Ventana software support forum

Posted by Tom Fiddaman <tom@vensim.com>
WAS Problem Solving versus Optimization (SD5226)
Jack Homer wrote:

>>Many ridiculous things have been said about LTG/SD, and this adds to that
>>list. But because the article is in a magazine with large circulation and
>>seems so dishonest and self-serving, I'm thinking that it deserves a
>>response, perhaps a letter to the editor of Scientific American, and perhaps
>>coming from Dennis Meadows and Jorgen Randers. (I haven't checked with
>>Dennis and Jorgen on this and have no idea whether they would have the time
>>or patience for it.) Does anyone second that idea, or have a different one?

I'll second the idea, with some preliminary thoughts. If Dennis and Jorgen
have grown too weary, or too wise, to take the bait, I'll cowrite with
anyone who cares to sign.

The SciAm article is a perfect illustration of something I said here a few
months ago (SD5057):

>>While there are valid things to criticize about LTG, I get the sense that
>>most critics haven't actually read the book; they're just repeating
>>something they heard about it years ago, probably from someone who also
>>didn't read it.

The authors claim that the runs presented in Limits to Growth were
forecasts. They were explicitly not. There are actually several kinds of
output in the book, from submodels on resources (see my earlier post,
clipping below) and other topics, and from runs of the full World3 model.
Reading from the introduction to the first World3 run presented in the book
(page 121 of my edition):

""... In many cases the information available is not complete. Nevertheless
we believe that the model based on this information is useful even at this
preliminary stage for several reasons.""

""First, we hope that by posing each relationship as a hypothesis, and
emphasizing its importance in the total world system, we may generate
discussion and research that will eventually improve the data we have to
work with. ... ""

""Second, even in the absence of improved data, information now available is
sufficient to generate valid basic behavior modes for the world system.
This is true because the model's feedback loop structure is a much more
important determinant of overall behavior than the exact numbers used to
quantify the feedback loops. ... Since we intend to use the world model
only to answer questions about behavior modes, NOT TO MAKE EXACT
PREDICTIONS [emphasis added], we are primarily concerned with the
correctness of the feedback loop structure and only secondarily with the
accuracy of the data.""

""Third, if decision-makers at any level had access to precise predictions
and scientifically correct analyses of alternate policies, we would
certainly not bother to construct or publish a simulation model based on
partial knowledge. ...""

How you get that from the paragraphs above to ""In presenting the analysis
as a forecast, the authors stretched the model beyond its limits..."" is
beyond me. I skimmed the text for the word ""forecast"" unsuccessfully. The
closest I could come is (pg. 126): ""We can thus say with some confidence
that, under the assumption of no major change in the present system,
population and industrial growth will certainly stop within the next
century, at the latest."" So the jury's out for six more decades.

Overall, the SciAm article strikes me as a brilliant piece of marketing for
RAND without much beef between the buns. Perhaps I'm overly focused on the
technical aspects of the problem; the methods they seem to describe don't
seem markedly different from anything else in decisionmaking under
uncertainty, which has been around for decades. The idea of robust policy
design is also nothing new. The authors hint at interactivity; if they've
developed an automated way to make guided strategy explorations in an
uncertain model that would be quite cool. If they have developed ways to
construct decision rules under uncertainty (as with the safety valve
policies around Kyoto targets) that would be a useful contribution as well.

The authors claim ""Our method thus reduces a complex problem to a small
number of simple choices. Decision makers make the final call. Instead of
fruitlessly debating models and other assumptions, they can focus on the
fundamental trade-offs, fully aware of the surprises that the future may
bring."" A few observations about this statement:

- Good tools for robust strategy development do not preclude the need for
robust models. Wonderland is a particularly unfortunate choice in this
case, as it has some formulation problems that are hard to detect by
inspection due to the lack of units of measure or even clear stock-flow
structure. They should be evident with testing under extreme conditions,
but somehow the scenario testing didn't reveal that here.

- The procedure described still operates in model-as-oracle mode. With
single deterministic simulations users get to argue about whose parameter
or equation is correct. Recognizing genuine scientific uncertainty (e.g.
the probability distribution of the sensitivity of climate to 2x CO2) helps
in some cases, but really just shifts the debate to ""whose distribution is
correct."" I suspect that this procedure is still susceptible to the growing
practice of sowing scientific disinformation to muddy the debate.

- If decision makers don't understand the problem, they're not likely to
act. It's clear that this is frequently the case. For example, Senator
Inhofe recently stated in a floor speech:

>>People are trying to say that the release of CO2 is the cause of climate
>>change. These people have to understand that historically it doesn't work
>>out that way. We went into a time right after World War II when we had an
>>85-percent increase in CO2 emissions. What happened there was that
>>precipitated not a warming period but a cooling period. Again, that is too
>>logical for some of the alarmists to understand. They want so badly to
>>feel a crisis is upon us.

Evidently he doesn't know - or doesn't want to know - that temperature
responds to the stock of CO2 in the atmosphere, not the flow of emissions.
As long as such broad misperceptions prevail, it's hard to see how one can
make much headway through the addition of another layer of complexity.

I believe that models can help solve problems by codifying knowledge that
can be agreed upon so that debate can focus on questions of genuine
uncertainty and differences of value. It seems the authors believe that
too, but I'm not able to fully appreciate their contribution from the SciAm
article. I hope someone will report back to this list on the full book at
http://www.rand.org/publications/MR/MR1626/ .

Tom
Posted by Tom Fiddaman <tom@vensim.com>
posting date Mon, 25 Apr 2005 13:17:58 -0600

Posted by George A Simpson <gsimpso4@csc.com>

WAS Problem Solving versus Optimization (SD5226)
Responding to points by Jack Homer and Bob Cavana:

I didn't see the points about ""Limits to Growth"" as an implied criticism of
the field at all. In fact, they are suggesting that the field of
Simulation Science do what the article is all about - use an incremental
adaptation strategy. So whereas in the naive early days of the field,
there may have been an implication of prediction, now we adopt a more
mature view that SD as part of simulation science in the large provides
tools that help us identify robust strategies for attaining the objectives
we are seeking.

Rather than a ""protest"" letter to the editors of Scientific American, how
about a full article on Simulation Science - covering systems dynamics,
discrete simulation, and recent advances in the field.

This exercise would help us deepen our understanding of core issues and
clear up misapprehensions.

Is there anyone out there with the gravitas to take on this challenge?

..george...

Dr. George Simpson, Principal Consultant, CSC
CSC Alliance: Performance Engineer
CSC House, Fleet
tel +44 1252 813930 mobile 07814 623518
email: gsimpso4@csc.com
Posted by George A Simpson <gsimpso4@csc.com>
posting date Tue, 26 Apr 2005 09:20:52 +0100

Posted by ""Jack Homer"" <jhomer@comcast.net>
George Simpson wrote:
>> I didn't see the points about ""Limits to Growth"" as an implied criticism of
>> the field at all....So whereas in the naive early days of the field,
>> there may have been an implication of prediction, now we adopt a more
>> mature view that SD as part of simulation science in the large provides
>> tools that help us identify robust strategies for attaining the objectives we are seeking.

I stand by my comments that the Popper et al criticisms of LTG were
flagrant. And I don't view the early days of SD as naive: The purposes are
still the same as Forrester laid them out in Industrial Dynamics, and the
way we work with clients still pretty similar to how Ed Roberts explained
the process in 1977 (""Strategies for Effective Implementation of Complex
Corporate Models,"" Interfaces 7(5).) Only difference I see from the 60s and
70s is that now we can do lots of runs quickly, and so can do a more
thorough job of model testing and more rapid cycling through model
iterations; and our graphics have become more client-friendly. Lots of good
work has been done in recent decades, of course, and we have gotten more
sophisticated about the roles of maps and models and group process in client
learning. But I don't find any fault in the writing of LTG: the backlash
was not due to naivete in the presentation, but due to a fundamental clash
of paradigms that lasts to this day. I believe the Popper gang have taken
the catty approach of doing pretty much what we have always done, but
packaging it as ""new"" and by distancing themselves from LTG. One look at
Wonderland (thanks much, Tom F.), and you will know that LTG more than 30
years ago was more robust than the Popper work is, and presented with much
greater sincerity.
- Jack Homer
Posted by ""Jack Homer"" <jhomer@comcast.net>
posting date Tue, 26 Apr 2005 10:56:09 -0400

Posted by ""Swanson, John"" <John.Swanson@sdgworld.net>
I think George Simpson's response is the wisest. The Scientific American
article is indeed a naked piece of marketing for RAND, but I'd have
thought it says a lot that SD people would agree with. The upset is
because (i) the authors re-cycle ideas that have been around for a long
time, claiming they are the first to think of them, and (ii) they
misrepresent, slightly, how LTG was intended to be used.

Rather than an angry counterblast, or another attempt to defend a book
that was published 40 years ago, all of which can look a bit defensive,
it would be better to do what George suggests: welcome the debate, put
it into a wider context, show how thinking and experience have developed
over the years, and slip in a mild correction, in a footnote, about LTG.

John Swanson
Associate
Steer Davies Gleave
28-32 Upper Ground, London SE1 9PD

[direct] +44 (0)20 7910 5542
[reception] +44 (0)20 7919 8500
[fax] +44 (0)20 7827 9850
[email] john.swanson@sdgworld.net

www.steerdaviesgleave.com
Posted by ""Swanson, John"" <John.Swanson@sdgworld.net>
posting date Tue, 26 Apr 2005 15:34:41 +0100

Posted by Ruth Fruland <fruland@hitl.washington.edu>

I found the SciAm article to be more constructive than not, and a golden
opportunity for system dynamicists to heighten awareness of their work.
However, there were subtle and disturbing representations related to LTG,
which is not the same as the ""field"" of SD, that deserve a response.

In addition to referring to the report itself as ""(in)famous,"" a gratuitous
signal that the report's findings are suspect, they also refer to its
originators as ""a group of scientists and opinion leaders,"" which evokes quite
a different image than, say, ""a group of MIT scientists commissioned by an
informal think tank of international scientists, economists, educators, and
industrialists"" (i.e., the Club of Rome).""

The emotion and disinformation associated with LTG warrants taking the
opportunity to set the record straight. [e.g., The Club of Rome was
organized by Dr. Aurelia Peccei, an Italian industrialist affiliated with Fiat
and Olivetti, and Dr. Alexander King, a distinguished English research chemist
who was Director General of Education and Science at the European Organization
for Economic Cooperation and Development. The study itself was funded by the
Volkswagen Foundation.]

These are not your ordinary ""opinion makers"" of today's lightweight variety.
Their prescient and visionary concerns and goals predate the current authors'
same concerns by more than 30 years. It is only to be
expected that credit should be graciously given where it is due (especially
given Scientific American's commitment to present ""fair and balanced science,""
April 2005, p.10). That better simulations are possible today, in no way
diminishes the pioneers in the field, nor their work.

This may also provide an opportunity to make visible the significant
continuation of that work, which, although completely relevant to the
issues raised in the article, has yet to command the attention of
researchers in other fields or such eminent and mainstream platforms as
SciAm.

The SciAm authors do seem to imply that their approach provides a new
way to manage uncertainty and that companies such as Volvo have ""already""
used their techniques to plan corporate strategy, thus do appear to
either disregard or dismiss the accumulation of accomplishments and
on-going research of the SD community.

Perhaps it could be gently pointed out that even K-12 school districts
are using SD models to generate ""what if"" scenarios (see the November 2004
issue of School Administrator). Similar simulations are being
productively used to demonstrate the counterintuitive nature of complex
social systems and guide strategic planning in a variety of business,
educational and government organizations. The time may, indeed, have
arrived for an article on simulation science.

The authors also perpetuate a dualistic model of thinking - economy versus the
environment - which is not only insufficient, but outdated: even the
engineering community recogizes the importance of the triple bottom line:
economy, ecology, and social justice.

The authors make the claim that their methods offer policy makers the ability
to answer: ""What actions today will best shape the future to our
liking?"" Ecological economists, notably, Robert Costanza, have previously
addressed this question, and in the context of democratic processes. For
example, there is an online paper and survey that allows people to rate how
comfortable/happy they would be living in one of four futures based on whether
decisions were made by technological optimists or pessimists, and then, which
one reflected the true state of the world.
[""Visions of alternative (unpredictable) futures and their use in policy
analysis. Conservation Ecology 2000: 4 (1):5, which is online at URL:
http://www.consecol.org/vol4/iss1/art5].

In a second paper, ""A vision of the future of science: reintegrating the study
of humans and the rest of nature"" [Futures 35 (2003) 651-671], Costanza argues
for the process of scientific ""synthesis"" to be undertaken as a necessary
counterbalance to traditional scientific analysis (reductivist) research. SD
provides the methods and tools to do both, as do the tools offered by the SciAm
authors. Both are qualitatively different (by virtue of being more
complete) from the traditional scientific method, and I believe this is
an important point that needs to be made explicitly and repeatedly.

Hopefully, people who have the necessary credentials and knowledge will have
the time to contribute to these and the other points that have been raised,
either by responding with a letter, if not a full-fledged article.

Appreciatively,
Ruth

******************************************************
--------------------------------------
Mail Address: Ruth M. Fruland, Ph.D.
Human Interface Technology Laboratory
215 Fluke Hall, Mason Road
Seattle, Washington 98195-2142
--------------------------------------
Email: fruland@hitl.washington.edu
******************************************************
Posted by Ruth Fruland <fruland@hitl.washington.edu>
posting date Tue, 26 Apr 2005 13:28:08 -0700 (PDT)

Posted by ""Louis Macovsky"" <dynbiosys@verizon.net>
Rather than a letter to the editor, wouldn't more be accomlished in the
debate of ideas by convincing Sci. Am. to publish an indepth article
dedicated to system dynamics... its underlying theories of system thinking,
its potential, and its successes where applied.

Louis
Dynamic BioSystems, LLC
Wilsonville, Oregon
Posted by ""Louis Macovsky"" <dynbiosys@verizon.net>
posting date Wed, 27 Apr 2005 08:51:27 -0700

Posted by =?iso-8859-1?Q?Jean-Jacques_Laubl=E9?= <jean-jacques.lauble@wanadoo.fr>

There is a Web site named http://www.evolvinglogic.com
supposed to explain the ideas behind the article in S.A. and that is
probably selling something but I do not know what.
I tried to understand what is meant by RAP (Robust adaptive planing) and CAR
(Computed aided reasoning) the original objective being nice, but there is
not the slightest explanation how it is achieved. No reference to any
software or any other help or service. RAP is what everybody is trying to do
all life long, I thought they has some original ideas about it.
I sent an e-mail asking for more details last saturday but got no answer
yet.
Maybe it is because I am in France, but that site is anyhow completely
illogical.
Regards.
J.J. Laublé Allocar
Strasbourg France
Posted by =?iso-8859-1?Q?Jean-Jacques_Laubl=E9?= <jean-jacques.lauble@wanadoo.fr>
posting date Wed, 27 Apr 2005 14:24:12 +0200

Posted by Paul Newton <plnwtn@gmail.com>
Here is my two cents worth on the article that Jack Homer commented on.

Maybe this is taking things too much out of context, but the portions that I
think are really suspect in the quote from the article in Jack Homer's email
are:

1) ""The (in)famous report The Limits to Growth from the early 1970s..."" If
the report is infamous, it is not the authors' fault, but rather the fault
of the sloppy reading and criticism of the report by many of its critics.

2) ""In presenting the analysis as a forecast, the authors stretched the
model beyond its limits and reduced the credibility of their entire research
program,"" and ""A group of scientists and opinion leaders... predicted that
the world would soon exhaust its natural resources unless it took immediate
action to slow their use."" I don't think the authors ""presented the analysis
as a forecast,"" or ""predicted,"" but rather presented a theory that has
undesirable tendencies toward overshoot and collapse, which tendencies their
simulation experiments illustrate are robust to a number of policy actions.
I think the LTG authors intended their theory to invite reasoned debate and
discussion, but for the most part it seems the ensuing discussion was not
well reasoned.

The portions I think are more reasonable are:

3) ""...the standard tools of analysis often fail to mediate such debates.""
In the usual meaning of the term ""mediate,"" LTG did fail in this respect,
but not because of the model itself or how it was presented. I suspect that,
as Sterman says in Section 1.3.8 of Business Dynamics, ""defensive routines
and interpersonal impediments"" got in the way and caused people, such as the
SA article authors, to fail to, as Homer writes, ""credit LTG/SD as a worthy
predecessor upon which they would like to build, instead of knocking it down
and promoting what they are doing as if it is something brand new."" So,
again, the failure is not with the LTG model or how it was presented, but
rather with how it was used by its critics.

4) ""the model was not wrong; it was just used incorrectly. Any computer
model is, by definition, a simplified mirror of the real world, its
predictions vulnerable to some neglected factor. The model developed for The
Limits to Growth revealed some important aspects of the challenges faced by
society."" This seems right on to me, except that, after the word ""factor,""
I'd add ""or feedback loop.""

I hope that if the SD Society responds, that the response mostly supports
the article. Indeed Jack writes that he is ""sympathetic with much of what it
(the article) says,"" and that the authors' ""approach sounds so similar to
what we do: evaluating alternative scenarios with simulation, and seeking to
inform decision-makers about tradeoffs and uncertainties rather than
forecasting single futures."" I hope that if we respond, the response conveys
our general agreement with how the SA article authors suggest models be
used. We should laud their main message, but point out the problems we see
with their criticism of LTG, and perhaps also point to some examples in SD
practice that illustrate that the approach they suggest is already being
used, and agree with them that the approach they suggest should be used more
than it is.

Paul Newton
pcn4@cornell.edu
paulnewton@StewardshipModeling.com
Posted by Paul Newton <plnwtn@gmail.com>
posting date Wed, 27 Apr 2005 07:09:58 -0400

Posted by Tom Fiddaman <tom@vensim.com>


>>Posted by George A Simpson <gsimpso4@csc.com>
>>
>>I didn't see the points about ""Limits to Growth"" as an implied criticism of
>>the field at all.


I think the reason the comments about LTG struck a nerve is that there have
been so many worse statements in the past, including many that declared the
death of SD in the same breath. In this case, I view a response as a
defense of LTG rather than a defense of SD. If nothing else, the author's
comments are perpetuate a mindless dismissal or ignorance of the basic
argument of LTG (rapid growth + delayed negative feedback = likelihood of
overshoot).

I think the authors also mischaracterize the fundamental reasons for the
""failure"" of LTG. It had little to do with neglect of uncertainty. In fact
LTG presented a dozen or so sensitivity scenarios illustrating that the
dominant behavior mode of overshoot was robust. Its prescription for a
viable future is extremely general (material stocks in a steady state
consistent with values and technology) and probably characterizes all
attractive policies in Wonderland as well. Had LTG used the methods the
authors describe, I seriously doubt the outcome would have been any
different. The more sentient arguments against LTG were mainly about
feedback structure - e.g. that positive feedback from technology would
perpetually overwhelm any environmental constraints. Others were really an
immune reaction against its policy implications, presaging the
well-orchestrated disinformation campaigns against evolution and climate
science today.

The attractive feature of the Wonderland model is that it captures the
techno-optimist world view as well as the overshoot view, and thus could
serve as a vehicle to set aside the facts both sides can agree on focus the
debate on questions of genuine uncertainty or value differences.
Unfortunately, as a representation of reality, it is rather flawed. I made
a quick & dirty replication of the model in the online appendix of the book
at RAND, which was challenging given the opaque functional forms, lack of
clear stock-flow definition, lack of units, and typos in the listing. It
seems to share the same problems as the original Wonderland: if the
environment is wiped out, the economy contracts at a fixed rate (only 3%
per year); population grows when per capita economic output is zero; and a
pristine environment is nearly invincible. Some of the integrated
assessment models that now exist for climate policy and the like are
clearly better, and have been used for equally or more sophisticated
uncertainty analysis.


>>In fact, they are suggesting that the field of
>>Simulation Science do what the article is all about - use an incremental
>>adaptation strategy. So whereas in the naive early days of the field,
>>there may have been an implication of prediction, now we adopt a more
>>mature view that SD as part of simulation science in the large provides
>>tools that help us identify robust strategies for attaining the objectives
>>we are seeking.


I was in daipers at the time, but I don't think naive predictive claims are
part of the early history of SD. If anything, SD was struggling against
naive predictive use of linear regression models in other fields, and
working hard to make basic points about structure and behavior. Robust
strategy design has always been at the core of SD, though generally
achieved through tests in extreme conditions rather than formal uncertainty
analysis. If anything SD has remained too averse to prediction, even where
technology and data permit it. Section 13.7 of Industrial Dynamics has a
good discussion of prediction.


>>Rather than a ""protest"" letter to the editors of Scientific American, how
>>about a full article on Simulation Science - covering systems dynamics,
>>discrete simulation, and recent advances in the field.
>>This exercise would help us deepen our understanding of core issues and
>>clear up misapprehensions.
>>Is there anyone out there with the gravitas to take on this challenge?


Great idea!

Tom
Posted by Tom Fiddaman <tom@vensim.com>
posting date Wed, 27 Apr 2005 11:13:54 -0600

Posted by Paul Martin <Paul_M@profitfoundation.com.au>
Economists continually forecast and model things that prove to be
incorrect, and then use this to justify (rightly) more investment in
their science.
More than 50% of market launches fail, and marketers use this to
justify (rightly) more investment in their science cum art.
Lawyers consistently come up against deficiencies in the effectiveness
of law, accountants still find accounting errors .. and so on.
Get over it - even attempting to prove that a discipline is 'right'
fails to reflect the reality that all disciplines are distortions of
reality, predictions are essential but unreliable, and most scientists
are attempting to absorb the lessons of history, not pretend that the
lessons are not there.Why get so upset of criticism of what was
brilliant work, but on a number of levels proved to be innaccurate
forecasting (which I am sure we would all strive to do better today).
I enjoy the list, but am constantly amazed at the defensiveness and
'preciousness' about what is, after all, only another (but still an
important) discipline.

Prof. Paul Martin
Australian Centre for Agriculture and Law
University of New England
Posted by Paul Martin <Paul_M@profitfoundation.com.au>
posting date Wed, 27 Apr 2005 22:23:08 +1000

Posted by Steven Roderick <sroderic@mac.com>
It is sometimes unfortunate the human mind works best in the realm of
sharply defined edges and near certainty for it makes consideration of
complexity difficult. I believe that is the point that Popper et al are
attempting to make in their article, but they too have fallen prey to the
trap of our human habits of thinking when they claim that LTG presented
""analysis as a forecast"".

I don't claim to be a practitioner of SD in the same way as most readers of
this list, but it is an important part of what I do in my high school
classroom and I have been involved in the systems world since the early
70's. I remember well the publication of LTG. I was doing an independent
undergraduate project with the MIT systems lab at the time. Criticism was
primarily around the link that readers of LTG made between ""assumptions""
that went into the model's structure (or lack of structure) and a perceived
""forecast of doom"". My memory is that the authors, Jay Forrester included,
went on and on about how their model was not meant to be predictive at all,
but rather to be a platform for discussion and debate that could lead to a
more complete and shared understanding of the complicated world system in
which we live. Jay, and others, spoke rather eloquently about their hopes
to develop new habits and reflexes of mind when considering complex issues.
Consideration of feedback, behavior over time, stocks and flows in the world
system, and limits to growth, were all seen as ways to bring more resolution
to the lenses that we use for decision making, not to predict where and when
to expect a crash in the world economy.

I think the authors of the Popper article are correct when they state ""the
model was not wrong; it was just used incorrectly"". They are incorrect
though, in their assumption that the users were the authors of LTG. I
believe the ""users"" were those who criticized the model, and that they used
it in just the way that the authors had hoped, to promote and sustain
discussion.

To see one's thinking placed down on paper or computer screen, with
accompanying equations to show emphasis among cause and effect
relationships, brings a reality to the thinking that is difficult to shake.
But as we all know, models are just models. In their attempt to bring focus
to long term problems, I think that the authors of ""Shaping the Future"" have
seen only one small facet of what LTG was all about.

Steven Roderick
Biology Teacher
Lincoln Sudbury Regional High School
Sudbury, MA USA
Posted by Steven Roderick <sroderic@mac.com>
posting date Thu, 28 Apr 2005 15:58:42 -0400

Posted by Erling Moxnes <Erling.Moxnes@ifi.uib.no>


Steven Roderick sroderic mac.com wrote:


>> I think the authors of the Popper article are correct when they state ""the
>> model was not wrong; it was just used incorrectly"". They are incorrect
>> though, in their assumption that the users were the authors of LTG. I
>> believe the ""users"" were those who criticized the model, and that they used
>> it in just the way that the authors had hoped, to promote and sustain
>> discussion.
>>


I found the first part of this comment very insightful and giving guidance
for how an eventual reply should be formulated.

As can be seen from the LtG book and much other material from Meadows et
al., the authors use the model correctly. The problem has been the readers
(those who liked it and those who did not) who have used it as a prediction
of doom. I guess this tendency has increased among those who have not read
the original, only commentaries to LtG.

What the authors could be criticised for is that they did not foresee how
the report would be misinterpreted and that they did not prevent this. This
may however be just as difficult as it is to get people to discuss the
structure of the climate problem and not only one-cause-only explanations of
temperature changes in the last century.

I am not sure that Steven R. is correct when stating that the authors of LtG
had hoped for the type of debate that took place after the publication.


Erling Moxnes


-- The System Dynamics Group University of Bergen, Norway http://www.ifi.uib.no/sd/
Posted by Erling Moxnes <Erling.Moxnes@ifi.uib.no>
posting date Mon, 02 May 2005 11:32:45 +0200

Posted by ""John Gunkler"" <jgunkler@sprintmail.com>
Erling Moxnes writes: ""What the authors could be criticised for is that
they did not foresee how the report would be misinterpreted and that they
did not prevent this. This may however be just as difficult as ...""

I would suggest that it's not only very difficult, it's truly impossible.
There is nothing that can be written or said, including this message, that
will not be misinterpreted -- through ignorance, laziness, or willfully to
further someone else's agenda.

It's one of the things that amazes me most about Jay Forrester's writings --
not that he's never misinterpreted, but how often he anticipated the
misinterpretations and criticisms and provided answers along with his
original presentation of ideas. This is truly a mark of his genius. I'm in
awe of this accomplishment and have never been able to duplicate it in my
own work.

Posted by ""John Gunkler"" <jgunkler@sprintmail.com>
posting date Tue, 3 May 2005 19:07:40 -0500

Posted by Arlen Wolpert <awolpert@TheWorld.com>


On Wed, 4 May 2005, John Gunkler jgunkler sprintmail.com wrote:
>> It's one of the things that amazes me most about Jay Forrester's writings --
>> not that he's never misinterpreted, but how often he anticipated the
>> misinterpretations and criticisms and provided answers along with his
>> original presentation of ideas. This is truly a mark of his genius. I'm in
>> awe of this accomplishment and have never been able to duplicate it in my
>> own work.
>>


Underlying Forrester's genius is the fact that
he is a no nonsense, straight shooting guy out
of the Sandhills of Nebraska.

Posted by Arlen Wolpert <awolpert@TheWorld.com>
posting date Mon, 9 May 2005 11:26:09 -0400

Posted by DavidPKreutzer@aol.com

Part of Jay's genius is his ability to articulate brilliant mathematical
insights with stories that express the essence of the critical issues at
hand. I was fortunate to have the opportunity to work with him full time for
about 15 years. He once told me that Gordon Brown, his own teacher from the
1930s and 1940s advised him that people of the world perceive in thousands of
shades of grey, so we if we want to make an impact, we sometimes need to
learn to speak in black and white.

My art teacher tells me that the human eye can actually only detect 17
shades of grey, but the point is still provactive.

David Kreutzer
Posted by DavidPKreutzer@aol.com
posting date Wed, 11 May 2005 15:19:13 EDT

Posted by Matzaball50@aol.com
Hi Folks,

Posted by DavidPKreutzer@aol.com
>Part of Jay's genius ...

Although I can agree with David and all the others who posted on this thread
about the greatness of the work done by Jay, I think we might be doing SD a
little disservice with all this praise and putting him up on a pedestal.

I believe, and I think Jay would agree with me, that science advances in
part by the critical method. By this I mean critical discussion, the testing of
ideas, and the creative, active search for error. Newton was a genius, and
his work was ultimately supplanted in part by Einstein. Einstein was ultimately
supplanted in part by Schrodinger, and Schrodinger's idea's will be
supplanted by some others in the future.

Our search for the 'truth' is an _infinite_, on-going, and never ending
process.

Making someone an 'authority' or god like figure makes it _extremely_
difficult to 'criticize' the ideas of that individual without incurring the wrath
of the 'faithful'. Just ask Galileo.

I think we need to keep in mind the importance of what SD the science
currently means to us all, and what it could mean to many others, _if_ they felt
more secure with the method.

How do we get others to feel more 'secure'?

I believe we need to engage others in dialogue, and _encourage_ the
continued critical examination of SD because although Jay and many others have done
some wonderful work in this area, _the ideas_ (not Jay himself ) are _not_
above reproach, criticisms, and even, eventual modifications.

If we are secure about ourselves and our ideas we should welcome the
criticisms as a chance to look at, and question our _own_ ideas and beliefs and see
if any modifications are warranted or necessary.

But even if the criticisms were found to be false or misguided, they
provide us with the opportunity, if we are smart about it, to both understand why
someone has some issues about SD and an opportunity to clarify them for that
individual.

When we are able to 'clarify' an issue for either ourselves, that is,
correct an error, or to help someone else correct their 'error', 'learning' will
have occurred for all involved and our knowledge improved.

Of course some folks are_not_ interested in 'learning' or finding out if
their beliefs are valid or not, and are simply interested in 'protecting' their
own turf and/or projecting their own ideas. I just don't happen to be one of
those people. Are you?

regards,

Marc
Posted by Matzaball50@aol.com
posting date Fri, 13 May 2005 06:23:50 EDT

Posted by ""John Gunkler"" <jgunkler@sprintmail.com>
Jay Forrester, although he strikes me as a modest man, IS an authority in
SD. That doesn't mean, of course, that he is infallible. But to deny him
the position of authority is absurd.

Moreover, I have been struck over the several years I've monitored this
list, by the number of times issues arise that not only did Jay anticipate
but that he provided awfully good answers to.

I think anyone working in SD would be a fool who didn't first refer to
Forrester's writings when confronted with an issue. After seeing what Jay
had to say, then go on to further elaboration or questioning -- but first
find out what our authority said.

Notice that the word ""authority"" contains the concept of ""author"" -- and
even you, Marc, must admit that Jay is the author of SD. But he's more than
that. He's still one of the most incisive and critical thinkers the field
has produced.


John
Posted by ""John Gunkler"" <jgunkler@sprintmail.com>
posting date Sat, 14 May 2005 13:01:34 -0500

Posted by Matzaball50@aol.com

[ Host's note - this post is the final one on this topic (no more will
be posted) and has been abridged ]

Hi John,

I'm afraid you misinterpreted my message. Let me try and clarify.

You say;
>Jay Forrester, although he strikes me as a modest man, IS an authority in
>SD. That doesn't mean, of course, that he is infallible. But to deny him
>the position of authority is absurd.


First,I [would never] deny that Jay did a masterful job with SD.

Doing science is not about reputation nor is it about 'authority'. Science
is about conjecturing and testing. _No_ one's theory's and ideas are above
reproach and testing; not Newton, Einstein, nor anyone else.

Jay, for all his wonderful work is not omnipotent. Jay's ideas and work have
been extremely important and influential in my thinking but there have been
others as well.

I think we need to be aware of that, and although recognizing the great work
he has done, be aware that others might be able to improve on it; and we
must be vigilante and creative in trying to find and correct error as we become
aware of it.

Discovering errors is a good thing. Nothing in this world is certain and
nothing is perfect, so when we do find an error we are just that much closer to
the truth if we can correct it.

I view the Popper & Co article as an _opportunity_ and as a challenge. We
really should be thankful that others are willing to spend _their_ time trying
to refute ideas we hold. If they did not think it important they would not
bother. Then the question comes down to whether we can show them that our
thinking is superior to theirs. This is quite a challenge, but one we should all
welcome, because if their ideas are shown to be superior, we would have to
change our currently held beliefs and we will have 'learned' something. On the
other hand if we can show through the 'critical method'; that our ideas are
superior, and we may be able to gain others who are willing to pursue the SD
method.

regards,

Marc
Posted by Matzaball50@aol.com
posting date Sun, 15 May 2005 11:20:38 EDT