Replying to LO4425,
In my book, A SCIENCE OF GENERIC DESIGN, I define the structure of a
science as follows: Foundations, which drive Theory, which drives
Methodology. These three components feed into something that is not part
of a science, but is part of the domain of a science. This model,
comprised of four parts, is called the Domain of Science Model. This
model is used to structure and present the science of generic design.
In Figure 4.4 I present what I call the "Poly-loop Model". This model
consists of four feedback loops in a nested pattern. The model involves
gradations.
The 0-Loop is entered with a process plan for doing something. It acts on
a problem and gets result A.
The 1-Loop extends to include a desired result B. This time the result A
is compared with the desired result B. If they are not adequately
similar, the error is introduced for further action, but with the same
process. This is like the conventional single loop control system.
If, instead of continuing to work with the 1-Loop, it is finally decided
that the job isn't getting done, because the process (i.e., the
controller) is incapable of resolving the problem, then one goes to the
next loop.
The 2-Loop modifies the process plan, so that the next foray against the
problem uses the new process. The process is comparable to the
Methodology derived from the Domain of Science Model, i.e., from the
science that underpins the attempt to resolve a problem or issue. This
loop is the same as the Argyris loop. In Argyris' double-loop scheme,
people get tired of trying the same old thing and change the process
(methodology).
If the comparison done in this new scheme still produces significant
errors between what is achieved and what is sought, we then give up on the
2-Loop and go to the 3-Loop.
When the 3-Loop is entered, one assumes that since the Methodology has not
proved adequate, there may be something wrong with the theory. In this
loop, the theory is changed, and the attack continues.
If none of this works, one assumes that the Foundations must be at fault.
Thus one no changes the Foundation as indicated by study of the prior
failures.
When the Foundations are changed, a new science is discovered, or the
older science is amended.
All of this is consistent with Peirce's view of science.
Chapters 5, 6, and 7 in the book present, respectively, the Foundations,
the Theory, and the Methodology for the science of generic design.
I'm not familiar with Bateson's paradigm shift, but I think my formulation
is consistent with Bateson's. However I like mine better, because it
connects science to action in ways that minimize the necessity for the
practitioner to go back to fundamentals, while keeping that open as a
possible requirement. This means that the science itself can be
constantly assessed in applications and, when appropriate, amended. The
paradigm shift concept seems a little too vague to suit me.
One of the main reasons for proposing the domain of science model is that
we now have a need to integrate sciences. This need cannot be
well-addressed if we only talk science; but if we can get scientists to
break up their sciences into the three components, we have a much easier
job of integrating at the level of foundations, where the body of material
is much smaller than if we took on the whole science.
I keep hoping that there is a future for this system of thinking.
-- JOHN WARFIELD Johnwfield@aol.com