Characterizing System Dynamics

Jean-Jacques Laublé jean-jacques
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

Characterizing System Dynamics

Post by Jean-Jacques Laublé jean-jacques » Sat Dec 03, 2005 12:58 pm

Posted by =?iso-8859-1?Q?Jean-Jacques_Laubl=E9?= <jean-jacques.lauble@wanadoo.fr> Hi Invig (or should I say Invigorator instead of Invigilator)

300 years ago, when the first results of calculus where known after Pascal, Leibniz, Newton, Euler and some others, the scientific community must certainly have been full of expectation about the power of this new universal paradigm. The concrete results did not come so quickly and now after 300 years there are new discoveries in the field every year. Calculus though widely used is not yet used by the majority of people, and many very successful business people do not even know its existence (I know plenty of them). It is still not the solution to all problems either, and is now mainly used in technical problems, which is in the present time important enough.

Assimilating SD to a sort of calculus applied to about any problem, one can consider too that it is a new promising technique, as it is too very intellectually elegant and looks very powerful.

Suppose that one follows the path that you propose, the first thing people will ask is: show us what is has already resolved in the past.

The second thing they will ask is considering the huge amount problems our world tries to solve: what sort of problems does it solve and what are the conditions to use it (mainly time, expertise and general cost)?

So before trying to make SD more known, to my opinion the SD community should make an effort to list the good and bad experiences ingrained from the use of the method to determine the optimal conditions for a satisfactory use of the method. It must not be easy, while useful experiences may be hidden, because of the competition and unsatisfactory experiences can be hidden too, because of fear of publicity. The good experiences will be used as publicity and both good and bad will be used to define the sort of problems it can solve.

When I am looking for customers, I focus on those that I have the more chance to satisfy because it drives me away from the competition. So should do the SD community as long as the conditions of success have been well established. Instead of taking the risk to deceive and that can be very counterproductive, it is better to find the place where to use the tool, and to use it afterwards as a publicity. Unless one is Microsoft or works in a monopolistic situation, the best way to success is to have a sufficient number of customers that are satisfied.

Pretending that the tool is the general unifying method is pretentious and is to my opinion wrong at least in the actual knowledge of the field. All of us, would like the field be more recognized but hope is not sufficient to make things happen. Regards. J.J. Lauble Allocar Strasbourg France Posted by =?iso-8859-1?Q?Jean-Jacques_Laubl=E9?= <jean-jacques.lauble@wanadoo.fr> posting date Fri, 2 Dec 2005 16:21:23 +0100

northsheep juno.com
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

Characterizing System Dynamics

Post by northsheep juno.com » Sat Dec 03, 2005 12:58 pm


Jim Duggan JAMES.DUGGAN NUIGALWA
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

Characterizing System Dynamics

Post by Jim Duggan JAMES.DUGGAN NUIGALWA » Sun Dec 04, 2005 3:12 pm

Posted by Jim Duggan <JAMES.DUGGAN@NUIGALWAY.IE>
Hi Bill,

I'd agree with your analysis on emergence, in that it
is the overall behaviour of a system that results from
the subset of policies, and as such ""emerges"" from the
inherent dynamic complexity.

I think there are good examples of this in SD, that
also are discussed in more popular readings
on chaos and complexity. For example, in John Casti's
book ""Complexification"", he discusses in considerable
depth the Beer Game, and relates that to complexity ideas.

Also, in Malcolm Gladwell's ""The Tipping Point"", the dynamics he refers to are well captured and explained in SI/SIR models, which I also think are a good example of SD models showing emergent properties and surprising behaviour.

regards,
Jim.
Posted by Jim Duggan <JAMES.DUGGAN@NUIGALWAY.IE>
posting date Sun, 04 Dec 2005 09:27:14 -0500 (EST)

the invigilator the_invigilator
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

Characterizing System Dynamics

Post by the invigilator the_invigilator » Mon Dec 05, 2005 12:24 pm

Posted by ""the invigilator"" <the_invigilator@hotmail.com>
Hi Francisco,

While I am inexpert in most things (including agent-based modelling) it would appear to me that we (us who take an objective view of systems) need to take a step back from our work and see it in context of that one great overarching examplar of a successful 'human activity system'...democracy.

the trouble with democracy however is if the voters cast votes on the basis of ignorance, then you won't get a good government voted in (or you will have one who panders to their ignorance...i am sure you can think of a few examples :)

if *any* modelling science is to have a role to play in *improving* the operation of the system called democracy (and strengthen people's understanding of 'reality') it must be able to *talk* to the people.

this works dualistically...

what i mean is, people have to see their own experiences reflected in the model. it must resonate with them, and be accessible on their home computer when they are trying to explain to their kids why such-and-such a policy is a good/bad idea.

the flip side is that the model must be able to extract the people's own experiences, in whatever industry, or part of the economy/society, and incorporate them into the structure of the model. SD needs the information in their heads...we can't remain sitting on the top of our mountain retreats formulating models.

i think that the fundamental simplicity of SD...its reliance on 3 types of objects that can pretty much cover any possible system and its' accompanying diagrammatic representativeness allow it to fulfill both these requirements. In so doing, SD can improve the functioning of our democratic system. ie, people can see the underlying thinking behind an SD model simply by looking at its structure, and from there contribute to its' formulation. Sure, some learning is needed, but no more than a week of study to get the basics. This will happen when SD has *STATUS* - not before. Fear is not the reason why people do not accept us...

This is in stark contrast to other modelling sciences such as agent based modelling, where everything is 'emergent' and thus 'magical' and difficult to connect cause and effect. Or economics (even the latest reincarnation -
'econophysics') which remains impossible for any layman to penetrate - and i even doubt the 'experts' know what the hell they are doing. Yet, econophysics is supplying our politicians with the excuses they need to continue with their socially-corrupting policies while SD remains on the sidelines.

So, some may believe in slowly, slowly. I say bugger that!

Assume our responsibility. People need us. The biosphere needs us. Communities need us. Dare I say it, politicians need us.

Regardless, it'll be fun ;)

Invig
Posted by ""the invigilator"" <the_invigilator@hotmail.com> posting date Mon, 5 Dec 2005 11:44:07 +1100

Kim Warren Kim strategydynamics.
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

Characterizing System Dynamics

Post by Kim Warren Kim strategydynamics. » Tue Dec 06, 2005 11:51 am

Posted by ""Kim Warren"" <Kim@strategydynamics.com>
A number of people have asked for the paper mentioned below and the model of management migration through the systems thinking/SD field. Unfortunately, the article itself can only be obtained from the journal, but brief notes on this specific issue, and a supporting model are available in a zip file at www.strategydynamics.com/tsd

Kim

[ The model runs on the mystrategy software. There is a free ""Reader"" at www.strategydynamics.com/reader. Download the exe (1.2 mb) to your PC then double click to install. Open from Start | Programs | mystrategy. ]

Posted by ""Kim Warren"" <Kim@strategydynamics.com>
posting date Mon, 5 Dec 2005 17:26:13 -0000

Finn Jackson finn.jackson tangle
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

Characterizing System Dynamics

Post by Finn Jackson finn.jackson tangle » Tue Dec 06, 2005 11:56 am

Posted by ""Finn Jackson"" <finn.jackson@tangley.com>
Hello,

Everyone else on this discussion list uses their real name.

I do not see why ""The Invigilator"" should be any different.

Who are you?


Finn Jackson
Posted by ""Finn Jackson"" <finn.jackson@tangley.com>
posting date Mon, 5 Dec 2005 12:53:15 -0000

Weaver Elise A eweaver WPI.EDU
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

Characterizing System Dynamics

Post by Weaver Elise A eweaver WPI.EDU » Sun Dec 11, 2005 1:34 pm

Posted by ""Weaver, Elise A"" <eweaver@WPI.EDU>
My understanding of ""emergence"" is that it has to do with what we perceive at different levels of analysis. What we see at one level of analysis (or
aggregation) is seen differently at another. The classic example is that quick moving molecules at the molecular level are experienced as heat when the molecules are experienced collectively (as a solid, liquid, or gas, for instance).

Because this example of emergence is bottom up (from the molecules to the aggregate), I don't understand it as a system dynamics concept. Also critical to emergence is the idea of what we perceive. We experience heat ""emerging"" from moving molecules because of how we perceive moving molecules.


On the other hand, perhaps the heat example represents a limited meaning of emergence. Perhaps emergence includes any stabilized patterned behavior that emerges over time from interactions within a complex system. In that sense, emergence could be a system dynamics concept.

My conclusion: Whether ""emergence"" is part of the characterization of system dynamics depends on your definition of emergence.

Elise
Posted by ""Weaver, Elise A"" <eweaver@WPI.EDU>
posting date Sun, 11 Dec 2005 01:20:46 -0500

Green Stephen usksxg13 WPI.EDU
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

Characterizing System Dynamics

Post by Green Stephen usksxg13 WPI.EDU » Sun Dec 11, 2005 1:44 pm

Posted by ""Green, Stephen"" <usksxg13@WPI.EDU>
As a new person to SD I would definitely say that the things that attracted me to SD were two-fold.

1) Peter Senge's The Fifth Discipline
2) The Systems Thinker Newsletter

It is truly the simplicity of these two items that draw people in because they can understand the language and it has application to daily life and projects. It was only after starting courses at WPI that I discovered the whole host of computer modeling and its language (exogenous, endogenous, aggregate, etc.). We need ambassadors to translate this language to the rest of the world even though it is so simple and basic to us. In my circles in the corporate world or amongst friends these terms never come up in discussion (unless I introduce them.)

Stephen Green
Posted by ""Green, Stephen"" <usksxg13@WPI.EDU>
posting date Sat, 10 Dec 2005 09:33:15 -0500

Martin Taylor mmt mmtaylor.net
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

Characterizing System Dynamics

Post by Martin Taylor mmt mmtaylor.net » Mon Dec 12, 2005 12:13 pm

Posted by Martin Taylor <mmt@mmtaylor.net>

>>Posted by ""Weaver, Elise A"" <eweaver@WPI.EDU>
>>My understanding of ""emergence"" is that it has to do with what we
>>perceive at different levels of analysis. What we see at one level of
>>analysis (or


Sorry if this repeats something said earlier in the thread, which I haven't been following closely.

I guess that what Elise says could be an example of ""emergence"", but it doesn't conform to the way I think of it. Think of a simple feedback loop, in which the properties of the different links in the chain are all precisely specified as open-loop equations output = f(input). Put the units together, and the properties of the conjoined system can be precisely specified by equations that finally have the form x = f(g(h....(x and other variables)...)). Such equations can be solved, numerically if not analytically.

Someone in this or a related thread said that this solvability meant that the system dynamic was not an emergent property. I think it is, because what was not specified in the individual equations that specify the elements is their connectivity. ""Connectivity"" is not a property of any element, nor of any set of them. It is a property only of the structure as a whole.

Connectivity itself does not determine the behaviour of the structure. Put different elements together in the same structure, and you get different behaviour. The behaviour of the connected structure depends on a property of the structure that is not inherent in its components. To me, that makes the behaviour an emergent property of the structure.

Martin
Posted by Martin Taylor <mmt@mmtaylor.net>
posting date Sun, 11 Dec 2005 10:33:14 -0500

Locked