Thanks Kim.
You caution that ""If we kill brainstorming now, we would be processing a badly
incomplete and partial set of prospective objectives.""
Then I didn't explain the goal setting process well enough. My apologies. What I
saw in 1979 and at numerous other goal setting or consensus building meetings
was
that brainstorming is not done once in one big batch. It's done as many times as
necessary, in short quick bursts. Between bursts the facilitator helps the group
evaluate what they have so far and how it can be refined into what the group
needs to end up with. The group alternates between divergent and convergent work
until it converges on the high quality set of items desired.
It's been a long time since 1979, but I'd estimate we broke up into small groups
about four times for the convergent part. The basic cycle was:
1. State the types of goals to be brainstormed. (A process management step)
2. Brainstorm goals as a group. (The all-important divergent step.)
At first these are created from scratch. Later new goals are created by looking
at a list of existing goals and brainstorming better ones or changing existing
ones. Toward the very end sometimes only a single phrase or word is changed. The
final result is great clarity, consensus, and motivation to go out and achieve
the goals.
3. Organize the goals into categories. Some discussion happens here. (A
convergent step)
4. Break up into small groups to refine, condense, clarify, improve the goals.
(Convergent step)
5. Report back to the full group on results. Discuss, judge the quality of the
goals, come to agreement. (Communication and judgment step)
6. Repeat as necessary until done.
In the very large, this is how international treaties are negotiated. The hard
work between meetings is the convergent phase of step 4, and the meetings are
for
everything else.
Step 1 was done in the opening post to this thread. Step 2 could have ended in
about two weeks, although I've not graphed the posts containing goals on a
timeline. If we want to try something like the above we could go for one week
cycles and get several done before the Society conference. We could start this
as
soon as you report on what we've got so far. I think we are all looking forward
to that.
Since this list's membership is large and dispersed, goals will trickle in. We
can simply add them as we go. But the bulk of the goals will come in step 2
bursts.
Brainstorming is a divergent, creative activity. Synthesis into a smaller number
of better stated, better directed goals is a convergent activity. Divergence and
convergence are best done as separate activities. See this short page for
definitions of these two terms:
http://en.wikipedia.org/wiki/Convergent ... production
In ""The Facilitator's Guide to Participatory Decision-Making"" on page 6 is a
great diagram of how divergent and convergent thinking is used, though only one
cycle is shown. I've copied the diagram to this page:
http://www.thwink.org/sustain/work/reference/index.htm
Notice the list of what the two types of thinking do:
Divergent Thinking versus Convergent Thinking:
- Generating a list of ideas versus Sorting ideas into categories
- Free-flowing open discussion versus Summarizing key points
- Seeking diverse points of view versus Coming to agreement
- Suspending judgment versus Exercising judgment
Hope this explains things,
SDMAIL Richard Stevenson wrote:
> Posted by Richard Stevenson <
rstevenson@valculus.com>
>
> I know this won't get published, Bob. Because you won't pass it.
>
> I have read the recent correspondence on strategy development with a very
limited degree of interest. This thread was instigated over six months ago and
initially received very little response. Those of us that did respond
(enthusiastically) were basically ignored.
>
Richard,
My humble opinion is this is due to an immature process in goal setting. I
posted
a message to address this problem on July 1.
> There is now very little in this ""second wave"" that is worthy of comment.
Intrinsically, it seems that system dynamics will always prefer to gaze at its
own navel rather than to do anything constructive.
>
In your post on April 24 in this thread I see this excellent suggestion:
""If the SDS is really serious about conducting a 'strategy review', then
it needs to adopt a more formal and controlled forum of business people,
headed by people who have appropriate business credentials as well as SD
experience. Reality is that the SDS is still a tiny - yes a tiny -
collection of academics who appear to revel in their academic isolation.""
I agree. Academics teach. Business consultants and managers solve problems. The
latter group is far more qualified to lead a strategy review. The rule of thumb
in business is if you put an academic in charge of a major project or business,
you are pretty much guaranteeing its failure. There are of course some
exceptions.
Another jewel: ""So if we really are serious about rethinking the future of SD -
we need to completely rethink the way that SD fits into the way that the
business world actually works.""
This is a great insight. But it is incomprehensible to most academics, who on
the
average do not understand the business world. (By the way, my career is 100%
business management and consulting, with 8 years off for an artistic endeavor.
How refreshing!)
A strong close: ""We must shift our emphasis away from promoting SD in its own
right - and take a market-based perspective in business. The primary customer
is
the CFO, who speaks the language of value - not of systems. The need
and the opportunity is to provide deep insights into the relationships
between strategy and long-term financial value, using the language of
resources - indeed the very language of SD!
""I suggest that our 'task force' should address these issues by engaging
significant external expertise and interest - not simply by reinforcing
entrenched SD ideas with internal SD 'experts'.""
Yes, yes, yes. But since the Society is academic centered, these thoughts have
been ignored.
Or have they? It's possible that ""significant external expertise and interest""
HAS been planned for inclusion in the very project this thread launched. I
wonder
if Jim Lyneis, Society president, could speak to this.
> I really am sorry. Get on with it - or get over it. Most of the rest of
world
has now moved on. And - SD people - it is your fault. Have a good time in
Athens.
>
> But - SD Society - please understand. Your academic fumbling and internal
censorship has effectively destroyed a great idea.
I feel your pain, Richard. But if people like you and me speak up, and calmly
point out these problems, offer constructive ways to surmount them, and exhibit
patience and empathy, then perhaps the Society can get past its growing pains.
By growing pains I mean the fact the the Society has done well in what
professional societies normally do: publish journals and host conferences. But
there are other areas in which societies can benefit their members and society.
Our Society is now trying to grow into those other areas, and is having the
usual
initial trouble. But if we can all work together constructively, we should be
able to ""develop a vision and strategy for the next 50 years"" that will astonish
the world as it is achieved, step by tiny step.
Gosh, hope this doesn't sound too artificially inspirational....
Jack
Posted by Jack Harich <
register@thwink.org>
posting date Wed, 02 Jul 2008 20:46:31 -0400
_______________________________________________
.