SD and system design

[geert nijland]

Dear system dynamicists,

Who of the SD-community could give me literature or personal information
about methods of designing new systems, departing from the framework of
system dynamics thinking?

Our department wants to develop a course in designing new sustainable
agricultural systems on an ecological base. The system should exhibit some
desired behavior with regard to selected goal-variables, and it should
exhibit that behaviour within a number of prior limiting conditions. (the
idea is that the problematic behaviour of the conventional agricultural
system cannot be overcome by policy-measures, so that complete new system
stuctures are needed).

The method of designing we look for should be as straight forward, systematic
and reproducible as possible.

In systems analysis we always ask for the system behaviour which can be
inferred from the system structure. We start by modelling a part of reallity
with some problem (and goals and instuments) in mind and look for the
behaviour of the structure and analyze which manipulations with which
decision variables may remove the problematic behaviour, and which are the
new side effects.

In system design the question is just the other way around. We start from a
desired behaviour of goal-variables and ask for a structure (mainly to be
composed of existing elements in the real ecological, social and technical
world), which exhibit that behaviour - within selected limiting conditions.

Up to now we found some systematical approaches to design:
In the technical sciences methods exist for a more systematic way of
designing machines, buildings, towns. Ill try to adapt such methods for our
own goals, but from the literature I got the impression that the dynamical
aspects are neglected somewhat. It is true that socio-ecological systems are
very different from technical machines, but their methods might be usefull
for designing ecologic and socio-economic systems as well.

Some of my collegues are working with linear programming and similar
optimization tools. Has any of you the opinion that for the design of systems
linear programming is a more promising tool than system dynamics ?

Other collegues are working much more intuitively and qualitatively. I look
for a more logical, quantitative approach.

I should be glad to get some fruitful ideas to proceed,

best wishes,

Geert Nijland,
Geert.Nijland@USERS.ECO.WAU.NL

Wageningen Agricultural University,
Department of Ecological Agriculture,
Haarweg 333 / 6709 RZ Wageningen, The Netherlands

["Robert M. Kane"]

Geert,

I wrote a net response to your message as it related to others just previously
transmitted, but also would like to send along a few other thoughts specific
to your effort.

I am a Systems Engineer with years of experience in military systems. This
perspective has taught me to view the problem of sustainable system design in
a little different light than my previous experience allowed.

The most serious problem in design of a system of people and life resources is
that it really cant be done, as such. I mean that the system in question is
in motion and has been in motion since the dawn of history. The system exists
and at best we might hope to induce it along a path. The question remains how
to identify the path and, having identified the path, how to induce the
autonomous players to choose that path.

This means that system design must be preceeded by a rigorous system analysis
and the design of system will have implications relative to intervention
through motivation on individuals. The alternative is totalitarian command
economies and weve already experimented in this century.

I have no answers but I believe that the problem must be worked interactively
through action based on analysis. Its too easy to build castles in the sky
and to discuss the molecular behaviors of imaginary particles. Human nature
reverts to comfort in the face of thorny issues and there is no more
comforting diversion than isolated and uncritical abstraction.

I am very pleased to see your objective to be so closely aligned with our own
and wish you the best of luck.

Thanks for your ideas and for sharing your plans.

Bob Kane, SIAT North America
robert_kane@corp.cubic.com