REPLY Models, Problems and Systems (SD7023)

[SDMAIL George Richardson]

Posted by  George Richardson <gpr at albany.edu>

The suggestion that we can "model systems" seems to me like the 
assignments that buffaloed me when I was an undergraduate:  "Write a 
2000-word essay on Hamlet," or "...19th century England," or "...the 
U.S. criminal justice system."  I remember doing very badly on such 
assignments and thinking it was because I couldn't write.

Later I realized it was not because I couldn't write but because I had 
nothing to say.  I needed a more targeted problem statement before I 
could have thoughts worth writing down, and as an undergraduate I didn't 
have tools to help me craft a more targeted writing purpose.

"Build a model of New York City" would be something like these undergrad 
essay assignments that left me clueless.  That's why I tend to be one of 
those that repeats the mantra that we can't model systems but must model 
problems.  We'd have to know what it is about NY City that we would be 
trying to address.  I think of whatever that is as "the problem."

We, and our clients and students, need tools to help us focus.  The 
tools that I suggest for my students to begin a system dynamics study 
are named in the following list:

*Problem focus
*Problem dynamics
*Context
*Audience
*Model purposes
Model boundaries
    Temporal - what's the time horizon?
    Conceptual - what's included and what's excluded?
    Causal - what's endogenous and what's exogenous?
Aggregation
Reference modes
Initial policy options
Model sectors
Important processes in each sector
Important levels and associated rates in each process and/or sector
Apparently important feedback loops
Next steps

The first five are starred because they seem to be the really crucial 
ones, involved in almost anyone's suggestions for how to begin to tackle 
a serious study.  The unstarred ones that follow are more specially 
useful for projects that are aimed at building a system dynamics model 
or a well-grounded stock-and-flow/feedback map.

These steps or stages, or something like them, move us pretty quickly 
from thoughts about modeling "a system" to thoughts about what it is 
that we or our clients really want to understand.

One last thought about "systems":  Systems theorists and thinkers have 
historically had a tough time defining what a "system" is.  It has 
helped me to think that the pictures, maps, and models that emerge from 
the stages listed above are "the system" we are talking about.  I'd even 
go so far as to say "the system" doesn't exist until we think it into 
existence.  As Adam Smith once so insightfully said, "a system is an 
imaginary machine invented to connect together in the fancy those 
different movements and effects which are already in reality 
performed."  We don't know "the system" until we think about something 
systemically.

We should be able to drop the discussion of modeling a problem versus 
modeling a system, in favor of talking wisely about how smart systems 
thinkers and modelers suggest how we should go about getting started on 
a study.  The wisdom we need is in those details, not in "problem" 
versus "system."

..George

George P. Richardson
Chair of public administration and policy
Rockefeller College of Public Affairs and Policy
University at Albany - SUNY, Albany, NY 12222
Posted by  George Richardson <gpr at albany.edu>
posting date  Sat, 10 May 2008 09:24:20 -0400