in the development of automated computation. Dynamic modeling was done before
then, but to escape the use by the skilled mathematicians, we had to wait for the
advent of machine computation, the development by Jay Forrester of the basic
computer logic, and the eventual emergence of this development to the personal
computer in the form of STELLA and other easy-to-use programming languages.
STELLA is a graphical programming process that evokes the most easily accessible form of symbolic understanding by humans, the use of icons. As you will find
throughout the text, the classification of variables is quite simple and the resulting
icons associated with them are quite appropriate for capturing all the parts that
influence a system’s behavior. An experienced STELLA modeler literally sees and
understands the dynamic process through the arrangement of these icons. This is part
of what we mean by the modeling art. Now the art can be practiced by anyone with
knowledge of basic mathematics and the ability to use a personal computer.
The only way to achieve the facile use of the icons of dynamic modeling is to use
them repeatedly in many different applications. Thus our books are a carefully
arranged set of models of biological processes that build on each other to various
degrees and make use of the same modeling tools in various contexts. We have tried
to arrange the models from simple to more complex and by scale, from small to
large—from simple growth models of a cell to a rather involved set of interacting
spatial ecosystems. As you work through these models you will not only become
fluent in the use of the modeling language STELLA but you will develop a new way
of thinking about dynamic systems. Practice is the foundation of the modeling art.
Throughout the examples, we try to show how the principal idea of one model
can be used again in a different application. The basic growth models are elements
of large models. The law of mass action in chemistry is used in epidemic and
ecosystem modeling. The play of analogy is dangerous in that it can be misleading
but the loss is never more than blush of embarrassment and a little electricity. Much
new science seems to ride on analogy: its skillful use is the final piece in the
construct of the modeler’s art.
The goals of dynamic modeling are to explain and, with enough effort and luck,
predict. The dynamic events occurring in the real world are multifaceted, interrelated and difficult, perhaps impossible, to understand. To reduce our worry and to
state our curiosity about such events, we pose and then try to answer specific
questions about the dynamic processes that seem to comprise these events. Of
necessity, we abstract from most of the details and attempt to concentrate on
some portion of the larger picture—a particular set of features of the real world.
The resulting models are true abstractions of reality. They force us to face the
results of the structural and dynamic assumptions we have used in our abstractions.
The modeling process is necessarily complicated and it is unending. Well-posed
questions lead us to develop a model—that model leads us to more questions. If we
are good at modeling, we improve our understanding of reality. If we are expert at
modeling, we approach reality asymptotically. To become expert, we make the
modeling process an iterative one: We observe what we call real events from
the world, we create an abstract version of these events, highlighting our view of
the important elements, then we build and run a dynamic model, we draw
1.1 Process and Art of Model Building
5
then, but to escape the use by the skilled mathematicians, we had to wait for the
advent of machine computation, the development by Jay Forrester of the basic
computer logic, and the eventual emergence of this development to the personal
computer in the form of STELLA and other easy-to-use programming languages.
STELLA is a graphical programming process that evokes the most easily accessible form of symbolic understanding by humans, the use of icons. As you will find
throughout the text, the classification of variables is quite simple and the resulting
icons associated with them are quite appropriate for capturing all the parts that
influence a system’s behavior. An experienced STELLA modeler literally sees and
understands the dynamic process through the arrangement of these icons. This is part
of what we mean by the modeling art. Now the art can be practiced by anyone with
knowledge of basic mathematics and the ability to use a personal computer.
The only way to achieve the facile use of the icons of dynamic modeling is to use
them repeatedly in many different applications. Thus our books are a carefully
arranged set of models of biological processes that build on each other to various
degrees and make use of the same modeling tools in various contexts. We have tried
to arrange the models from simple to more complex and by scale, from small to
large—from simple growth models of a cell to a rather involved set of interacting
spatial ecosystems. As you work through these models you will not only become
fluent in the use of the modeling language STELLA but you will develop a new way
of thinking about dynamic systems. Practice is the foundation of the modeling art.
Throughout the examples, we try to show how the principal idea of one model
can be used again in a different application. The basic growth models are elements
of large models. The law of mass action in chemistry is used in epidemic and
ecosystem modeling. The play of analogy is dangerous in that it can be misleading
but the loss is never more than blush of embarrassment and a little electricity. Much
new science seems to ride on analogy: its skillful use is the final piece in the
construct of the modeler’s art.
The goals of dynamic modeling are to explain and, with enough effort and luck,
predict. The dynamic events occurring in the real world are multifaceted, interrelated and difficult, perhaps impossible, to understand. To reduce our worry and to
state our curiosity about such events, we pose and then try to answer specific
questions about the dynamic processes that seem to comprise these events. Of
necessity, we abstract from most of the details and attempt to concentrate on
some portion of the larger picture—a particular set of features of the real world.
The resulting models are true abstractions of reality. They force us to face the
results of the structural and dynamic assumptions we have used in our abstractions.
The modeling process is necessarily complicated and it is unending. Well-posed
questions lead us to develop a model—that model leads us to more questions. If we
are good at modeling, we improve our understanding of reality. If we are expert at
modeling, we approach reality asymptotically. To become expert, we make the
modeling process an iterative one: We observe what we call real events from
the world, we create an abstract version of these events, highlighting our view of
the important elements, then we build and run a dynamic model, we draw
1.1 Process and Art of Model Building
5