Management Transients

["Pedro Mendes"]

Dear Colleague,

My request is twofold.

1. Do you know a paper I can cite expressing opinions similar to Tom
Fiddamans? ("REPLY SD a special case of Complexity/Chaos? (SD1168),"
October 97, this list.)

.QUOTE:
>Some of the management literature Ive seen makes appealing >metaphorical
use of insights that come from complexity research. >However, when you try
to actually tie down assertions about management >principles to modeling
particulars, you find nothing but rubbish. I
>know several colleagues who could go on at greater length about this.
>The same can be said of some management allusions to quantum physics,
>which sometimes seem to me to imply that, because Schrodingers cat
>can exist in multiple states, I can step in front of a speeding bus
>without being killed.
>
>This is damaging to all modeling disciplines; it erodes our hard-won
>credibility by using pseudoscience as a bogus source of authority. If
>the espoused management principles are really worthwhile, then it
>should be possible to demonstrate them explicitly and subject them to
>formal criticism.
:ENDQUOTE.

Id like to use the idea behind Toms comment as part of the argument for a
paper, but Id like to use a more formal citation rather than refer to a
"private communication." Tom couldnt immediately help because his
materials were still boxed after his move, but he agreed I could restream
his wisdom into the list.

2. Are you familiar with SD work on new venture startups? Are there
specific variables and/or structures you would suggest me to include in a
model?

Im after describing the transient behavior associated to starting a
technology new business. The only SD work in the area that Im familiar
with is the 1991 SDR paper "Modeling growth strategy in a biotechnology
startup firm" by John Morecroft/David Lane/Paul Viita. However, this paper
kind of starts where I want to finish, that is, with an already established
firm.

Your help will be appreciated. Thank you.

Pedro
From: "Pedro Mendes" <pedro@ati.utexas.edu>

[Jim Hines]

Pedro Mendes wrote:

> 1. Do you know a paper I can cite expressing opinions similar to Tom
> Fiddamans?

I certainly agree with Toms assessment of the sorry use thus far of
complexity research in talking about management and other human
endeavors. There will be a conference in December 98 in England on
"ORGANISATIONS AS COMPLEX EVOLVING SYSTEMS", I think that this
conference will contain the state of the art on the topic and so the
procedings will probably be of interest to you.

An example of the use of complexity results to make what seems to me to
be shallow analogies to human endeavors is the otherwise excellent book
__Growing Artificial Societies: Social Science from the Bottom Up__ by
Epstein and Axtell (1996, MIT Press). This remarkably well written book
descibes their very interesting work on Sugarscape. However, the
connections to human societies and activities are, well, disappointing.

>> 2. Are you familiar with SD work on new venture startups?

Martin Grossman is doing terrificly interesting system dynamics work on
biotech firms -- how they can finance their research, what strategies
they can take to allocate their research resources to different
categories of research, and what kind of alliances with other firms make
sense. Martin can be reached at mugrossm@MIT.EDU.

Regards,
Jim Hines
LeapTec and MIT
JimHines@interserv.com

[Tom Fiddaman]

An article in the March 98 Scientific American is an interesting read on
the debate over the (mis)use of chaos, complexity, and quantum theory. It
focuses on Alan Sokol, a physicist who succeeded in publishing a spoof
article, "Transgressing the Boundaries: Towards a Transformative
Hermeneutics of Quantum Gravity," in the journal Social Text. The article
critiqued scientists who "cling to the dogma ... that there exists an
external world, whose properties are independent of any individual ... and
that human beings can obtain reliable, albeit imperfect and tentative,
knowledge of [its] laws by hewing to the "objective" procedures ...
prescribed by the (so-called) scientific method."

My personal concerns are less epistemological and more practical. I think
two questions should be asked about model-based suggestions for policy or
strategy.

First, did the modeler do a good job ensuring that the conclusions follow
from the model? Its usually not possible to have the perfect knowledge an
analytical solution provides, but the modeler should have at least explored
the local parameter space and conducted extreme conditions tests to debug
the model and ensure that outcomes are not merely an artifact of fortuitous
parameter choices. The modeler needs to understand the model. An obvious
way to generate understanding is to apply the scientific method to the
exploration of model behavior. If, for example, "leverage" is attributed to
changing a single parameter, that insight should come from changing one
parameter at a time, not several at once. I think there are examples of
defective model exploration in all disciplines (including my own work, Im
sure).

Second, the model must correspond in some way with the system modeled.
Obviously theres no way to establish this with certainty. But there ought
to be at least some attempt to demonstrate that model constructs correspond
with elements of the real system. I think the #1 problem here is with
behavioral assumptions. Ive seen neat articles documenting interesting
articles using genetic algorithms to represent agents learning process,
without any justification of the use of the GA or attempt to explore
alternative, simpler hypotheses like gradient search. Assumptions of global
intertemporal optimization in economic models fall into the same trap.

I dont want to imply a blanket condemnation of complexity theory, which
has contributed enormously to my own thinking. But I think we should be
clear about which results are really model-based, and which are just
attractive metaphors. I think it was economist Bob Solow who pointed out
that the real strength of metaphors comes from knowing where they break down.

- Tom

****************************************************
Thomas Fiddaman, Ph.D.
Ventana Systems http://www.vensim.com
34025 Mann Road Tel (360) 793-0903
Sultan, WA 98294 Fax (360) 793-2911
Tom@Vensim.com http://home1.gte.net/tomfid/
****************************************************

[Khalid Saeed]

Tom,

See also:

Saeed, K. and Bach, N. L. Chaos Out of Stiff Models. Proceedings of 1992
International System Dynamics Conference. Utrecht: July 1992

Saeed, K. and Bach, N. L. Is Deterministic Chaos only a Property of Models?
Proceedings of 1990 International System Dynamics Conference, Boston, MA:
July 1990

Khalid Saeed
From: Khalid Saeed <saeed@WPI.EDU>

Khalid Saeed
Professor and Department Head
Social Science and Policy Studies
Worcester Polytechnic Institute
100 Institute Road
Worcester, MA 01609, USA

Ph: 508-831-5563; fax: 508-831-5896

http://www.wpi.edu/Academics/Depts/SSPS ... saeed.html

[Bradd Libby]

The value of the application of complexity theory to business is analyzed
from a critical point of view in the current issue of Strategy & Business
magazine. The article, entitled "Complexity Theory: Fact-Free Science or
Business Tool?," is available on the Web in full-text form at

http://www.strategy-business.com/strategy/98104

Also available in that issue is an interview with Chris Argyris, whose
work some on this list may be familiar with.

Sincerely,
Bradd Libby.
From: Bradd Libby <bradd@WPI.EDU>

["Roberto Vacca"]

Rome, March 24, 1998

Dear Tom:

Alans last name is Sokal, not Sokol.

Appreciate your views on modelling integrity. We are working on a paper
on quality of data and grievous consequences on forecasting of quality
being low; also methods for assessing noise. In chaos theory one can tell
noise from chaos only if there quite a few hundred thousands data in a
time sequence.

Our work is using many different models concurrently: SD, Input/Output,
logistic substitution.

Best
Roberto Vacca, ISIS, Rome, Italy
For more detailed information email "Roberto Vacca" <mc4634@mclink.it>