IFSR International Series in Systems Science and Systems

Engineering

Jeffrey Yi-Lin Forrest

General
Systems
Theory
Foundation, Intuition and Applications
in Business Decision Making
IFSR International Series in Systems Science
and Systems Engineering

Volume 32

Editor-in-chief
George E. Mobus, Institute of Technology, University of Washington Tacoma,
Tacoma, WA, USA
More information about this series at http://www.springer.com/series/6104
Jeffrey Yi-Lin Forrest

General Systems Theory

Foundation, Intuition and Applications
in Business Decision Making

123
Jeffrey Yi-Lin Forrest
School of Business
Slippery Rock University
Slippery Rock, PA, USA

ISSN 1574-0463
IFSR International Series in Systems Science and Systems Engineering
ISBN 978-3-030-04557-9 ISBN 978-3-030-04558-6 (eBook)
https://doi.org/10.1007/978-3-030-04558-6
Library of Congress Control Number: 2018961727

Mathematics Subject Classification (2010): 93

© Springer Nature Switzerland AG 2018

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface of the Current Edition

Since the first edition, entitled “General Systems Theory: A Mathematical

Approach,” was initially published in 1999, many magnificent and important
developments have been made at various fronts of systems science. So, at this very
special moment of inaugurating the new leadership for the IFSR International Series
on Systems Science and Engineering, it is definitely a good time for me to update
the previous edition by enriching it with some of the most important works pub-
lished since the start of this new century.
Because of my background and training in mathematics and my strong curiosity
on a wide range of topics, as soon as I had my hands on the idea of and work on
systems science by some of the best minds of our modern time in the early 1980s, I
immediately became and have ever since been a faithful follower of the systems
movement. I have been firmly convinced that systems research will provide a new
dimension to the existing spectrum of science. Such belief was later argued com-
prehensively and spelled out clearly by George Klir in the early 1990s. However,
with nearly one hundred years of development, systems science still needs to
enhance itself in many different aspects in order to fully meet the challenges posed
by several scholars, including David Berlinski (in his book On Systems Analysis,
MIT Press, Cambridge, Massachusetts, 1976). In particular, other than addressing a
set of new problems that were not possible to be considered in the past within the
scope of the traditional science, systems science has to be able to provide solutions
to problems, be them age old or modern, considered in various disciplines of the
traditional science with its unique intuition, logic of thinking, and technical
approach. Only with such a scientific capability, systems science will be affec-
tionately embraced by the entire world of learning.
That means, in other words, that systems science needs to develop its own
intuition and playground, like the one in modern science, known as Cartesian
coordinate system, on which most important conclusions and results, be them
systems science specific or not, can be visually seen by all educated eyes. For
example, in modern science basic methods of analytical reasoning are all estab-
lished on the Cartesian coordinate system of various dimensions; all statistical tools
are developed on the top of various combinations, again Cartesian coordinate

v
vi Preface of the Current Edition

system based, of algebra and geometry. With such a playground and tool for
intuitive reasoning in place, systemic logic of thinking can be systematically
developed so that particular technical methods of reasoning will be established
accordingly. In other words, to realize the great promise of systems science, there is
a need to make this new science possess a glorious and long-lasting life. That means
this new science needs to satisfy the following four conditions (Lin, Y. (2009),
Systemic Yoyo: Some Impacts of the Second Dimension, CRC Press, New York):
1. It must be readable by as many people as possible.
2. It must coincide with people’s intuition.
3. It must possess a certain kind of beauty, which can be easily felt.
4. It must be capable of producing meaningful results and insights that excite the
population.
In the traditional science, calculus satisfies these conditions with its beauty
manifested and intuition played out on the Cartesian coordinate system. Euclidean
geometry is also a long living theory, because it is intuitive, possesses a logical and
visual beauty, and can be employed in people’s daily lives. In comparison, systems
science did not have its unique playground and intuition, before at least of the start
of this new century, through which the beauty of the science can be easily felt, and
on which meaningful results and insights can be established.
To improve the general systems theory, as presented in the first edition published
in 1999, this volume will introduce the systemic yoyo model as the desired play-
ground and intuition for systemic reasoning and thinking. Other than presenting the
elementary properties of this model and how general systems would follow the laws
on state of motion, we will look at the theoretical reasons why such a model for a
general system would hold true, followed by several empirical supports.
To satisfy Condition 4 listed above, this book will consider several issues of
business decision making and see how meaningful results and insights that excite
the population can be produced by using the systemic yoyo model and related
systems reasoning. In particular, we will closely look at the following topics:
1. Why sustainable competitive advantages are becoming transient, while markets
change faster and customers are less patient than ever before?
2. How nonlinearity appears in demand/supply interactions;
3. When new market competitions appear;
4. Some mysteries of the family;
5. Why no fixed value system will ever be perfect;
6. How to potentially achieve management efficiency;
7. Why organizational inefficiency always exists;
8. How to deal with customers’ indecisiveness through pricing strategies;
9. How to heighten the competitive spirits of sales’ associates; etc.
The reason why I choose various scenarios of business decision making as the
showcase of successful and exciting applications is because such words as “be-
lieve,” “should,” and “would” are widely used in the literature of economics and
Preface of the Current Edition vii

business. To me, what that reality means is that the theoretical reasoning that led to
the conclusions in the first place is not reliable, and the author knows about it. On
the contrary, these words are not generally used in natural sciences and mathe-
matics. For example, scientists generally do not use these words when they speak of
scientific facts. For instance, no one in science would speak as follows: When a
cubic solid piece of iron is dropped in the water in a container, I believe that the
metal block would sink to the bottom of the container. Instead, people simple state
the fact as follows: The metal block will sink to the bottom of the container (without
using the words believe and would)!!! What happens in the research of economics
and business when compared to that of natural sciences and mathematics is that
1. When a rigorous tool of reasoning is used, the author does not know for sure
whether or not he/she has considered all related factors.
2. A lot of conclusions are drawn based on one or a few anecdotes without any
follow-up rigorous analysis.
3. A lot of so-called theories are really established on data mining.
One of the reasons for Situation 1 to occur is that there is no readily available
playground or intuition for large-scale or global economic reasoning. So, when a
rigorous or analytical tool is employed, it tends to only model a regional, local
phenomenon so that conclusions of a larger scale can only be conjectured. When
Situation 2 is the case, the conclusions are really derived inductively instead of
deductively. To this end, there are many counterexamples in mathematics that show
the fact that observations based on anecdotes generally do not lead to reliable
theories. For example, from the anecdote 02 ¼ 0 can one conclude that n2 ¼ n, for
any natural number n? Of course not. Situation 3 above is similar to the case that we
look at the clouds in the sky and ask each other: Do you see a person in the clouds
who is smiling at us right now? Even when all people agree with you on the
observation, one still cannot conclude that there is indeed a person in the sky who is
smiling at people on earth. As a matter of fact, whatever methods used in Situations
2 and 3 above are only employed to reveal potential facts. These potential facts still
need to be shown scientifically. For details about this end, see Lin Y. and
OuYang S. C. (2010), Irregularities and Prediction of Major Disasters, New York:
CRC Press.
If we identify the current state of research affairs in economics and business with
that in the history of natural sciences right around the time when Isaac Newton
developed his laws of physics, one can see such a possibility of great likelihood to
happen: All the facts and nonfacts discovered in the literature of economics and
business are similar to those uncovered by various scholars in natural sciences a few
hundred years ago, and the present opportunity created by the development of
systems thinking and the systemic intuition in the last century is quite parallel to
that used by Newton to develop his laws of physics, because of the matrimony of
Euclidean geometry and algebra, as a natural consequence of the introduction of the
Cartesian coordinate system. So, considering the massive amount of facts and
nonfacts discovered currently, one can expect a major breakthrough in areas of
viii Preface of the Current Edition

economics and business is coming soon. And, because of the fact of how systems
science and the yoyo model can provide effective means for large-scale analysis and
intuitive thinking, I expect that when such a breakthrough appears, it will be sys-
tems science based.
Summarizing what is discussed above, this volume is a follow-up edition of my
previous book, entitled “General Systems Theory: A Mathematical Approach,”
initially published in 1999. And more than two thirds of the contents in this current
book are based on recent developments in the relevant areas of research. In par-
ticular, this new edition consists of three parts, entitled The Foundation, The
Systemic Intuition, and Applications in Business Decision Making, respectively.
Other than some slight changes and additions, the first part is on the foundation of
general systems theory and is mainly based on the earlier edition. The second part
introduces the systemic yoyo model to satisfy the desperate need for a practically
useful intuition for reasoning in systems science specific contents, as mentioned
previously. The role to be played by this model is expected to be similar to that
played by the Cartesian coordinate system in the classical science, on top of which
all statistical and analytical methods, widely and heavily employed in modern
science, are developed. The third part looks at how systemic thinking and the yoyo
model can be beautifully applied to address many important problems facing
decision makers in business by organically combining methods of the classical
science, the first dimension of the 2D spectrum of science, with those of systems
science, the second dimension, as argued by George Klir in the 1990s. As of this
writing, important decisions in business are mostly drawn based on data mining or
anecdotes. Scientifically speaking, such processes of decision making have been
time and again shown to be flawed. Because of this reason, this part of the book is
expected to open up a brand new territory of research valuable for decision-making
managers and working economics professionals.
By presenting a rigorously developed foundation, a tool for intuitive reasoning
that is supported by both theory and empirical evidence, and extremely fruitful
applications in economics and business, this book demonstrates the theoretical
value and practical significance of systems science and its thinking logic. By
making use of this science and by employing the systemic intuition, one can pro-
duce interesting, convincing, and scientifically sound results. As shown in this
book, many of the conclusions drawn on the basis of systems science can be
practically applied to produce tangible economic benefits.
To comprehend the following chapters in this book, here are the necessary
mathematical knowledge. Since Part 1 of this book is mostly from the first edition,
entitled “General Systems Theory: A Mathematical Approach,” published by
Kluwer Academic Publishers in 1999, as mentioned before in that edition, the
mathematical background needed here is a beginner’s basic understanding of the
naïve and axiomatic set theories. On the other hand, if the reader likes to make
contributions along the lines given in this part of the proposed book, then he/she
will need a better comprehension of these set theories in order to derive insightful
conclusions. To satisfy this in-depth need, the reader is advised to consult with the
previous editor of this book, where more than sufficient depth of these set theories is
Preface of the Current Edition ix

provided in a single place. And for the rest of this book, calculus and some basic
knowledge of game theory will be sufficient for the reader to understand what is
presented.
By studying this book and by referencing back to it regularly, you, the reader,
will master a brand new tool to resolve your problems and an intuition from which
important insights can be acquired intuitively and useful decisions can be made
relatively quickly without wasting unnecessarily your valuable time and a lot of the
limited financial resources.
I hope you will enjoy reading and referencing this book in your scientific
exploration, academic pursuit, and real-life practice. If you have any comments or
suggestions, please let me hear from you by joining several thousands of other
colleagues who have communicated with either me or other members of my research
groups. I can be reached at jeffrey.forrest@sru.edu or jeffrey.forrest@yahoo.com.

Slippery Rock, USA Jeffrey Yi-Lin Forrest

Preface of the First Edition

As suggested by the title of this book, I will present a collection of coherently

related applications and a theoretical development of a general systems theory.
Hopefully, this book will invite all readers to sample an exciting and challenging
(even fun!) piece of interdisciplinary research, that has characterized the scientific
and technological achievements of the twentieth century. And, I hope that many
of them will be motivated to do additional reading and to contribute to topics along
the lines described in the following pages.
Since the applications in this volume range through many scientific disciplines,
from sociology to atomic physics, from Einstein’s relativity theory to Dirac’s
quantum mechanics, from optimization theory to unreasonable effectiveness of
mathematics to foundations of mathematical modeling, from general systems theory
to Schwartz’s distributions, special care has been given to write each application in
a language appropriate to that field. That is, mathematical symbols and abstractions
are used at different levels so that readers in various fields will find it possible to
read. Also, because of the wide range of applications, each chapter has been written
so that, in general, there is no need to reference a different chapter in order to
understand a specific application. At the same time, if a reader has the desire to go
through the entire book without skipping any chapter, it is strongly suggested to
refer back to Chapters 2 and 3 as often as possible.
The motivation to write this book came from the strong influence of historical
works by L. von Bertalanffy, George Klir, and M. D. Mesarovic, and the book On
Systems Analysis, by David Berlinski (MIT Press, Cambridge, Massachusetts,
1976). Berlinski’s book and challenges from several scholars really made me decide
to write such a book with strong applications in different scientific fields in order to
justify the very meaning of existence for a general systems theory. At the same
time, one of the important lessons we have learned from the several decade-old
global systems movement, started and supported by many of the most powerful
minds of our modern time, is that senseless transfer of statements (more specifi-
cally, theoretical conclusions or results) from one discipline to another makes
people feel that general systems theory is a doubtful subject. To keep such
unnecessary situations from occurring, we develop each application with rigorous

xi
xii Preface of the First Edition

logical reasoning. Whenever a bold conclusion is deduced, some relevant gaps in

the reasoning process will be pointed out right on the spot or in the final chapter
(“Some Unsolved Problems in General Systems Theory”). On the other hand,
doubtful people will be as doubtful as they can no matter what facts or evidence are
out there to show them their doubt is unfounded. For example, more than 100 years
ago, when naive set theory was first introduced and studied, many first-class
mathematicians did not treat it as a serious theory at all. Furthermore, Cantor, the
founder, was personally attacked by these scholars. As a consequence, he was
hospitalized and eventually died in a psychiatric hospital. Today, set theory has
succeeded in a great many areas of modern science, including the entire spectrum of
mathematics, when the central idea of infinity is employed in systems science, we
can still hear doubters saying things like: Infinity? One can be sure that in an
infinitely long period of time, a monkey will produce the great Beethoven’s music!
(A note: according to results in set theory, this statement is not true!)
The structure of my theoretical development in this book is the “top-down”—
formalization—approach, launched in 1960 by Mesarovic. This approach is char-
acterized by the following: (1) All concepts are introduced with minimal mathe-
matical structures. (2) Additional mathematical conditions are added when
necessary to display the richness of systems properties. At the same time, appli-
cability is always used to test the mathematical conditions added.
Calculus is all that is needed to comprehend this book, since all other mathe-
matical techniques are presented at appropriate levels.
Finally, I would to express my sincere appreciation to many individuals, too
many to list. My thanks go to President Robert Aebersold and Vice President and
Provost Charles Foust, Deans Charles Zuzak and Jay Harper of Slippery Rock
University, Pennsylvania, whose academic support for the past several years was
essential to finishing this book. I thank Dr. Ben Fitzpatrick, my Ph.D. supervisor,
for his years’ teaching and academic influence, Prof. Lotfi Zadeh, the father of fuzzy
mathematics, for his keen encouragement, Prof. Xavier J. R. Avula, President of the
International Association for Mathematical and Computer Modeling, for his per-
sonal influence and education on professional perfection for the past several years.
I hope you enjoy reading and referencing this book, and your comments and
suggestions are welcome! Please let me hear from you—my e-mail address is
jeffrey.forrest@sru.edu.

Jeffrey Yi-Lin Forrest

Acknowledgements

This book contains many research results previously published in various sources,
and I am grateful to the copyright owners for permitting me to use the material.
They include the International Association for Cybernetics (Namur, Belgium),
Gordon and Breach Science Publishers (Yverdon, Switzerland, and New York),
Hemisphere (New York), International Federation for Systems Research (Vienna,
Austria), International Institute for General Systems Studies, Inc. (Slippery Rock,
Pennsylvania), Kluwer Academic and Plenum Publishers (Dordrecht, The
Netherlands, and New York), MCB University Press (Bradford, UK), Pergamon
Journals, Ltd. (Oxford), Springer Nature, Taylor and Francis, Ltd., World Scientific
Press, Wroclaw Technical University Press (Poland), and Emerald Publishing, UK.
I express my sincere appreciation to many individuals who have helped to shape
my life, career, and profession. Because there are so many of these wonderful
people from all over the world, I will just mention a few. Even though Dr. Ben
Fitzpatrick, my Ph.D. degree supervisor, had left this material world, he will forever
live in my works. His teaching and academic influence will continue to guide me
for the rest of my professional life. My heartfelt thanks go to Shutang Wang, my
M.S. degree supervisor. Because of him, I always feel obligated to push myself
further and work harder to climb high up the mountain of knowledge and to swim
far into the ocean of learning. To George Klir—from him I acquired my initial sense
of academic inspiration and found the direction in my career. To Mihajlo
D. Mesarovic and Yasuhiko Takaraha—from them I affirmed their chosen endeavor
in my academic career. To Lotfi A. Zadeh—with personal encouragements and
appraisal words from which I was further inspired to achieve high scholastically. To
Shoucheng OuYang and colleagues in our research group, named Blown-Up
Studies, based on our joint works, Yong Wu and I came up with the systemic yoyo
model, which eventually led to the completion of the earlier book Systemic Yoyos:
Some Impacts of the Second Dimension (published by CRC Press as an Auerbach

xiii
xiv Acknowledgements

book, an imprint of Taylor and Francis in 2008) and this book. To Zhenqiu Ren—
with him I established the law of conservation of informational infrastructure. To
Gary Becker, a Nobel laureate in economics, his rotten kid theorem has brought me
deeply into economics, finance, and corporate governance, from which some of
contents of this book are created jointly by Dillon Forrest and me.
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Historical Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Whole Evolutions—Blown-Ups and Revolving Currents . . . . . . 6
1.3 Systems Philosophy and Methodology . . . . . . . . . . . . . . . . . . . 9
1.4 The Reason for Looking at Applications in Economics
and Business Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 12
1.5 Organization of Contents in This Book . . . . . . . . . . . . . ..... 16
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 19

Part I The Foundation

2 The Concept of General Systems . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1 Various Attempts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1.1 The Concept of Systems . . . . . . . . . . . . . . . . . . . . . . . 26
2.1.2 The Set-Theoretical Definition of General Systems . . . . 30
2.1.3 Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.1.4 L-Fuzzy Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.1.5 G-Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.2 Internalities of Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.3 The Edifice of Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3 Mappings and Constructions of Systems . . . . . . . . . . . . . . . . . . . . . 41
3.1 Mappings from Systems into Systems . . . . . . . . . . . . . . . . . . . 41
3.2 Free Sums . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.3 Products and Cartesian Products . . . . . . . . . . . . . . . . . . . . . . . 48
3.4 One-Level Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

xv
xvi Contents

4 Connectedness . . . . . . . . . . . . . . . . ......... . . . . . . . . . . . . . . . . 57
4.1 Connected Systems . . . . . . . . ......... . . . . . . . . . . . . . . . . 57
4.2 Connectedness of Constructed Systems . . . . . . . . . . . . . . . . . . 61
References . . . . . . . . . . . . . . . . . . . ......... . . . . . . . . . . . . . . . . 67
5 Hierarchies—Construction of Large-Scale Systems . . . . . . . . . . . . . 69
5.1 Existence of Linkages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.2 Mappings from Hierarchies into Hierarchies . . . . . . . . . . . . . . . 74
5.3 Time Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
6 Controllabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.1 Controllabilities of Input–Output Systems . . . . . . . . . . . . . . . . . 81
6.2 General Controllabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
6.3 Controllability of Time Systems . . . . . . . . . . . . . . . . . . . . . . . 89
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
7 Limit Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
7.1 Existence of Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
7.2 Induced Mappings Between Limits . . . . . . . . . . . . . . . . . . . . . 97
7.3 Centralizability of Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
8 Systems of Single Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
8.1 Chaos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
8.2 Attractors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
9 The Feedback Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
9.1 Linear Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
9.2 Feedback Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
10 Properties Invariant Under Feedback . . . . . . . . . . . . . . . . . . . . . . . 121
10.1 Linearity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
10.2 Mesarovic–Takahara Time Systems . . . . . . . . . . . . . . . . . . . . . 123
10.3 Time-Invariable Realization . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
11 Decoupling of Single-Relation Systems . . . . . . . . . . . . . . . . . . . . . . 131
11.1 The Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
11.2 Transferrable Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
11.3 Decoupling Through Feedback . . . . . . . . . . . . . . . . . . . . . . . . 137
11.4 An Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Contents xvii

12 Decomposability Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

12.1 Global and Complex System Representations . . . . . . . . . . . . . . 145
12.2 Isomorphisms and Monomorphisms . . . . . . . . . . . . . . . . . . . . . 149
12.3 Input–Output System Covering . . . . . . . . . . . . . . . . . . . . . . . . 152
12.4 Surjective Complex System Representation of Order n . . . . . . . 156
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162

Part II The Systemic Intuition

13 How Systems Could Be Intuitively Seen . . . . . . . . . . . . . . . . . . . . . 165
13.1 The Systemic Yoyo Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
13.2 Properties of Systemic Yoyo Fields . . . . . . . . . . . . . . . . . . . . . 172
13.3 Interactions of Systemic Fields . . . . . . . . . . . . . . . . . . . . . . . . 176
13.4 Laws on State of Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
13.5 Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
13.5.1 How Calculus, a Pinnacle Achievement
of Science, Was Made Possible? . . . . . . . . . . . . . . . . . 187
13.5.2 Usefulness of Intuition . . . . . . . . . . . . . . . . . . . . . . . . 188
13.5.3 Validity of Figurative Analysis . . . . . . . . . . . . . . . . . . 189
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
14 Whole Evolutions, Where Systemic Yoyos Come from . . . . . . . . . . 193
14.1 Blown-Ups: Moments of Transition in Evolutions . . . . . . . . . . 193
14.2 Properties of Blown-Ups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
14.2.1 Blown-Ups of Quadratic Models . . . . . . . . . . . . . . . . . 196
14.2.2 Blown-Up Properties of nth Degree Polynomial
Models, n  3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
14.2.3 Blown-up of Nonlinear Time–Space Evolution
Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
14.3 Issues Facing Quantitative Infinity . . . . . . . . . . . . . . . . . . . . . . 203
14.3.1 An Implicit Transformation Between Euclidean
Spaces and Curvature Spaces . . . . . . . . . . . . . . . . . . . 204
14.3.2 Motions Represented by General Dynamic System . . . . 207
14.3.3 Wave Motions Versus Eddy Motions . . . . . . . . . . . . . 209
14.4 Equal Quantitative Movements and Effects . . . . . . . . . . . . . . . . 212
14.5 Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
15 Some Empirical Justifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
15.1 Eddy Effects Caused by Changes in Density of Media
of Movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
15.2 Conservation of Informational Infrastructure . . . . . . . . . . . . . . . 226
xviii Contents

15.2.1 Physical Essence of Dirac’s Large Number

Hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
15.2.2 The Angular Momentum Mystery of the Solar
System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
15.2.3 Movements of the Atmosphere on Earth . . . . . . . . . . . 232
15.2.4 Conjecture of a Law of Conservation of Informational
Infrastructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
15.2.5 Impacts of the New Law of Conservation . . . . . . . . . . 235
15.3 Bodily Communications that Can Be Felt Undoubtedly . . . . . . . 237
15.3.1 Visual Stares . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
15.3.2 Reciprocal Impressions of People . . . . . . . . . . . . . . . . 238
15.4 Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240

Part III Applications in Business Decision Making

16 Economic Events and Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
16.1 How Market Demand and Supply Interact with Each Other . . . . 244
16.2 Model Economic Cycles by Using Systemic Yoyos . . . . . . . . . 247
16.3 Emergence of Market Competitions . . . . . . . . . . . . . . . . . . . . . 249
16.4 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
17 Issues of Competition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
17.1 Some Background Information . . . . . . . . . . . . . . . . . . . . . . . . 258
17.2 The Market Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
17.3 Coordinated Monopoly and Stagnation in Expected Profits . . . . 264
17.4 Markets Change Faster and Customers Less Patient . . . . . . . . . 268
17.5 Once Sustainable Competitive Advantages Become
Transient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
17.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
18 Some Issues of the Family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
18.1 How Could Selfish People Behave Unselfishly . . . . . . . . . . . . . 277
18.2 Two Mysterious Phenomena of the Family . . . . . . . . . . . . . . . . 285
18.3 Why Is Any Established Value System Imperfect? . . . . . . . . . . 290
18.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
19 Some Remarks on Organizational Efficiencies . . . . . . . . . . . . . . . . . 299
19.1 The Question that Is to Be Addressed . . . . . . . . . . . . . . . . . . . 300
19.2 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
19.3 Individuals’ Underlying Philosophical and Value Systems . . . . . 302
19.4 Forever Existence of Inefficiency . . . . . . . . . . . . . . . . . . . . . . . 306
Contents xix

19.5 One Possible Way to Achieve Managerial Efficiency . . . . . . . . 312

19.6 Principle of Organizational Inefficiency . . . . . . . . . . . . . . . . . . 313
19.7 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
20 Dealing with Indecisive Customers Through Pricing . . . . . . . . . . . . 321
20.1 One Strategic Firm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
20.2 All Firms Are Strategic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
20.3 Which Price Strategy Leads to Higher Expected Profits? . . . . . . 331
20.4 Some Final Words . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
21 An Idea on How to Heighten the Competitive Spirits of Sales
Associates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
21.1 Sales Associates’ Compensation . . . . . . . . . . . . . . . . . . . . . . . 339
21.2 Associates’ Sales at Cournot Equilibrium . . . . . . . . . . . . . . . . . 341
21.3 Associates’ Payoffs at Cournot Equilibrium . . . . . . . . . . . . . . . 344
21.4 Private Cost Information and Expected Incomes . . . . . . . . . . . . 345
21.5 Associates’ Incentive Weights . . . . . . . . . . . . . . . . . . . . . . . . . 347
21.6 Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
22 Reducing the Adverse Effects of Speculative Capital . . . . . . . . . . . 351
22.1 Categorizing Purchasing Power . . . . . . . . . . . . . . . . . . . . . . . . 352
22.2 Economic Sectors Defined Through Feedback Mechanism . . . . 354
22.3 The Main Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
22.4 A Strategy for National Defense . . . . . . . . . . . . . . . . . . . . . . . 359
22.5 A Few Final Words . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
About the Author

Dr. Jeffrey Yi-Lin Forrest also known as Yi Lin,

holds all his educational degrees in pure mathematics
and had one-year postdoctoral experience in statistics at
Carnegie Mellon University. He had been a guest
professor of economics, finance, and systems science at
several major universities in China, including Nanjing
University of Aeronautics and Astronautics. And
currently, he is a professor of mathematics at Slippery
Rock University, Pennsylvania, and the president of the
International Institute for General Systems Studies,
Inc., Pennsylvania. He serves either currently or in the
past on the editorial boards of eleven professional
journals, including Kybernetes: The International
Journal of Systems, Cybernetics and Management
Science, Journal of Systems Science and Complexity,
International Journal of General Systems, etc. Some of
his research was funded by United Nations, State of
Pennsylvania, National Science Foundation of China,
and German National Research Center for Information
Architecture and Software Technology. As of the end
of 2016, he has published well over 300 research papers
and nearly 50 monographs and special topic volumes.
Some of these monographs and volumes were pub-
lished by such prestigious publishers as Springer,
World Scientific, Kluwer Academic, Academic Press,
etc. Over the years, his scientific achievements have
been recognized by various professional organizations
and academic publishers. In 2001, he was inducted into
the Honorary Fellowship of the World Organization of

xxi
xxii About the Author

Systems and Cybernetics. His research interests are

wide-ranging, covering areas like economics, finance,
management, marketing, data analysis, predictions,
mathematics, systems research and applications, phi-
losophy of science, etc.
Synopsis

This book is a follow-up edition of my previous book, entitled “General Systems

Theory: A Mathematical Approach,” initially published in 1999. More than two
thirds of the contents in this current volume are based on recent developments in the
relevant area of research.
In particular, this new edition consists of three parts, entitled The Foundation,
The Systemic Intuition, and Applications in Business Decision Making, respectively.
Other than some slight changes and additions, the first part is mainly based on the
original edition. The second part introduces the systemic yoyo model to satisfy the
desperate need for a practically useful intuition for reasoning in systems science
specific contents. The role to be played by this model is expected to be similar to
that played by Cartesian coordinate system in the classical science, on top of which
all statistical and analytical methods, widely and heavily employed in modern
science, are developed. The third part looks at how systemic thinking and the yoyo
model can be beautifully applied to address many important problems facing
decision makers in business by organically combining methods of the classical
science, the first dimension of the two-dimensional science, with those of systems
science, the second dimension, as argued by George Klir in the 1990s. As of this
writing, important decisions in business are mostly drawn based on data mining or
anecdotes. Scientifically speaking, such processes of decision making have been
time and again shown to be flawed. Because of this reason, this part of the book is
expected to open up a brand new territory of research valuable for working man-
agers and economics professionals.
By presenting a rigorously developed foundation, a tool for intuitive reasoning
that is supported by both theory and empirical evidence, and extremely fruitful
applications this book demonstrates the theoretical value and practical significance
of systems science and its thinking logic. By making use of this science and by
employing the systemic intuition, one can produce interesting, convincing, and
scientifically sound results. As shown in this book, many of the conclusions drawn
on the basis of systems science can be practically applied to produce tangible
economic benefits.

xxiii
xxiv Synopsis

By studying this book and by referencing back to it regularly, the reader, who
would be either a graduate student, a researcher, or a practitioner in the areas of
mathematics, either theoretical or applied, systems science or engineering, eco-
nomics, and decision science, will master a brand new tool to resolve his/her
problems and an intuition from which useful decisions can be made relatively
quickly without wasting unnecessarily the valuable time and a lot of the limited
financial resources.
Chapter 1
Introduction

In our modern times, science develops quickly, technology renews itself in front of
our very eyes, and management theories evolve in different directions. The most
important and most obvious characteristic of the quickly developing science is that
the forest of specialized and interdisciplinary disciplines can be easily seen, yet
the boundaries of disciplines become blurred. In terms of the speedy renewing of
technology, the most important and most obvious characteristic is that generations of
newer gadgets with better functionalities appear with shortening time intervals, while
unexpected products are designed and introduced into the marketplace in abundance.
Regarding modern management techniques, an increasing number of new theories
are developed with forever different insights provided.
That is, the overall trend of modern science is to synthesize all areas of knowledge
into a few major blocks, as evidenced by the survey of an American national commit-
tee on scientific research in 1985 (Mathematical Sciences 1985), and that of modern
technology is to make everything imaginable “smart” through using computer-related
technologies and the knowledge of artificial intelligence. On the contrary, the theo-
ries of management seem to be more diverse than ever before without any sign of
immediate convergence.
Because of the rapid progress of science, technology, and management theories,
researchers and practitioners of scientific research and decision-making managers
have been faced with unprecedented problems: How can they equip themselves with
the newest knowledge? How can they handle the geometrically growing amount
of new knowledge? And how can they cope with the faster-changing market envi-
ronment and much less patient customers? If traditional methods of exploration
and learning are employed without modification, then today’s unpublished scientific
results could become outdated tomorrow, and today’s hot products could be out of
fashion tomorrow. Facing these challenges, researchers, practitioners, and decision-
making managers evidently need to acquire new skills and knowledge regarding how
the concepts of wholes and parts and their relationships can be employed to investi-
gate similar problems in different scenarios so that the points of view of interconnec-
tion and interdependence can be employed to observe the world in which we live.
© Springer Nature Switzerland AG 2018 1
J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_1
2 1 Introduction

In the second decade of the twentieth century, von Bertalanffy (1924) wrote:
Since the fundamental character of living things is its organization, the customary investi-
gation of the single parts and processes cannot provide a complete explanation of the vital
phenomena. This investigation gives us no information about the coordination of parts and
processes. Thus the chief task of biology must be to discover the laws of biological sys-
tems (at all levels of organization). We believe that the attempts to find a foundation for the
theoretical point are a fundamental change in the world picture. This view, considered as a
method of investigation, we call “organismic biology” and, as an attempt at an explanation,
“the system theory of the organism.”

(Note: Similarly, in the study of economics, Rostow (1960) wrote:

The classical theory of production is formulated under essentially static assumptions which
freeze – or permit only once-over change – in the variables most relevant to the process of
economic growth. As modern economists have sought to merge classical production theory
with Keynesian income analysis they have introduced the dynamic variables: population,
technology, entrepreneurship, etc. But they have tended to do so in forms so rigid and
general that their models cannot grip the essential phenomena of growth, as they appear to
an economic historian. We require a dynamic theory of production which isolates not only
the distribution of income between consumption, savings, and investment (and the balance
of production between consumers and capital goods) but which focuses directly and in some
detail on the composition of investment and on developments within particular sectors of
the economy.)

From von Bertalanffy’s statement, it can be seen that parallel to the challenge of
modern science, technology, and management, formally proposed was a new concept
of systems, which can be potentially used by researchers, practitioners, and decision-
making managers to face their respective challenges. Tested in the past ninety some
years, this concept of systems has been widely accepted by the entire spectrum of
science and technology and by working economists and business professionals; for
details, see Lin (1999) and Klir (1985).
This chapter consists of five sections. Section 1.1 analyzes the historical back-
ground for how systems science has appeared and why it has been widely accepted.
Section 1.2 looks at how the evolution of systems should be seen holistically.
Section 1.3 explains what systems philosophy and methodology entail. Section 1.4
addresses the reason why we closely look at applications in business areas in this
book, and Section 1.5 presents how this book is organized.

1.1 Historical Background

Since the time when the concept of systems was formally introduced in the second
decade of the twentieth century in biology (von Bertalanffy 1924), the concept has
been widely applied either implicitly or explicitly in the entire spectrum of science,
be it natural or social, and technology. Although it seems to be a quite recent phe-
nomenon, as all new concepts studied in the world of learning, the ideas and thinking
logic of systems have a long history. For example, Chinese traditional medicine, treat-
ing each human body as a whole, can be traced to the time of Yellow Emperor about
1.1 Historical Background 3

5,000 years ago, and Aristotle’s statement that “the whole is greater than the sum
of its parts” has been a fundamental problem in systems theory. In other words,
throughout the history, mankind has been studying and exploring nature by using the
thinking logic of systems. It is only in modern times that have some new contents
been added to the ancient systems thinking. The methodology of studying systems
as wholes adequately agrees with the development trend of modern exploration of
nature and man itself, namely divide the object, event, and process of consideration
into parts as small as possible while studying each and every of them carefully, seek
interactions and connections between phenomena, and to observe and comprehend
more and bigger pictures of nature than ever before.
In the recorded history, although the word “system” was never emphasized, we can
still find many explanatory terms concerning the concept of systems. For example,
Nicholas of Cusa, that profound thinker of the fifteenth century, linking medieval
mysticism with the first beginning of modern science, introduced the notion of
coincidentia oppositorum, the opposition or indeed fights among the parts within
a whole which nevertheless forms a unity of higher order. Leibniz’s hierarchy of
monads looks quite like that of modern systems; his mathesis universalis presages
an expanded mathematics which is not limited to quantitative or numerical expres-
sions and is able to formulate much conceptual thought. Hegel and Marx emphasized
the dialectic structure of thought and of the universe it produces: the deep insight that
no proposition can exhaust reality but only approaches its coincidence of opposites
by the dialectic process of thesis, antithesis, and synthesis. Gustav Fechner, known
as the author of the psychophysical law, elaborated, in the way of the natural philoso-
phers of the nineteenth century, supraindividual organizations of higher order than
the usual objects of observation, for example, life communities and the entire earth,
thus romantically anticipating the ecosystems of modern parlance. Here, only a few
names are listed. For a more comprehensive study, see von Bertalanffy (1972).
Even though Aristotelian teleology was eliminated in the development of modern
science, problems considered in it, such as the one that “the whole is greater than
the sum of its parts,” the order and goal directedness of living things, etc., are still
among the problems of today’s systems research. For example, what is a “whole”?
What does “the sum of its parts” mean? Could the sum of parts sometimes be greater
than the whole? Although decision-making managers and working professionals
know the importance of these and other related problems, none of these problems
has been actually studied and addressed in any of the classical branches of science.
These classical branches of science have been established on Descartes’ second
principle and Galileo’s method, where Descartes’ second principle says to divide
each problem of concern into as small parts as possible so that each part could
possibly be investigated, and Galileo’s method implies to simplify a complicated
process into as basic portions and processes as possible so that each portion and
process could potentially be comprehended; for details, see Kuhn (1962).
From this superficial discussion, it can be seen that the concept of systems we
are studying today is not simply a product of yesterday. As a matter of fact, it is
a reappearance of some ancient thought and a modern form of an ancient quest.
This quest has been recognized in human struggle for survival with nature and has
4 1 Introduction

been studied at various points in time by using the languages of different historical
moments.
Ackoff (1959) comments that during the past two decades, we witnessed the
appearance of the key concept “systems” in scientific research; however, with the
appearance of the concept, what changes have occurred in modern science? Under the
name “systems research,” many branches of modern science have shown the trend of
synthetic development; research methods and results in various disciplines have been
intertwined to influence the overall research progress, so one feels the tendency of
synthetic development in scientific activities. This synthetic development requires the
introduction of new concepts and new thoughts in the entire spectrum of science. In a
certain sense, all of this can be considered as the center of the concept of “systems.”
Hahn (1967, p. 185), an ex-Soviet expert, describes the progress of modern science
as follows: Refining specific methods of systems research is a widespread tendency
in the exploration of modern scientific knowledge, just as science in the nineteenth
century with forming natural theoretical systems and processes of science as its
characteristics.
von Bertalanffy (1972) described the scientific revolution in the sixteenth–sev-
enteenth centuries as follows: “(The Scientific Revolution) replaced the descriptive-
metaphysical conception of the universe epitomized in Aristotle’s doctrine by the
mathematical—positivistic or Galilean conception. That is, the vision of the world
as a teleological cosmos was replaced by the description of events in causal, mathe-
matical laws.” Based on this description, we can describe the change in today’s sci-
ence and technology as follows: At the same time of continuously using Descartes’
second principle and Galileo’s method, systems methodology is introduced to deal
with problems of order and organization.
Should we continue to use Descartes’ second principle and Galileo’s method?
The answer is yes for two reasons. First, they have been extremely effective in
scientific research and administration, where all problems and phenomena could be
decomposed into causal chains, which could be treated individually. That has been
the foundation for all basic theoretical research and modern laboratory activities. In
addition, they won victories for physics and led to several technological revolutions.
Second, modern science and technology are not Utopian projects as described by
Popper (1945), reknitting every corner for a new world, but based on the known
knowledge, they are progressing in all directions with more depth, more applicability,
and a higher level of difficulty.
On the other hand, instead of treating the world as a pile of infinitely many iso-
lated objects, on which the concept of numbers is introduced with internal structures
ignored and modern science is established, not every problem or phenomenon, espe-
cially when humans and social events are concerned with, can be simply described by
causal relations. The fundamental characteristics of this world are its organizational
structure and connections of interior and exterior relations of different matters. The
study of either isolated parts or causalities of events and organizations can hardly
explain completely or relatively globally our surrounding world. At this junction,
the research progress of the three-body problem in mechanics, where the problem
is about how three or more bodies would interact with each other, as compared to
1.1 Historical Background 5

the two-body scenario as described in Newton’s second law of motion, is an ade-

quate example that shows why systems thinking is badly needed. So, as human race
advances, studying problems with multi-causality or multi-relation will become more
and more significant than ever before.
In the history of scientific development, the exploration of nature has always
moved back and forth between specific matters or phenomena and generalities. Sci-
entific theories need foundations and stimulations rooted deeply inside real-life prac-
tice, while being used to explain natural phenomena so that human understanding can
be greatly enhanced and to provide guidelines for decision makers to take educated
actions. In the following, we discuss the technological background for the devel-
opment of systems science—that is, the need for such a science that emphasizes
on order and organization which arose in the development of technology and from
the requirement for increasingly higher levels of production and better efficiency of
management.
There have had been many advances in technology in the modern history: ener-
gies produced by various devices, such as steam engines, motors, computers, and
automatic controllers; self-controlled equipment from domestic temperature con-
trollers to self-directed missiles; and the information highway and internet of knowl-
edge and things, which have resulted in increased communication of new scientific
results and practical successes. On the other hand, increasing speed of communi-
cation furthers scientific development to a different level and helps to intensify the
market competition globally. Additionally, societal changes have brought forward
more pressing demands for new construction materials and different management
strategies. From these examples, it can be seen that the development of technology
forces mankind to consider not only a single machine or matter or phenomenon, but
also “systems” of machines and “systems” of matters and phenomena. The design of
steam engines, automobiles, cordless equipment, etc., can all be handled by specially
trained engineers. However, when dealing with the designs of missiles, aircraft, or
new construction materials, for example, a collective effort, combining many differ-
ent aspects of knowledge and know-how, has to be in place. Such joint, collective
efforts include the combination of various techniques, machines, electronic technol-
ogy, chemical reactions, people, etc. Here the relation between people and machines
becomes more obvious, and uncountable financial, economic, social, and political
problems are intertwined to form a giant, complicated system, consisting of men,
machines, and many other components. It was the great political, technical, and per-
sonnel arrangement success of the American Apollo project—landing humans on the
moon—that hinted about the fact that history has reached such a particular point that
all aspects of science and technology have been developed maturely enough so that
each rational combination of information or knowledge could result in unexpected
consumable products and that different combinations of men and personalities could
lead to various unforeseen outcomes.
A great many problems in production require locating the optimal point of the
maximum economic effect and minimum cost in an extremely complicated network.
This kind problem not only appears in industry, agriculture, military affairs, and
business, but politicians are also using similar (systems) methods to seek answers
6 1 Introduction

to problems like air and water pollution, transportation blockages, decline of inner
cities, and crimes committed by teenage gangs.
In manufacturing production, there exists a tendency that bigger or more accu-
rate products with more profits are designed and produced. In fact, under different
interpretations, all areas of learning have been faced with complexity, totality, and
“systemality.” This tendency betokens a sharp change in scientific thinking. By com-
paring Descartes’ second principle and Galileo’s method with systems methodology
and considering the development tendency, as described previously, appearing in
the world of learning and production, it is not hard to see that because of systems
concepts, another new scientific and technological revolution will soon appear. To
this end, for example, each application of systems theory points out the fact that
the relevant classical theory needs to be further enriched somehow; see Klir (1970),
Berlinski (1976); Lilienfeld (1978). However, not all scientific workers have this
kind of optimistic opinion. Some scholars believe that it is an omen that systems
science itself is facing a crisis; see Wood-Harper and Fitzgerald (1982) for details.
Of course, who is right and who is wrong only time and history will have the ability
to tell.

1.2 Whole Evolutions—Blown-Ups and Revolving Currents

When we enjoy the splendid fruits of modern science and technology, we still need
to think about the limitations of and issues existing in the science and technology we
inherited from the generations before us. From the primeval to modern civilizations,
mankind has gone through a history of development for over several millions of
years. However, a relatively well-recorded history goes back only as far as about
3,000 years. During this time frame, the development of science mirrors that of
human civilizations. And, each progress of pursuing after the ultimate truth is that
of getting rid of the stale and taking in the fresh and that of making new discoveries
and new creativities. Each time when the authority was repudiated, science is reborn
again. Each time when the “truth” was questioned, a scientific prosperity appears.
That is, each time when people praise authorities, they are in fact praising ignorance.
At the turn of the twenty-first century, Shoucheng OuYang (Wu and Lin 2002)
proposed the blown-up theory for purpose of addressing and resolving nonlinear
evolution problems based on a reversed thinking logic, factual evidence, and over
30 years of reasoning and practice. It is found that in terms of the formalism, nonlin-
ear evolution models represent the singularity problems of mathematical blown-ups
of uneven formal evolutions; and in terms of physical objectivity, nonlinear evo-
lution models describe mutual reactions of uneven internal structures of materials,
organizations, and events. They no longer stand for a problem of formal quanti-
ties. Since uneven structures and organizations symbolize eddy sources, they lead
to eddy motions, the mystery of nonlinearity, instead of waves. That mystery has
been bothering mankind for over 300 years, is resolved at once both physically and
mathematically by OuYang (Wu and Lin 2002). What is shown is that the essence of
1.2 Whole Evolutions—Blown-Ups and Revolving Currents 7

nonlinear evolutions is the destruction of the initial-value automorphic structures and

the appearance of discontinuity. It provides a tool of theoretical analysis for studying
objective transitional and reversal changes of materials, organizations, and events.
In the ancient scientific history of the Western civilizations, there existed two oppo-
site schools on the structure of materials. One school believed that materials were
made up of uncountable and invisible particles, named “atoms,” the other school that
all materials were continuous. A representative of the former school is the ancient
Greek philosopher Democritus (about 460–370 BC) and the latter the ancient philoso-
pher Aristotle (384–322 BC). Since abundant existence of solids made it easy for
people to accept the Aristotelian continuity, the theory of “atoms” was not treated with
any validity until the early part of the nineteenth century when J. Dalton (1766–1844)
established relevant evidence. In principle, Leibniz and Newton’s calculus were orig-
inated in the Aristotelian thoughts. Along with calculus, Newton constructed his laws
of motion on the computational basis of calculus and accomplished the first successes
in applications in celestial movements under unequal-quantitative effects. With over
two hundred years of development, the classical theory of mechanics has gradually
evolved into such a set of methods for analysis based on the concept of continuity
that even nearly a century after quantum mechanics and relativity theory were estab-
lished; the thinking logic and methods developed on continuity are still widely in
use today.
Due to differences in environmental conditions and living circumstances, where
the west was originated from castle-like environments and the east from big river
cultures with agriculture and water conservation as the foundation of their national
prosperities, the ancient eastern civilizations were different or even opposite of those
of the west. So, naturally, Chinese people have been more observing about reversal
and transitional changes of weathers and rivers. Since fluid motions are irregular
and difficult to compute exactly, that was why the Book of Changes (Wilhalm and
Baynes 1967) and Lao Tzu (time unknown) appeared in China. The most important
characteristic of the Book of Changes is its way of knowing the world through images
of materials and organizations with uneven internal structures and analyzing changes
in events through figurative structures with an emphasis placed on irregularities,
discontinuities, transitional, and reversal changes.
As pointed out by Wu and Lin (2002), in terms of symbolism, due to escapes
in uneven forms from continuity, the evolution of any nonlinear evolution model is
no longer a problem of simply extrapolating the given initial values or the past into
the future. What is significant here is that through nonlinear evolutions, the concept
of blown-ups can represent Lao Tzu’s teaching that “all things are impregnated by
two alternating tendencies, the tendency toward completion and the tendency toward
initiation, which, acting together, complete each other,” and agrees with non-initial-
value automorphic evolutions as what the Book of Changes describes: “At extremes,
changes are inevitable. As soon as a change occurs, things will stabilize and stability
implies long lasting.”
Since nonlinearity describes eddy motions (Lin 2007), there must exist different
eddy vectorities and consequent irregularities. That is, the phenomenon of “order-
lessness” is inevitable. When looking at fluid motions from the angle of eddies, one
8 1 Introduction

can see that the corresponding quantitative irregularities, orderlessness, multiplici-

ties, complexities, etc., are all about the multiplicity of rotating matters. Therefore,
there exist underlying reasons for the appearance of quantitative irregularities, mul-
tiplicities, and complexities. Those underlying reasons are the unevenness of “time
and space” within the structures of the evolutionary materials, organizations, and
events.
From the blown-up theory (Wu and Lin 2002), one can see that all eddy motions,
as described symbolically with nonlinearities, are irregular when compared to the
standardized tools of analysis. And, all standardized methods become powerless in
front of the challenge of solving discontinuously quantified deterministic problems
of nonlinear evolution models (Lin and OuYang 2010).
One important concept studied in the blown-up theory is that of equal-quantitative
effects, which describes the conclusion of quantitative analysis under quasi-equal
acting forces. Although this concept was initially introduced in the study of fluid
motions, it represents the fundamental and universal characteristic of materials’
movements, organizational developments, and events’ evolutions. What is impor-
tant about this concept is that it reveals the fact that nonlinearities are originated
from the structures of materials instead of nonstructural quantities.
On the basis of the blown-up theory and the theory’s successes of real-life disaster
predictions (Lin and OuYang 2010; OuYang 1994; OuYang et al. 2009), the concepts
of inputs, outputs, and converging and diverging eddy motions are coined together
in the model as shown in Figure 1.1. This model was initially established in (Wu and
Lin 2002) for each object and every system imaginable. In particular, each system or
object considered in a study is a multi-dimensional entity that spins about its invisible
axis. If we fathom such a spinning entity in our three-dimensional space, we will
have a structure as shown in Figure 1.1. The side of inputs sucks in all things, such
as materials, information, energy, investment, human resources, etc. After funneling
through the short narrow neck, all things are spit out in the form of an output. Some
of the materials, spit out from the end of outputs, never return to the other side and
some will. For the sake of convenience of communication, such a structure, as shown
in Figure 1.1, is called a (Chinese) yoyo due to its general shape. More specifically,
what this model says is that each physical entity in the universe, be it a tangible or
intangible object, a living being, an organization, a culture, a civilization, etc., can
all be seen as a kind of realization of a certain multi-dimensional spinning yoyo with
an invisible spin field around it. It stays in a constant spinning motion as depicted
in Figure 1.1. If it does stop its spinning, it will no longer exist as an identifiable
system.
The theoretical justification for such a model is the blown-up theory (see Chapter
14 for more details). It can also be seen as a practical background for the law of
conservation of informational infrastructures (see Chapter 15 for more details). More
specifically, based on empirical data, the following law of conservation is proposed
(Ren et al. 1998): For each given system, there must be a positive number a such that

AT × BS × CM × DE  a (1.1)
1.2 Whole Evolutions—Blown-Ups and Revolving Currents 9

Figure 1.1 Yoyo model of

general systems

where A, B, C, and D are some constants determined by the structure and attributes
of the system of concern, and T stands for the time as measured in the system, S the
space occupied by the system, M and E the total mass and energy contained in the
system.
Because M (mass) and E (energy) can exchange to each other and the total of them
is conserved, if the system is a closed one, Equation (1.1) implies that when the time T
evolves to a certain (large) value, the space S has to be very small. That is, in a limited
space, the density of mass and energy becomes extremely high. So, an explosion
(outputs) is expected. Following the explosion, the space S starts to expand. That is,
the time T starts to travel backward or to shrink. This end gives rise of the well-known
model for the universe derived from Einstein’s relativity theory. In terms of systems,
we have the following model of systems’ evolution: Each system goes through such
cycles as: … → expanding → shrinking → expanding → shrinking → … Now, the
geometry of this model from Einstein’s relativity theory is given in Figure 1.1.
Practically, the yoyo model in Figure 1.1 is manifested in different areas of life.
For example, each human being, organization, or civilization, as we now see them,
is a three-dimensional realization of such a spinning yoyo structure of a higher
dimension. Human feelings, impressions, and silent communications are some of
the manifestations of the systemic yoyo structure beyond the entities of human. For
details, see Chapter 15 and Lin and Forrest (2011).
The presentation of this section is mainly based on Einstein (1983), Haken (1978),
Jarmov (1981), Kline (1972), Liang (1996), Lin and OuYang (1996), Lin (1988, 1989,
1990), Lin et al. (1990), OuYang (1994), Prigogine (1967), Thom (1975), Zhu (1985).
For more details, please consult with these references.

1.3 Systems Philosophy and Methodology

Philosophically, systems science is concerned with the concepts of “wholes” and

“parts” and their relationships by focusing on general systems as its objects in order
to uncover isomorphic structures of systems. That is, systems science attempts to
unearth the unified laws that govern forms of existence and forms of motion of
systems of various characteristics, and the laws all systems comply with. According
10 1 Introduction

to the difference between thinghood and systemhood, systems science focuses on

the properties of systemhood of systems instead of properties of thinghood as what
traditional science studies.
Systems science attempts to understand the world by using its unique logic of
thinking, view of nature, view of time and space, and meaning of life. It explores the
territories the traditional science has never touched before. It challenges to reveal a
much greater and much more truthful world than the one the traditional science has
shown.
The essence of all sciences is, to a great extent, to reveal orders within the uncer-
tainties of the world. To this end, the cognitive approach of systems science provides
a brand new way to comprehend the world in its original characteristics: structures,
organizations, interconnectedness, and interdependence. And the development of
systems science is mainly motivated by societal developments and backgrounds pre-
pared by prior scientific, technological, and managerial advances. As a matter of fact,
the dialectical principles of universal connections, movements, and developments are
the premises of systems science that need to be clearly understood.
As for what the methodology of systems science is, scholars in the field understand
it in different ways. However, their fundamental points are the same, loosely speaking.
In the following, let us first look at typical opinions of two important figures in modem
science: H. Quastler and L. Zadeh.
Quastler (1965) wrote: Generally speaking, systems methodology is essentially
the establishment of a structural foundation for four kinds of theories of organiza-
tion: cybernetics, game theory, decision theory, and information theory. It employs
the concepts of a black box and a white box (a white box is a known object, which
is formed in certain way, reflecting the efficiency of the system’s given input and
output), showing that research problems, appearing in the afore-mentioned theo-
ries on organizations, can be represented as white boxes, and their environments as
black boxes. The objects of systems are classified into several categories: motiva-
tors, objects needed by the system to produce, sensors, and effectors. Sensors are the
elements of the system that receive information, and effectors are the elements of
the system that produce real reactions. Through a set of rules, policies, and regula-
tions, sensors and effectors do what they are supposed to do. By using these objects,
Quastler proved that the common structure of the four theories of organization can
be described by the following laws:
• Interactions are between systems and between systems and their environments.
• A system’s efficiency is stimulated by its internal movements and the reception of
information about its environment.
Zadeh (1962) listed some important problems in systems theory as follows: sys-
tems characteristics, systems classification, systems identification, signal represen-
tation, signal classification, systems analysis, systems synthesis, systems control and
programming, systems optimization, learning and adaptation, systems liability, sta-
bility, and controllability. The characteristics of his opinion are that the main task
of systems theory is the study of general properties of systems without considering
their physical specifics. Systems methodology in Zadeh’s viewpoint is that systems
1.3 Systems Philosophy and Methodology 11

science is an independent scientific endeavor whose job is to develop an abstract

foundation with concepts and frames in order to study various behaviors of different
kinds of systems. Therefore, systems science should contain a theory of mathematical
structures of systems with the purpose of studying the foundation of organizations
and systems structures.
Since the time when these authors made their important marks in the development
history of systems science, nearly half a century has gone by quickly. During these
past decades, systems science has witnessed major development. So, by summarizing
the newest achievements, we will be able to better describe the methodology of
systems science as follows.
First, the landscape of knowledge is now seen as two-dimensional with the tradi-
tional science as the first dimension, which focuses on the study of thinghood, and
systems science as the second dimension, which emphasizes on the investigation of
systemhood, as argued by Klir (2001). That is why systems science and research could
be and have been naturally transcending all the disciplines of the classical science
and become a force making the existing disciplinary boundaries totally irrelevant
and superficial. The cross-disciplinary feature of the systems science and research
implies that
(1) Researches, done in systems science, can be applied to virtually all disciplines
of the classical science.
(2) Issues involving systemhood, studied in individual specialization of the classical
science, can be studied comprehensively and thoroughly at the height of systems
science.
(3) A unifying influence on the traditional science where a growing number of
narrower and narrower disciplines appears is created.
With the advantage of the newly discovered second dimension, the world of learn-
ing will be expectedly able to resolve some of the unsettled age-old problems that
have been difficult or impossible for the traditional science to conquer.
Second, as the totality of knowledge expands and human understanding of nature
deepens, it is discovered that many systems interact nonlinearly with each other and
do not satisfy the commonly assumed property of additivity. Systems’ emergent irre-
versibility, sensitivity, and complexity cannot be analyzed and understood by using
the methodology of the traditional reductionism. Facing this challenge, systems sci-
ence takes a hard look at the concept of “emergence”: The whole that consists of a
large number of individuals that interact with each other according to some elemen-
tary rules possesses some complicated properties. That is, the whole is greater than
the sum of its parts. By discovering elementary laws that govern complex systems of
different kinds, one will be able to explain many complicated and numerous matters
and events of the kaleidoscopic world and introduce different control mechanisms.
Third, the rigor of the foundation of systems science is now ideally based on
set theory, because such an abstract theory is very flexible in terms of its power to
describe and can be readily specified to address particular issues of concern. At the
same time, the qualitative theory of differential equations, game theory, gray systems
theory are employed to derive various properties of general systems, to study how
12 1 Introduction

systems interact with each other, and to discover patterns from available data, be
it a very small sample or a flood of information, in order to produce practically
useful guidelines. Because of the use of these repeatable and analytical methods, the
mostly philosophical works of the earlier stage of systems science have been made
rigorous, visible, and tangible, leading to a good amount of convincingly successful
applications in the traditional disciplines.
Fourth, based on an understanding of the history of the systems movement and a
creative comparison between systems science and calculus, the yoyo model is now
available for us to intuitively see how systems would evolve and interact with each
other. In the recent development of systems science, this simple model has played
several crucial roles; for more details about this model, please consult with (Lin
2009):
(1) It is the intuition and playground for systemic thinking to take place in a fashion
similar to that of Cartesian coordinate systems in modern science. To use the
model effectively, of course each of the characteristics of the issue of concern
needs to be identified with some aspect of the yoyo model.
(2) It provides an intuitive ground for logic thinking so that one could see how to
establish models in the traditional fields in an unconventional sense in order to
resolve the problem in hand.
(3) It helps understand the issue of concern from an angle that is not achievable
from the traditional science alone.
As a matter of fact, the recent development of systems science has clearly shown
that this yoyo model has successfully played the role of intuition and playground,
as expected, for scholars to investigate the world and explore new knowledge holis-
tically, just as what the Cartesian coordinate system did for the traditional science
(Lin 2009; Lin and Forrest 2011; Forrest 2013, 2014; Forrest and Tao 2014; Ying
and Forrest 2015). In particular, this yoyo model of general systems has been suc-
cessfully applied in the investigation of Newtonian physics of motion, the concept of
energy, economics, finance, history, foundations of mathematics, small-probability
disastrous weather forecasting, civilization, business organizations, the mind, among
others.

1.4 The Reason for Looking at Applications in Economics

and Business Areas

From the presentation in the previous section, it can be readily seen that systems
science has made sufficient progress in the past century regarding its foundation,
intuitive reasoning, and methodology for rigorous analysis. It now possesses all the
necessary characteristics and capabilities to enjoy a glorious and long-lasting life,
similar to the lives enjoyed by Euclidean geometry and calculus.
Although all aspects of systems science still need to be developed further, as what
is happening in the traditional science, some of the major scientific labors devoted
1.4 The Reason for Looking at Applications in Economics and Business Areas 13

to this area of learning have to be now directed toward producing tangible results
and insights that will excite the public and resolve challenges facing the mankind.
In other words, it is important for systems scientists and practitioners to locate some
weaknesses and incapabilities in the traditional science that affect the effectiveness of
practical decision making and challenges the traditional science has not been able to
meet successfully. To this end, there are many wonderful choices of materials we can
use for this book, as what was done in the first edition of this volume. In particular,
in the first edition, we looked at successful applications of systems research and
systems thinking in such diverse areas that range from sociology to atomic physics,
from Einstein’s relativity theory to Dirac’s quantum mechanics, from optimization
theory to unreasonable effectiveness of mathematics to foundations of mathematical
modeling, from general systems theory to Schwartz’s distributions. In each of these
specific applications, at least one important niche was found and then addressed by
using results or methods of analysis of systems science.
Contrary to what is done in the first edition of this volume with a diverse scope
of applications, this book will focus on several applications only in areas of eco-
nomics and business decision making. The reason for me to take this choice is that
the currently available theories of the traditional science cannot meaningfully pre-
dict near-zero probability disastrous events in advance other than making live reports
(Lin and OuYang 2010). For instance, in the area of weather forecasts, the meteorol-
ogist relies on satellites, radar systems, and other technology to inform the residents
and organizations located on the downstream of an observed moving weather sys-
tem what to expect and when to expect based on either visual observations or some
mathematical extrapolations of the data of rolling availability or both. Even with the
employment of all the most advanced technologies and all well-developed scien-
tific theories (of the first dimension), each of his weather forecasts is still attached
with a probability in order to theoretically guarantee the accuracy of his forecast. It
is because the attached probability basically means no matter what turns out to be
the case in reality, his earlier forecast is always correct (probabilistically speaking)
(OuYang 1994). In a similar fashion, in economics and business decision making,
people need to make educated projections about what to expect and what possible
courses of actions to take, unfortunately, without much technology available for live
reports to rely on. What is more challenging than the situation of weather forecasting
is that in business the decision maker has to deal with people and the possible con-
sequences of their collective endeavors. In particular, when forecasting the possible
appearance of a natural event, the accuracy of the forecast does not affect the actual
appearance of the event; however, when working with systems involving people, pre-
dictions on the consequences of collective actions tend to alter how people behave
and what different courses of actions they would actually take. That end practically
makes the earlier predictions often incorrect (Soros 2003). Because of this reason, it
is no doubt to me that helping economists and business decision makers to improve
their abilities to make better predictions and then better decisions on their courses of
actions, which has been a challenge impossible for the traditional science to meet, is
both practically important and theoretically significant.
14 1 Introduction

Speaking more specifically, in order to produce meaningful results and insights

that successfully address challenges facing areas of the traditional science, in this
book I will showcase fruitful resolutions of some problems related to business deci-
sion making while demonstrating how the systemic yoyo model and related systems
reasoning can be practically applied. In particular, we will closely look at why com-
petitive advantages that were sustainable in the past are becoming transient, while
markets change faster and customers are less patient than ever before; how nonlin-
earity appears in demand/supply interactions so that new methods of analysis are
needed; what market conditions actually invite new competitions; how mysterious
family members behave even under the altruism of the head of the household; why no
value system can stay unchanging for long if it is expected to provide the necessary
guidance for desirable behaviors; how a leader can potentially achieve management
efficiency; why any organization is expected to always experience inefficiencies;
how the management can effectively deal with customers’ indecisiveness through
applying pricing strategies; how the manager can heighten the competitive spirits of
her sales’ associates, etc.
The literature in economics and business decision making tends to show the fre-
quent use of such words as “believe,” “should,” and “would.” For example, Kotler
et al. (2010), Krauss (2011), Stengel (2011) believe that today’s customers want to
be treated as whole human beings and be acknowledged that their needs go beyond
pure consumerism. The application of the word “believe” here really indicates, sci-
entifically speaking, that these authors are not quite sure what they are talking about.
Considering the fact that Philip Kotler is one of the most influential marketing thinkers
(Kaul 2012), what is stated above simply means that most decisions in the area of
marketing are made based on anecdotes or data mining or both. However, neither
anecdotes nor data mining are recognized to be reliable ways to produce dependable
theories beyond potential facts finding in natural science and mathematics, where
a pattern or theory derived from anecdotes and data mining cannot be trusted even
though the observed pattern can be shown to be true for a very large number of empir-
ical cases, no matter how large the number is, until it is proven for the general case.
In terms of data mining, other than possibly discovering facts, the related methods
also frequently lead to “realities” that only exist with the particular sets of data used
in the analyses. That is why I often describe the approach of using data analysis
to develop theories as looking at the clouds in the sky and try to find patterns. For
example, let us look at the clouds in Figure 1.2. Do we see an image of two people
deeply in love? If you do not see the image, then I or someone else can easily help
highlight the pattern out of the background. Now, no matter what story we can make
out of this observation about two heavenly figures, no one will believe it—period!
Speaking generally, when a discipline or an area of study frequently uses such
words as “believe,” “should,” and “would,” it means that
1. When a rigorous tool of reasoning is used, the author does not know for sure
whether or not he/she has considered all related factors.
2. A lot of conclusions are drawn based on one or a few anecdotes without any
follow-up rigorous analysis.
1.4 The Reason for Looking at Applications in Economics and Business Areas 15

Figure 1.2 A pair of

lovebirds in the sky

3. A lot of so-called theories are really established on data mining.

One of the most important reasons for Situation 1 to occur is that there is no read-
ily available playground or intuition for large-scale or global economic reasoning.
For example, all calculus-related methods and all statistical approaches are devel-
oped and established on the Cartesian coordinate system. However, the Cartesian
coordinate system does not really exist in real life. That renders the applied methods
and approaches questionable or at least they are only useful locally. That explains
why when an analytical or statistical tool is employed, it only models a regional,
local phenomenon so that conclusions of any larger scale can only be conjectured.
When Situation 2 is the case, the conclusions are really derived inductively instead of
deductively. As we mentioned earlier, such conjectured conclusions in general cannot
be trusted as reliable. For example, from the anecdotes that 1 < 100, 2 < 100, 3 < 100,
can we conclude that any natural number is less than 100? As another example, from
the anecdotes that 02  0, 12  1, can one conclude that for any chosen natural
number, its square is equal to itself? Of course, no one can draw such conclusions
scientifically from the few anecdotes. Then how can such practice be commonly
accepted in business decision making? It is simply because an adequate tool for
analysis is desperately needed in these areas, where calculus-related methods and
statistical approaches are shown to be inadequate, for more detailed discussion about
this end, see Lin and OuYang (2010).
Because we now see where systems science can potentially make a breakthrough,
let us next see how we can make at least some meaningful progress little by little. To
this end, let us identify the current state of research affairs in economics and business
with that in the history of natural sciences right around the time when Isaac Newton
developed his laws of physics. By doing so, we can see the similarity between all
the facts, partial facts, and nonfacts discovered in the literature of economics and
business which are similar to those uncovered by various scholars in natural sciences
a few hundred years ago; and the present opportunity created by the development
of systems thinking and the systemic intuition in the recent past is quite parallel
to that used by Newton and his contemporaries to develop their laws of natural
16 1 Introduction

science because of the matrimony of Euclidean geometry and algebra, as a natural

consequence of the introduction of the Cartesian coordinate system.
Considering the massive amounts of facts, partial facts, and nonfacts unearthed
currently through analyzing anecdotes and data mining, we can confidently expect
based on educated analyses of evidence and history that a major breakthrough in
areas of economics and business is coming soon. And, because systems science and
the yoyo model can provide effective means for large-scale analysis and intuitive
thinking, I expect quite surely that when such a breakthrough appears, it will be
systems science and systems thinking based, as a consequence of combining the
Western thinking with that of the Eastern logic of reasoning.
Consequent to what is discussed in the previous paragraphs in the presentation of
the contents in this book special care is given to write the theory and each application
in a language appropriate to that field. In particular, the part of the foundation of
systems science is written in the language of pure mathematics; the part on the
systemic yoyo model is presented in such a way that it is familiar to scholars who are
used to developing and employing mathematical models for the purpose of resolving
real-life problems; and then each application in economics and business decision
making is offered with abstractions and explanations at different levels so that readers
in various fields will find it possible to understand and to appreciate the beauty and
power of systems science.
Additionally, because of the wide range of topics covered in this book, I try to
make each chapter self-contained so that in general, there is no need for the reader
to reference a different chapter for his/her reading of any particular chapter.

1.5 Organization of Contents in This Book

As an update and expansion of my previous book, entitled “General Systems The-

ory: A Mathematical Approach,” initially published in 1999. This volume contains
more than two thirds of the contents on recent developments in the relevant areas of
research. In particular, this new edition consists of three parts, entitled The Founda-
tion, The Systemic Intuition, and Applications in Business Decision Making, respec-
tively.
The first part on the foundation of systems science contains only some minor
changes and additions compared to the previous edition. It consists of Chapters 2–
12, where Chapters 2–7 focus on the rigorous development of the concept of general
systems, discovery of elementary properties of systems, construction of large-scale
systems, interactions between systems, controllabilities of systems, etc. And Chap-
ters 8–12 look at single-relation systems, where such concepts as chaos, attractors,
feedback transformations, decoupling, decomposition, etc., are investigated in depth.
The second part introduces the systemic yoyo model, the desperately needed
intuition that is specific to systems science. This part consists of three Chapters 13–15.
Here, Chapter 13 introduces the yoyo model, explains what the yoyo field is composed
of, and shows how systems interact with each other through their individual yoyo
1.5 Organization of Contents in This Book 17

fields. Then the chapter considers how Newton’s laws of motions can be rewritten
in the language of systems science. Chapter 14 provides several different theoretical
supports for why each and every system possesses its specific yoyo structure, and
Chapter 15 lists a few empirical evidences for the existence of the yoyo structure
behind each and every system.
The third part of the book looks at how systemic thinking and the yoyo model
can be beautifully applied to address many important problems facing economists
and business decision makers by organically combining methods of the traditional
science, the first dimension of the two-dimensional spectrum of science, with those
of systems science, the second dimension, as argued by George Klir in the 1990s.
In particular, this part of the book consists of Chapters 16–21, where Chapter 16
looks how nonlinearity appears in the interaction of market demand and supply, how
economic cycles can be seen through the lens of the systemic yoyo model, and how
market competition emerges. Chapter 17 focuses on several issues on competition,
explains why coordinated monopoly tends to mean stagnation in expected profits,
markets change faster and customers become less patient, and once sustainable com-
petitive advantages become transient. Chapter 18 presents when the famous Becker’s
rotten kid theorem holds true along with two other mysteries of the household eco-
nomics, as well as the important never-perfect value system. Chapter 19 shows the
reasons why management efficiency is possible in real life, while organizational inef-
ficiency always exists in any business firm. Chapter 20 investigates how indecisive
customers can be potentially handled with pricing strategies. And Chapter 21 looks
at the relationship between private information and the performance of sales asso-
ciates. Different of the traditional research in the relevant areas of economics and
business, all the conclusions established in this part of the book are first seen intu-
itively through the yoyo model and then proved rigorously using different methods
of the traditional science.
To conclude this chapter, let me remind you, the reader, that the foundation of
systems science, presented in Part 1 of this book, is entirely based on a set-theoretical
definition of the concept of general systems in order for us to derive rigorously all the
elementary properties of systems. As discussed earlier in this chapter, the concept of
systems has been investigated in different ways using various languages throughout
the recorded history. However, like the concept of Tao (English and Feng 1972;
Lao Tzu, time unknown), whenever the concept of systems is defined in a specific
fashion, what is described is not exactly the concept, see Section 15.2 for more details.
This fact means that each specific technical definition of systems only captures a
particular aspect of the desired concept of systems. So, to reach the desired level
of understanding, one needs to combine several different versions of the concept in
order to achieve the required level of mastery. For our purpose in this volume, the
listed versions of definition of systems in Section 15.2 will be adequate.
Along with the convention, established in Chapter 13, that all systems, be they
physical objects, or organizations of biological beings, or imaginary entities, can be
investigated as rotational pools of “fluids” together with a deep understanding of
the concept of systems, the applications in economics and business decision making
presented in Chapters 16–22 become quite intuitive and straightforward, developed
18 1 Introduction

on the sound scientific means of analytic analysis, laboratory experiments, and formal
logic reasoning. To this end, some readers might prefer to lay the applications on a
more thorough analysis of human behaviors and economic phenomena themselves.
If you are one of such readers, I can fully understand for several reasons:
• All applications are presented in a relatively (natural) scientific fashion. In other
words, after a conclusion or model is validated, it is employed throughout each
of the applications without repeatedly explaining why the conclusion once again
employed is correct.
• My stance is that if reliable thorough analysis of human behaviors and economic
phenomena already existed, why do we still have trouble explaining many organi-
zational behaviors and economic events, such as the recent financial crisis? To me,
the answer is quite simple: As of this writing, there is no reliable thorough analysis
of human behaviors and economic phenomena, especially no analyses have been
developed for large-scale phenomena, processes, and events.
• If you, the reader, feel that without enough reliable information, it is impossible to
know whether or not specific human phenomena and economic events conform to
the yoyo model, the truth of the matter is that in the (natural) scientific history if a
model is shown to hold true under certain conditions, then as long as a phenomenon,
be it physical or human or economic, satisfies all the listed conditions, it will
definitely fit the model or the model applies to the phenomenon. In particular to
the applications in Part 3 of this book, as long as we are convinced that the human
phenomena and economic events considered herein are systems, then they can be
seen as and modeled by the yoyo model without any slightest doubt. As for the
reliability of information, it has been shown in practice (Soros 1998) and theory
(Liu and Lin 2006) that the more reliable a piece of information is the more chaotic
situation it creates. That is, reliability is only relative instead of being absolute.
• Could this study be seen as secondary developed based on the existing scientific
literature? The answer to this question is both yes and no. It is yes, because many
detailed areas of the applications dig deeply into the existing literature and try
to elevate what have been known onto much higher levels. In this sense, all the
applications listed in this book in fact should not be treated as secondary to what has
already been known in the literature. At the same time, the answer is no, because
from my reasonably intensive search of the available literature, it is quite clear that
most of the relevant publications are generated on the bases of either anecdotes
without much scientific rigor or based on data mining that is really an art. In this
sense, what is presented in this volume is more fundamental with scientific rigor
than most of the relevant literature and lays the theoretical and intuitive foundation
for achieving sound understanding of and arriving at reliable decisions regarding
various large-scale organizations, economic processes, and events.
To conclude this chapter, let us look at how the parts and chapters of this book
are connected and coordinated with one another.
The three parts present a sequential order of learning and dynamic development
of a new theory. In particular, Part 1 presents how a cohesive and systematic theory
of general systems can be developed by using a unified language—set theory. From
1.5 Organization of Contents in This Book 19

the role successfully played by set theory in mathematics and in science, it is clear
that the language of sets is very versatile and universal in terms of its elevated level of
abstraction and its superlative power to describe very different phenomena observed
in all walks of life. Because of the use of set theory, this first part of this book presents
only results and conclusions that can be rigorously derived from a few basic concepts.
In other words, results and conclusions presented in this part are universally true no
matter which specific area of learning or application is concerned with.
Continuing what is presented in the previous part, while filling a gap existing in
the reasoning process of systemic logic, Part 2 establishes a badly needed systemic
intuition—the yoyo model—based on a holistic analysis of different understandings
(or formulations) of the concept of (general) systems. With this systemic intuition in
place, the magnificent successes of Euclidean geometry and calculus throughout the
scientific history clearly indicate that the cohesive body of results and conclusions
given in Part 1 now becomes a potentially useful knowledge in terms of its capability
to resolve problems in real life and to meet challenges in various areas of human
endeavor. To demonstrate the fact that what is developed in the previous two parts has
become a practically useful theory, as expected, Part 3 focuses on various successful
applications in business decision making. And at the end of this book, to show how
the large-scale systems, studied in Chapters 3, 5, 7, 9–12 of Part 1, can be practically
used jointly with the systemic yoyo model of Part 2 in the area of business decision
making, Chapter 22 looks at how to manipulate the flow of speculative hot money
and how to introduce monetary policies in order to reduce the severity of the currency
attacks. In other words, this final chapter of the book not only shows how the theory
developed in this volume can be employed for practical purposes, as what has been
done in other chapters of Part 3, but also demonstrates why what is presented in this
book is indeed a holistic, organic body of knowledge.

References

Ackoff, R.L.: Games, decisions and organizations. General Syst. 4, 145–150 (1959)
Berlinski, D.: On systems analysis. MIT Press, Cambridge, MA (1976)
Einstein, A: Complete Collection of Albert Einstein (translated by L. Y. Xu). Commercial Press,
Beijing (1983)
English, J., Feng, G.F.: Tao De Ching. Vintage Books, New York (1972)
Forrest, J.Y.L.: A Systemic Perspective on Cognition and Mathematics. CRC Press, an imprint of
Taylor and Francis, Balkema, The Netherlands (2013)
Forrest, J.Y.L.: A Systems Perspective on Financial Systems. CRC Press, an imprint of Taylor and
Francis, Balkema, The Netherlands (2014)
Forrest, J.Y.L., Tao, L.X.: Investment and Employment Opportunities in China. CRC Press, an
imprint of Taylor and Francis, New York (2014)
Hahn, E.: Aktuelle entwicklungstendenzen der soziologischen theorie. Deutsche Z. Phil. 15,
178–191 (1967)
Haken, H.: Synergetics, 2nd edn. Springer, Berlin (1978)
Jarmov, G.: Development History of Physics (Chinese translation). Commercial Press, Beijing
(1981)
20 1 Introduction

Kaul, V.: Beyond Advertising: Philip Kotler Remains One of the Most Influential Marketing
Thinkers. The Economic Times, 29 Feb 2012
Kline, M.: Mathematical Thought from Ancient to Modern Times. Oxford University Press, Oxford
(1972)
Klir, G.: An Approach to General Systems Theory. Van Nostrand, Princeton, NJ (1970)
Klir, G.: Architecture of Systems Problem Solving. Plenum Press, New York (1985)
Klir, G.: Facets of Systems Science. Springer, New York (2001)
Kotler, P., Kartajaya, H., Setiawan, I.: Marketing 3.0: From Products to Customers to the Human
Spirit. Wiley, Hoboken, NJ (2010)
Krauss, M.: Evolution of an Academic: Kotler on Marketing 2.0. Marketing News, 30 Jan 2011
Kuhn, T.: The Structure of Scientific Revolutions. University of Chicago Press, Chicago (1962)
Lao Tzu (time unknown). Tao Te Ching: The Classic Book of Integrity and the Way. An Entirely New
Translation Based on the Recently Discovered Ma-Wang-Tui Manuscript. Translated, annotated,
and with an Afterword by Mair, V.H., Bantam Books, New York (1990)
Liang, H.M.: Lao Tzu (in Chinese). Liaoning People’s Press, Liaoning (1996)
Lilienfeld, D.: The Rise of Systems Theory. Wiley, New York (1978)
Lin, J., OuYang, S.C.: Exploration of the mystery of nonlinearity. Res. Nat. Dialectics 12–13, 34–37
(1996)
Lin, Y.: On periodic linked time systems. Int. J. Syst. Sci. 19, 1299–1310 (1988)
Lin, Y.: Multi-level systems. Int. J. Sys. Sci. 20, 1875–1889 (1989)
Lin, Y.: The concept of fuzzy systems. Kybernetes: Int. J. Syst. Cybernet. 19, 45–51 (1990)
Lin, Y.: General Systems Theory: A Mathematical Approach. Kluwer Academic and Plenum Pub-
lishers, New York (1999)
Lin, Y.: Systemic yoyo model and applications in Newton’s, Kepler’s laws, etc. Kybernetes: Int. J.
Cybern. Syst. Manag. Sci. 36, 3–4, 484–516 (2007)
Lin, Yi: Systemic Yoyos: Some Impacts of the Second Dimension. Auerbach Publications, an
imprint of Taylor and Francis, New York (2009)
Lin, Y., Forrest, B.: Systemic Structure Behind Human Organizations: From Civilizations to Indi-
viduals. Springer, New York (2011)
Lin, Y., Ma, Y.H., Port, R.: Several epistemological problems related to the concept of systems.
Math. Comput. Model. 14, 52–57 (1990)
Lin, Y., OuYang, S.C.: Irregularities and Prediction of Major Disasters. Auerbach Publications, an
imprint of Taylor and Francis, New York (2010)
Liu, S.F., Lin, Y.: Grey Information: Theory and Practical Applications. Springer, London (2006)
Mathematical sciences: a unifying and dynamical resource: Notices of the American Mathematical
Society 33, 463–479 (1985)
OuYang, S.C.: Break-Offs of Moving Fluids and Several Problems of Weather Forecasting. Press
of Chengdu University of Science and Technology, Chengdu (1994)
OuYang, S.C., Chen, Y.G., Lin, Y.: Digitization of Information and Prediction. China Meteorological
press, Beijing (2009)
Popper, K.: The Open Society and Its Enemies. Routledge, London (1945)
Prigogine, I.: Structure, dissipation and life. In: Communication presented at the First International
Conference of the Oretical Physics and Biology, Versailles (1967)
Quastler, H. General principles of systems analysis. In: Waterman T.H., Morowitz H.J. (eds.) The-
oretical and Mathematical Biology. New York (1965)
Ren, Z., Lin, Q.Y., OuYang, S.C.: Conjecture on law of conservation of informational infrastructures.
Kybernetes: Int. J. Syst. Cybern. 27, 543–552 (1998)
Rostow, W.W.: The Stages of Economic Growth: A Non-Communist Manifesto. Cambridge Uni-
versity Press, Cambridge, UK (1960)
Soros, G.: The Crisis of Global Capitalism: Open Society Endangered. Public Affairs, New York
(1998)
Soros, G.: The Alchemy of Finance. Wiley, New York (2003)
References 21

Stengel, J.: Grow: How Ideals Power Growth and Profit at the World’s Greatest Companies. Crown
Business, New York (2011)
Thom, R.: Structural Stability and Morpho-Genesis. Benjamin Reading, Massachusetts (1975)
von Bertalanffy, L.: Einführung in Spengler’s Werk. Literaturblatt Kolnische Zeitung, May (1924)
von Bertalanffy, L.: The history and status of general systems theory. In: Klir, G. (ed.) Trends in
General Systems Theory. New York, pp. 21–41 (1972)
Wilhalm, R., Baynes, C.: The I Ching or Book of Changes, 3rd edn. Princeton University Press,
Princeton, NJ (1967)
Wood-Harper, A.T., Fitzgerald, G.: A taxonomy of current approaches to systems analysis. Comput.
J. 25, 12–16 (1982)
Wu, Y., Lin, Y.: Beyond Nonstructural Quantitative Analysis: Blown-Ups, Spinning Currents and
the Modern Science. World Scientific, River Edge, New Jersey (2002)
Ying, Y.R., Forrest, J.Y.L.: Capital Account Liberation: Methods and Applications. CRC Press, an
imprint of Taylor and Francis, New York (2015)
Zadeh, L.: From circuit theory to systems theory. Proc. IRE 50, 856–865 (1962)
Zhu, Y.Z.: Albert Einstein: The Great Explorer. Beijing People’s Press, Beijing (1985)
 Part I
The Foundation
Chapter 2
The Concept of General Systems

This chapter focuses on the basic concept of general systems and then the fundamen-
tal properties of general systems. In particular, Section 2.1 introduces the concept of
general systems that will be used throughout the rest of this book along with a brief
review of other relevant concepts studied by various scholars in different contexts.
Section 2.2 investigates the internal properties of general systems. Section 2.3
considers the set-theoretical structures of general systems and L-fuzzy systems
along with the existence of centralized systems.

2.1 Various Attempts

von Bertalanffy (1934) introduced the concept of systems in biology in the late 1920s,
and it has been developed a great deal since then. Starting in the early 1960s, scholars
began to establish a theory of general systems in order to lay down a theoretical
foundation for all approaches of systems analysis developed in different disciplines.
Led by von Bertalanffy and supported by several of the most powerful minds of our
time, a global systems movement has been going on for several decades.
Along the development path of the systems movement, many scholars from dif-
ferent disciplines have tried to first define the concept of general systems and then
apply the concept and the theory, developed on the concept of general systems, to
solving problems in various research areas. Technical difficulties of introducing an
ideal definition for general systems, and notably unsuccessful applications of general
systems theory [for details, see Berlinski (1976; Lilienfeld 1978)], have taught many
painful lessons. Among them are:
(1) There might not exist an ideal definition for general systems, upon which a
general systems theory could be developed so that this theory would serve as
the theoretical foundation for all approaches of systems analysis, developed in
various disciplines.

© Springer Nature Switzerland AG 2018 25

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_2
26 2 The Concept of General Systems

(2) Without an ideal definition of general systems in place, “general systems theory”
would be meaningless. In this circumstance, under the name of systems analysis,
senseless transfer of results from one discipline to another has really made many
scholars doubt the value of “general systems theory.”
(3) Based on historical experience and the “unreasonable effectiveness” of mathe-
matics (Wigner 1960), several fruitful definitions of general systems have been
introduced since the early 1960s.
Even though a theory of abstract general systems has been established through the
efforts of many scholars, including Mesarovic and Takahara, the real challenge—-
namely, to show the need and power of the concept of general systems and its theory,
it is necessary to obtain results, which have not been nor could be, achieved without
using the concept and its theory—has not been very well met.
To meet the challenge of showing the need and power of the concept and theory
of general systems, Lin (1999) considers the following problems:
• Infinite or finite divisibility of the material world;
• Why the concept of systems was not studied more intensively until now;
• Meaning of a system with contradictory relations;
• Existence of absolute truths and unreasonable effectiveness of mathematics;
• Philosophy of mathematical modeling;
• The need to study an applied set theory;
• Multilevel structure of the world;
• The law of conservation of matter–energy;
• The material foundation of human thought;
• The origin of the universe; and
• The need to develop mathematics of multi-levels and its application in Einstein’s
relativity theory.
From what is done by Lin (1999), one can see that the study of general systems
theory will surely impact the entire spectrum of human knowledge. How the sys-
tems movement is regarded is each scholar’s opinion and position. No matter how
suspicious some scholars are, history will still proceed at its own pace. To this end,
the development histories of Cantor’s set theory, Einstein’s relativity theory, etc.,
have taught us that no new theory would be born without suffering from suspicions,
prejudices, and calumnies. In fact, meeting challenges and self-renewing are among
the most important motivations for a newborn theory to grow, to mature, and to gain
a visible position in scientific history.

2.1.1 The Concept of Systems

The fundamental characteristics of a system are the unification of “isolated” objects,

relations between the objects, and the structure of layers.
2.1 Various Attempts 27

The idea of systems appeared at least as early as the one-element Ionians

(624–500 B.C.). For example, in the search for order, Ionians had a naturalistic
and materialistic bent. They sought causes and explanations in terms of the eternal
working of things themselves rather than in any divine, mythological, or supernatural
intervention. Looking for a single basic reality, each Ionian believed that all things
have their origin in a single knowable element: water, air, fire, or some indeterminate,
nebulous substance (Perlman 1970).
In history, the idea of systems has been used by many great thinkers to study
different problems. For example, Aristotle’s statement “the whole is greater than the
sum of its parts” could be an early definition of a basic systems problem. Nicholas of
Cusa, a profound thinker of the fifteenth century, linking medieval mysticism with
the first beginnings of modern science, introduced the notion of the coincidentia
oppositorum. Leibniz’s hierarchy of monads looks quite like that of modern systems.
Gustav Fechev, known as the author of the psychophysical law, elaborated, in the way
of the naive philosophers of the nineteenth century, supraindividual organizations of
higher order than the usual objects of observation, thus romantically anticipating the
ecosystems of modern parlance (von Bertalanffy 1972). For our purpose here, we
list only a few of the great thinkers.
In the second decade of the twentieth century, von Bertalanffy (1934) wrote:
Since the fundamental character of the living thing is its organization, the customary inves-
tigation of the single parts and processes cannot provide a complete explanation of the vital
phenomena. This investigation gives us no information about the coordination of parts and
processes. Thus the chief task of biology must be to discover the laws of biological systems
(at all levels of organization). We believe that the attempts to find a foundation for theo-
retical biology point at a fundamental change in the world picture. This view, considered
as a method of investigation, we shall call “organismic biology” and, as an attempt at an
explanation, “the system theory of the organism.”

Here, a new concept, “system” was introduced. After nearly one hundred years
of testing, this concept has been widely accepted by scientists in all disciplines
(Blauberg et al. 1977). To establish a theoretical foundation for all approaches of
systems analysis, developed in different disciplines, Mesarovic, in the early 1960s,
introduced the mathematical definition of (general) systems, based on Cantor’s set
theory, as follows (Mesarovic and Takahara 1975): A (general) system S is a relation
on nonempty sets Vi :

S⊆ {Vi : i ∈ I }

where I is an index set.

This structure of general systems reflects the unification of “isolated” objects,
relations between the objects, and the structure of layers. Here, elements in the
nonempty sets Vi , i ∈ I , are the objects of the system, S represents a relation between
objects, and elements of the sets Vi , i ∈ I , could be systems too. Using this idea of
layers inductively, we can see that an element S1 of S could be a system, an element
S2 of the system S1 could be a system, … At this junction, a natural question arises:
Can this process go on forever? If the answer is “yes,” then “the world would be
28 2 The Concept of General Systems

infinitely divisible”! If the answer to the question is “no,” does that mean the world
is made up of fundamental elements?
The idea that the world is infinitely divisible can be found in Anaxagoras’ “seeds”
philosophy (510–428 B.C.). He was unwilling to submerge the tremendous varieties
of things into any common denominator and preferred to accept the immediate diver-
sity of things as is. With his philosophy, every object is infinitely divisible. No matter
how far an object is divided, what is left would have characteristics of the original sub-
stance. The opposite idea that the world is made up of fundamental particles appeared
during 500–55 B.C. The Leucippus–Democritus atom combined features of the Ioni-
ans’ single element, Anaxagoras’ seeds, and some thoughts of other schools and yet
was an improvement over all of them. “Atom” means “not divisible” in Greek. This
term was intentionally chosen by Democritus to emphasize a particle so small that
it could no longer be divided. To Leucippus and Democritus, the universe originally
and basically consisted entirely of atoms and a “void” in which atoms move.
Upon consideration of the interrelationship between the systems under concern
and some environments of the systems, Bunge (1979) gave a model of systems as
follows: For nonempty set T , the ordered triple W  (C, E, S) is a system over T
if and only if C and E are mutually disjoint subsets of T and S is a nonempty set of
relations on C ∪ E; C and E are called the composition and an environment of the
system W , respectively. Klir (1985) introduced a philosophical concept of general
systems, which reads: “a system is what is distinguished as a system.” Theoretically,
Klir’s definition contains the most general meaning of the concept of systems origi-
nally posed by von Bertalanffy. Starting in 1976, Xuemou Wu and his followers have
studied many different theories, under the name “pansystems.” Pansystems analysis
is a new research of multi-levels across all disciplines. The theory deals with gen-
eral systems, relations, symmetry, transformation, generalized calculus, and shengke
(means survival and vanquishing), called the emphases of pansystems. Based on the
research of these emphases, the theory blends philosophical reasoning, mathematical
logic, and mechanical structures into one solid body of knowledge (Wu 1990).
The discussion here gives rise naturally to the following question: Why was the
concept of systems not studied more before now? One answer is that the develop-
ment of modern technology, e.g., computer technology, designs of satellites, climate
control of giant buildings, etc., reveals one fact: History is in a special moment,
where the accumulation of knowledge has reached such a level that each discovery
of a relation between different areas of knowledge will materially produce a useful
or consumable product.
Closely related to the concept of systems are the following questions:
(1) What is the meaning of a system with two contradictory relations, e.g., {x > y,
x ≤ y} as the relation set?
(2) How can we know whether there exist contradictory relations in a given realistic
system?
From Klir’s definition of systems, which is one of the definitions with the widest
meaning among all definitions of systems, it follows that generally these two ques-
tions cannot be answered. Concerning this, a theorem of Gödel shows that it is
2.1 Various Attempts 29

impossible to show whether the systems representation of mathematics, developed

on the ZFC axioms, is consistent. For a detailed discussion of the systems represen-
tation of mathematics, see Lin (1989a). This fact implies that there are systems, for
example, the system of set theory on ZFC, so that we do not know if propositions
with contradictory meanings exist or not. This means that maybe not every system
is consistent or that not every system has no contradictory relations.
By rewriting Russell’s famous paradox in the language of systems theory, the
following is true: There does not exist a system whose object set consists of all
systems, where S is a system, if and only if S is an ordered pair (M, R) of sets such
that R is a set of some relations defined on the set M (Lin 1987). The elements in M
are called the objects of S. The sets M and R are called the object set and the relation
set of S, respectively. Now, for any given system S  (M , R), each relation r ∈ R
can be understood as an S-truth. That is, r is true among the objects in S. Hence,
Russell’s paradox implies that there does not exist any statement (or relation) which
describes a fact among all systems. If any structure in the world can be studied as a
system, as Klir’s definition of systems says, does Russell’s paradox say that there is
no universal truth or absolute truth? It is well known in the mathematics community
that no theorem is true in all axiom systems.
Descartes and Galileo developed individually the following methods about sci-
entific research and administration: Divide the problem under consideration into as
many small parts as possible, study each isolated part (Kline 1972), and simplify the
complicated phenomenon into basic parts and processes (Kuhn 1962). In the history
of science and technology, Descartes’ and Galileo’s methods have been very suc-
cessfully applied. They guaranteed that physics would win great victories one by one
(von Bertalanffy 1972), and not only those, because their methods are still widely
used for research of fundamental theories and in modern laboratories. At the same
time, due to the tendency of modern science toward synthesis and the transverse
development of modem technology, more and more scientists and administrators
are forced to study problems with many cause–effect chains (i.e., systems). Some
examples are the many-body problem in mechanics, the body structure of living
things (but not the particles constituting the living things), etc. In the study of such
problems, Descartes’ and Galileo’s methods need to be modified somehow, because
their methods emphasize separating the problem and phenomenon under consider-
ation into isolated parts and processes. But according to von Bertalanffy, it should
be recognized that the world we live in is not a pile of innumerable isolated “parts”,
and any practical problem and phenomenon cannot be described perfectly by only
one cause–effect chain. The basic character of the world is its organization and the
connection between the interior and the exterior of different things. Thus, the chief
task of modem science should be a systemic study of the world.
30 2 The Concept of General Systems

2.1.2 The Set-Theoretical Definition of General Systems

In order to produce useful results for theoretical and practical purposes, let us from
here on focus on the definition of general systems introduced by Lin (1987).
Let A be a set and n be an ordinal number. An n-ary relation r on A is a subset of
the Cartesian product An . In the sequel, An means either the Cartesian product of n
copies of A or the set of all mappings from n into A. From set theory (Lin 1999), it
is easy to see that the two definitions of An are equivalent.
If r is a relation on A, there then exists an ordinal n  n(r), a function of r, such
that

r ⊆ An .

The ordinal number n  n(r) is called the length of the relation r. When r is the
empty relation ∅, assume that n(∅)  0; i.e., the length of the relation ∅ is 0.
S is a (general) system if S is an ordered pair (M, R) of sets, where R is a set of
some relations on the set M. Each element in M is called an object of S, and M and
R are called the object set and the relation set of S, respectively. The system S is
discrete if R  ∅ or R  ∅ and M  ∅. It is trivial if M  ∅.
The idea of systems, roughly speaking, appeared as long ago as the time of Yel-
low Emperor in China and the time of Aristotle. For example, Chinese traditional
medicine can be dated back to the time of Yellow Emperor and could be seen as
the earliest theory of systems, and Aristotle’s statement that “the whole is greater
than the sum of its parts” could be treated as the first definition of a basic systems
problem. Later, many great thinkers used the languages of their times to study certain
systems problems. Nicholas of Cusa, for example, a profound thinker of the fifteenth
century, linking medieval mysticism with the first beginnings of modern science,
introduced the notion of the coincidentia oppositorum. Leibniz’s hierarchy of mon-
ads looks quite like that of modem systems. Gustav Fechner, known as the author of
the psychophysical law, elaborated in the way of the native philosophers of the nine-
teenth century supraindividual organizations of higher order than the usual objects
of observation, thus romantically anticipating the ecosystems of modem parlance;
for details, see von Bertalanffy (1972).
However, the concept of systems was not introduced formally until the 1920s
when Von Bertalanffy began to study the concept of systems formally in biology.
Since then, more and more scholars have studied the concept of systems and related
topics. For example, Tarski (1954–1955) defined the concept of relational systems.
Hall and Fagan (1956) described a system as a set of objects and some relations
between the objects and their attributes. They did not define mathematical meanings
for objects nor for attributes of the objects. In 1964, Mesarovic began to study the
model of general systems in the language of set theory. His final model (Mesarovic
and Takahara 1975) reads as follows: A system S is a relation on nonempty sets:

S⊆ {Vi : i ∈ I }. (2.1)
2.1 Various Attempts 31

After considering the interrelationship between the systems under concern and
some environments of systems, Bunge (1979) gave a model of systems as follows:
Let T be a nonempty set. Then the ordered triple W  (C, E, S) is a system over T
if and only if C and E are mutually disjoint subsets of T and S is a nonempty set of
relations on C ∪ E; the sets C and E are called the composition and an environment of
the system W , respectively. Klir (1985) introduced a philosophical concept of general
systems. The concept of general systems discussed intensively in this chapter and
the rest of this book was first introduced by Lin (1987) and Lin and Ma (1987a, b).
In the following, we will conclude this section with three mathematical structures
related to the concept of general systems: structure, L-fuzzy system, and G-system.

2.1.3 Structures

Let A be a set and n a nonnegative integer. An n-ary operation on A is a mapping

f from An into A. An n-ary relation r on the set A is a subset of An . A type τ of
structures is an ordered pair
 
(n0 , . . . , nν , . . .)ν<00(τ ) , (m0 , . . . , mν , . . .)ν<01(τ ) , (2.2)

where 00(τ ) and 01(τ ) are fixed ordinals and nν and mν are nonnegative integers.
For every ν < 00(τ ) , there exists a symbol fν of an nν -ary operation, and for every
ν < 01(τ ) , there exists a symbol rν of an mν -ary relation.
A structure U (Gratzer 1978) is a triplet (A, F, R), where A is a nonempty set. For
every ν < 00(τ ) , we realize fν as an nν -ary operation (fν )U on A; for every ν < 01(τ ) ,
we realize rν as an mν -ary relation (rν )U on A, and
 
F  (f0 )U , . . . , (fν )U , . . . , ν < 00(τ ) , (2.3)
R  {(r0 )U , . . . , (rν )U , . . .}, ν < 01(τ ) . (2.4)

If 01(τ )  0, U is called an algebra and if 00(τ )  0, U is called a relational system.

It can be seen that the concept of algebras is a generalization of the concepts of
rings, groups, and the like. Any n-ary operation is an (n + 1)-ary relation. Therefore,
the concept of relational systems generalizes the idea of algebras, and the concept
of structures combines those of algebras and relational systems. The fact that any
topological space is a relational system shows that the research of relational systems
will lead to the discovery of properties that topologies and algebras have in common.
Example 2.1. We show that any topological space is a relational system, thus a
general system. Let (X, T ) be a topological space; for details, see Engelking (1975).
Then for each open set U ∈ T , there exists an ordinal number n  n (U)  1 such
that U ∈ X n . Therefore, (X, T ) is also a system with object set X, and relation set T ,
and the length of each relation in T is 1. Now, applying the Axiom of Choice implies
that the topology T can be well-ordered as
32 2 The Concept of General Systems

T  {U0 , U1 , . . . , Uν , . . .}, ν < |T |, (2.5)

where each open set is a 1-ary relation on X. So (X , ∅, T ) becomes a structure, which

is a relational system.
A systematic introduction to the study of structures can be found in (Gratzer 1978).
Many great mathematicians worked in the field, including Whitehead, Birkhoff,
Chang, Henkin, Jonsson, Keisler, Tarski, etc.

2.1.4 L-Fuzzy Systems

The concept of fuzzy systems (Lin 1990) is based on the concept of fuzzy sets intro-
duced in 1965 by Zadeh. An ordered set (L, ≤) is called a lattice if for any elements
a, b ∈ L, there exists exactly one greatest lower bound of a and b, denoted by a ∧ b,
and there exists exactly one least upper bound of a and b, denoted by a ∨ b, such that

a ∨ a  a, a ∧ a  a, a ∨ b  b ∨ a, a ∧ b  b ∧ a,
a ∨ (b ∨ c)  (a ∨ b) ∨ c, a ∧ (b ∧ c)  (a ∧ b) ∧ c,
a ∧ (a ∨ b)  a, a ∨ (a ∧ b)  a

We call a lattice distributive if the lattice, in addition, satisfies

a ∧ (b ∨ c)  (a ∧ b) ∨ (a ∧ c) and a ∨ (b ∧ c)  (a ∨ b) ∧ (a ∨ c).

We call a lattice complete if for any nonempty subset A ⊆ L, there exists exactly
one least upper bound of the elements in A, denoted by ∨A or supA, and there exists
exactly one greatest lower bound of the elements in A, denoted by ∧A or infA.
We call a lattice completely distributive if for any a ∈ L and any subset
{bi : i ∈ I } ⊆ L , where I is an index set, ∨{bi : i ∈ I } and ∧{bi : i ∈ I } exist and
satisfy

a ∨ (∧{bi : i ∈ I })  ∧{a ∨ bi : i ∈ I } (2.6)

and

a ∧ (∨{bi : i ∈ I })  ∨{a ∧ bi : i ∈ I }. (2.7)

We say that L is a lattice with order-reversing involution if there exists an

operator
defined on L such that for any a, b ∈ L, if a ≤ b, then a
≥ b
.
In the following, a completely distributive lattice L with order-reversing involution
is fixed; assume that 0 and 1 are the minimum and maximum elements in L.
Let X be a set. A is an L-fuzzy set on X if and only if there exists a mapping
μA ∈ LX such that A  {(x, μA (x)) : x ∈ X }. Without confusion, we consider that A
2.1 Various Attempts 33

is the mapping μA in LX . If μA , or say A, has at most two values, 0 and 1, then the
L-fuzzy set A can be seen as a subset of X.
Suppose that A and B are two L-fuzzy sets on X. Then A  B if and only if
μA (x)  μB (x) for all x ∈ X. A ⊂ B if and only if μA (x) ≤ μB (x), for all x ∈ X. If Y
⊂ X, then A|Y  {(x, μA (x)) : x ∈ Y }.
Let {Ai : i ∈ I } be a set of L-fuzzy sets on X. Then C  ∪i∈I Ai and D  ∩i∈I Ai
are defined by
 
μC (x)  sup μAi (x) : i ∈ I (2.8)

and
 
μD (x)  inf μAi (x) : i ∈ I , (2.9)

and for all x ∈ X.

S is an L-fuzzy system if and only if S is an ordered pair (M, R) of sets, where R
is a set of some L-fuzzy relations defined on the set M; i.e., for any L-fuzzy relation
r ∈ R, there exists an ordinal number n  n(r) such that r is an L-fuzzy set on the
Cartesian product M n . The ordinal n is called the length of the L-fuzzy relation r.

2.1.5 G-Systems

In this subsection, any set A will be seen as a partially ordered set (A, ≤A ), where the
ordering is defined by ≤A  {(x, x) : x ∈ A}.
S is a G-system (Lin and Ma 1989) of type K  {Ki }i∈I , where I is an index set
and, for each i ∈ I, Ki is a set partially ordered by ≤i if and only if S is an ordered
pair of sets M , {ri }i∈I such that for any i ∈ I , ri ⊆ M Ki . The set Ki is called the
type of the relation ri , and the set K is called the type of the system S.
The concept of G-systems can be used to study more applied problems because
it contains a more general meaning of relations. The interested reader is encouraged
to consult with (Lin and Ma 1989).
In the rest of this book, our discussion is mainly focused on the definition that S
is a (general) system if S is an ordered pair (M, R) of sets, where R is a set of some
relations on the set M. Each element in M is called an object of S, and M and R are
called the object set and the relation set of S, respectively.

2.2 Internalities of Systems

Let Si  (Mi , Ri ), i  1, 2, be two systems. We say that they are equal or identical,
denoted S1  S2 , if
34 2 The Concept of General Systems

M1  M2 and R1  R2 .

The system S1 is a partial system of the system S2 , if M1 ⊆ M1 , and for each

relation r1 ∈ R1 , there exists a relation r2 ∈ R2 such that r1  r2 |M1 , where r2 |M1 is
the restriction of r2 on M1 , which is defined by

r2 |M1  r2 ∩ M1n(r2 ) .

The system S1 is a subsystem of S2 if M1 ⊆ M2 , and for each relation r1 ∈ R1 ,

there exists a relation r2 ∈ R2 , satisfying r1 ⊆ r1 |M1 .

Theorem 2.1. There is no such a system that the object set of the system consists of
all systems.

This is a restatement of the Zermelo–Russel paradox.

Proof: Consider the class V of all sets. For each x ∈ V, define a system by (x, ∅).
Then a 1–1 correspondence h is defined such that h(x)  (x, ∅). If there is a system
S  (M, R) whose object set contains all systems, then M is a set. Therefore, the
Axiom of Comprehension shows that V is a set, contradiction. QED

A system Sn  (Mn , Rn ) is an nth-level object system of S0  (M0 , R0 ) if there

exist systems Si  (Mi , Ri ), for i  1, 2, . . . , n−1, such that Si is an object in
Mi−1 , 1 ≤ i ≤ n. Each element in Mn is called an nth-level object of S0 .

Theorem 2.2 [ZFC]. Let S0  (M0 , R0 ) be a system and Sn  (Mn , Rn ) an nth-level

object system of S0 . Then S0 cannot be a subsystem of Sn for each natural number
n ∈ N.

Proof: The theorem will be proved by contradiction. Suppose that for a certain
natural number n, the system S0 is a subsystem of the nth-level object system Sn .
Let Si  (Mi , Ri ) be the systems, for i  1, 2, …, n − 1, such that Si ∈ Mi−1 , i 
1, 2, . . ., n . Define a set X by

X  {Mi : 0 ≤ i ≤ n} ∪ {Si : 0 ≤ i ≤ n} ∪ {{Mi } : 0 ≤ i ≤ n}.

From the Axiom of Regularity, it follows that there exists a set Y ∈ X such that
Y ∩ X  ∅. There now exist three possibilities:
(i) Y  Mi for some i;
(ii) Y  Si for some i; and
(iii) Y  {Mi } for some i.
If possibility (i) holds true, then we have Si+1 ∈ Y ∩ X , for i ≤ n − 1, and
S1 ∈ Y ∩ X if Y  Mn . Therefore, Y ∩ X  ∅, a contradiction. If (ii) holds true,
then we have Y  Si  (Mi , Ri )  {{Mi }, {Mi , Ri }} and {Mi } ∈ Y ∩ X  ∅, a
contradiction. If (iii) holds true, then we have Mi ∈ Y ∩ X  ∅, a contradiction.
These contradictions show that S0 cannot be a subsystem of Sn . QED
2.2 Internalities of Systems 35

A chain of object systems of a system S is a sequence {Si : i < α}, for some
ordinal number α, of different-level object systems of S, such that for each pair i,
j < α with i < j, there exists an integer n  n(i, j), a function of i and j, such that Sj
is an nth-level object system of Si . A similar proof to that of Theorem 2.2 gives the
following.
Theorem 2.3 [ZFC]. Each chain of object systems of S must be finite.
Theorem 2.4 [ZFC]. For any system S, there exists exactly one set M(S) consisting
of all fundamental objects in S, where a fundamental object in S is a level object of
the system S which is no longer a system.
Proof: Suppose that S  (M0 , R0 ). Define
∗
M0  {x ∈ M0 : x is not a system}

and
 
0 
M Mx : x  (Mx , Rx ) ∈ M0 − ∗ M0

Then ∗ M0 and M0 are sets.

 
Suppose that for a natural number n, two sequences ∗ Mi : i  0, 1, 2, . . . , n
i : i  0, 1, 2, . . . , n} have been defined such that
and {M
 
∗ i−1 : x is not a system
Mi  x ∈ M

and
 
i 
M i−1 − ∗ Mi .
Mx : x  (Mx , Rx ) ∈ M

We then define
 
∗ i : x is not a system and
Mi+1  x ∈ M
 
i+1  ∪ Mx : x  (Mx , Rx ) ∈ M
M i − ∗ Mi+1 .
 
∗By mathematical
 induction, a sequence ∗ Mi : i ∈ ω can be defined. Let M (S) 
∪ Mi : i ∈ ω . Then Theorem 2.3 guarantees that M(S) consists of all fundamental
objects in S, and from the uniqueness of each ∗ Mi , it follows that M(S) is uniquely
determined by S. QED

2.3 The Edifice of Systems

The following result describes the overall structure of general systems when each
relation is binary.
36 2 The Concept of General Systems

Theorem 2.5. Let S  (M, R) be a system such that, for any r ∈ R, the length n(r) 
2. Then
 
S ⊆ p2 (M ) ∪ p p(M ) ∪ p4 (M ) , (2.10)

where p(X) is the power set of X and pi+1 (X )  p(pi (X )) for each i  1, 2, ….
Proof: For each element, (x, y) ∈ r, (x, y)  {{x}, {x, y}} ∈ p2 (M
 ) and r ∈ p3 (M
 ).
Therefore, R ∈ p (M ). Since {M } ∈ p(M ) and (M , R) ∈ p p(M ) ∪ p (M ) , it
4 4

follows that the inclusion in equation (2.10) holds true. QED

Question 2.1. Give a structural representation for general systems similar to that in
Theorem 2.5.
A system S  (M, R) is called centralized if each object in S is a system, and
there exists a nontrivial system C  (MC ,RC ) such that for any distinct elements x,
y ∈ M, say x  (Mx , Rx ) and y  My , Ry , then

MC  Mx ∩ My and RC ⊆ Rx |MC ∩ Ry MC ,
 
where Rx |MC  {r|MC : r ∈ Rx } and Ry MC  r|MC : r ∈ Ry . The system C is
called a center of the centralized system S.
Theorem 2.6 [ZFC]. Let κ be any infinite cardinality and θ > κ a regular cardi-
nality such that, for any α < θ , |α <κ | < θ . Assume that S  (M, R) is a system
satisfying |M | ≥ θ and each object m ∈ M is a system with m  (Mm , Rm ) and
|Mm | < κ. If there exists an object that

is contained
 in at least θ objects in M, then

there exists a partial

system S  M , R of S such that S
forms a centralized
system and M ≥ θ .

Proof: Without loss of generality, we assume that |M |  θ and that there is a

common element in all the object systems in M. Then we have

|∪{Mx : x  (Mx , Rx ) ∈ M }| ≤ θ.

Since the specific objects in each Mx , for each object x  (Mx , Rx ) ∈ M , are
irrelevant, we may assume that

∪{Mx : x  (Mx , Rx ) ∈ M } ⊆ θ.

Then, for each x  (Mx , Rx ) ∈ M , the object set Mx has some order type < κ
as a subset of θ . Since θ is regular and θ > κ, there exists a ρ < κ such that
M1  {x ∈ M : Mx has order type ρ} has cardinality θ . We now fix such a ρ and
deal only with the partial system S1  (M1 , R1 ) of S, where R1 is the restriction of
the relation set R on M1 .
For each α < θ , |α <κ | < θ implies that less than θ objects of the partial system
S1 have object sets as subsets of α. Thus, ∪{Mx : x  (Mx , Rx ) ∈ M1 } is cofinal in θ .
2.3 The Edifice of Systems 37

If x ∈ M1 and ξ < ρ, let Mx (ξ ) be the ξ th element of Mx . Since θ is regular, there is

some ξ such that {Mx (ξ ) : x ∈ M1 } is cofinal in θ . Now, fix ξ0 to be the least such ξ .
Then the condition that there exists a common element in each system in M1 implies
that we can guarantee that ξ0 > 0. Let

α0  ∪{Mx (η) + 1 : x ∈ M1 and η < ξ0 }.

Then α0 < θ and Mx (η) < α0 for all x ∈ M1 and all η < ξ0 .
By transfinite induction on μ < θ , pick xμ ∈ M1 so that Mxμ (ξ0 ) > α0 and
Mxμ (ξ0 ) is above all elements of earlier xv ; i.e.,

 
Mxμ (ξ0 ) > max α0 , Mxυ (η) : η < ρ and ν < μ .
 
Let M2  xμ : μ < θ . Then |M2 |  θ and Mx ∩ My ⊆ α0 whenever x 
(Mx , Rx ) and y  My , Ry are distinct objects in M2 . Since for each α < θ , we have
<κ
α < θ,

there exists an r ⊂ α0 and a B ⊂ M2 with |B|  θ and for each x ∈ B, Mx ∩ α0  r,

S2  (B, RB ) forms a centralized system, where RB is the restriction of the relation
set R on B. QED

Corollary 2.1. If S is a system with uncountable (or nondenumerable) object set, if

each object in S is a system with a finite object set, and if there exists an object that
is contained in at least ℵ1 objects in S, then there exists a partial system S ∗ of S with
an uncountable object set and S ∗ forms a centralized system.
 
Proof: It suffices from Theorem 2.6 to show that ∀α < ℵ1 α <ℵ0 < ℵ1 . This is
clear because for each α < ℵ1 , |α| ≤ ℵ0 . So, we have
<ℵ
α 0  ℵ0 < ℵ1 . QED (2.11)

A system S  (M, R) is strongly centralized if each object in S is a system and there

is a nondiscrete system C  (M  C , RC ) such that for any distinct elements x and y ∈
M, say x  (Mx , Rx ) and y  My , Ry , MC  Mx ∩ My and RC  Rx |MC ∩ Ry |MC .
The system C is called an S-center of S.

Question 2.2. Give conditions under which a given system has a partial system
which is strongly centralized and has an object set of the same cardinality as that of
the given system.

In terms of L-fuzzy systems, we have the following result in terms of their struc-
tures when each relation is binary.

Theorem 2.7. Suppose that an L-fuzzy system S  (M, R) is such that, for any r ∈
R, n(r)  2. Then we have the following:
38 2 The Concept of General Systems
  
S ⊆ p M ∪ p3 p2 (M ) ∪ L . (2.12)

Proof: S  (M, R)  {{M}, {M, R}} ⊆ p(M ∪ R), and for any (x, y, i) ∈ r ⊆ M 2 ×L ,
we have

(x, y, i)  ((x, y), i)  {{(x, y)}, {(x, y), i}},

 
{(x, y), i}  {{{{x}, {x, y}}}, {i, {{x}, {x, y}}}} ∈ p p3 (M ) ∪ p2 (M ∪ L) 
    
 p p p2 (M ∪ L)  p2 p2 (M ) ∪ L .
   
So r ∈ p3 p2 (M ) ∪ L and R ⊆ p3 p2 (M ) ∪ L . Thus, the inclusion in equa-
tion (2.12) holds true. QED
Suppose that Si  (Mi , Ri ), i  1, 2, are L-fuzzy systems. The L-fuzzy system
S1 is a subsystem of the L-fuzzy system S2 if M1 ⊆ M2 , and for any L-fuzzy relation
r1 ∈ R1 , there exists an L-fuzzy relation r2 ∈ R2 such that r1 ⊆ r2 |M1 .
Example 2.2. An example is constructed to show that the equation S1  S2 will not
follow from the condition that the L-fuzzy systems S1 and S2 are subsystems of each
other.
Suppose that the completely distributive lattice with order-reversing involution
lattice L is the closed unit interval with the usual order relation. Let N  {1, 2, . . .}
be the set of all natural numbers, and {ai : i ∈ N} a sequence defined by ai  1−1/i.
We now define two L-fuzzy systems S1  (N, R1 ) and S2  (N, R2 ) as follows:

R1  {r11 , r12 , . . . , r1i , . . .} and R2  {r20 , r21 , . . . , r2i , . . .},

where

r1i (x)  ai , for each x ∈ N (2.13)

and for any i, x ∈ N


ai+1 , ifx  1
r2i (x)  (2.14)
ai , ifx  1

and

r20 (x)  0, for all ∈ N. (2.15)

Then it is easy to show that S1 and S2 are subsystems of each other, but S1  S2 .
Remarks on references: All the discussions, presented above, are based on the fol-
lowing works by various scholars, Bunge (1979), Engelking (1975), Gratzer (1978),
Hall and Fagan (1956), Klir (1985), Lin (1987, 1990, 1999), Lin and Ma (1987a; b,
1989), Mesarovic (1964), Mesarovic and Takahara (1975), Tarski (1954–1955), von
Bertalanffy (1972), and Zadeh (1965).
References 39

References

Berlinski, D.: On Systems Analysis. MIT Press, Cambridge, MA (1976)

Blauberg, I.V., Sadovsky, V.N., Yudin, E.G.: Systems Theory, Philosophical and Methodological
Problems. Progress Publishers, Moscow (1977)
Bunge, B.: Treatise on Basic Philosophy, vol. 4. A World of Systems. Reidel, Dordrecht, Holland
(1979)
Engelking, R.: General Topology. Polish Scientific Publishers, Warszawa (1975)
Gratzer, G.: Universal Algebra. Springer, New York (1978)
Hall, A.D., Fagen, R.E.: Definitions of systems. Gen. Syst. 1, 18–28 (1956)
Kline, M.: Mathematical Thought from Ancient to Modern Times. Oxford University Press, Oxford
(1972)
Klir, G.: Architecture of Systems Problem Solving. Plenum Press, New York (1985)
Kuhn, T.: The Structure of Scientific Revolutions. University of Chicago Press, Chicago (1962)
Lilienfeld, D.: The Rise of Systems Theory. Wiley, New York (1978)
Lin, Y.: A model of general systems. Math. Modell. Int. J. 9(2), 95–104 (1987)
Lin, Y., Ma, Y.: Remarks on the Definition of Systems. Syst. Anal. Model Simu. 6(11), 923–931
(1989)
Lin, Y.: A multi-relation approach of general systems and tests of applications. Syn. Int. J. Epistem.
Methodol. Phil. Sci. 79, 473–488 (1989a)
Lin, Y.: The concept of fuzzy systems. Kybernetes: Int. J. Systems. Cybernet. 19, 45–51 (1990)
Lin, Y.: General Systems Theory: A Mathematical Approach. Kluwer Academic and Plenum Pub-
lishers, New York (1999)
Lin, Y., Ma, Y.H.: Remarks on analogy between systems. Int. J. Gen. Syst. 13, 135–141 (1987a)
Lin, Y., Ma, Y.H.: On the stabilities of input-output systems. Cybern. Syst. Int. J. 18(4), 285–298
(1987b)
Mesarovic, M.D.: Views on general systems theory. In: Mesarovic, M.D. (ed.) Proceedings of the
2nd Systems Symposium at Case Institute of Technology. Wiley, New York (1964)
Mesarovic, M.D., Takahara, Y.: General Systems Theory: Mathematical Foundations. Academic
Press, New York (1975)
Perlman, J.S.: The Atom and the Universe. Wadsworth, Belmont, CA (1970)
Tarski, A.: Contributions to the theory of models I, II, III. Nederl. Akad. Wetensch. Proc. Ser. A.
57, 572–581, 582–588; 58, 56–64 (1954, 1955)
von Bertalanffy, L.: Modern Theories of Development (Woodge, J.H., Trans.). Oxford Univer-
sity Press, Oxford (1934); Harper Torch Books, New York (1962); German Original: Kritische
Theories der Formbildung, Borntäger, Berlin (1928)
von Bertalanffy, L.: The history and status of general systems theory. In: Klir, G. (ed.) Trends in
General Systems Theory. New York, pp. 21–41 (1972)
Wigner, E.P.: The unreasonable effectiveness of mathematics in the natural sciences. Comm. Pure
Appl. Math. 13, 1–14 (1960)
Wu, X.-M.: The Pansystems View of the World. People’s University of China Press, Beijing (1990)
Zadeh, L.: Fuzzy sets. Inf. Control 8(3), 338–353 (1965)
Chapter 3
Mappings and Constructions of Systems

This chapter considers interactions between systems through mappings and the con-
struction of large-scale systems. In particular, Section 3.1 looks at various mappings
from one general system into another. Section 3.2 studies the construction of free
sums of systems and related mapping properties. Section 3.3 shows why product and
Cartesian product systems exist. And Section 3.4 considers the concept of one-level
product systems.

3.1 Mappings from Systems into Systems

Let Si  (Mi , Ri ), i  1, 2, be two systems and h: M 1 → M 2 a mapping. By transfinite

 

induction, two classes M̂i , i = 1, 2, and a class mapping ĥ: M 1 → M 2 can be defined
with the following properties:
 
Mi  Min , i  1, 2, (3.1)
n∈ON

where ON stands for the collection of all ordinals, and for each x


(x0 , x1 , . . . , xα , . . .) ∈ M 1 ,

ĥ(x)  (h(x0 ), h(x1 ), . . . , h(xα ), . . .). (3.2)

 
For each relation r ∈ R1 , ĥ(r)  ĥ(x): x ∈ r is a relation on M2 with length
n(r). Without confusion, h will also be used to indicate the class mapping ĥ and h is
a mapping from S1 into S2 , denoted h: S1 → S2 . When h: M1 → M2 is surjective,
injective, or bijective, h: S1 → S2 is also called surjective, injective, or bijective,
respectively. Let X be a subsystem of S1 . Then h|X indicates the restriction of the
mapping on X. Without confusion, we use h for h|X.

© Springer Nature Switzerland AG 2018 41

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_3
42 3 Mappings and Constructions of Systems

The systems Si are similar, if there exists a bijection h: S1 → S2 such that h(R1 ) 
{h(r): r ∈ R1 }  R2 . The mapping h is a similarity mapping from S1 onto S2 . Clearly,
the inverse mapping h−1 is a similarity mapping from S2 onto S1 . A mapping h:
S1 → S2 is a homomorphism from S1 into S2 if h(R1 ) ⊆ R2 .
Let E be an equivalence relation on the object set M of a system S = (M, R). The
quotient system S/E is defined by

S/E  (M /E, R/E), (3.3)

where μ: M → M/E is the canonical mapping defined by μ(m)  [m]  the equiva-
lence class of m. Then R/E = {μ(r): r ∈ R}.

Theorem 3.1. Suppose that h: S1 → S2 is a homomorphism from S1 into S2 .

Then h̄: S1 /Eh → h(S1 )  (h(M1 ), h(R1 )) is a similarity mapping, where Eh 
{(x, y): h(x)  h(y)} and h̄([m])  h(m), for each [m] ∈ M1 /Eh .

Proof: It is easy to check that Eh is an equivalence relation on the object set M1 .

We next show that h̄: M1 /Eh → h(M1 ) is well defined. Let [m1 ], [m2 ] ∈ M1 /Eh be
such that [m1 ]  [m2 ], then m2 ∈ [m1 ], and so h(m2 )  h(m1 ). This implies that
h̄([m1 ])  h̄([m2 ])
Finally, we show that h̄ is a similarity mapping. First, h̄: M1 /Eh → h(M1 ) is
a bijection. In fact, let [m1 ] and [m2 ] ∈ M1 /Eh be such that [m1 ]  [m2 ] and
h̄([m1 ])  h̄([m2 ]). Therefore, h(m1 )  h(m2 ). This implies that (m1 , m2 ) ∈ Eh , and
so [m1 ]  [m2 ], a contradiction.
For any relation r ∈ R1 /Eh , there exists a relation s ∈ R1 such that r = μ(s), where
μ is the canonical mapping from M1 into M1 /Eh . Then we have

h̄(r)  h̄(μ(s))  h(s). (3.4)

Thus, h̄(r) ∈ h(R1 ).

For any relation r ∈ h(R1 ), there exists a relation s ∈ R1 such that r  h(s).
Therefore, h̄−1 (r)  μ(s) ∈ R1 /Eh . This completes the proof that h̄ is a similarity
mapping from S1 /Eh onto h(S1 ). QED

Proposition 3.1. Suppose that A, B, and C are systems.

(a) If A is n-level homomorphic to B and B is m-level homomorphic to C, then A is
min{n, m}-level homomorphic to C.
(b) If A is n-level similar to B and B is m-level similar to C, then A is min{n, m}-level
similar to C.

The proof is straightforward and is omitted.

A system S is centralizable if it is one-level homomorphic to a centralized system
SC under a homomorphism h: S → SC such that for each object m in S, the object
systems m and h(m) are similar. Each center of SC is also called a center of S.
3.1 Mappings from Systems into Systems 43

Theorem 3.2. Let S  (M, R) be a system such that each object in M is a system.
Then S is centralizable, if and only if there exists a nontrivial system C  (MC , RC )
such that C is embeddable in each object of S, meaning that C is similar to a partial
system of each object of S, and each similarity mapping here is called an embedding
mapping from C into the object of S.
Proof: Necessity. If the system S  (M , R) is centralizable, there exists a centralized
system S ∗  (M ∗ , R∗ ) such that S is one-level homomorphic to S ∗ . Let C be a center
of S ∗ . Then C is embeddable in each object of S.
Sufficiency. Without loss of generality, we can assume that the collection of all
object sets of the objects in S consists of disjoint sets. For each x  (Mx , Rx ) ∈ M ,
the system C is embeddable in x; so there exists a one-to-one mapping hx : MC → Mx
 hx (C)  (hx (MC ), hx (RC )) is a partial system of x. Define a new system
such that
x∗  ∗ Mx , ∗ Rx as follows:
∗
Mx  MC ∪ (Mx − hx (MC )) (3.5)

and let gx : Mx →∗ Mx be the bijection defined by

m, if m ∈ Mx − hx (MC )
gx (m)  −1
. (3.6)
hx (m), if m ∈ hx (MC )

Then define
∗
Rx  {gx (r): r ∈ Rx } (3.7)

So C is a partial system of x∗ . Let h∗ be a mapping defined by letting h∗ (x)  x∗

for each x ∈ M. Then the system h∗ (S) is centralized with the system C as a center.
Therefore, S is a centralizable system. QED
Example 3.1. An example is constructed to show that the hypothesis in Theorem 2.6
that there exists an object belonging to at least θ objects in S is essential for the result to
follow and that not every centralizable system is a centralized system. Consequently,
the concepts of centralized systems and centralizable systems are different.
Let ω1 be the first uncountable ordinal number, and M a collection of n-element
of ω1 , for some natural number n > 0, such that |M |  ω1 and for any x, y
subsets

∈ M, x y  ∅ if x  y.
We now consider a system S with object set M ∗ such that for any x∗ ∈ M ∗ , x∗ is
a system with some element in M as its object set, and for any x ∈ M there exists
exactly one x∗ ∈ M ∗ such that x is the object set of the system x∗ . It is clear that S
has no partial system which forms a centralized system and that S is a centralizable
system.
Theorem 3.3. Let κ and θ be cardinalities satisfying the conditions in Theorem 2.6.
Assume that S is a system with an object set of cardinality ≥θ and each object in S
44 3 Mappings and Constructions of Systems

is a system with an object set of cardinality <κ. There then exists a partial system S 
of S such that the object set of S  is of cardinality ≥θ and S  forms a centralizable
system.

The proof follows from Theorems 2.6 and 3.2.

A mapping h from a system S1  (M1 , R1 ) into a system S2  (M2 , R2 ) is
S-continuous if for any relation r ∈ R2 ,
 
h−1 (r)  h−1 (x): x ∈ r ∈ R1 . (3.8)

Proposition 3.2. Suppose that h: S1 → S2 is a mapping. Then the following hold

true:

(i) If the systems Si , i  1, 2, are topological spaces, h is S-continuous, if and only

if it is continuous from S1 into S2 .
(ii) If the systems Si , i  1, 2, are groups, the S-continuity of h implies that h is a
homomorphism from S1 into S2 .
(iii) If the systems Si , i  1, 2, are rings, the S-continuity of h implies that h is a
homomorphism from S1 into S2 .

Proof: Conclusion (i) follows from the definition of S-continuity of mappings and
Example 2.1.
For (ii), let (X, +) be a group. A system (X , R) is defined as follows: R  {r+ }
and r+  (x, y, z) ∈ X 3 : x + y  z . The system (X, R) will be identified with the
group (X, +),and the relation
 r+ with +.
Let Si  Mi , r+i , i  1, 2, be two groups and h: S1 → S2 an S-continuous
mapping. Then h−1 r+2  r+1 . Thus, for any x, y, and z ∈ M1 , if x +1 y  z, then
h(x) +2 h(y)  h(z). This implies that h is a homomorphism from S1 into S2 .
The proof of (iii) is a modification of that of (ii) and is omitted. QED

Theorem 3.4. A bijection h: S1 → S2 is a similarity mapping if and only if h:

S1 → S2 and h−1 : S2 → S1 are S-continuous.

The proof is straightforward and is left to the reader.

Theorem 3.5. Suppose that f : S1 → S2 and g: S2 → S3 are S-continuous mappings.

Then the composition g ◦ f : S1 → S3 is also S-continuous.

Proof: For each relation r in the system S3 , from the hypothesis that the mapping g
is S-continuous we have

g −1 (r) ∈ R2 . (3.9)

From the hypothesis that the mapping ƒ is S-continuous, it follows that

 
f −1 g −1 (r) ∈ R1 . (3.10)
3.1 Mappings from Systems into Systems 45

Therefore, from equation (3.10), we have

 
(g ◦ f )−1 (r)  f −1 ◦ g −1 (r)  f −1 g −1 (r) ∈ R1 .

Thus, g ◦ f is S-continuous. QED

3.2 Free Sums

Let {Si  (Mi , Ri ): i ∈ I } be a set of systems such that Mi ∩ Mj  ∅ for any distinct i
and j ∈ I . The free sum of the systems Si , denoted ⊕{Si : i ∈ I } or S1 ⊕ S2 ⊕ . . . ⊕ Sn ,
if I {1, 2, …, n} is finite, is defined to be the system

⊕{Si : i ∈ I }  (∪{Mi : i ∈ I }, ∪{Ri : i ∈ I }). (3.11)

The concept of free sums of systems can be defined for any set {Si : i ∈ I } of
systems such that the collection of all object sets of the systems in the set consists of
not necessarily pairwise disjoint sets. To show this, we simply take a set {∗ Si : i ∈ I }
of systems with pairwise disjoint object sets such that ∗ Si is similar to Si , for each i
∈ I, and define
 
⊕{Si : i ∈ I }  ⊕ ∗ Si : i ∈ I . (3.12)

Up to a similarity, the free sum of the systems Si is unique. Therefore, it can

be assumed that any set of systems has a free sum (uniquely determined up to a
similarity), but in the proofs of some theorems in the future it will be tacitly assumed
that the set under consideration consists of systems with pairwise disjoint object sets.
A system S  (M , R) is an input–output system if two sets X and Y are given
such that X ∪ Y  M and for each relation r ∈ R there are nonzero ordinals n  n(r)
and m  m(r) such that r ⊆ X n × Y m . The sets X and Y are called the input space
and the output space of the system S, respectively.

Proposition 3.3. Let S1 and S2 be two similar systems. If S1 is a free sum of some
systems, then so is S2 .
 i 
 i i Suppose that S1  ⊕ S : i ∈ I , where {S : i ∈ I} is a set of systems S 
i i
Proof:
M , R , and h: S1 → S2 is a similarity mapping. Then we have
   
S2  ⊕ h S i : i ∈ I , (3.13)
 
because h Ri ⊆ R2 for each i ∈ I, and for each r ∈ R2 there exists a relation s ∈ Ri ,
for some i ∈ I, such that h(s)  r. QED
46 3 Mappings and Constructions of Systems

Theorem 3.6. Suppose that {Si  (Mi , Ri ): i ∈ I } is a set of systems, where I is an

index set. Then the free sum ⊕{Si : i ∈ I } is an input–output system if and only if each
system Si  (Mi , Ri ) is an input–output system, i ∈ I.

Proof: Necessity. Suppose that ⊕{Si : i ∈ I } is an input–output system with input

space X and output space Y . Then for any fixed i ∈ I and any relation r ∈ Ri , there
are nonzero ordinals n  n(r) and m  m(r) such that r ⊆ X n × Y m . Hence, Si is an
input–output system with input space Mi ∩ X and output space Mi ∩ Y .
Sufficiency. Suppose that Si  (Mi , Ri ) is an input–output system with input space
Xi and output space Yi . Then for each relation r in ⊕{Si : i ∈ I }, there exists an i ∈ I
such that r ∈ Ri ; therefore, we have

r ⊆ Xin(r) × Yim(r) ⊆ (∪i∈I Xi )n(r) × (∪i∈I Yi )m(r) . (3.14)

Hence, the free sum ⊕{S i : i ∈ I } is an input–output system with input space
X
i∈I i and the output space i∈I Yi . QED

Theorem 3.7. If a system S is embeddable in the free sum ⊕{Si : i ∈ I } of some

systems Si , then S is also a free sum of some systems.

The proof is straightforward and is omitted.

Theorem 3.8. If the free sum ⊕{Si : i ∈ I } of nontrivial systems Si is a centralized

system, then each Si is centralized, for i ∈ I.

Proof: The theorem follows from the fact that each center of the free sum ⊕{Si : i ∈ I }
is a center of each system Si , i ∈ I. QED

Theorem 3.9. Suppose that {Si : i ∈ I } is a set of centralizable

 systems
 such that
each system Si has a center Ci  MCi , RCi , i ∈ I , and that MCi   MCj , for all i
and j ∈ I. Then the free sum ⊕{Si : i ∈ I } is a centralizable system.
 
Proof: Let MCi   ℵ, for some i ∈ I. Then the system (ℵ, ∅) is embeddable in
each object in the free sum ⊕{Si : i ∈ I }. Theorem 3.2 implies that the free sum is
centralizable. QED

It is easy to construct an example to show that the free sum of two centralized
systems may not be a centralizable system.

Theorem 3.10. Let h: S → ⊕{Si : i ∈ I } be a homomorphism. Then the system S is

also a free sum of some systems.

Proof: Suppose that S  (M , R) and Si  (Mi , Ri ) for each i ∈ I. Then

 
M  ∪ h−1 (Mi ): i ∈ I . (3.15)
3.2 Free Sums 47

Define
∗
Ri  {r ∈ R: h(r) ∈ Ri }. (3.16)

From the hypothesis that h is a homomorphism it follows that

 
∗
R Ri : i ∈ I . (3.17)
  
Therefore, S  ⊕ h−1 (Mi ), ∗ Ri : i ∈ I . QED

Theorem 3.11. Let ƒ: ⊕{Si : i ∈ I } → S be an onto S-continuous mapping. Then

there are partial systems {∗ Si : i ∈ I } of S such that
   ∗ 
∗
S Mi : i ∈ I , Ri : i ∈ I , (3.18)

where ∗ Si  (∗ Mi ,∗ Ri ), for each i ∈ I, and ƒ: Si →∗ Si is S-continuous.

Proof: For each i ∈ I, we define a system ∗ Si  (∗ Mi ,∗ Ri ) as follows:

∗
Mi  f (Mi ) (3.19)

and
∗
 
Ri  r ∈ R: f −1 (r) ∈ Ri . (3.20)

From the hypothesis that ƒ is surjective, it follows that M  ∪{∗ Mi : i ∈ I }. Since ƒ

is S-continuous, R  ∪{∗ Ri : i ∈ I }. Then equation (3.18) follows and the S-continuity
of the mapping ƒ: Si →∗ Si follows from equation (3.20). QED

Question 3.1. Under what conditions will the system S in Theorem 3.11 be a free
sum of some systems?

Theorem 3.12. Suppose that h: S1 → S2 is a homomorphism from a system S1 into

an input–output system S2 . Then S1 is also an input–output system.

Proof: Suppose that S2 has input space X and output space Y . Define

X ∗  h−1 (X ) and Y ∗  h−1 (Y ). (3.21)

Then S1 is an input–output system with input space X ∗ and output space Y ∗ . In

fact, let r be a relation in S1 . Then h(r) ∈ R2 . Hence, there are nonzero ordinals n
and m such that h(r) ⊆ X n × Y m . This implies that r ⊆ (X ∗ )n × (Y ∗ )m . QED

Theorem 3.13. Let ƒ: S1 → S2 be an S-continuous mapping from an input–output

system S1 into a system S1 . Then S2 is also an input–output system.
48 3 Mappings and Constructions of Systems

Proof: Suppose that S1 has input space X1 and output space Y1 . Define

X2  f (X1 ) and Y2  f (Y1 ). (3.22)

Then S2 is an input–output system with input space X2 and output space Y2 . In

fact, for each relation r ∈ R2 , f −1 (r) ∈ R1 . Hence, there are nonzero ordinals n and
m such that f −1 (r) ⊆ X1n × Y1m . Therefore, r ⊆ X2n × Y2m . QED

3.3 Products and Cartesian Products

Let S1 and S2 be two systems. We define

Hom(S1 , S2 )  {h: S1 → S2 : h is a homomorphism}. (3.23)

Suppose that {Si : i ∈ I } is a set of systems. A system S  (M , R) is a product of

the systems Si if there exists a family of homomorphisms
  {i ∈ Hom(S,
 Si ): i ∈ I }
such that for each system S  and a family i ∈ Hom S  , Si : i ∈ I , there exists a
unique homomorphism λ ∈ Hom(S  , S) so that

i  i ◦ λ, (3.24)

for all i ∈ I.

Theorem 3.14. For each set {Si : i ∈ I } of systems, a product

 of the systems
 Si always
exists, which is unique up to a similarity and denoted by i∈I Si or {Si : i ∈ I }.

Proof: For the set of systems {Si : i ∈ I }, we define a system S  (M , R) as follows:


M  Mi , (3.25)
i∈I

   
R  r̂  (ri )i∈I ∈ Ri : n(ri )  n rj , ∀i, j ∈ I , (3.26)
i∈I

and
  

R  r: r̂ ∈ R , (3.27)

 
where for any r̂  (ri )i∈I ∈ R̂, let n r̂  n(ri ), for i ∈ I, and

    
r  x ∈ M n(r̂ ) : ∃(xi ) ∈
i∈I ri ∀α < n r̂ x(α)  (xi (α)) i∈I . (3.28)
i∈I
3.3 Products and Cartesian Products 49


Let pi : i∈I Mi → Mi be the projection  for each i ∈ I. Then pi ∈ Hom(S,
 S i ). 
We now suppose that S   M  , R is a system and i ∈ Hom S  , Si : i ∈ I .
Define
  a mapping
 λ: S  → S as follows: For any object m ∈ M  , λ(m ) 
i m : i ∈ I ∈ M . It then can be shown that λ ∈Hom(S  , S).    
If λ is another such homomorphism, then i m  pi ◦ λ m  mi if λ m 
(mi )i∈I ∈ M . Thus, λ  λ as mappings from the object set M  into the object set M.
Therefore, they induce the same homomorphism. Therefore, S is a product of the
systems Si , i ∈ I. QED
Let {Si : i ∈ I } be a set of systems. A system S  (M , R) is a Cartesian product of
the systems Si if there exists a family of S-continuous
 mappings
 {pi : S → Si : i ∈ I }
such that for each system S  and a family qi : S  → Si : i ∈ I of S-continuous map-
pings, there exists a unique S-continuous mapping λ: S  → S so that qi  pi ◦ λ, for
each i ∈ I.

Theorem 3.15. For each set {Si : i ∈ I } of systems, a Cartesian product

 of the systems
Si always exists, which is unique up to a similarity and denoted by CP {Si : i ∈ I }.

Proof: Let Si  (Mi , Ri ) for each i ∈ I. Define a system S  (M , R) as follows:


M  {Mi : i ∈ I }, (3.29)

and
   
R  r: ∃j ∈ I ∃rj ∈ Rj r  (xα ) ∈ M n(rj ) : (xα (j))α<n(rj ) ∈ rj , (3.30)
    
and define mappings pi : S → Si by letting pi mj j∈I  mi for any mj j∈I ∈ M ,
for each i ∈ I. Then it can be readily checked that each mapping pi : S → Si is
S-continuous.   
Let S   M  , R be another system, and qi : S  → Si : i ∈ I a family of S-
continuous mappings. Define a mapping λ: S  → S by
    
λ m  qi m i∈I , (3.31)

for each object m ∈ M  . Then λ is S-continuous. In fact, for an arbitrary relation r

∈ R, let rj ∈ Rj be a relation such that
    
r  (mik )i∈I k<n(rj ) ∈ M n(rj ) : mjk k<n(rj ) ∈ rj .

Then we have
 
λ−1 (r)  λ−1 (m): m ∈ r
      
 λ−1 (mik )i∈I k<n(rj ) : n  (mik )i∈I k<n(rj ) ∈ r
         
 λ−1 (mik )i∈I : k < n rj : (mik )i∈I k<n(rj ) ∈ r
50 3 Mappings and Constructions of Systems

Figure 3.1 Uniqueness of

the Cartesian product

        
 qi−1 (mik ): i ∈ I : k < n rj : (mik )i∈I k<n(rj ) ∈ r
         
 qj−1 mjk : k < n rj : (mik )i∈I k<n(rj ) ∈ r
     
 qj−1 mj : mj  mjk k<n(rj ) ∈ r
 
 qj−1 rj
∈ R .

To complete the proof, we show that all Cartesian products of the systems Si are
similar to each other.
Let S ∗  (M ∗ , R∗ ) be another Cartesian product of Si , i ∈ I. Then there are S-
continuous mappings λ1 : S → S ∗ and λ2 : S ∗ → S such that the diagram in Figure 3.1
commutes for each i ∈ I. That is, for each i ∈ I, we have

qi  pi ◦ λ2 and pi  q ◦ λ1 . (3.32)

It suffices to show that

λ2 ◦ λ1 (m)  m (3.33)

and
 
λ1 ◦ λ2 m∗  m∗ (3.34)

for each m ∈ M and m∗ ∈ M ∗ . In fact, equation (3.33) implies that the mapping λ1 is
1–1, and equation (3.34) implies that λ2 is 1–1; λ1  λ−1 −1
2 and λ2  λ1 . Therefore,
λ1 and λ2 are similarity mappings.
Now, we prove equations (3.33) and (3.34). For equation (3.33), let m  (mi )i∈I ∈
M , then

pi (λ2 (λ1 (m)))  pi ◦ λ2 (λ1 (m))

 qi ◦ λ1 (m)
 pi (m)
 mi .
3.3 Products and Cartesian Products 51

Therefore, λ2 ◦ λ1 (m)  λ2 (λ1 (m))  (mi )i∈I . A similar argument shows equa-
tion (3.34). QED

The reason why the system S in Theorem 3.15 is called a Cartesian product of the
systems Si is that when Si are topological spaces, S is the Cartesian product of the
spaces Si . The system {S i : i ∈ I} in Theorem 3.14 is the box product of the spaces
Si . When the systems Si are groups, {S i : i ∈ I} in Theorem 3.14 is the direct sum
of the groups Si . The mappings pi in the proofs of Theorems 3.14 and 3.15 are called
projections from the product systems into the factor systems.

3.4 One-Level Products

Let {Si : i ∈ I } be a set of systems such that each object in any one system
 Si in the
set is a system again. The one-level product system, denoted by 1
{S i : i ∈ I }, is
defined by 1 {Si : i ∈ I }  (M , R), where the  object set M and the relation set R are
defined as follows: There exists a bijection h: {Mi : i ∈ I } → M such that
  
h (mi )i∈I  {mi : i ∈ I } (3.35)

and
  
R  h(r): r is a relation in {Si : i ∈ I } . (3.36)


The one-level Cartesian product, denoted by 1 CP {Si : i ∈ I }, is the system (M,
R) defined by the following: There exists a bijection h: {Mi : i ∈ I } → M such
that
  
h (mi )i∈I  {mi : i ∈ I } (3.37)
CP

and


R  h(r): r is a relation in {Si : i ∈ I } . (3.38)
CP

Theorem 3.16. Let {Si : i ∈ I } be a set of centralizable systems. Then the one-
level
 product system 1
{Si : i ∈ I } and the one-level Cartesian product system
1
CP {Si : i ∈ I } are also centralizable systems.
 
Proof: For each i ∈ I suppose  that Ci  MCi , RCi is a center of the system Si . For
the one-level product system
1
{Si : i ∈ I }, it suffices by Theorem
 3.2 to show that
the product  system {C i : i ∈ I } is embeddable in each product {mi : i ∈ I } for all
{mi }i∈I ∈ {Mi : i ∈ I }.
52 3 Mappings and Constructions of Systems

From Theorem 3.2, we will not lose generality by assuming that each system
Ci is embeddable
 in
 each object in M i for i ∈ I. For each {m }
i i∈I ∈ {Mi : i ∈ I },
let mi  Mmi , Rmi and hi : Ci → mi be an embedding mapping. Then the product
mapping
  
hi : {Ci : i ∈ I } → {mi : i ∈ I }, (3.39)
i∈I

defined by
 
  
hi (xi )i∈I  (hi (xi ))i∈I , (3.40)
i∈I


for each object (xi )i∈I in {Ci : i ∈ I } is an embedding mapping. In fact, first, the
product
 mapping h
i∈I i is 1–1, because if (xi )i∈I and (yi )i∈I are two distinct objects
in {Ci : i ∈ I }, there exists an i ∈ I such that xi  yi . Therefore, since hi is an
embedding mapping it follows that hi (xi )  hi (yi ). Therefore,
   
     
hi (xi )i∈I  (hi (xi ))i∈I  (hi (yi ))i∈I  hi (yi )i∈I . (3.41)
i∈I i∈I

 
Second, for each relation r in {Ci : i ∈ I }, there exists a relation s in {mi : i ∈ I }
such that
 
    

hi (r)  s hi MCi : i ∈ I . (3.42)
i∈I


In fact, for each relation r in {Ci : i ∈ I }, there exist relations ri ∈ RCi , for each
i ∈ I, such that n(r)  n(ri ) for all i ∈ I and

 n(r)    
r x∈ MCi : i ∈ I : ∃(xi )i∈I ∈ ri ∀α < n(r) x(α)  (xi (α))i∈I
i∈I
(3.43)

From the hypothesis that each hi is an embedding mapping, it follows that there
exists a relation si ∈ Rmi , for each i ∈ I, such that
  
hi (ri )  si hi MCi . (3.44)

 Equation (3.44) implies that n(si )  n(r) for each i ∈ I. Define a relation s in
{mi : i ∈ I } as follows:
3.4 One-Level Products 53

 n(r)    
s x∈ Mmi : i ∈ I : ∃(xi )i∈I ∈ si ∀α < n(r) x(α)  (xi (α))i∈I
i∈I
(3.45)

We now show that s satisfies equation (3.42). In fact,

    
 
hi (r)  hi (x): x ∈ r
i∈I i∈I
⎧   ⎫
⎨  ⎬
 hi (xα ) : x  (xα )α<n(r) ∈ r
⎩ ⎭
i∈I α<n(r)
        
 hi xαi i∈I α<n(r) : x  (xα )α<n(r) ∈ r&∀α < n(r) xα  xαi i∈I ∈ M Ci : i ∈ I

       
 hi xαi i∈I α<n(r) : ∃(xi )i∈I ∈ ri ∀α < n(r) xαi  (xi (α))i∈I
i∈I

     
 (yα )α<n(r) : ∃(yi )i∈I ∈ si hi MCi ∀α < n(r) yα  (yi (α))i∈I , (3.46)
i∈I

  
where yα  hi xαi i∈I , for each α < n(r) and i ∈ I.
We complete the proof of the theorem by observing that
   

The end in equation (3.46)  s hi MCi : i ∈ I . (3.47)


CP {Si : i ∈ I}
1
The centralizability of the one-level Cartesian product system
can be proved in a similar way. QED

Theorem  3.17. A mapping h from a system S  (M , R) into a Cartesian product

system CP {Si : i ∈ I } is S-continuous if and only if the composition (pi ◦ h) is S-
continuous for each i ∈ I.

Proof: Necessity. Suppose that h: S → CP {Si : i ∈ I } is S -continuous. From The-
orem 3.5 and the proof of Theorem 3.15, it follows that each composition (pi ◦ h) is
S-continuous, for i ∈ I.
Sufficiency. Suppose that each composition (p i ◦ h) is S-continuous, i ∈ I. Let
Si  (Mi , Ri ) for each i ∈ I. For each relation r in CP {Si : i ∈ I }, there exist an i ∈ I
and a relation ri ∈ Ri satisfying that
  
r  (xα )α<n(ri ) : xα ∈ {Mi : i ∈ I }&(xα (j))α<n(ri ) ∈ ri

Therefore,

(pi ◦ h)−1 (ri )  h−1 (r). (3.48)

This implies that h−1 (r) ∈ R; i.e., h is S-continuous. QED

54 3 Mappings and Constructions of Systems

Theorem
 3.18. A mapping h from a system S  (M , R) into a product system
{Si : i ∈ I } is a homomorphism if and only if the composition (pi ◦ h) is a homo-
morphism for every i ∈ I.

Proof: Necessity. Suppose that  h: S → {Si : i ∈ I } is a homomorphism. From the
fact that each projection pi : {Si : i ∈ I } → Si is a homomorphism, it follows that
(pi ◦ h) is a homomorphism for every i ∈ I.
Sufficiency. Suppose that (pi ◦ h) is a homomorphism
 for every i ∈ I. Let r ∈ R.
We need to show that h(r) is a relation in {Si : i ∈ I }. For each i ∈ I we have
(pi ◦ h)(r) ∈ Ri , since (pi ◦ h) is a homomorphism. Therefore,

h(r)  pi−1 ((pi ◦ h)(r)) (3.49)

i∈I

and for each i ∈ I and s ∈ Ri ,

!  n(s) "
pi−1 (s)  x ∈ {Mi : i ∈ I } : pi (x) ∈ s
!  n(s) "
 (xα )α<n(s) ∈ {Mi : i ∈ I } : (xα (i))α<n(s) ∈ s . (3.50)

Equation (3.50) implies that for every si ∈ Ri with n(s)  n(S i ),

!  n(s) "
 
pi−1 (si )  (xα )α<n(s) ∈ {Mi : i ∈ I } : ∀i ∈ I (xα (i))α<n(s) ∈ si .
i∈I
(3.51)

So equation (3.49) implies that

!  n(r) "
 
h(r)  x ∈ {Mi : i ∈ I } : ∀i ∈ I (xα (i))α<n(s) ∈ (pi ◦ h)(r) . (3.52)


This implies that the relation h(r) belongs to the relation set of {Si : i ∈ I }. QED

Remarks on references: All the discussions above in this chapter are based on
the works of various scholars, such as Lin (1987, 1989a, b, 1999), Lin et al. (1997),
Lin and Ma (1993), Mesarovic (1964), Mesarovic and Takahara (1975, 1989), and
Mickens (1990).

References

Lin, Y.: A model of general systems. Math. Modell. Int. J. 9(2), 95–104 (1987)
Lin, Y.: A multi-relation approach of general systems and tests of applications. Synth. Int. J. Epistem.
Methodol. Philos. Sci. 79, 473–488 (1989a)
References 55

Lin, Y.: Multi-level systems. Int. J. Syst. Sci. 20, 1875–1889 (1989b)
Lin, Y.: General Systems Theory: A Mathematical Approach. Kluwer Academic and Plenum Pub-
lishers, New York (1999)
Lin, Y., Ma, Y.H.: System—a unified concept. Cybern. Syst. 24, 375–406 (1993)
Lin, Y., Hu, Q.P., Li, D.: Some unsolved problems in general systems theory I and II. Cybern. Syst.
28, 287–303, and 591–605 (1997)
Mesarovic, M.D.: Views on general systems theory. In: Mesarovic, M.D. (ed.) Proceedings of the
2nd Systems Symposium at Case Institute of Technology. Wiley, New York (1964)
Mesarovic, M.D., Takahara, Y.: General Systems Theory: Mathematical Foundations. Academic
Press, New York (1975)
Mesarovic, M.D., Takahara, Y.: Abstract Systems Theory. Springer, Berlin (1989)
Mickens, R.E.: Mathematics and Science. World Scientific, Singapore (1990)
Chapter 4
Connectedness

This chapter looks at how systems can be connected or disconnected and then
investigates the connectedness of systems under various conditions. In particular,
Section 4.1 introduces the concept of connected systems and then Section 4.2 stud-
ies under what conditions constructed systems are connected.

4.1 Connected Systems

Let S  (M , R) be a system and r ∈ R a relation. The support of r, denoted Supp(r),

is defined by

Supp(r)  {m ∈ M : ∃x ∈ r∃β < n(r)(x(β)  m)}. (4.1)

The system S is connected if it cannot be represented in the form S1 ⊕ S2 , where S1

and S2 are nontrivial subsystems of S.
Theorem 4.1. For every system S  (M , R), the following conditions are equiva-
lent:
(i) The system S is connected.
(ii) For any two objects x and y ∈ M , there exist a natural number n > 0 and n
relations ri ∈ R such that

x ∈ Supp(r1 ) and y ∈ Supp(rn ) (4.2)

and

Supp(ri ) ∩ Supp(ri+1 )  ∅, (4.3)

for each i  1, 2, …, n − 1
© Springer Nature Switzerland AG 2018 57
J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_4
58 4 Connectedness

Proof: (i) → (ii). We prove by contradiction. Suppose that the system S is connected
and there exist two objects x and y ∈ M such that there do not exist relations ri ∈ R,
i = 1, 2, …, n, for any natural number n ≥ 1, such that

x ∈ Supp(r1 ) and y ∈ Supp(rn ) (4.4)

and

Supp(ri ) ∩ Supp(ri+1 )  ∅, (4.5)

for each i = 1, 2, …, n − 1. From the hypothesis that S is connected, it follows that

there must be relations r1 , s1 ∈ R such that x ∈ Supp(r1 ) and y ∈ Supp(s1 ). Then our
hypothesis implies that

Supp(r1 ) ∩ Supp(s1 )  ∅. (4.6)

Let U0  Supp(r1 ) and V0  Supp(s1 ), and for each natural number n ∈ N let

Un  {Supp(r): r ∈ R & Supp(r) ∩ Un−1  ∅} (4.7)

and

Vn  {Supp(s): s ∈ R & Supp(s) ∩ Vn−1  ∅}. (4.8)

Then U0 ⊆ U1 ⊆ · · · ⊆ Un ⊆ · · · , V0 ⊆ V1 ⊆ · · · ⊆ Vn ⊆ · · · , and Un ∩ Vm  ∅,
for all natural numbers n, m ∈ N.
We now define two subsystems of Si  (Mi , Ri ) such that
∞

M1  Un , (4.9)
n0
M2  M − M1 , (4.10)
R1  {r ∈ R: Supp(r) ∩ M1  ∅}, (4.11)

and

R2  {r ∈ R: Supp(r) ⊆ M2 }, (4.12)

Then R1 ∪ R2  R and R1 ∩ R2  ∅. In fact, for each relation r ∈ R, if r ∈ / R1 ,

Supp(r) ∩ M1  ∅ and so Supp(r) ⊆ M2 ; thus, r ∈ R2 . Therefore, S  (M , R) 
(M1 ∪ M2 , R1 ∪ R2 )  S1 ⊕ S2 , a contradiction.
(ii) → (i). The proof is again by contradiction. Suppose condition (ii) holds
true and S is disconnected. Thus, there exist nontrivial subsystems S1 and S2 of
S such that S  S1 ⊕ S2 . Suppose that Si  (Mi , Ri ), i  1, 2. Pick an object
4.1 Connected Systems 59

mi ∈ Mi , i =1, 2. Then there are no relations rj ∈ R, j = 1, 2, …, n, for any fixed n ∈ N,

such that

m1 ∈ Supp(r1 ) and m2 ∈ Supp(rn ) (4.13)

and
   
Supp rj ∩ Supp rj+1  ∅, (4.14)

for each j = 1, 2, …, n − 1, a contradiction. QED

Let M ∗ ⊆ M , where S  (M , R) is a system. Define the star neighborhood of the
subset M ∗ , denoted Star (M ∗ ), as follows:
    
Star M ∗  Supp(r): r ∈ R Supp(r) ∩ M ∗  ∅ . (4.15)

By applying induction on n ∈ N, we can define

   
Star1 M ∗  Star M ∗ (4.16)

and
    
Starn+1 M ∗  Star Starn M ∗ . (4.17)

Corollary 4.1. For every system S  (M , R), the following conditions are equiva-
lent:

(i) The system S is connected.

(ii) For each object x ∈ M , Star∞ ({x})  ∪∞
n1 Star ({x})  M .
n

Proof: (i) → (ii) follows from Theorem 4.1 directly.

(ii) → (i). For any two objects x, y ∈ M , from the hypothesis that Star∞ ({p})  M ,
for every object p ∈ M , it follows that there exist n, m ∈ N and relations r1 , r2 , . . . , rn
and s1 , s2 , . . . sm ∈ R such that

p ∈ Star(r1 ) and p ∈ Star(s1 ), (4.18)

x ∈ Star(rn ) and y ∈ Star(sm ), (4.19)

Supp(ri ) ∩ Supp(ri+1 )  ∅, for i  1, 2, . . . , n − 1,

and

Supp(si ) ∩ Supp(si+1 )  ∅, for i  1, 2, . . . , m − 1,

60 4 Connectedness

Thus, it follows that Theorem 4.1(ii) holds true and from Theorem 4.1 that S is
connected. QED
An object m in a system S  (M , R) is isolated if there is not relation r ∈ R such
that the cardinality |Supp(r)| > 1 and m ∈ Supp(r).

Corollary 4.2. There is no isolated object in a connected system.

Theorem 4.2. Suppose that S  (M , R) is a connected system and D  (MD , RD ) 

(M1 , R1 ) ⊕ (M2 , R2 ) is a disconnected system. If there exists a mapping f : S → D with
f −1 (RD )  {f −1 (r): r ∈ RD } ⊇ R, then only one of the following holds:
(i)

f (M ) ⊆ M1 .
(ii)

f (M ) ⊆ M2 .

Proof: We prove by contradiction. Suppose there exists a mapping f : S → D such

that f (M ) ∩ M1  ∅  f (M ) ∩ M2 and f −1 (RD ) ⊇ R. Pick elements x ∈ f −1 (M1 )
and y ∈ f −1 (M2 ). From Theorem 4.1, it follows that there exist n ∈ N and n relations
ri ∈ R, i = 1, 2, …, n, such that

x ∈ Supp(r1 ) and y ∈ Supp(rn ) (4.20)

and

Supp(ri ) ∩ Supp(ri+1 )  ∅, (4.21)

for each i = 1, 2, …, n − 1. Pick an i with 1 ≤ i ≤ n such that

Supp(ri ) ∩ f −1 (M1 )  ∅  Supp(ri ) ∩ f −1 (M2 ). (4.22)

Since f −1 (RD ) ⊇ R, it follows that there exists a relation r ∈ RD such that f (ri ) 
r. Now there are two possibilities: (1) r ∈ R1 ; (2) r ∈ R2 . If (1) holds true, we
must have Supp(ri ) ∩ f −1 (M2 )  ∅, a contradiction. If (2) holds true, we will have
Supp(ri ) ∩ f −1 (M1 )  ∅, again a contradiction. The contradictions imply that every
mapping f : S → D with f −1 (RD ) ⊇ R must satisfy either f (M ) ⊆ M1 or f (M ) ⊆ M2 .
QED

Corollary 4.3. Let h: S1 → S2 be a homomorphism from a connected system S1 into

a system S2 . Then h(S1 )  (h(M1 ), h(R1 )) must be a connected partial system of S2 .

Theorem 4.3. Suppose that h: S1 → S2 is an S-continuous mapping from a system

S1 into a system S2 . If S2 is connected, then S1 is also.
4.1 Connected Systems 61

Proof: Suppose that Si  (Mi , Ri ), i  1, 2. Let us pick two arbitrary objects x,

y ∈ M1 . There are relations rj ∈ R2 , j  1, 2, …, n, for some fixed n ∈ N such that
h(x) ∈ Supp(r1 ), h(y) ∈ Supp(rn ), and Supp(ri ) ∩ Supp(ri+1 )  ∅, for each i  1, 2,
…, n − 1. From the hypothesis that h is S-continuous, it follows that h−1 (ri ) ∈ R1 ,
for i  1, 2, …, n. Thus, we have

   
x ∈ Supp h−1 (r1 ) and y ∈ Supp h−1 (rn ) , (4.23)

and
   
Supp h−1 (ri ) ∩ Supp h−1 (ri+1 )  ∅, (4.24)

for i = 1, 2, …, n − 1. From Theorem 4.1, it follows that S1 is connected. QED

Example 4.1. We construct an example to show that the connectedness of the system
S1 cannot guarantee the connectedness of the system S2 in Theorem 4.3.

Let S2  (Z, R2 ) be a system where Z is the set of all integers and R2 

{ri : i  ±1, ±2, ±3, . . .}, where for each i = ±1, ±2, ±3, …, ri  {(i, i + 1)}. Then
the system S2  (Z− , R− −
2 ) ⊕ (Z , R2 ) is disconnected, where Z  {0, −1, −2, . . .},
+ +
−
Z {1, 2, 3, . . .}, R2  {ri : i  −1, −2, . . .} and R2  {ri : i  1, 2, . . .}.
+ +

Define a system S1  (Z, R1 ) as follows: R1  {ri : i  0, ±1, ±2, . . .}, where

ri  {(i, i + 1)} for each i ∈ Z. Then S1 is connected. In fact, for any two objects i,
j ∈ Z, there exist relations ri , ri+1 , …, rj , where we assumed that i < j, such that i ∈
Supp (ri )  {i, i + 1}, j ∈ Supp (rj )  {j, j + 1}, and Supp (rk ) ∩ Supp (rk+1 )  {k + 1},
for k = i, i + 1, …, j − 1. Therefore, it follows from Theorem 4.1 that S1 is connected.
Let h: S1 → S2 be the identity mapping defined by h(i)  i for each i ∈ Z. Then h
is S-continuous, and S1 is connected, and S2 is disconnected.

4.2 Connectedness of Constructed Systems

As the title suggests, this section looks at the concept of connected systems and
various constructed systems.

Theorem 4.4. If a subsystem C of a system S is connected, and for every pair S1 ,

S2 of subsystems of S with disjoint object sets such that C is a subsystem of S1 ⊕ S2 ,
then C is a subsystem of either S1 or S2 .

Proof: We show the result by contradiction. Suppose that C is neither a subsystem

of S1 nor of S2 . Let C  (MC , RC ) and Si  (Mi , Ri ), i = 1, 2. Then

MC ∩ M1  ∅  MC ∩ M2 . (4.25)
62 4 Connectedness

Pick elements x ∈ MC ∩ M1 and y ∈ MC ∩ M2 . Since C is connected and from

Theorem 4.1, it follows that there exist a natural number n > 0 and n relations ri ∈ RC ,
for i = 1, 2, …, n, such that

x ∈ Supp(r1 ) and y ∈ Supp(rn ), (4.26)

and

Supp(ri ) ∩ Supp(ri+1 )  ∅, for each i  1, 2, . . . , n − 1. (4.27)

Therefore, there must be an i satisfying 0 < i ≤ n such that

Supp(ri ) ∩ M1  ∅  Supp(ri ) ∩ M2 . (4.28)

From the definition of subsystems, it follows that there exists a relation r in the free
sum S1 ⊕ S2 such that ri ⊆ r|MC . It follows from the definition of free sums that r is
contained either in R1 or in R2 , but not both. Therefore, if r ∈ R1 , Supp(ri ) ∩ M2  ∅;
if r ∈ R2 , Supp(ri ) ∩ M1  ∅, a contradiction. This implies that C is a subsystem of
either S1 or S2 . QED

Theorem 4.5. Let {Ci  (Mi , Ri ): i ∈ I } be a set of connected subsystems

of a system S   (M , R). If there exists an index i0 ∈ I such that each
system (Mi ∪ Mi0 , RMi ∪ Mi0 ) is connected, for each i ∈ I , then the system
(∪i∈I Mi , R|∪i∈I Mi ) is connected.

Proof: Let us pick two arbitrary objects x and y from ∪i∈I Mi . There are two indices i
and j such that x ∈ Mi andy ∈ Mj . Choose an object z ∈ Mi0 . From the hypothesis that
the systems (Mi ∪ Mi0 , RMi ∪ Mi0 ) are connected, it follows that there are natural
numbers n and m and relations rk and pq ∈ R, for k = 1, 2, …, n, and p = 1, 2, …, m,
such that

x ∈ Supp(r1 ) and y ∈ Supp(rn ), (4.29)

Supp(rk ) ∩ Supp(rk+1 )  ∅, for each k  1, 2, . . . , n − 1. (4.30)

and

y ∈ Supp(q1 ) and z ∈ Supp(qm ), (4.31)

   
Supp qp ∩ Supp qp+1  ∅, for each p  1, 2, . . . , m − 1. (4.32)

Therefore, Theorem 4.1 implies that the system (∪i∈I Mi , R|∪i∈I Mi ) is connected.
QED

Corollary 4.4. If the set {Ci  (Mi , Ri ): i ∈ I } of connected subsystems of the

system S  (M , R) satisfies ∩i∈I Mi  ∅, then the subsystem (∪i∈I Mi , R|∪i∈I Mi ) is
connected.
4.2 Connectedness of Constructed Systems 63

Let S  (M , R) be a system and D ⊂ M a subset. The subset D is dense in the

object set M if the following conditions hold:
(i) For any relation r ∈ R with r  Ø, Supp(r) ∩ D  Ø.
(ii) Each isolated object in M belongs to D.
A subsystem S ∗  (M ∗ , R∗ ) of S is a dense subsystem of S if the object set M ∗ is
dense in M.
Theorem 4.6. If a system S  (M , R) contains a connected dense subsystem, then
S is connected.
Proof: Suppose that S ∗  (M ∗ , R∗ ) is a connected dense subsystem of S. To show
that S is connected, we pick two arbitrary objects x, y ∈ M. There are three possi-
bilities: (1) x, y ∈ M − M ∗ ; (2) one of x and y is contained in M ∗ and the other is
contained in M − M ∗ ; (3) x, y are contained in M ∗ .
If possibility (1) holds true, x and y are then not isolated objects in M. Therefore,
there are relations rx and ry ∈ R such that
 
x ∈ Supp(rx ) and y ∈ Supp ry . (4.33)

It follows from the definition of dense subsets that there must be an object x∗ ∈
Supp(rx ) ∩ M ∗ and an object y∗ ∈ Supp(ry ) ∩ M ∗ . Theorem 4.1 implies that there
are natural number n and n relations ri ∈ R∗ , i = 1, 2, …, n, such that

x∗ ∈ Supp(r1 ) and y∗ ∈ Supp(rn ) (4.34)

and

Supp(ri ) ∩ Supp(ri+1 )  ∅, for each p  1, 2, . . . , n − 1. (4.35)

For each i = 1, 2, …, n, let qi be a relation in R such that ri ⊆ qi |M ∗ . Then it has to

be that
 
x ∈ Supp(rx ) and y ∈ Supp ry , (4.36)
 
Supp(rx ) ∩ Supp(q1 )  ∅  Supp ry ∩ Supp(qn ), (4.37)

and

Supp(qi ) ∩ Supp(ri+1 )  ∅, for each i  1, 2, . . . , n − 1. (4.38)

Theorem 4.1 implies that S is connected.

If possibility (2) holds true, without confusion, let x ∈ M ∗ and y ∈ M – M ∗ .
Then y is not an isolated object in M. Let ry ∈ R be a relation such that
 
y ∈ Supp ry . (4.39)
64 4 Connectedness

Pick y∗ ∈ Supp(ry ) ∩ M ∗ . From the hypothesis that S ∗ is connected, it follows

that there are a natural number n > 0 and n relations ri ∈ R∗ , i = 1, 2, …, n, such that

x ∈ Supp(r1 ) and y∗ ∈ Supp(rn ), (4.40)

and

Supp(ri ) ∩ Supp(ri+1 )  ∅, for each i  1, 2, . . . , n − 1. (4.41)

For each i = 1, 2, …, n, let qi be a relation in R such that ri ⊂ qi |M ∗ . Then we have

 
x ∈ Supp(q1 ) and y ∈ Supp ry , (4.42)
Supp(qi ) ∩ Supp(qi+1 )  ∅, for each i  1, 2, . . . n−1, (4.43)

and
 
Supp(qn ) ∩ Supp ry  ∅. (4.44)

Applying Theorem 4.1, we have showed that S is connected. QED

A connected subsystem S ∗ of S is maximal in S if there is not connected subsystem
S of S such that S ∗ is a subsystem of S + and S ∗  S + . Each maximal connected
+

subsystem of S is called a component. Then, for each object m ∈ M , the partial

system (Star∞ ({m}), R|Star∞ ({m}) ) of S is a component.

Theorem 4.7. Suppose that S  (M , R) is a system and E an equivalence relation

on the object set M. Then the quotient system S/E  (M /E, R/E) is connected if
and only if for any two components X  (MX , RX ) and Y  (MY , RY ) of S, there
exist a natural number n and n components Si  (Mi , Ri ) of S, i = 1, 2, …, n, such that
there exist x ∈ MX , y ∈ MY and mxi , myi ∈ Mi , i =1, 2, …, n, satisfying (x, mx1 ) ∈ E,
(mxi+1 , myi ) ∈ E, i = 1, 2, …, n − 1, and (myn , y) ∈ E.

Proof: Necessity. Suppose that the quotient system /E  (M /E, R/E) is connected,
and there exist two components X  (MX , RX ) and Y  (MY , RY ) of S such that
there do not exist components Si  (Mi , Ri ) of S, i = 1, 2, …, n, for each natural
number n, such that there exist x ∈ MX and y ∈ MY and mxi , myi ∈ Mi , for each i,
satisfying (x, mx1 ) ∈ E, (mxi+1 , myi ) ∈ E, i = 1, 2, …, n − 1, and (myn , y) ∈ E.

Pick objects x ∈ MX and y ∈ MY . We then make the following claim.

Claim. The subsystems
  
Star∞ ({μ(x)}), R/E Star∞ ({μ(x)})

and
  
Star∞ ({μ(y)}), R/E Star∞ ({μ(y)})
4.2 Connectedness of Constructed Systems 65

of the quotient system S/E are not identical, where μ: M → M/E is the canonical
mapping.
This claim implies that the quotient system S/E is not connected. In fact, from the
hypothesis that the quotient system S/E is connected, it follows that for each object
μ(m) ∈ M/E, the connected component of S/E containing the object μ(m) must equal
the whole system S/E.
We now see the proof of Claim 1. It suffices to show that Star∞ ({μ(x)}) 
Star∞ ({μ(y)}). To this end, it suffices to show that

Star∞ ({μ(x)}) ∩ Star∞ ({μ(y)})  ∅, (4.45)

which we do by contradiction. Suppose that

Star∞ ({μ(x)}) ∩ Star∞ ({μ(y)})  ∅, (4.46)

for some x ∈ MX and some y ∈ MY . Choose [z] ∈ Star∞ ({μ(x)}) ∩ Star∞ ({μ(y)}),
where z is an object in M. Then there exist components Sxi  (Mxi , Rxi ) and Syi 
(Myi , z), for i = 1, 2, …, n, and j = 1, 2, …, m, for some natural numbers n and m, such
that there are objects xi , mxi ∈ Mxi and yj , myj ∈Myj , for each i = 1, 2, …, n, and j = 1,
2, …, m, such that (x, x1 ), (xi+1 , mxi ) ∈ E, for i = 1, 2, …, n − 1, and (mxn , z), (mym , z),
(y, y1 ), (yj+1 , myj ) ∈ E, for j = 1, 2, …, m − 1. This contradicts the assumption in the
first paragraph of the necessary part of the proof.
Sufficiency. Suppose that for any two components X  (MX , RX ) and Y 
(MY , RY ) of S, there exist a natural number n and components Si  (Mi , Ri ) of S,
i =1, 2, …, n, such that there exist x ∈ MX , y ∈ MY , and mxi , myi ∈ Mi , for each i = 1,
2, …, n, satisfying (x, mx1 ), (myn , y), (mxi+1 , myi ) ∈ E, i =1, 2, …, n − 1. To show that
S/E is connected, it suffices to verify that, for each object x ∈ M ,

Star∞ ({μ(x)})  M /E, (4.47)

by Corollary 4.1.
It is clear that Star∞ ({μ(x)}) ⊂ M/E. Suppose M/E  Star∞ ({μ(x)}). We can then
pick an object [m] ∈ M/E – Star∞ ({μ(x)}), where m is an object in M. This implies that
there are two connected components X  (MX , RX ) and Y  (MY , RY ), such that
x ∈ MX and m ∈ MY and for any two objects mx ∈ MX and my ∈ MY there do not exist
connected components Si  (Mi , Ri ) of S, i =1, 2, …, n, for any n ∈ N, such that there
exist mxi , myi ∈ Mi , i =1, 2, …, n, satisfying (mx , mx1 ), (mxi+1 , myi ), (myn , my ) ∈ E,
i = 1, 2, …, n − 1, a contradiction, which implies that M /E  Star∞ ({μ(x)}). QED

Theorem 4.8. Let {Si : i ∈ I } be a set of systems such that there existsan i0 ∈ I such
that the system Si0 is connected. Then the Cartesian product system CP {Si : i ∈ I }
is connected.

Proof: Pick objects x  (xi )i∈I and y  (yi )i∈I from the Cartesian product system

CP {Si : i ∈ I }. Then pi0 (x)  xi0 and pi0 (y)  yi0 ∈ Mi0 . From Theorem 4.1, it
66 4 Connectedness

follows that there are a natural number n > 0 and n relations rj ∈ Ri0 , j = 1, 2, …, n,
such that

xi0 ∈ Supp(r1 ) and yi0 ∈ Supp(rn ), (4.48)

and

Supp(ri ) ∩ Supp(ri+1 )  ∅, for i  1, 2, . . . , n − 1. (4.49)


Therefore, the relations (pi0 )−1 (rj ) in CP {Si : i ∈ I }, for j = 1, 2, …, n, satisfy
 
 

−1 −1
x ∈ Supp pi0 (r1 ) and y ∈ Supp pi0 (rn ) , (4.50)

and
 
 

−1 −1
Supp pi0 (ri ) ∩ Supp pi0 (ri+1 )  ∅, (4.51)

for each i = 1, 2, …, n − 1. Applying Theorem 4.1, we have proved the theorem.

QED

Theorem 4.9. Let {Si : i ∈ I } be a set of systems such that the product {Si : i ∈ I }
is connected. Then each factor system Si is connected.

Proof: Pick two arbitrary objects xi and yi from a system Si for any fixed i ∈ I .
There then are two objects x and y in the product system such that pi (x)  xi and
pi (y)  yi . From the hypothesis that the product system is connected, it follows that
there are a natural number n > 0 and n relations rj from the product system, j = 1,2,
…, n, such that

x ∈ Supp(r1 ) and y ∈ Supp(rn ) (4.52)

and
   
Supp rj ∩ Supp rj+1  ∅, (4.53)

for each j = 1, 2, …, n − 1. Then the relations pi (rj ) ∈ Rj , j = 1, 2, …, n, satisfy

xi ∈ Supp(pi (r1 )) and yi ∈ Supp(pi (rn )) (4.54)

and
     
Supp pi rj ∩ Supp pi rj+1  ∅, (4.55)

for each j = 1, 2, …, n − 1. Theorem 4.1 implies that Si is connected. QED

4.2 Connectedness of Constructed Systems 67

Remarks on references: The discussions in this chapter are mainly based on Comac-
chio (1972), Lin (1990, 1999), Lin and Ma (1987a, b), Ma and Lin (1990), and Wu
(1990).

References

Comacchio, J.V.: Topological concepts in the mathematical theory of general systems. In: Klir, G.
(ed.) Trends in General Systems Theory. Wiley-Interscience, New York (1972)
Lin, Y.: Connectedness of general systems. Syst. Sci. 16, 5–17 (1990)
Lin, Y.: General Systems Theory: A Mathematical Approach. Kluwer Academic and Plenum Pub-
lishers, New York (1999)
Lin, Y., Ma, Y.H.: Remarks on analogy between systems. Int. J. General Syst 13, 135–141 (1987a)
Lin, Y., Ma, Y.H.: On the stabilities of input-output systems. Cybern. Syst. Int. J 18(4), 285–298
(1987b)
Ma, Y.H., Lin, Y.: Continuous mappings between general systems. Math. Comput. Model 14, 58–63
(1990)
Wu, X.-M.: The Pansystems View of the World. People’s University of China Press, Beijing (1990)
Chapter 5
Hierarchies—Construction
of Large-Scale Systems

This chapter investigates the existence and construction of very large systems, named
hierarchies of systems and their mappings properties. In particular, Section 5.1 looks
at when linkages between state systems of different time moments could exist.
Section 5.2 considers the mappings from hierarchies of systems into hierarchies
of systems and related properties. Section 5.3 focuses on the special case of time
systems.

5.1 Existence of Linkages

Let (T ,≤) be a partially ordered set with order type α. An α-type hierarchy S of
systems over the partially ordered set (T ,≤) is a function defined on T such that for
each t ∈ T , S(t)  St  (Mt , Rt ) is a system, called the state of the α-type hierarchy
S at the moment t. Without causing confusion, we omit the words “over the partially
ordered set (T ,≤).”
For an α-type hierarchy S of systems, let tr : Sr → St be a mapping from the
system Sr into the system St , for any r, t ∈ T , with r ≥ t such that

ts  tr ◦ rs and tt  idSt , (5.1)

where r, s, t are arbitrary elements in T satisfying s ≥ r ≥ t, and idSt  idMt is the

identity mapping on the set Mt . The family {ts : t, s ∈ T , s ≥ t} is a family of linkage
mappings of the α-type hierarchy S, and each mapping ts a linkage mapping from
Ss into St .
An α-type hierarchy of systems S, denoted {S, ts , T } or {S(t), ts , T }, is referred
to as a linked α-type hierarchy (of systems) if a family {ts ct, s ∈ T , s ≥ t} of linkage
mappings is given.

© Springer Nature Switzerland AG 2018 69

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_5
70 5 Hierarchies—Construction of Large-Scale Systems

Theorem 5.1 [ZFC]. Let S be a nontrivial α-type hierarchy of systems; i.e., each
state St is a nontrivial system. Then there exists a family {ts : t, s ≤ T , s ≥ t} of
linkage mappings of S.

Proof: Suppose St  (Mt , Rt ), for each t ∈ T. From the Axiom of Choice, it follows
that there exists a choice function C: T → ∪{Mt : t ∈ T } such that C(t) ∈ Mt for
each t ∈ T. A family {ts : t, s ∈ T , s ≥ t} of linkage mappings of S can now be
defined as follows. For any s, t ∈ T with s > t, let

ts (x)  C(t), (5.2)

for all x ∈ Ms and tt  idMt . QED

Question 5.1. How many families of linkage mappings are there for a given α-type
hierarchy of systems?
Suppose {S, ts , T } is a linked α-type
 hierarchy of systems , where St  (Mt , Rt )
for all t ∈ T. An element {xw }w∈T ∈ {Mt : t ∈ T } is called a thread of the linked
α-type hierarchy S if ts (xs )  xt for all s, t ∈ T with s ≥ t. Let  ⊂ {Mt : t ∈ T }
be the subset of all threads, which will be called the thread set of the hierarchy
{S, ts , T }.

Theorem 5.2 [ZFC]. Let κ be a cardinal number and S an α-type hierarchy of

systems such that |Mt | ≥ κ for each t ∈ T. Then for any nonzero cardinal number
κ
≤ κ, there exists a family {ts : t, s ∈ T , s ≥ t} of linkage mappings of S such that
|| ≥ κ
.

Proof: From the hypothesis that for any t ∈ T , |Mt | ≥ κ

, and the Axiom of Choice,
it follows that, for any t ∈ T , there exist At ⊂ Mt satisfying |At |  κ
, and an
equivalence relation Et on the object set Mt such that Mt /Et  {[a]Et : a ∈ At } and
that for any distinct a, b ∈ At ,

[a]Et  [b]Et . (5.3)

Again, from the Axiom of Choice it follows that there exists a set of well-order
relations {≤t : t ∈ T } such that, for each t ∈ T , the relation ≤t well-orders At as
follows:

a0t , a1t , . . . , aαt , , . . . , α < κ

(5.4)

For any s, t ∈ T with s > t, we now define a mapping ts : Ss → St . For any x ∈ Ms ,
there exists exactly one α < κ such that x ∈ [aαs ]Es . Then define

ts (x)  aαt . (5.5)

Then {ts : t, s ∈ T , s ≥ t} is a family of linkage mappings, where tt  idSt , and

|Σ| ≥ |At |, for any t ∈ T. Thus || ≥ κ
. QED
5.1 Existence of Linkages 71

Question 5.2. Suppose that S is an α-type hierarchy of systems over a partially

ordered set (T ,≤) and {ts : s, t ∈ T ∗ , s ≥ t} is a family of linkage mappings of the
hierarchy of systems {St : t ∈ T ∗ }, where T ∗ is a proper subset of T. Under what
conditions does there exist a family {¯ts : t, s ∈ T , s ≥ t} of linkage mappings for S
such that

¯ts  ts , (5.6)

for all s, t ∈ T ∗ with s ≥ t?

Lemma 5.1 [ZFC]. Let D ⊂ T be a cofinal and coinitial subset in a partially ordered
set (T ,≤). Then there exists a subset T ∗  D−
∗
∪ D+∗ of D such that
 
(a) D+∗  ∪ Dα+ : α < η and D− ∗
 ∪{Dα− : α < ξ } for some ordinals η and ξ such
that

D0+  D0− . (5.7)

(b) Let Eβ+  {x ∈ T : ∃d ∈ ∪{Dα+ : α < β}(x ≤ d )}, then Dβ+ is a maximal antichain
in D − Eβ+ , for any β < η.
(c) Let Eβ−  {x ∈ T : ∃d ∈ ∪{Dα− : α < β}(x ≤ d )}. Then Dβ− is a maximal
antichain in D − Eβ− , for each β < ξ .
(d) T ∗ is cofinal and coinitial in T.

Proof: First, note the construction of D+∗ . From the Axiom of Choice, it follows that
there exists an order relation ≤∗ on T which well-orders T as

t0 , t1 , . . . tα , . . . , α < |T |, (5.8)

where ≤ and ≤∗ are generally different. Let d0 be the ≤∗ -initial element of D, and
D01 ⊂ D such that
 
D01  d ∈ D: d0 ≤∗ d & d0 and d are incomparable under ≤ . (5.9)

If D01  ∅, let D01  {d0 }; otherwise, let d1 be the initial element of (D01 , ≤∗ ).

Suppose that sequences {dα : α < η} and {D0α : 0  α < η} have been defined, for
some ordinal η > 1, such that
(i) {dα : α < η} is an antichain under the order <.
β
(ii) ∅  D0α ⊂ D0 for all α, β < η with α > β.
(iii) For each α < η, dα is the initial element of the well-ordered set (D0α , ≤∗ ).
β
(iv) For each α < η, dα <∗ d , for all d ∈ D0 , and dα is incomparable with d under
β
the relation < , for each d ∈ D0 , for each β with α < β < η.
If η is a limit ordinal, let
72 5 Hierarchies—Construction of Large-Scale Systems

η
 
D0  D0α : α < η . (5.10)

η
If D0  ∅, let D0+  {dα : α < η}. Then D0+ is a maximal antichain in D under ≤.
η η
If D0  ∅, let dη be the initial element of (D0 , <∗ ). By (iv) {dα : α < η + 1} is an
η
antichain under the relation ≤; and for each α < η, dα <∗ d for each d ∈ D0 , dα and
η
d are incomparable under ≤ for each d ∈ D0 .
η ξ
If η  ξ + 1 is isolated, let D0 ⊂ D0 be defined by
 
η ξ η η
D0  d ∈ D0 : d0 <∗ d , d0 and d are incomparable under ≤ (5.11)

η
If D0  ∅, let D0+  {dα : α < η}. Then D0+ is a maximal antichain in D under the
η η
order ≤. If D0  ∅, let dη be the initial element of (D0 , ≤∗ ). Then {dα : α < η + 1}
η β
is an antichain in D under ≤, and D0 ⊂ D0 for all β < η. For each α < η + 1, dα <∗ d ,
β β
for all d ∈ D0 , and dα is incomparable with d ∈ D0 under ≤ for any β, with α < β < η
+ 1.
By transfinite induction, sequences {dα : α < η} and {D0α : α < η} with properties
η
(i)–(iv) can be defined, such that if a subset D0 of ∩{D0α : α < η} with property (iv)
η
can be defined in the way discussed earlier, it must be that D0  ∅.
Let D0  {dα : α < η}. Then D0 is a maximal antichain in (D, ≤). Using transfinite
+ +

induction to consider sets U  {x ∈ T : ∃d ∈ D0+ (x > d )} and V  {x ∈ T : ∃d ∈

D0+ (x < d )} and applying the same technique we used before, we can show that
ordinals η and ξ and subsets Dα+ ⊂ U and Dβ+ ⊂ V , for 0  α < η and 0  β < ξ ,
exist such that properties (a)–(d) in the lemma hold. QED

Theorem 5.3 [ZFC]. Suppose that S is an α-type hierarchy of systems over a par-
tially ordered set (T ,≤) such that there exists a cofinal and coinitial subset D ⊂ T
such that {ts : t, s ∈ D, S ≥ t} is a family of linkage mappings for the hierarchy of
systems {St : t ∈ D} satisfying

ts (Ms )  tr (Mr ), (5.12)

for any s, r, t ∈ D with s ≥ r > t. Then there exists a subset T ∗ ⊂ D such that
(i) T ∗ is cofinal and coinitial in T.
(ii) There is a family {¯ts : t, s ∈ T , s ≥ t} of linkage mappings for S.
(iii) For any s, t ∈ T * with s ≥ t, ¯ts  ts .

Proof: Let T ∗ ⊂ D be the subset defined in the proof of Lemma 5.1. Let us rewrite
T  TL ∪ TR , where
 
TL  x ∈ T : ∃d ∈ D0− (x ≤ d ) (5.13)

and
 
TR  x ∈ T : ∃d ∈ D0+ (x ≤ d ) , (5.14)
5.1 Existence of Linkages 73

and pick one t ∈ T. There are two possibilities: (1) t ∈ TL and (2) t ∈ TR .
Case 1: t ∈ TL . There then exists an ordinal α < ξ such that there exists eα ∈ D0−
with eα ≤ t, and assume that α is the initial one in ξ with this property. If eα  t, let
¯tt  tt . Now we assume that eα < t.
If α = δ + 1 is isolated, there exists eδ ∈ Dδ− such that eα < t < eδ . From the
Axiom of Choice, it follows that for any  s ∈ T satisfying eα < t < eδ , there exist an
As ⊂ Ms such that |As |  eα eδ (Meδ ) ≤ κ; an equivalence relation Es on Ms such
that Ms /Es  {[a]Es : a ∈ As } and satisfies, for any distinct a, b ∈ As ,

[a]Es  [b]Es (5.15)

and a set {<s : s ∈ T , eα < s < eδ } of well-order relations such that, for any s ∈ T
satisfying eα < s < eδ , the relation <s well-orders Ms /Es as follows:

as,0 Es , as,1 Es , . . . , as,i Es , , . . . , i < |As |, (5.16)

where as,i ∈ As , for all i < |As |, and the relation <eα well-orders eα eδ (Meδ ) as
follows:
  
m0 , m1 , . . . , mα , . . . , α < eα eδ Meδ . (5.17)

We now define mappings ¯st : Mt → Ms for all s, t ∈ T , with eα ≤ s < t < eδ

as follows: For any x ∈ Mt , there exists exactly one ati ∈ At such that x ∈ [ati ]Et . If
s > eα , let

¯st (x)  asi . (5.18)

If s  eα , let

¯st (x)  mi . (5.19)

Define mappings ¯seδ : Meδ → Ms for all s ∈ T with eα < s < eδ as follows:

¯seδ (x)  asi . if ¯eα eδ (x)  mi . (5.20)

If α is a limit ordinal, then by definition α > 0. Let

Tα  {s ∈ T : eα < s < eδ , ∀δ < α}. (5.21)

From the Axiom of Choice, it follows that for any s ∈ Tα , there exists an As ⊂ Ms
satisfying |As |  eα q (Mq ), where q ∈ D0− satisfies t < q. An equivalence relation Es
on Ms exists such that Ms /Es  {[a]Es : a ∈ As } satisfies, for any distinct a, b ∈ As ,
[a]Es  [b]Es . Let {<s : s ∈ {eα } ∪ Tα } be a set of order relations such that, for any
s ∈ Tα , the relation <s well-orders Ms /Es as follows:
74 5 Hierarchies—Construction of Large-Scale Systems

[as0 ]Es , . . . , [asi ]Es , . . . , i < |As |, (5.22)

where asi ∈ As , for all i, and the relation <eα well-orders eα q (Mq ) as follows:
  
m0 , m1 , . . . , mi , . . . , i < eα q Mq . (5.23)

We now define mappings ¯ij : Mj → Mi , for all i, j ∈ {eα } ∪ Tα , as follows: For any
x ∈ Mj , there exists exactly one ajk ∈ Aj such that x ∈ [ajk ]Ej . If j > eα , let

¯ij (x)  aik . (5.24)

If j  eα , let

¯ij (x)  mk . (5.25)

Define mappings ¯seδ : Meδ → Ms for all s ∈ Tα and all δ < α by

¯seδ (x)  ask , if ¯eα eδ (x)  mk . (5.26)

Then it is easy to check that {¯ts : s, t ∈ TL , t ≥ s} is a family of linkage mappings,

where ¯ss  idMs . Furthermore, for any s, t ∈ TL with s > t, if ¯ts is undefined
above, then there exist α, β < ξ such that there exist dα ∈ Dα− and dβ ∈ Dβ− with
dβ < t < dα < s. Let α and β be the least ordinals with this property. (1) If β = δ +
1, then there exists exactly one dδ ∈ Dδ− such that dβ < t ≤ dδ . Define

¯ts  ¯sdδ ◦ ¯dδ dα ◦ ¯dα s . (5.27)

(2) If β is a limit ordinal, define

¯ts  ¯sdα ◦ ¯dα s . (5.28)

Case 2: t ∈ TR . A similar method can be used to define a family {¯ts : s, t ∈ TR , t ≥ s}

of linkage mappings such that ¯st  st for any s, t ∈ D+∗ with s ≤ t.
Now let s ∈ TR and t ∈ TL satisfying s > t. There exists exactly one q ∈ D0+ such
that t ≤ q ≤ s. Define ts  tq ◦ qs . Then {¯ts : s, t ∈ TR , t ≥ s} is a family of linkage
mappings for S such that ¯ts  ts for all s, t ∈ T ∗ with s ≥ t. QED

5.2 Mappings from Hierarchies into Hierarchies

Let H  {S, ts , T } and H

 {S
, t
s
, T
} be two hierarchies of systems. A mapping
from the hierarchy H into the hierarchy H
is a set {ϕ, ft
} of mappings, where ϕ is a
5.2 Mappings from Hierarchies into Hierarchies 75

Figure 5.1 A hierarchy

mapping guarantees the
diagram is commutative

nondecreasing mapping from T

into T with ϕ(T
) cofinal in T , and for each t
∈ T
,
ft
is a mapping from Sϕ(t
) into St

such that

t
s
◦ fs
 ft
◦ ϕ(t
)ϕ(s
) (5.29)

i.e., such that the diagram in Figure 5.1 commutes for any s
, t
∈ T
satisfying s
≥ t
.
A mapping {ϕ, ft
} from H into H
is a similarity mapping if ϕ is a similarity
mapping from T
onto T and each ft
: Sϕ(t
) → St

is a similarity mapping. It is easy

to show in this case that ϕ−1 , ft −1
is a similarity mapping from H
into H.
When the partially ordered set (T ,≤) is a pseudotree (i.e., for each t ∈ T, the set
{y ∈ T : y ≤ x} is a chain), each hierarchy S of systems over T is called a tree-like
hierarchy of systems; each linked hierarchy {S, ts , T } of systems is called a linked
tree-like hierarchy of systems. A tree-like hierarchy B of systems over a tree T is a
partial tree-like hierarchy of a tree-like hierarchy A of systems over the same tree T
if for each t ∈ T , the state Bt is a partial system of the state At .
Theorem 5.4 [ZFC]. Let {S, ts , T } be a linked tree-like hierarchy of systems such
that the partially ordered set (T ,≤) satisfies
(i) T has a subset T0 consisting of all minimal elements in T such that for each t ∈
T , there exists t0 ∈ T0 such that t0 ≤ t.
(ii) There exists a cofinal subset D ⊂ T such that ts is an onto mapping for all s, t
∈ D ∪ T0 with s ≥ t.
Then there exists a nontrivial partial linked hierarchy of {S
, ˜ts , T } of {S, ts , T }
such that the state S
(t) is embeddable in the state S
(s) for all t, s ∈ T with s ≥ t.
Proof: From Lemma 5.1, it follows that there exists D∗ ⊂ D such that
(a) D∗  ∪{Dα∗ : α < η} for some ordinal η > 0.
(b) Dβ∗ is a maximal antichain in D−Eβ , for each β < η, where
   
Eβ  x ∈ T : ∃d ∈ ∪ Dα∗ : α < β (x ≤ d )

(c) D∗ is cofinal in T.
Let ≤∗ indicate the order relation on T defined in the proof of Lemma 5.1. From
the hypothesis that ts is onto for any s, t ∈ D ∪ T0 with s ≥ t, it follows that each
linkage mapping t0 r is onto, for all r ∈ T and t0 ∈ T0 satisfying r ≥ t0 . Let d0 be the
initial element of the well-ordered set (D∗ , ≤∗ ). We define an equivalence relation
Ed0 on Md0 such that the equivalence classes are given by
76 5 Hierarchies—Construction of Large-Scale Systems
  
[m]Ed0  x ∈ Md0 : ∀t0 ∈ T0 t0 ≤ d0 → t0 d0 (x)  t0 d0 (m) , (5.30)

for each m ∈ Md0 . By transfinite induction, for each element d in (D∗ , ≤∗ ) an

equivalence relation Ed on the object set Md of the state Sd can be defined such
that the equivalence classes are given by
  
[m]Ed  x ∈ Md : ∀t0 ∈ T0 t0 ≤ d0 → t0 d (x)  t0 d (m) . (5.31)

The linkage mapping tr , for any t, r ∈ D∗ with r ≥ t, can induce a mapping ¯tr
from Sr /Er onto St /Et such that
 
¯tr [mr ]Er  [mt ]Et , (5.32)

where [mt ]Et ∈ Mt /Et satisfies tr (mr )  mt . In fact, for each [mr ]Er ∈ Mr /Er , there
exists exactly one [mt ]Et ∈ Mt /Et such that
 
∀t0 ∈ T0 t0 ≤ t → t0 t (mt )  t0 r (mr ) . (5.33)

Therefore, tr ([mr ]Er ) ⊂ [mt ]Et , and ¯tr is well defined.

For any maximal linearly ordered set B ⊂ D∗ ∪ T0 , called a branch, Lemma 5.1
implies that (B, ≤) is well-ordered. Let b be the initial element in B. For any fixed
element x ∈ Mb from the fact that (B, ≤) is well-ordered
 and the Axiom of Choice,
it follows that there exists an element {[nd ]}d ∈B ∈ {Md /Ed : d ∈ B} such that

bd ([nd ])  {x} (5.34)


and that there exists an element {gd }d ∈B ∈ {[nd ] : d ∈ B} such that

ji (gi )  gj , (5.35)

for all i, j ∈ B with i ≥ j.

From
 the Axiom of Choice, it follows that there exists a subset M (B) ⊂
∪{ {[nd ]d ∈B : ∀d ∈ B(bd (nd )  x)}: x ∈ Mb } such that, for each x ∈ Mb ,
 
M (B) ∩ {[nd ]d ∈B : ∀d ∈ B(bd (nd )  x)}  (gd )d ∈B , (5.36)

where ts (gs )  gt , for all s, t ∈ B with s ≥ t. Let pd : M (B) → Md , for each d ∈ B,
be the projection, and define

Sd
(B)  (pd (M (B)), {∅}). (5.37)

Then {Sd
(B), ts Ss
(B), B} is a hierarchy of systems and a partial linked hierarchy of
{S, ts , B}.
5.2 Mappings from Hierarchies into Hierarchies 77

Let X be the set of all branches in the set D∗ ∪ T0 . By transfinite induction,

hierarchies {Sd
(B), ts , B} of systems, for each B ∈ X, can be defined such that for
any branches C, B ∈ X and d ∈ C ∩ B,

Sd
(B)  Sd
(C). (5.38)

Here we used the hypothesis that T is a pseudotree. Therefore, for each d ∈ D∗ ∪ T0 ,

we can define

Sd
 Sd
(B), (5.39)

where B is a branch containing d. Then Sd
is a partial system of Sd , and {S
, ts Ss
, D∗ ∪
T0 } is a nontrivial linked hierarchy of systems such that the state S
(t)  St
is
embeddable in the state S
(s)  Ss
, for any t, s ∈ D∗ ∪ T0 satisfying t ≤ s.
For any t ∈ T −(D∗ ∪ T0 ), choose s ∈ D∗ such that s ∈ Dα∗ , where α is the least
ordinal in η with this property, and s > t. We take a partial system St
of the system St
defined by
   
St
 ts Ms
, {∅} . (5.40)

Then it is not hard to check that {S
, ts Ss
, T } is a nontrivial partial linked hierarchy
of {S, ts , T } such that the state S
(t) is embeddable in the state S
(s) for any s, t ∈ T
with s ≥ t. QED

5.3 Time Systems

When the partially ordered set (T ,≤) is a total ordering, each hierarchy S of systems
over T is called a time system; each linked hierarchy {S, ts , T } of systems is called
a linked time system. In this case, T is called the time set (or time axis) of the time
system S.
An ordered triple (T , +, ≤) is called an ordered Abelian group if (T , +) is an
Abelian group and (T ,≤) is a linearly ordered set such that for any elements ti , si ∈ T ,
i  1, 2,

t1 ≤ t2 and s1 ≤ s2 → s1 + t1 ≤ s2 + t2 . (5.41)

We conclude this chapter with a discussion of time systems over ordered Abelian
groups. This kind of time system is called a system over a group . If the time system
is also a linked time system, we call the system a linked system over a group.
Suppose (T , +, ≤) is an ordered Abelian group. We let 0 be the identity in T , and,
for any x ∈ T , |x| indicates the absolute value of x, defined by
78 5 Hierarchies—Construction of Large-Scale Systems

|x|  max{x, −x}, (5.42)

where –x is the additive inverse of x. For any integer n,

⎧
⎪
⎪ 0 if n  0
⎪
⎪
⎪
⎨  x + x +
 . . . + x
 if n>0
nx  n times . (5.43)
⎪
⎪
⎪
⎪ −x − x − . . . − x if n < 0
⎩
⎪  
n times

The following notation will be used:

T[a, b)  {x ∈ T : a ≤ x < b}

T(a, b]  {x ∈ T : a < x ≤ b}

T(←, a)  {x ∈ T : x < a}

T(←, a]  {x ∈ T : x ≤ a}

T(a, →)  {x ∈ T : x > a}

T[a, →)  {x ∈ T : x ≥ a}

Let S be a system over a group and t0 ∈ T a fixed element. The right transfor-
mation of S, denoted S −t0 , is defined by S −t0 (t)  S(t−t0 ) for each t ∈ T ; the left
transformation of S, denoted S +t0 , is defined by S +t0 (t)  S(t + t0 ) for each t ∈ T.
Theorem 5.5. Suppose that S is a system over a group. Then for any a, b ∈ T,
 +a +b  +b
S  S +(a+b) and S −a  S −(a−b) .

A system S over a group T is periodic if there exists p ∈ T such that p  0 and

S +p  S. The element p is called a period of S.
Theorem 5.6. Let S be a system over a group T. If S is periodic with p ∈ T as its period,
then for each integer n, S np  S, where S np  S +np if n ≥ 0, and S np  S −[(−n)p] if
n < 0.
Proof: We show the theorem by induction on n.
Case 1: n ≥ 0. If n  0, the result is clear. Suppose S +kp  S for each integer k
satisfying 0 ≤ k < n. Then it follows that

S np  S +[(n−1)p+p]
 +p
 S +[(n−1)p]
5.3 Time Systems 79

 S +p
 S.

Thus, by induction, we have shown S +np  S for each n ≥ 0.

Case 2: n < 0. If n = –1, in the condition that S(t + p)  S(t), for each t ∈ T , let t 
x − p. Then S(x)  S(x − p) for each x ∈ T. This says that S −p  S.
Suppose S −kp  S for each integer k satisfying 0 ≤ k < –n. Then it follows that

S np  S −[(−n)p]
 S −[(−n−1)p+p]
 −p
 S −[(−n−1)p]
 S −p
 S.

Hence, by induction we have showed that S np  S for each integer n < 0.

Combining cases 1 and 2 completed the proof. QED
Corollary 5.1. If a system S over a group is periodic, there then is a positive period
for S.
A linked system {S, ts , T } over a group is periodic, if S is a periodic system over
a group with a period τ such that ts  t−τ,s−τ for any t, s ∈ T satisfying s ≥ t. This
element τ in T is called a period of the linked system {S, ts , T }.
Theorem 5.7. If τ is a period of a linked system {S, ts , T } over a group, then so is
n τ for each integer n  0.
The proof is straightforward and is omitted.
Theorem 5.8. Suppose that a linked system {S, ts , T } over a group is periodic with
τ > 0 as a period. Then the hierarchy H  {S, ts , T[0,τ ] } is similar to the hierarchy
K  {S, ts , T[nτ,(n+1)τ ] } for any integer n.
Proof: First, we show that nτ < (n + 1)τ for each integer n. In fact, if nτ ≥ (n + 1)τ
for some integer n, then

0  nτ − nτ ≥ (n + 1)τ − nτ  τ. (5.44)

This contradicts the hypothesis that τ > 0.

Second, we define a mapping ϕ: T[nτ,(n+1)τ ] → T[0,τ ] by

ϕ(t)  t − nτ. (5.45)

Then ϕ is a similarity mapping from the ordered set T[nτ,(n+1)τ ] onto the ordered
set T[0,τ ] . We next define mappings fx : Sϕ(x) → Sx , for each x ∈ T[nτ,(n+1)τ ] , as
identity mappings. Then {ϕ, fx } is a similarity mapping from the hierarchy H into
the hierarchy K. QED
80 5 Hierarchies—Construction of Large-Scale Systems

Corollary 5.2. In Theorem 5.8, the ordered sets T[0,τ ] and T[nτ,(n+1)τ ] can be replaced
by each of the following pairs:
(1) T(0 , τ ] and T(nτ , (n+1)τ ] ;
(2) T(0 , τ ) and T(nτ , (n+1)τ ) ; and
(3) T[0 , τ ) and T[nτ , (n+1)τ ) .
Generally, for a periodic system {S, ts , T } over a group, the hierarchies
{S, ts , T(←, a) } and {S, ts , T(0, →) } are not similar.
Remarks on references: What are presented in this chapter are mainly based on
Lin (1988, 1999), Ma and Lin (1987), and Mesarovic and Takahara (1975). In terms
of time systems, the concept has been studied by scholars in many publications. For
example, Schetzen (1980) studied continuous time systems of multi-linearity; Rugh
(1981) gave a systematic theory on nonlinear continuous and discrete time systems;
a more general setting of the concept of time systems can be found in Mesarovic and
Takahara (1975); and their definition was generalized by Ma and Lin (1987).

References

Lin, Y.: On periodic linked time systems. Int. J. Systems Sci. 19, 1299–1310 (1988)
Lin, Y.: General Systems Theory: A Mathematical Approach. Kluwer Academic and Plenum Pub-
lishers, New York (1999)
Ma, Y.H., Lin, Y.: Some properties of linked time systems. Int. J. General Systems 13, 125–134
(1987)
Mesarovic, M.D., Takahara, Y.: General Systems Theory: Mathematical Foundations. Academic
Press, New York (1975)
Rugh, W.J.: Nonlinear Systems Theory: The Volterra/Wiener Approach. Johns Hopkins University
Press, Baltimore (1981)
Schetzen, M.: The Volterra and Wiener Theories of Non-Linear Systems. Wiley, New York (1980)
Chapter 6
Controllabilities

This chapter focuses on the study of the various concepts of controllability. This
concept has found a wide range of applications in theory and practice. In partic-
ular, Section 6.1 looks at the concepts of controllability of input–output systems.
Section 6.2 generalizes these concepts to the scenarios of general systems which
might not be input–output. And Section 6.3 studies the concepts of controllability of
time systems.

6.1 Controllabilities of Input–Output Systems

The concept of controllability of systems was originally introduced as a property of

input–output systems. The main idea behind the concept says that an input–output
system S is controllable provided that, if we need to obtain some output of the
system with certain predetermined properties, we then can find some input such that
the corresponding output satisfies the required properties. In this section, we study
different concepts of controllability not only of input–output systems but also of
general systems.
Let S  (M , R) be an input–output system
 with input space X and output space  Y.
In the following, the symbols (x, y) and x0 , x1 , . . . , xα , . . . , y0 ,y1 , . . . , yβ , . . . will

be used congruently, where x  (x0 , x1 , . . . , xα , . . .) and y  y0 , y1 , . . . , yβ , . . . .
We now define
  
o(X )  n: ∃r ∈ R∃m ∈ ON r ⊆ X n × Y m , (6.1)
  
o(Y )  m: ∃r ∈ R∃n ∈ ON r ⊆ X n × Y m , (6.2)

where ON stands for the collection of all ordinals, for any i ∈ o(X),
  
X i,D  x ∈ X i : ∃r ∈ R∃j ∈ o(Y ) r(X ) ⊆ Y j , (6.3)

© Springer Nature Switzerland AG 2018 81

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_6
82 6 Controllabilities

for any j ∈ o(Y ),

 
Yj,R  y ∈ Y j : ∃r ∈ R∃x ∈ X i,D ((x, y) ∈ r) , (6.4)

and
 
Dom(X )  ∪ X i,D : i ∈ o(X ) , (6.5)

and
 
Ran(Y )  ∪ Yj,R : j ∈ o(Y ) . (6.6)

Suppose V is a set and G a mapping such that

G: Dom(X ) × Ran(Y ) → V. (6.7)

For an element v ∈ V , the system S is v-controllable relative to G if and only if

there exists a mapping cg: Dom(X) → Dom(X) such that, for any x ∈ Dom(X), there
exists a yx ∈ Ran(Y ) such that

G(cg(x))  v. (6.8)

The mapping cg is called a control function of S.

For example, suppose we are processing a chemical reaction. To reach a collection
of desired results (denoted by yx ) which satisfy an assigned target (denoted by v),
we could choose various methods (denoted by x) for achieving the wanted results.
But, in order to control the whole reaction procedure, we may add relative materials,
denoted by cg(x), together with those “choices” we have had. We would like to deem
the “input” (the relative materials) as “control” for the whole reaction procedure.

Theorem 6.1. For each element v ∈ V , the input–output system S is v-controllable

relative to the mapping G if and only if v ∈ G(Dom(X) × Ran(Y )).

Proof: The necessity part of the proof follows from equation (6.8).
Sufficiency. Let v ∈ G(Dom(X) × Ran(Y )). Then there exists (x, y) ∈ Dom(X) ×
Ran(Y ) such that G(x, y)  v. Define a control function of S, cg: Dom(X) → Dom(X),
as follows: cg(w)  x for all w ∈ Dom(X). Then for each w ∈ Dom(X) there exists yw
 y ∈ Ran(Y ) such that

G(cg(w), yw )  G(x, y)  v.QED

Let V  ⊂ V be a subset. The input–output system S is controllable on V  rel-

ative to G: Dom(X) × Ran(Y ) → V if and only if there exists a mapping cg:
Dom(X) → Dom(X) such that, for each v ∈ V  and any x ∈ Dom(X), there exists
yz  y(v, x) ∈ Ran(Y ) such that
6.1 Controllabilities of Input–Output Systems 83

G(cg(x), yz )  v. (6.9)

The mapping cg is called a control function of S.

Theorem 6.2. For a subset V  ⊂ V , the input–output system S is controllable on V 
relative to the mapping G if and only if there exists an element x ∈ Dom(X) such that
G({x} × Ran(Y )) ⊃ V  .
Proof: Necessity. Let cg: Dom(X) → Dom(X) be a control function of the system
S. Then, for each v ∈ V  and each x ∈ cg(Dom(X)), there exists a y ∈ Ran(Y ) such
that G(x, y)  v. Hence, G(x, Ran(Y )) ⊃ V  .
Sufficiency. Suppose there exists an element x ∈ Dom(X) such that G({x} ×
Ran(Y )) ⊃ V  . Define a control function cg: Dom(X) → Dom(X) by letting cg(w) 
x for all w ∈ Dom(X). Clearly, each v ∈ V  and any w ∈ Dom(X), there exists y
∈ Ran(Y ) such that

G(cg(w), y)  G(x, y)  v.

Therefore, S is controllable on V  relative to G. QED

For V  ⊂ V , the input–output system S is collectively controllable on V  relative
to the mapping G: Dom(X) × Ran(Y ) → V if and only if there exists a mapping cg:
Dom(X) → p(Dom(X)), where p(Dom(X)) is the power set of Dom(X), such that for
any v ∈ V  and any x ∈ Dom(X), there exists y = y(v, x) ∈ Ran(Y ) such that

G(z, y)  v, for some z ∈ cg(x). (6.10)

Theorem 6.3. For a subset V  ⊂ V , the input–output system S is collectively control-

lable on V  relative to the mapping G: Dom(X) × Ran(Y ) → V if and only if there
exists a nonempty subset A ⊆ Dom(X) such that G(A × Ran(Y )) ⊇ V  .
Proof: Necessity. Let cg: Dom(X) → p(Dom(X)) be a control function for S. Then
for any v ∈ V  and each x ∈ Dom(X), there exists a y ∈ Ran(Y ) such that G(z, y) 
v for some z ∈ cg(x). Hence, there exists an x ∈ Dom(X) such that G(cg(x) ×
Ran(Y )) ⊇ V  , where cg(x)  .
Sufficiency. Suppose there exists a nonempty subset A ⊆ Dom(X) such that G(A ×
Ran(Y )) ⊇ V  . Let us define a control function cg: Dom(X) → p(Dom(X)) as follows:
cg(x)  A for all x ∈ Dom(X). Then for any v ∈ V  there exists (z, y) ∈ A × Ran(Y )
such that G(z, y)  v. This implies that, for any v ∈ V  and any x ∈ Dom(X), there
exists y = y(v, x) ∈ Ran(Y ) such that G(z, y)  v. This completes the proof of the
theorem. QED

6.2 General Controllabilities

Examples can be given to show that not all systems are input–output systems. In the
following, we study the concept of controllability of general systems.
84 6 Controllabilities

Suppose S  (M , R) is a general system. For any relation r ∈ R and any ordinal

number μ > 0, we define the μ-initial section of r, denoted by I (rμ ), as follows: For
any x ∈ I (rμ ) ⊂ M μ , there exists y ∈ M ν such that (x, y) ∈ r, where v is a nonzero
ordinal number such that μ  +v = n(r) such that r ⊂ M . We
n(r)
 define the μ-final
section of r, denoted by E rμ , as follows: for any x ∈ E rμ ⊂ M ν , there exists
y ∈ I (rμ ) such that (y, x) ∈ r.
Let
   
I
r  x: ∃μ ∈ n(r) μ  0 ∧ x ∈ I rμ , (6.11)
   
E
r  x: ∃μ ∈ n(r) μ  0 ∧ x ∈ E rμ , (6.12)

and
   
I
R  ∪ I r: r ∈ R andE R  ∪ E r: r ∈ R . (6.13)

If M ∗ ⊂ M ,
     
I
RM ∗  I R M ∗ and E RM ∗  E R M ∗ (6.14)

are used to indicate all the elements in I R and all the elements in E R, respectively, in
which only the elements of M ∗ appear as coordinates.
Let V be another set and V  ⊂ V . The general system S is generally controllable
on V  relative to a mapping

G: I R × E R → V, (6.15)

if, for each v ∈ V  , there exist a set Wv , an Mv ⊂ M , and a mapping

 
Cv : Wv → p I R(Mv ) , (6.16)

such that for each u ∈ Wv there exists a yu ∈ E R such that

G(z, , yu )  v, for some z ∈ Cv (u). (6.17)

The mapping Cv is called a control function of S.

The main idea of general controllability of systems is the following. Suppose that
we are given a set of desired targets V . For each v ∈ V , there exists a set of “commands”
given to a system S so that S can realize the desired target v. When S receives the
commands, there will be some “well-organized” objects of S to accomplish the task.
Therefore, there should be some “receivers” and “accomplishers” in S such that the
whole process of executing the task is under a good control with a “fine” cooperation
between the “receivers” and the “accomplishers.”
Theorem 6.4. Let S be a general system and V  ⊂ V a subset. Then S is generally
controllable on V  relative to a mapping G: I R×E R → V , if and only if there exists a
6.2 General Controllabilities 85
 
nonempty
 subset
 G1 : I R1 × E R1 → V such that for some nonempty A ∈ p I R(MV  ) ,
G A × E R ⊃ V .

Proof: Necessity. Suppose S is generally controllable onV  relative

 to G. Then for
each v ∈ V’ there exist MV  ⊂ M and a nonempty Av ∈ p I R(Mv ) such that
 
G Av × E R ⊃ {v}. (6.18)
   
Let MV   ∪ Mv : v ∈ V  and A  ∪ Av : v ∈ V  . Then
    
A ∈ p ∪ I R(Mv ): v ∈ V  ⊂ p I R(MV  ) , (6.19)
 
and from equation (6.18), we have G A × E R ⊃ V  .
Sufficiency. Suppose
 there
 exists a nonempty
 subset MV  ⊂ M such that for some
nonempty A ∈ p I R(MV  ) , G A × E R ⊃ V  . Thus, for each v ∈ V  there are subsets
Av ⊂ A and Bv ⊂ E R such that

G(Av × Bv )  {v}. (6.20)

  
Define Mv ⊂ M by Mv  m ∈ M : ∃x ∈ Av ∃μ ∈ ON m  xμ . Let Wv be any
nonempty set and Cv : Wv → p I R(Mv ) a mapping defined by Cv (u)  Av for each
u ∈ Wv . Then for each u ∈ Wv there exists a yu ∈ Bv ⊂ E R such that

G(z, yu )  v, for some z ∈ Cv (u).

This proves the assertion. QED

Proposition 6.1. Suppose that S  (M , R) is an input–output system with input

space X and output space Y , and that S is v-controllable relative to a mapping
G: Dom(X) × Ran(Y ) → V . Then S is generally controllable on {v} relative to the
mapping G ∗ : I R × E R → V , where G ∗ is an extension of G.

Proof: Because Dom (X ) ⊂ I R and Ran (Y ) ⊂ E R, we can define an extension

G ∗ : I R × E R → V for the mapping G.
We now show the general controllability of S. For v ∈ {v}, there exist a set
Wv  Dom(X ) and a mapping
 
Cv : Wv → p I R , (6.21)

defined by Cv (u)  {cg(u)}, where cg: Dom(X) → Dom(X) is a control function for
S. Hence, for each u ∈ Wv , there exists a yu ∈ Ran(Y ) ⊂ E R such that

G ∗ (cg(u), yu )  G(cg(u), yu )  v, (6.22)

where cg(u) ∈ Cv (u). QED

86 6 Controllabilities

Proposition 6.2. Let V  ⊂ V be a subset such that an input–output system S  (M , R)

is controllable on V  relative to a mapping G: Dom(X) × Ran(Y ) → V , where X and Y
are the input and the output spaces of S, respectively. Then S is generally controllable
on V  relative to each extension G ∗ of G.

Proof: For each v ∈ V  , there exist sets Wv  Dom(X ) and Mv  M and a mapping
 
Cv : Wv → p I R , (6.23)

defined by Cv (u)  {cg(u)} for each u ∈ Wv . Then for each u ∈ Wv , there exists
yu  y(v, u) ∈ Ran(Y ) ⊂ E R such that

G ∗ (cg(u), yu )  G(cg(u), yu )  v, (6.24)

where cg(u) ∈ Cv (u). QED

Proposition 6.3. Suppose that S  (M , R) is an input–output system with input

space X and output space Y such that S is collectively controllable on V  relative to
a mapping G: Dom(X) × Ran(Y ) → V . Let G ∗ : I R × E R → V be an extension of G.
Then S is generally controllable on V  relative to G ∗ .

Proof: For each v ∈ V  , there exist sets Wv  Dom(X ) and Mv  M and a mapping
 
Cv : Wv → p I R , (6.25)

defined by Cv (u)  cg(u) such that, for each u ∈ Wv , there exists yu  y(v, u) ∈
Ran(Y ) ⊂ E R such that

G ∗ (z, yu )  G(z, yu )  v, (6.26)

for some z ∈ Cv (u)  cg(u). QED

Theorem 6.5. Suppose that h: S1  (M1 , R1 ) → S  (M , R) is an S-continuous

mapping from a system S1 into a system S such that S is generally controllable on
V  relative to a mapping G: I R × E R → V . Then S1 is generally controllable on V 
relative to a mapping G1 : I R1 × E R1 → V .

Proof: First, we construct a desired mapping G1 : I R1 × E R1 → V . Let v0 ∈ V be a

fixed element. For any x ∈ I R1 and y ∈ E R1 , if there exist relations rx , ry ∈ R such
that
    
x ∈ I h−1 (rx ) and y ∈ E h−1 ry , (6.27)

then define

G1 (x, y)  G(h(x), h(y)). (6.28)

6.2 General Controllabilities 87

If either rx or ry or both do not exist, we define

G1 (x, y)  v0 . (6.29)

It can be shown that G1 : I R1 × E R1 → V is well defined.

Second, we show that S1 is generally controllable on V  relative to G1 : I R1 ×E R1 →
V . Suppose that
I for each
 v ∈ V  , there exist sets Wv and Mv ⊂ M and a mapping
Cv : Wv → p R(Mv ) such that, for each u ∈ Wv , there exists a yu ∈ E R such that
G(z, yu )  v for some z ∈ Cv (u). Wenow let h−1 (Mv ) ⊂ M1 and define a control
function 1 Cv : Wv → p I R1 h−1 (Mv ) as follows: For each u ∈ Wv ,
   
1
Cv (u)  x ∈I R1 h−1 (Mv ) : h(x) ∈ Cv (u) . (6.30)

Now it only remains to show that 1 Cv is a control function of S1 . In fact, for each
u ∈ Wv , there exists yu ∈ E R such that

G(z, yu )  v, for some z ∈ Cv (u), (6.31)

so there exists 1 yu ∈ h−1 (yu ) ⊂ E R1 such that

      
G1 z  , 1 yu  G h z  , h 1 yu  G(z, yu )  v,

where z  ∈ h−1 (z) ⊂ 1 Cv (u). Therefore, S1 is generally controllable on V  relative

to G1 . QED

Theorem 6.6. Let {Si : i ∈ I } be a set of systems, where I is an index set, such that
there exists an i0 ∈ I such that the system Si0 is generally controllable onV  relative to
a mapping Gi0 : I Ri0 × E Ri0 → V . Then the Cartesian product system CP {Si : i ∈ I }
is also generally controllable onV  (relative
 to adifferent mapping G: R× R → V ),
I E

where it is assumed that Si0  Mi0 , Mi0 and CP {Si : i ∈ I }  (M , R).

The proof follows from Theorems 3.15 and 6.5.

Theorem 6.7. Suppose that h: S  (M , R) → S1  (M1 , R1 ) is an embedding

mapping from a system S into a system S1 such that S is generally controllable on
V  relative to a mapping G: I R × E R → V . Then, S1 is generally controllable on V 
relative to a mapping G1 : I R1 × E R1 → V .

Proof: We first construct a desired mapping G1 : I R1 × E R1 → V . Let v0 ∈ V be a

fixed element. For any x ∈ I R1 and y ∈ E R1 , if there exist x∗ ∈ I R and y∗ ∈ E R such
that h(x∗ )  x and h(y∗ )  y, we define
 
G1 (x, y)  G x∗ , y∗ . (6.32)

From the hypothesis that h is 1–1, we see that equation (6.32) is well-defined, for
the existence of x∗ and y∗ is unique. If one or both of the elements x∗ and y∗ do not
exist, then we define
88 6 Controllabilities

G1 (x, y)  v0 . (6.33)

We then have defined a mapping G1 : I R1 × E R1 → V .

Second, we show that the system S1 is generally controllable on V  relative to the
mapping G1 . Suppose that for each v ∈ V  , there exist sets Wv and Mv ⊂ M and a
mapping Cv : Wv → p I R(Mv ) such that, for each u ∈ Wv , there exists a yu ∈ E R
 z ∈ Cv (u).We now let h(Mv ) ⊂ M1 and define a
such that G(z, yu )  v for some
control function 1 Cv : Wv → p I R1 h−1 (Mv ) as follows: For each u ∈ Wv ,
1
Cv (u)  {h(z): z ∈ Cv (u)}. (6.34)

It remains to show that 1 Cv is indeed a control function for S1 . For each u ∈ Wv

there exists a yu ∈ E R such that

G(z, yu )  v, for some z ∈ Cv (u). (6.35)

Therefore,

G1 (h(z), h(yu ))  G(z, yu )  v, (6.36)

where h(z) ∈ 1 Cv (u) and h(yu ) ∈ E R1 . QED

Corollary 6.1. If a subsystem A  (MA , RA ) of a system S  (M , R) is generally

controllable on V  relative to a mapping GA : I RA × E RA → V , then S is generally
controllable on V  relative to a mapping G: I R × E R → V .

Theorem 6.8. Let {Si  (Mi , Ri ): i ∈ I }. be a set of systems and {Vi ⊂ V : i ∈ I } a

set of nonempty subsets of V . If each system Si is generally controllable on the subset
Vi relative to a mapping Gi : I Ri × E Ri → V , then ⊕{Si : i ∈ I }  (M , R) is generally
controllable on ∪{Vi : i ∈ I } relative to a mapping G: I R × E R → V , where for each
i ∈ I and any (x, y) ∈ I Ri × E Ri , G(x, y)  Gi (x, y).

Proof: From the definition of free sums of systems, it follows that

   
I
R  ∪ I Ri : i ∈ I and E R  ∪ E Ri : i ∈ I , (6.37)

where I Ri ∩ I Rj   E Ri ∩ E Rj for any distinct i, j ∈ I. Let G be an extension of

the mappings {Gi : i ∈ I}; i.e., G is a mapping from I R × E R into V such that for any
(x, y) ∈ I R × E R, if (x, y) ∈ I Ri × E Ri , then G(x, y)  Gi (x, y).
To show that the free sum ⊕{Si : i ∈ I } is generally controllable on ∪{Vi : i ∈ I }
relative to G, we pick an arbitrary v ∈ ∪{Vi : i ∈ I }. Then there exists an i ∈ I such that
v ∈ Vi . Therefore, there exist sets Wv and Mv ⊂ Mi ⊂ ∪{Mi : i ∈ I } and a mapping
   
Cv : Wv → p I R(Mv )  p I Ri (Mv ) , (6.38)

such that for each u ∈ Wv , there exists a yu ∈ E Ri ⊂ E R such that

6.2 General Controllabilities 89

G(z, yu )  v, for some z ∈ Cv (u). (6.39)

This proves the theorem. QED

6.3 Controllability of Time Systems

In the following, we conclude this chapter with a discussion of controllabilities of

time systems.
Theorem 6.9. Suppose that {S, ts , T } is a linked time system. If there exists a t ∈ T
such that
(a) St is a nondiscrete input–output system,
(b) st (Rt ) ∩ Rs  , for each s < t,
(c) ts (Rs ) ∩ Rt  , for each s > t,
then each state Ss of the linked time system has a nondiscrete partial system which
is an input–output system.
Proof: Let St  (Mt , Rt ) and Xt and Yt be the input and output spaces of the state
St , respectively. If s < t, let S̄s  M̄s , R̄s be a partial system of the system Ss defined
by M̄s  st (Mt ) and R̄s  st (Rt ) ∩ Rs . Then S̄s is a nondiscrete input–output system
with input space st (X t ) and output space st (Yt ). Finally, if s > t, let S̄s  M̄s , R̄s
be a partial system of Ss defined by M̄s  Ms and {r ∈ Rs : ts (r) ∈ Rt }. Then S̄s is
a nondiscrete input–output system with input space X s  {x ∈ Ms : ts (x) ∈ X t } and
output space Ys  {y ∈ Ms : ts (x) ∈ Yt }. QED
Theorem 6.10. Suppose that {S, ts , T } is a linked time system such that
(a) T has an initial element t0 ;
(b) Each t0 s is surjective with t0 s (Rs ) ∩ Rt0  ;
(c) St0 is a nondiscrete input–output system with input space X t0 and output space
Yt0 ; and  
(d) St0 is v-controllable relative to a mapping G: Dom X t0 × Ran (Yt0 ) → V .

Then there exists a linked time system {S̄, ¯ts , T } such that each state S̄t is a
nondiscrete input–output system and is v-controllable relative to some mapping Gt .
Proof: First we see the construction of the linked time system {S̄, ¯ts , T }. By the
Axiom of Choice, we well order the time set T as follows:

t0 , t1 , . . . , tα , . . . , α < |T |, (6.40)

where t0 is the initial element in T . Suppose that the original order on T is ≤ , and in
equation (6.40) T is well ordered by a well-order relation ≤∗ . Generally, ≤ and ≤∗
are different.
90 6 Controllabilities

Define S̄t0  St0 and ¯t0 t0  t0 t0 . Suppose that for some α < |T | the system S̄tk
and the mappings ¯tj tr have been defined such that ¯tj ti  ¯tj tr ◦ ¯tr ti , for each k < α
 
and any i, r, j < α such that ti ≥ tr ≥ tj . Now define S̄tα and ¯ti tα : iαandtα ti and
 
¯tα ti : i < αandtα < ti asfollows: Let Etα be the equivalence
 relation on the object
set Mtα defined by Etα  (x, y) ∈ Mt2α : t0 tα (x)  t0 tα (y) and
 ∗
S̄tα  Stα /Etα , (6.41)
 ∗    
where Stα  Mtα , ∗ Rtα and ∗ Rtα  r ∈ Rtα : t0 tα (r) ∈ Rt0 . For each β < α, if
tα > tβ , define

¯tβ tα ([mα ])  mβ , (6.42)
 
for each [mα ] ∈ Mtα /Etα , where t0 tα (mα )  t0 tβ mβ . If tα < tβ , we define
 
¯tα tβ mβ  [mα ], (6.43)
  
for each mβ ∈ Mtβ /Etβ , where t0 tβ mβ  t0 tα (mα ). It is then not hard to prove
  
that ¯tj ti : i, j ≤ α ti ≥ tj is a family of linkage mappings.
By transfinite induction, a linked time system {S̄, ¯ts , T } is defined such that
S̄t0  St0, and  each
 state S̄t is a nondiscrete input–output system with input

−1 −1
space
  t0 t X t0  [mt ] ∈ M /E :  (m
 t t t0 t t ) ∈ X t0 and output space t0 t Yt0 
[mt ] ∈ Mt /Et : t0 t (mt ) ∈ Yt0 .
Second, we show that each state S̄t is v-controllable relative to some mapping
     
Gt : Dom −1
t0 t X t0 × Ran −1
t0 t Yt0 → V. (6.44)

−1
  −1
 
Let Dom (X t ) 
 t0t X t0 and Ran (Yt )  t0 t Yt0 . For any xt ∈ Dom (X t ), there
exists xt0 ∈ Dom X t0 such that

xt0  cgt0 ◦ ¯t0 t (xt ), (6.45)

where cgt0 is a control function of St0 . Thus, there exists a yt0 ∈ Ran(Yt0 ) such that

G(xt0 , yt0 )  v. (6.46)

 
¯−1 ¯−1 ¯
 Let cgt   t0 t ◦ cgt0 ◦ t0 t and yt  t0 t yt0 . If we assign t0 t (xt , yt ) 
t0 t (xt ), t0 t (yt ) , then

Gt  Gt0 ◦ ¯t0 t : Dom(X t ) × Ran(Yt ) → V, (6.47)

and Gt (cgt (xt ), yt )  Gt0 ◦ ¯t0 t (cgt (xt ), yt )  Gt0 (xt0 , yt0 )  v. Therefore, the system
S̄t is v-controllable relative to the mapping Gt . QED
6.3 Controllability of Time Systems 91

Theorem 6.11. Assume hypotheses (a)–(c) of Theorem 6.10 and that St0 is con-
trollable on V  ⊂ V relative to a mapping G: Dom (X t0 ) × Ran (Yt0 ) → V . Then
there exists a linked time system {S̄, ¯ts , T } such that each state S̄t is a nondiscrete
input–output system and is controllable on V  relative to some mapping Gt .
The proof is a slight modification of that of Theorem 6.10 and is omitted.

Theorem 6.12. Assume hypotheses (a)–(c) of Theorem 6.10 and that St0 is col-
lectively controllable on V ∗ ⊂ V relative to a mapping G: Dom (X t0 ) × Ran (Yt0 )
→ V . Then there exists a linked time system {Q, ¯ts , T } such that each state Qt is
a nondiscrete input–output system and is collectively controllable on V ∗ relative to
some mapping Gt .
The proof is a slight modification of that of Theorem 6.10 and is left to the reader.

Remarks on references: Studies on various controllabilities of systems can be

found in almost all books on systems. For example, Deng (1985) studied the concept
of gray control, Wonham (1979) presented a concept of controllability of linear
systems of differential equations, and a study of controllabilities of nonlinear systems
can be found in Casti (1985). Not all of the concepts of controllabilities are equivalent.
The concepts of controllabilities presented here are based on research contained in
Mesarovic and Takahara (1975), Ma and Lin (1987).

References

Casti, J.: Non-linear Systems Theory. Academic Press, Orlando (1985)

Deng, J.-L.: Grey Control Systems. Hua-zhong Institute of Technology Press, Wuhan (1985)
Ma, Y.H., Lin, Y.: Some properties of linked time systems. Int. J. General Syst. 13, 125–134 (1987)
Mesarovic, M.D., Takahara, Y.: General Systems Theory: Mathematical Foundations. Academic
Press, New York (1975)
Wonham, W.M.: Linear Multi-variable Control: A Geometric Approach. Springer, New York (1979)
Chapter 7
Limit Systems

7.1 Existence of Limits

We say that C is a category if C consists of a class of objects, denoted by objC,

pairwise disjoint sets of morphisms, denoted by HomC (A, B), for every ordered pair
of objects (A, B), and a composition HomC (A, B) × HomC (B, C) → HomC (A, C),
denoted by ( f, g) → g f , satisfying the following axioms:
(i) For each object A, there exists an identity morphism I A ∈ HomC (A, A) such
that f I A  f , for all f ∈ HomC (A, B) and I A g  g, for all g ∈ HomC (C, A).
(ii) Associativity of composition holds whenever possible; that is, if f ∈
HomC (A, B), g ∈ HomC (B, C), and h ∈ HomC (C, D), then

h(g f )  (hg) f. (7.1)

Let us see some examples of categories.

a. Suppose that GC c is the category that consists of all general systems and all S-
continuous mappings between systems. Thus, for any two systems S1 and S2 , the
set HomGC c (S1 , S2 ) equals the set of all S-continuous mappings from S1 into S2 .
b. Suppose that GC h is the category that consists of all general systems and all
homomorphisms between systems. Thus, for two arbitrary systems S1 and S2 ,
the set HomGC h (S1 , S2 ) of morphisms equals the set of all homomorphisms from
S1 into S2 .
Let {S, ts , T } be an α-type hierarchy of systems in GC c (respectively, in GC h );
i.e., each linkage mapping ts is an S -continuous mapping (respectively, a homo-
morphism). The inverse limit of this hierarchy, denoted by lim←c {S, ts , T } or
lim←c St (respectively, lim←h {S, ts , T } or lim←h St ), is a system and a family of
S-continuous mappings αt : lim←c {S, ts , T } → St (respectively, homomorphisms
αt : lim←h {S, ts , T } → St ) with αi  i j ◦ α j whenever i ≤ j satisfying the univer-
sal mapping property in Figure 7.1 (respectively, the universal mapping property in
© Springer Nature Switzerland AG 2018 93
J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_7
94 7 Limit Systems

Figure 7.1 Universal

mapping property of limit
systems

Figure 7.2 Universal

mapping property of limit
systems

Figure 7.2) satisfying: For every system X and S-continuous mappings (respectively,
homomorphisms) f i : X → Si making the corresponding diagram commute when-
ever i ≤ j, there exists a unique S-continuous mapping β: X → lim←c St (respectively,
a homomorphism β: X → lim←h St ) making the diagram commute.
From the definition of inverse limits, it can be seen that if a linked hierarchy of
systems has an inverse limit, the limit is unique up to a similarity.
Theorem 7.1. Let {S, ts , T } be a linked α-type hierarchy of systems in the category
GC c . Then the inverse limit lim←c {S, ts , T } exists.
Proof: Let St  (Mt , Rt ), for each t ∈ T. Define a desired system lim ←c St  (M, R)
as follows:
 

M  (m t )t∈T ∈ Mt : ∀s, t ∈ T (t ≤ s → ts (m s )  m t ) , (7.2)
t∈T

and
    

R  r : ∃ j ∈ T ∃r j ∈ R j x ∈ r ↔ x ∈ M n (r j ) ∧ x jk k<n (r j ) ∈ r j , (7.3)

and define mappings αi : lim←c St → Si by letting

 
αi (m t )t∈T  m i , (7.4)

for every object (m t )t∈T ∈ M and any i ∈ T. Then it can be checked that the mappings
αt are S-continuous and satisfy αt  ts ◦ αs whenever t ≤ s.
We now let Y  (MY , RY ) be a system and { f t : Y → St : t ∈ T } a family of
S-continuous mappings such that f t  ts ◦ f s whenever t ≤ s. Define a mapping
β: Y → lim←c St by letting

β(y)  ( f t (y))t∈T , (7.5)

7.1 Existence of Limits 95

for each y ∈ MY . Then from the hypothesis that f t  ts ◦ f s whenever t ≤ s, it

follows that β : Y → lim←c St is well defined.
We now show that β: Y → lim←c St is S-continuous. For any relation
r ∈ R, there exist j ∈ T and
r j ∈ R j such that β: Y → lim←c St r 
n (r j )
 
x∈M : x jk k<n (r j ) ∈ r j . Therefore,
  

β −1 (r )  β −1 (x): x ∈ M n (r j ) ∧ x jk k<n (r j ) ∈ r j
 

 (yk )k<n (r j ) ∈ M n (r j ) : f j (yk )k<n (r j ) ∈ r j

 
 f j−1 r j ∈ RY .

From the definition of S-continuity, it follows that β: Y → lim←c St is S-

continuous. The uniqueness of β follows from the commutative diagram in the
theorem or Figure 7.1. QED
Corollary 7.1. Suppose that {S, ts , T } is a linked α-type hierarchy of systems such
that each linkage mapping ts is S-continuous. Then lim←c St is a partial system of

C P {S: t ∈ T }.

The proof follows from the proofs of Theorems 7.1 and 3.15.
Theorem 7.2. Let {S, ts , T } be a linked α-type hierarchy of systems in the category
GC h . Then the inverse limit lim←h {S, ts , T } exists.
Proof: We define a system lim←h {S, ts , T }  (M, R) as follows:
 

M  (m t )t∈T ∈ Mt : ∀s, t ∈ T (t ≤ s → ts (m s )  m t ) , (7.6)
t∈T

and
 

∗
R  (rt )t∈T ∈ Rt : ∀t, s ∈ T (t ≤ s → ts (rs )  rt ) , (7.7)
t∈T

and for any r ∗  (rt )t∈T ∈ R ∗ , define n = n(rt ), for t ∈ T, and

 
  
r  x ∈ M : ∃(xt )t∈T ∈
n
rt ∀α < n x(α)  (xt (α))t∈T , (7.8)
t∈T

and

R  r : r ∗ ∈ R∗ . (7.9)
 
Let αt : M → Mt , for each t ∈ T, be the projection; i.e., αt (m t )t∈T  m t . Then,
for any relation r ∈ R, there exists r ∗  (rt )t∈T ∈ R ∗ such that r is defined by
96 7 Limit Systems

equation (7.8). Thus, αt (r )  {αt (x): x ∈ r }  rt for each t ∈ T. That implies that
each αt is a homomorphism from (M, R) into St .
We now show that the system (M, R) is the inverse limit of the linked  hierarchy

{S, ts , T }. Only two things need to be proven: (i) For any system S̄  M̄, R̄ and

any set of homomorphisms f t : S̄ → St : t ∈ T such that

f t  ts ◦ f s , (7.10)

whenever t ≤ s, and (ii) there exists a unique homomorphism β: S̄ → (M, R) so that

αt ◦ β  f t , (7.11)

for each t ∈ T. To prove (i), define β: M̄ → M by β(m̄)  ( f t (m̄))t∈T . Then β is a

n(r̄ )
homomorphism from  S̄ into (M, R). Since for any relation r̄ ∈ R̄, x̄ ∈ r̄ ⊆ M̄ ,
we have β ◦ x̄  ( f t (x̄(α)))t∈T : α < n(r̄ ) ; so β(r̄  )  {β ◦x̄: x̄ ∈ r̄ } ∈ R; so
β(r̄ )  {β ◦ x̄: x̄ ∈ r̄ } ∈ R.Since αt ◦ β(m̄)  αt ( f t (m̄))t∈T  f t (m̄), for all
m̄ ∈ M̄, we have αt ◦ β  f t , for each t ∈ T, as mappings between sets. On the other
hand, for any relation r̄ ∈ R̄ and any x̄ ∈ r̄ ⊆ M̄ n(r̄ ) , we have
 
αt (β ◦ x̄)  αt { f t (x̄(α)): α < n(r̄ )}t∈T  { f t (x̄(α)): α < n(r̄ )}t∈T

so

αt ◦ β(r̄ )  { f t (x̄(α)): α < n(r̄ ) ∧ x̄ ∈ r̄ }  f t (r̄ ).

This implies that αt ◦ β  f t as mappings from the relation set R̄ into the relation
set Rt for each t ∈ T. Hence, αt ◦ β  f t , for each t ∈ T, as homomorphisms from
systems into systems.

(ii) The homomorphism β is unique. Suppose that β  : S̄ → (M, R) is another

homomorphism satisfying αt ◦ β   f t , for each t ∈ T. Let β  (m̄)  (m t )t∈T ∈
M. Then m t  αt ◦ β  (m̄)  f t (m̄)  αt ◦ β(m̄) for all t ∈ T. Thus, β  (m̄) 
β(m̄) for all m̄ ∈ M̄. This implies that β   β on the set M̄, so β  and β induce
the same homomorphism. Hence, β   β as homomorphisms. QED

Theorem 7.3 Each mapping {g, fs  } of a linked hierarchy H  {S, ts , T } of sys-
tems into a linked hierarchy H   S  , t  s  , T  of systems induces a mapping from
lim← H into lim← H  , where lim← indicates either lim←c or lim←h .

Proof: For an object x  (xt )t∈T in lim← H and every s  ∈ T  , define

 
ys   f s  x g(s  ) . (7.12)

The element y  {ys  }s  ∈T  thus obtained is an object in lim← H  . Indeed, for any
s , t ∈ T’ satisfying t ≤ s , then we have

7.1 Existence of Limits 97
 
t  s  (ys  )  t  s  ◦ f s  x g(s  )
 
 f t  ◦ g(t  )g(s  ) x g(s  )
 
 f t  x g(t  )
 yt 

For the object x in lim← H define ƒ(x)  y in lim← H  . Then ƒ is a mapping from
lim← H into lim← H  . QED

7.2 Induced Mappings Between Limits

The mapping ƒ: lim← H → lim← H  in the proof of Theorem 7.3 is called the limit
mapping induced by {g, f s  } and is denoted by f  lim← {g, f s  }.

Theorem 7.4. Let {g, f s  } be a mapping of a linked

 hierarchy of systems H 
{S, ts , T } to a linked hierarchy of systems H   S  , t  s  , T  . If all mappings f s 
are one-to-one, the limit mapping f  lim← {g, f s  } is also one-to-one. If, moreover,
all mappings f s  are onto, ƒ is also an onto mapping.

Proof: Let x  {xs }s∈T and y  {ys }s∈T be two distinct objects of the limit system  
lim← H . Take an s0 ∈ T such that xs0  ys0 and an s’ ∈ T’ satisfying s0 ≤ g s  .
Clearly,  x g(s )  yg(s ) , so it follows from the hypothesis that f s  is one-to-one, that
is, f s  x g(s  )  f s  yg(s  ) , which shows that f (x)  f (y). Therefore, the limit
mapping f is one-to-one.
We now suppose that f s  is a one-to-one mapping from Sg(s  ) onto Ss  for every  
s ∈ T  . Take an object y  {ys  }s  ∈T  from lim← H  . It follows that g s   g t 

−1 −1   
   fs  (y
implies  s  )  f t  (yt  ) for any s , t ∈ T satisfying s’ ≥ t’, so that for any
g s ∈ g T , the element

z g(s  )  f s−1
 (ys  ) ∈ M g(s  ) (7.13)
 
is well defined. For any s ∈ T, choose an element s’ ∈ T’ such that s ≤ g s  and
define
 
xs  sg(s  ) z g(s  ) ∈ Ms . (7.14)

It can be verified that xs does not depend on the choice of s’ that x  {xs }s∈T is
an element in lim← H and that f (x)  y. That shows ƒ is an onto mapping. QED

Theorem 7.5 Suppose that {g, f s  } is a mapping of a linked hierarchy of systems

H  {S, ts , T } into a linked hierarchy of systems H   S  , t  s  , T  , where H and
H’ are in the category GC c . If all mappings f s  are S-continuous, the limit mapping
f  lim← {g, f s  }: lim←c H → lim←c H  is also S-continuous.
98 7 Limit Systems
 
Proof: Suppose lim←c H  (M, R) and  lim 
←c H   M  , R  . Theorem 7.3 implies
that the limit mapping f : (M, R) → M  , R  is well defined. Let r be a relation in

lim←c H  . There exist j ∈ T’ and r j ∈ R j such that
   n r  

r  (m ik )i∈T  k<n (r j ) ∈ M  ( j ) : m jk k<n (r j ) ∈ r j . (7.15)

Then we have
    
f −1 (r )  f −1 (m ik )i∈T  : k < n r j :
   n (r j )  

(m ik )i∈T  k<n (r j ) ∈ M  ∧ m jk k<n (r j ) ∈ r j

     

 (m ik )i∈T k<n (r j ) ∈ M n (r j ) : m jk k<n (r j ) ∈ f j−1 r j ∈ R.

Thus, from the definition of S-continuity, the limit mapping ƒ is S-continuous.

QED

Theorem 7.6. Suppose that H  {S, ts , T } and H   S  , t  s  , T  are two linked
hierarchies of systems in the category GC c and {g, f s  } is a mapping from H into H’. If
all mappings f s  are similarity mappings, the limit mapping f : lim←c H → lim←c H 
is also a similarity mapping.

Proof: Applying Theorem 7.4, we know that the limit mapping f is a bijection from
lim←c H onto lim←c H  . Theorems 3.4 and 7.5 imply that f is also S-continuous.
Now, in order to show that f −1 is S-continuous from lim←c H  onto lim←c H , it
suffices to show that f is a homomorphism from lim←c H into lim←c H  .
Let r ∈ R be a relation such that there exist j ∈ T and r j ∈ R j satisfying
   

r  (m ik )i∈T k<n (r j ) ∈ M n (r j ) : m jk k<n (r j ) ∈ r j . (7.16)

Then, we have the following:

      

f (r )  f (m ik )i∈T k<n (r j ) : (m ik )i∈T k<n (r j ) ∈ M n (r j ) ∧ m jk k<n (r j ) ∈ r j

       

 f i m g(i)k i∈T  k<n r : (m ik )i∈T k<n (r j ) ∈ M n (r j ) ∧ m jk k<n (r j ) ∈ r j

( j)
    n (r j )    

 (m ik )i∈T  k<n (r j ) ∈ M : m jk k<n (r j ) ∈ f i r j ∈ R 

Therefore, the limit mapping f is a similarity mapping from lim←c H onto

lim←c H  . QED

Theorem 7.7. Suppose that H  {S, ts , T } and H   S  , t  s  , T  are two linked
hierarchies of systems in the category GC h and {g, f s  } is a mapping from H into H’.
If all mappings f s  are homomorphisms, the limit mapping f : lim←h H → lim←h H 
is also a homomorphism.
7.2 Induced Mappings Between Limits 99

Proof: It suffices to show that for any relation  r inthe system lim←h H  (M, R),
ƒ(r) is a relation in the system lim←h H   M  , R  . To this end, let r ∗  (rt )t∈T ∈
R ∗ , where R ∗ is defined by equation (7.7), such that
 
  ∗  
n(r ∗ )
r x∈M : ∃(xt )t∈T ∈ rt ∀α < n r x(α)  (xt (α))t∈T . (7.17)
t∈T

Then we have
  n(r ∗ )   
f (r )  x ∈ M  : ∃(xt  )t  ∈T  ∈ f t  r g(t  )
t  ∈T 
  
∀α < n r ∗ x(α)  (xt  (α))t  ∈T  . (7.18)

Hence, f (r) ∈ R’. QED

Theorem 7.8. Under the same hypotheses as in Theorem 7.7, if all mappings f s  ,
in addition, are similarity mappings, the limit mapping f : lim←h H → lim←h H  is
also a similarity mapping.
Proof: By Theorems7.4 and 7.7, it suffices to show that for any relation r’ in the
system lim←h H   M  , R  , there exists a relation r in the system lim←h H 

(M, R) such that f (r)  r’. Pick an element f : lim←h H → lim←h H  , where R ∗ is

defined by equation (7.7) with Rt , T, and ts replaced by Rt  , T’, and t  s  , respectively,
such that
 


  n r  ∗
r  x∈ M : ∃(xt  )t  ∈T  ∈ rt 
t  ∈T 
   

∀α < n r ∗ x(α)  (xt  (α))t  ∈T  . (7.19)


From the hypothesis that each f t  is a similarity mapping from Sg(t  ) onto St  , it
follows that there exists exactly one relation r g(t  ) ∈ Rg(t  ) such that
 
f t  r g(t  )  rt  . (7.20)
 
Now for any t ∈ T, choose a t’ ∈ T’ such that g t  ≥ t. Define a relation rt ∈ Rt
by
 
rt  tg(t  ) r g(t  ) . (7.21)

It can be shown that the definition of rt does not depend on the choice of t’. Then
an element r ∗  (rt )t∈T ∈ R ∗ is well defined, where R ∗ is defined by equation (7.7).
The relation r ∈ R, defined by
 ∗
   
r  x ∈ M n(r ) : ∃(xt )t∈T ∈ rt ∀α < n r ∗ x(α)  (xt (α))t∈T , (7.22)
t∈T
100 7 Limit Systems

satisfies f (r)  r ∗ .QED

Corollary 7.2 Let {S, ts , T } be a linked α-type hierarchy of systems in category
 GC c
or GC h , T’ a cofinal subset of T. Then lim← {S, ts , T } is similar to lim← S, ts , T  ,
where lim← indicates either lim←h or lim←c .
The result follows from Theorems 7.6 and 7.8.
Corollary 7.3 Let {S, ts , T } be a linked α-type hierarchy of systems in category
GC c or GC h . If the partially ordered set T has a maximum element t0 ∈ T , then the
systems lim←h {S, ts , T }, lim←c {S, ts , T }, and St0 are similar.
The result follows from Corollary 7.2.
Corollary 7.4. Suppose that {S, ts , T } is a linked α-type hierarchy  of systems in
the category GC h . Then the limit system lim←h St is a partial system of {St : t ∈ T }.
The result follows from the proofs of Theorems 7.2 and 3.14.
Theorem 7.9. Let {S, ts , T } be a linked α-type hierarchy of systems in the cate-
gory GC c such that one of the states St is connected. Then the inverse limit system
lim←c {S, ts , T } is connected.
The result follows from Theorems 7.1 and 4.3.

7.3 Centralizability of Limits

Let {S, ts , T } be a linked α-type hierarchy of systems

 such that for every element
m  (m t )t∈T in the thread set , there exists a set mts : m s → m t : t, s ∈ T (s ≥ t)
of linkage mappings, where each m s is again a system. The one-level inverse limit,
denoted by lim1←c {S, ts , T } or lim1←c St when all relevant linkage mappings are
S-continuous, or lim1←h {S, ts , T } or lim1←h St when all relevant linkage mappings
are homomorphisms, is defined by lim1← St  (M, R), where. indicates lim1←c or
lim1←h , and the object set M and the relation set R are defined as follows: There
exists a bijection h:  → M such that
 
h (m t )t∈T  lim1← m t ∈ M, (7.23)

and

R  h(r ): r is a relation in lim1← St ,

where lim1← indicates either lim1←c or lim1←h .

Theorem 7.10. Let {S, ts , T } be a linked α-type hierarchy of centralizable systems
in the category GC c . If the one-level inverse limit system lim1←c {S, ts , T } exists, then
the system is centralizable.
7.3 Centralizability of Limits 101

Proof: The proof follows from the fact that if Ct is a center of the system St , for
each t ∈ T, then the inverse limit lim←c Ct is a center of the one-level inverse limit
lim1←c {S, ts , T }. QED

Theorem 7.11. Let {S, ts , T } be a linked α-type hierarchy of centralizable systems
in the category GC h . If the one-level inverse limit system lim1←h {S, ts , T } exists, then
the limit system lim1←h {S, ts , T } is also centralizable.

Proof: This theorem follows from the fact that if Ct is a center of the system St , for
each t ∈ T, then the inverse limit lim←h Ct is a center of the one-level inverse limit
lim1←h {S, ts , T }. QED

Theorem 7.12. Suppose that {S, ts , T } is a linked α-type hierarchy of  systems in
the category GC c such that for some index t0 ∈ T , the state St0  Mt0 , Rt0 is
generally controllable on V’ relative to a mapping G : I Rt0 × E Rt0 → V . Then
the inverse limit lim←c {S, ts , T } is also generally controllable (but relative to a
different mapping G ∗ ).

Proof: From Theorem 6.5, it follows that we need only show that the mapping
αt0 : lim←c {S, ts , T } → St0 is S-continuous. Theorem 7.1 implies this. QED

Remarks on references: The construction of limit systems, were first given by Lin
(1989a, 1999), Ma and Lin (1988). Periodic and order structures of families of sys-
tems were studied intensively in Lin (1988b, 1989a).

References

Lin, Y.: Can the world be studied in the viewpoint of systems? Math. Comput. Model. 11, 738–742
(1988)
Lin, Y.: Order structures of families of general systems. Cybern. Syst. 20, 51–66 (1989)
Lin, Y.: General Systems Theory: A Mathematical Approach. Kluwer Academic and Plenum Pub-
lishers, New York (1999)
Ma, Y.H., Lin, Y.: Limit properties of hierarchies of general systems. Appl. Math. Lett. 1, 157–160
(1988)
Chapter 8
Systems of Single Relations

As studied in the previous chapters, a (general) system consists of a set of objects and
a set of relations between the objects. However, in practice, many systems, which can
be described with only one relation, have many complicated structures and properties
that we still do not completely understand. To address such concerns, this chapter
concentrates two fascinating properties—chaos and attractors—of single-relation
systems. In particular, Section 8.1 studies the concept of chaos and related properties
and characterizations. And Section 8.2 considers the concept of attractors and its
characterizations.

8.1 Chaos

A system S = (M, {r}) is called an input–output system if there are nonzero ordinal
numbers n and m such that

∅  r ⊂ M n × M m . (8.1)

Without loss of generality, we let X  M n and Y  M m ; and instead of using the

ordered pair (M, {r}), we will think of S as the binary relation r such that

∅  S ⊂ X × Y , (8.2)

where the sets X and Y are called the input space and the output space of S, respec-
tively. In the real world, most systems we see are input–output systems. For example,
each human being and each factory are input–output systems.
If we let Z  X ∪ Y , the input–output system S in equation (8.2) is a binary
relation on Z. Let D ⊂ Z be an arbitrary subset. If D2 ∩ S  ∅, then D is called a
chaos of S. Intuitively, the reason why D is known as a chaos is because S has no
control over the elements in D.
© Springer Nature Switzerland AG 2018 103
J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_8
104 8 Systems of Single Relations

Theorem 8.1. A necessary and sufficient condition under which an input–output

system S over a set Z has a chaotic subset D  ∅, is

S is not a subset of I , (8.3)

where I is the diagonal of the set Z defined by I  {(x, x): x ∈ Z}.

Proof: Necessity. Suppose that the input–output system S over the set Z has a
nonempty chaos D. Then D2 ∩ S  ∅. Therefore, for each d ∈ D, (d , d ) ∈ / S.
This implies that S ⊂ I .
Sufficiency. Suppose that equation (8.3) holds true. Then there exists d ∈ Z such
/ S. Let D  {d }. Then the nonempty subset D is a chaos of S. QED
that (d , d ) ∈

Let S ⊂ Z 2 and define

S ∗ D  {x ∈ Z: ∀y ∈ D, (x, y) ∈
/ S}, (8.4)

where D ⊂ Z.

Theorem 8.2. Let D ⊂ Z. Then D is a chaos of the system S if and only if D ⊂ S ∗D.

Proof: Necessity. Suppose D is a chaos of S. Then for each object d ∈ D and any
y ∈ D,

(d , y) ∈
/ S. (8.5)

Therefore, d ∈ S ∗ D. That is, D ⊂ S ∗ D.

Sufficiency. Suppose D satisfies D ⊂ S ∗ D. Then D2 ∩ S  ∅. Therefore, D is a
chaos of S. QED

Let COS(S) denote the set of all chaotic subsets on Z of the input–output system
S. If S ⊂ I , then

SI  {x ∈ Z: (x, x) ∈ S}  Z. (8.6)

We denote the complement of SI by SI−  Z − SI . Then Theorem 8.1 implies that

each chaos of S is a certain subset of SI− . Therefore, we have

|COS(S)| ≤ 2|SI | .
−
(8.7)

The following theorem is concerned with determining how many chaotic subsets
an input–output system has.

Theorem 8.3. Suppose that an input–output system S ⊂ Z 2 satisfies the following

conditions:
(i) S is symmetric, i.e., (x, y) ∈ S implies (y, x) ∈ S,
8.1 Chaos 105

(ii)S is not a subset of I.

(iii) S −   m < ω0 , and
I
(iv) ∀x ∈ SI− , there exists at most one y ∈ SI− such that either (x, y) ∈ S or
(y, x) ∈ S.
 
Let n   x ∈ SI− : ∃y ∈ SI− ((x, y) ∈ S) . Then
 k
3
|COS(S)| ≤ 2m × , (8.8)
4

where k  n/2.

Proof: According to the discussion prior to the theorem, it is known that every
chaotic subset of Z is a subset of SI− . The total number of all subsets of SI− is 2m .
Under the assumption of this theorem, a subset of SI− is a chaos of S if the subset
does not contain two elements of SI− which have S-relations. Among the elements of
SI− , there are n/2 pairs of elements with S-relations, and every two pairs do not have
common elements. Let {x, y} be a pair of elements in SI− with an S-relation, then
every subset of SI− − {x, y} ∪ {x, y} forms a subset of SI− not belonging to COS(S).
There are a total of 2m−2 of this kind of subsets. There are n/2 pairs of elements in
SI− with S-relations; we subtract these (n/2) × 2m−2 subsets from the 2m subsets of
SI− . But, some of the (n/2) × 2m−2 subsets then are in fact the same ones. Naturally,
every subset of SI− , containing two pairs
 of elements in SI− with S-relations, has
k
been subtracted twice. There are × 2m−4 such sets. We add back the number
2
of subsets which have been subtracted twice and obtain
 
k
2 −k ×2
m m−2
+ × 2m−4 . (8.9)
2
 
k
However, when we add × 2m−4 subsets, the subsets of SI− containing three
2
pairs of elements of SI− with S-relations have been added once more than they should
have. Therefore, this number should be subtracted. Continuing this process, we have
   
k k
|COS(S)|  2m − k × 2m−2 + × 2m−4 − × 2m−6 + · · ·
2 3
 
k
k
 × 2m−2i × (−1)i
i0
i
 
k
k
2 m−n
× 2n−2i × (−1)i
i0
i
106 8 Systems of Single Relations
 
k
i k
2 m−n
(−1) × 22k−2i
i0
i
 
k
i k
2 m−n
(−1) × 4k−i
i0
i
 2m−n (4 − 1)k
 2m−n × 3k
 2m × (3/4)k . QED

By using the same method, the constraint in Theorem 8.3 that the system S is
symmetric can be relaxed.
Theorem 8.4. Suppose that an input–output
 system S ⊂ Z 2 satisfies conditions


(ii)–(iv) of Theorem 8.3. Let n  x ∈ SI : ∃y ∈ SI− ((x, y) ∈ S or(y, x) ∈ S) , and
−

k  n/2. Then
 k
3
|COS(S)|  2m × . (8.10)
4

8.2 Attractors

Let S be an input–output system on a set Z. A subset D ⊂ Z is called an attractor of S

if for each x ∈ Z − D, S(x) ∩ D  ∅, where S(x)  {y ∈ Z: (x, y) ∈ S}. Figures 8.1
and 8.2 show the geometric meaning of the concept of attractors. When S is not a
function, Figure 8.1 shows that the graph of S outside the vertical bar D × Z overlaps
the horizontal bar Z × D. When S is a function, Figure 8.2 shows that the graph of
S outside the vertical bar D × Z must be contained in the horizontal region Z × D.
Theorem 8.5. Suppose that S ⊂ Z 2 is an input–output system over the set Z and
D ⊂ Z. Then D is an attractor of S if and only if D ⊃ S ∗ D.
Proof: Necessity. Suppose that D is an attractor of S. Let d ∈ S ∗ D be an arbitrary
element. From equation (8.4), we have that for every x ∈ D, (d , x) ∈
/ S. Therefore,
d ∈ D. That is, S ∗ D ⊂ D.

Figure 8.1 D is an attractor

of S, and S(x) contains at
least one element for each x
in Z−D
8.2 Attractors 107

Figure 8.2 D is an attractor

of S, and S is a function from
Z to Z

Sufficiency. Suppose that D satisfies D ⊃ S ∗ D. It follows from equation (8.4)

that for each object x ∈ Z−D, x ∈ / S ∗ D; so there exists at least one object d ∈ D
such that (d , x) ∈ S. That is, S(d ) ∩ D  ∅. This implies that D is an attractor of S.
QED
Theorem 8.6. A necessary and sufficient condition under which an input–output
system S ⊂ Z 2 has an attractor D ⊂ Z, satisfying D  Z, is that S is not a subset of
I.
Proof: Necessity. Suppose S has an attractor D such that D ⊂ Z, satisfying D  Z.
By contradiction, suppose S ⊂ I . Then for each D ⊂ Z, Z−D ⊂ S ∗ D. Therefore,
when Z−D  ∅, D ⊂ S ∗ D; i.e., S, according to Theorem 8.5, does not have any
attractor not equal to Z. This tells that the hypothesis that S has an attractor D such
that D ⊂ Z such that D  Z implies that S ⊂ I .
Sufficiency. Suppose S ⊂ I . Choose distinct objects x, y ∈ Z such that (x, y) ∈ S
and define D  Z − {x}. Then, according to equation (8.4), we have D ⊃ S ∗ D.
Applying Theorem 8.5, it follows that S has attractors. QED
Theorem 8.7. Suppose that an input–output system S ⊂ Z 2 satisfies the condition
that if (x, y) ∈ S then, for any z ∈ Z, (y, z) ∈
/ S. Let

SI −  {x ∈ Z: ∃y ∈ Z(y  x&(x, y) ∈ S)}, (8.11)

n  |SI − |, and ATR(S) be the set of all attractors of the system S. Then

|ATR(S)|  2n . (8.12)

Proof: For each subset A ⊂ SI − , let DA  Z−A. Then DA ⊃ S ∗ DA . In fact, for

each object d ∈ S ∗ DA , (d , x) ∈
/ S for every x ∈ DA . Therefore, d ∈ / A. Otherwise,
there exists an element y ∈ Z such that y  d and (d , y) ∈ S. The hypothesis,
stating that if (x, y) ∈ S then for any z ∈ Z, (y, z) ∈ / S, implies that the element
y ∈ Z−A, contradiction; so d ∈ Z−A  DA . This says that DA ⊃ S ∗ DA . Applying
Theorem 8.5, it follows that the subset DA is an attractor of S. Now it can be seen that
there are 2n many such subsets DA in Z. This completes the proof of the theorem.
QED
A subset D ⊂ Z is a strange attractor of an input–output system S ⊂ Z 2 if D is
both a chaos and an attractor of S. Figure 8.3 shows the case when a subset D of Z is
108 8 Systems of Single Relations

Figure 8.3 D is a strange

attractor of the system S

a strange attractor of an input–output system S, where the square D × D is the only

portion of the band Z × D over which the graph of S does not step.

Theorem 8.8. Let D ⊂ Z. Then D is a strange attractor of the input–output system

S if and only if D  S ∗ D.

The proof follows from Theorems 8.2 and 8.5.

Theorem 8.9. Suppose that S ⊂ Z 2 is an input–output system and D ⊂ Z. Let us

denote

S ◦ D  S −1 (D)  {x ∈ Z: ∃y ∈ D((x, y) ∈ S)}. (8.13)

Then the following statements hold:

(i) D is a chaos of S if and only if D ⊂ Z−S ◦ D.
(ii) D is an attractor of S if and only if D ⊃ Z−S ◦ D.
(iii) D is a strange attractor of S if and only if D  Z−S ◦ D.

Proof: It suffices to show that S ∗ D  Z−S ◦ D. Let x ∈ S ∗ D be an arbitrary

element. Then for any element y ∈ D, (x, y) ∈ / D. Thus, x ∈
/ S ◦ D; i.e., x ∈ Z−S ◦ D.
If an element x ∈
/ S ∗ D, there exists at least one element y ∈ D such that (x, y) ∈ S.
This implies that x ∈ S ◦ D; therefore, x ∈ / Z−S ◦ D. This completes the proof that
S ∗ D  Z−S ◦ D. The rest of the proof of the theorem follows from Theorems 8.2,
8.5, and 8.8. QED

We conclude this section with a discussion of the existence of strange attractors

of input–output systems.

Theorem 8.10. Suppose that S ⊂ Z 2 is a symmetric input–output system. Then a

necessary and sufficient condition under which there exists a strange attractor of the
system S is that there exists a subset D ⊂ Z such that D2 ∩ S  ∅ and for each
x ∈ SI , D ∩ S(x)  ∅.

Proof: Necessity. Suppose S has a strange attractor D. From the definition, it follows
that D2 ∩ S  ∅. At the same time, if there exists an element x ∈ SI satisfying
D ∩ S(x)  ∅, then x ∈ S ∗ D. By Theorem 8.8, x ∈ D. However, x ∈ SI , we must
/ S ∗ {x}, which implies that x ∈
have x ∈ / S ∗ D. This leads to a contradiction.
8.2 Attractors 109

Sufficiency. Suppose there exists D ⊂ Z such that D2 ∩ S(x)  ∅ and for each
x ∈ SI , D ∩ S(x)  ∅. Let A be the collection of subsets of Z defined by
 
A  A ⊂ Z: A ⊃ D and A2 ∩ S  ∅ . (8.14)

 Since D∈ A, A is not empty. It is clear thatfor each linearly

 ordered subset of
Aξ : ξ < α in A, ordered by inclusion, A  ∪ Aξ : ξ < α is also contained in A.
Therefore, A contains all maximal elements under the order relation of inclusion.
Let B be a maximal element in A. By Theorem 8.2, it follows that B ⊂ S ∗ B. If
B  S ∗ B, pick an element x ∈ S ∗ B−B and define

B∗  B ∪ {x}. (8.15)

According to the definition of the element x, for each y ∈ B, (x, y) ∈

/ S. From the
hypothesis that S is symmetric, it follows that (y, x) ∈/ S for every y ∈ B. Besides,
since for each z ∈ SI , A ∩ S(z)  ∅, for each A ∈ A, we have SI ∩ S ∗ B  ∅; i.e.,
x∈/ SI . Therefore, B∗2 ∩ SI  ∅. This contradicts the assumption that B is a maximal
element in A. QED
We next discuss the conditions under which a composition of input–output systems
has strange attractors.
Theorem 8.11. Suppose that f and g are two input–output systems on a set Z such
that g ◦ f  f −1 ◦ g −1 . Then a necessary and sufficient condition under which the
composition system g ◦ f has a strange attractor is that there exists a subset D ⊂ Z
such that the following conditions hold:
(i) (D × f ◦ D) ∩ g  ∅.

(ii) For every x ∈ (g ◦ f )I , f ◦ D is not a subset of Z 2 −g (x).
Proof: Applying Theorem 8.10, the composition system g ◦ f has a strange attractor
if and only if there is a subset D ⊂ Z such that
(i ) Z 2 ∩ g ◦ f  ∅.
(ii ) For each x ∈ (g ◦ f )I , D ∩ (g ◦ f )(x)  ∅.
It now suffices to show that (a) conditions (i) and (i ) are equivalent and (b)
conditions (ii) and (ii ) are equivalent.
Proof of (a). Condition (i ) is true if and only if for any x, y ∈ D, (x, y) ∈
/ g ◦ f if
and only if there does not exist a z ∈ Z such that (x, z) ∈ g and (z, y) ∈ f if and only
if for any x ∈ D and any z ∈ f ◦ D, (x, z) ∈ / g if and only if (D × f ◦ D) ∩ g  ∅;
i.e., (i) holds.
Proof of (b). Condition (ii ) holds if and only if for each x ∈ (g ◦ f )I there exists
d ∈ D such that (x, d ) ∈ g ◦ f if and only if for each x ∈ (g ◦ f )I there exist d ∈ D
and z ∈ Z such that (x, z) ∈ g and (z, y) ∈ f if and only if for each x ∈ (g ◦ f )I there
exists z ∈ f ◦ D such that (x, z) ∈ g if and only if for each x ∈ (g ◦ f )I there exists
z ∈ f ◦ D
such that
 (x, z) ∈ / Z−g; if and only if for each x ∈ (g ◦ f )I , f ◦ D is not a
subset of Z 2 − g (x); i.e., (ii) is true. QED
110 8 Systems of Single Relations

The concept of general systems of single relations was originated by Mesarovic,

who began to study this concept in 1964 (Mesarovic 1964). Its final form was con-
tained in the book by Mesarovic and Takahara (1975).
The concepts of chaos and attractors of single-relation systems were initially
introduced by Zhu and Wu (1987). They named the concepts as panchaos and panat-
tractor and systematically studied them under the name pansystems theory. That
theory appeared in the late 1970s. For more information, consult Wu’s work (1981,
1982a, b, c, 1983a, b, 1984a, b, c, d, 1985).

References

Lin, Y.: General Systems Theory: A Mathematical Approach. Kluwer Academic and Plenum Pub-
lishers, New York (1999)
Mesarovic, M.D.: Views on general systems theory. In: Mesarovic, M.D. (ed.) Proceedings of the
2nd Systems Symposium at Case Institute of Technology. Wiley, New York (1964)
Mesarovic, M.D., Takahara, Y.: General Systems Theory: Mathematical Foundations. Academic
Press, New York (1975)
Wu, X.M.: Pansystems analysis—a new exploration of interdisciplinary investigation (with appen-
dices on discussions on medicine, chemistry, logic and economics). Sci. Explor. 1(1), 125–164
(1981)
Wu, X.M.: Pansystems methodology: concepts, theorems and applications I. Sci. Explor. 2(1),
33–56 (1982a)
Wu, X.M.: Pansystems methodology: concepts, theorems and applications II. Sci. Explor. 2(2),
93–106 (1982b)
Wu, X.M.: Pansystems methodology: concepts, theorems and applications III. Sci. Explor. 2(4),
123–132 (1982c)
Wu, X.M.: Pansystems methodology: concepts, theorems and applications IV. Sci. Explor. 3(1),
125–137 (1983a)
Wu, X.M.: Pansystems methodology: concepts, theorems and applications V. Sci. Explor. 3(4),
97–104 (1983b)
Wu, X.M.: Pansystems methodology: concepts, theorems and applications VI. Sci. Explor. 4(1),
107–116 (1984a)
Wu, X.M.: Pansystems methodology: concepts, theorems and applications VII. Sci. Explor. 4(4),
1–14 (1984b)
Wu, X.M.: Investigation and applications of pansystems recognition theory and pansystems oper-
ations research of large-scale systems. Appl. Math. Mech. 5(1), 995–1010 (1984c)
Wu, X.M.: Approximation Transformation Theory and the Pansystems Concepts in Mathematics.
Hunan Science and Technology Press (1984d)
Wu, X.M.: Pansystems methodology: concepts, theorems and applications VIII. Sci. Explor. 5(1),
9–23 (1985)
Zhu, X.-D., Wu, X.-M.: An approach on fixed pansystems theorems: panchaos and strange panat-
tractor. Appl. Math. Mech. 8(4), 339–344 (1987)
Chapter 9
The Feedback Mechanism

Single-relation systems have seen practical applications in a wide range of scientific

areas. In this chapter, we focus on general linear systems, the concept of feedback,
and characterizations of feedback systems.

9.1 Linear Systems

Let A be a field, X and Y linear spaces over A, and S an input–output system such
that
(i) ∅  S ⊂ X × Y .
(ii) s ∈ S and s ∈ S imply s + s ∈ S.
(iii) s ∈ S and α ∈ A imply α · s ∈ S.
Here, + and · are addition and scalar multiplication in X × Y , respectively, and
defined as follows: For any (x1 , y1 ), (x2 , y2 ) ∈ X × Y and any α ∈ A,

(x1 , y1 ) + (x2 , y2 )  (x1 + x2 , y1 + y2 ), (9.1)

α(x1 , y1 )  (αx1 , αy1 ). (9.2)

The input–output system S is then called a linear system.

Theorem 9.1. Suppose that X and Y are linear spaces over the same field A. Then
the following statements are equivalent:
(1) S ⊂ X × Y is a linear system.
(2) There exist a linear space C over A and a linear mapping ρ: C × D(S) → Y
such that (x, y) ∈ S if and only if there exists a c ∈ C such that ρ(c, x)  y.
(3) There exist a linear space C over A and linear mappings R1 : C → Y and
R2 : D(S) → Y such that the mapping ρ in (2) is such that ρ(c, x)  R1 (c) +
R2 (x), for every (c, x) ∈ C × D(S).
© Springer Nature Switzerland AG 2018 111
J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_9
112 9 The Feedback Mechanism

Here, D(S) is the domain of the input–output system S.

Proof: First, the domain D(S) is a linear subspace of the space X. For any x1 , x2 ∈
D(S) and any α ∈ A, there exist y1 , y2 ∈ Y such that (x1 , y1 ), (x2 , y2 ) ∈ S. From the
hypothesis that S is linear, it follows that

(x1 + x2 , y1 + y2 )  (x1 , y1 ) + (x2 , y2 ) ∈ S, (9.3)

and

α(x1 , y1 )  (αx1 , αy1 ) ∈ S, (9.4)

and thus x1 + x2 and αx1 ∈ D(S).

The proofs that (3) implies (1) and (2) implies (3) are clear because we can define
the desired mappings R1 and R2 by R1 (c)  ρ(c, 0x ) and R2 (x)  ρ(0C , x), where
0x and 0C are the additive identities in the linear spaces X and C, respectively.
Proof that (1) implies (2). Suppose S is a linear system and 0x and 0y are the
additive identities in X and Y , respectively. Define

C  {b ∈ Y : (0x , b) ∈ S}. (9.5)

 
Then C  ∅; in fact, 0y ∈ C because 0x , 0y ∈ S, and C is a linear subspace over
A. Consider the quotient space Y /C  {b + C: b ∈ Y } with addition +∗ and scalar
multiplication ·∗ defined by (b1 + C) +∗ (b2 + C)  b1 + b2 + C and α ·∗ (b1 + C) 
αb1 + C, for any α ∈ A and any (b1 + C) and (b2 + C) ∈ Y /C.
Define a mapping σ : D(S) → Y /C by letting

σ (x)  y + C, (9.6)

if (x, y) ∈ S for every x ∈ D(S). Then σ is well defined. In fact, for each x ∈ D(S) and
any y1 , y2 ∈ Y satisfying (x, y1 ), (x, y2 ) ∈ S, (x, y1 ) − (x, y2 )  (0x , y1 − y2 ) ∈ S,
and thus y1 − y2 ∈ C, so y1 + C  y2 + C. This means that σ (x)  y1 + C  y2 + C.
Define a linear mapping π : C × D(S) → Y /C ⊕ C as follows:

π (c, x)  (σ (x), c). (9.7)

We now choose an isomorphism I : Y /C ⊕C → Y satisfying I (y, C)  y for every

y ∈ Y /C. Now, the desired mapping ρ: C × D(S) → Y can be defined by ρ  I ◦ π .
Then (x, y) ∈ S if and only if there exists c ∈ C such that ρ(c, x)  y. In fact, if
(x, y) ∈ S, there exists c ∈ C such that I (σ (x), c)  y, so ρ(c, x)  I ◦ π (c, x)  y.
On the other hand, if (x, y) ∈ X × Y is such that there exists c ∈ C with ρ(c, x)  y,
then ρ(c, x)  I ◦ π (c, x)  I (σ (x), c)  y, so y ∈ σ (x). Suppose σ (x)  y1 + C.
Then y  y1 + c1 for some c1 ∈ C; hence, we have

(x, y)  (x, y1 + c1 )  (x, y1 ) + (0x , c1 ) ∈ S. (9.8)

9.1 Linear Systems 113

This completes the proof of the theorem. QED

Example 9.1. Three examples are given here to show that the concept of linear
systems is an abstraction of different mathematical structures.

(a) Suppose S is a system described by the following linear equations:

a11 x1 + a12 x2 + · · · + a1n xn  b1

a21 x1 + a22 x2 + · · · + a2n xn  b2
......
am1 x1 + am2 x2 + · · · + amn xn  bm

where xi , i  1, 2, . . ., n, are n variables, m is the number of equations, aij , (i 

1, 2, . . ., m; j  1, 2, . . ., n) are the coefficients of the system, and bj , j 
1, 2, . . ., m, are the constraints of the system. The system S can then be rewritten as
a linear system (M , {r}), where the object set M is the set R of all real numbers, and

r  (x1 , x2 , . . . , xn ) ∈ M n : ai1 (x1 + y1 ) + ai2 (x2 + y2 )
+ · · · + ain (xn + yn )  bi , i  1, 2, ..., m}

where (y1 , y2 , . . . , yn ) ∈ M n is fixed such that ai1 y1 + ai2 y2 + · · · + ain yn  bi , for

each i  1, 2, . . . , m.

(b) Suppose that a given system S is described by the differential equations

d nx d n−1 x dx
+ a1 (t) + · · · + an−1 (t) + an (t)x  f (t)
dt n dt n−1 dt
where ai (t), i  1, 2, . . . , n, and f (t) are continuous functions defined on the interval
[a, b]. The system S can be rewritten as a linear system (M , {r}) in the following
way: The object set M is the set of all continuous functions defined on [a, b], and
the relation r is defined by

d n (x + y) d n−1 (x + y)
r  x ∈ M: n
+ a1 (t)
dt dt n−1

d (x + y)
+ · · · + an−1 (t) + an (t)(x + y)  f (t)
dt

where y ∈ M is a fixed function such that

d ny d n−1 y dy
+ a1 (t) + · · · + an−1 (t) + an (t)y  f (t).
dt n dt n−1 dt
114 9 The Feedback Mechanism

Figure 9.1 Structure of a

feedback system

(c) Suppose X is a linear space over a field A, and Y a set of some linear trans-
formations on X. Then Y can be studied as a linear system (X , R), where, for
each relation r ∈ R, there exists a transformation y ∈ Y such that for any
(x1 , x2 ) ∈ X 2 , (x1 , x2 ) ∈ r if and only if y(x1 )  x2 .
In Example 9.1(b), S is not an input–output linear system. Generally, the concept
of general linear systems is the following: Let A be a field and S  (M , R) a system
such that the object set M is a linear space over A. Then S is a linear system if for each
relation r ∈ R there exists an ordinal number n  n(r) such that r is a linear subspace
of M n . We will study only input–output linear systems defined in the beginning of
this chapter.

9.2 Feedback Systems

An input–output system S ⊂ X × Y is a functional system if S is a function from the

input space X into the output space Y . Let S ⊂ X × Y and Sf : Y → X be a linear
system and a linear functional system, respectively. Then the feedback system of S
by Sf is defined as the input–output system S  such that
 
(x, y) ∈ S  ↔ (∃z ∈ X ) (x + z, y) ∈ S and (y, z) ∈ Sf . (9.9)

The systems S and Sf are called an original system and a feedback component
system, respectively.
Let

S  {S ⊂ X × Y : S is a linear system}, (9.10)

Sf  {Sf : Y → X : Sf is a linear functional system}, (9.11)

and

S   {S  ⊂ X × Y : S  is a subset}. (9.12)

Then a feedback transformation F: S × Sf → S  is defined by equation (9.9),

which is called the feedback transformation over the linear spaces X and Y . Figure 9.1
shows the geometric meaning of the concept of feedback systems.
The following theorem gives a concrete structure of each feedback system.
9.2 Feedback Systems 115

Theorem 9.2. Suppose that S ⊂ X × Y is a linear system and Sf : Y → X a linear

functional system. Then a subset S  ⊂ X × Y is the feedback system of S by Sf , if
and only if
  
S  x − Sf (y), y : (x, y) ∈ S . (9.13)

Proof: Necessity. Suppose that S  is the feedback system of S by Sf . Then, by

the definition of feedback systems, for each (x, y) ∈ S  , (x + Sf (y), y) ∈ S. This
implies that (x, y) ∈ {(z − Sf (y), y): (z, y) ∈ S}. Again,
 the relation
 in equa-

tion
 (9.9) implies
 that for
 each element (x, y) ∈ S, x−Sf (y), y ∈ S . That is,
z−Sf (y), y : (z, y) ∈ S ⊂ S  .
Sufficiency. It is clear because equation (9.13) implies the relation in equa-
tion (9.9). QED
Theorem 9.3. Suppose that S, S  ⊂ X × Y are linear systems and Sf : Y → X is a
linear functional system such that
 
F S, Sf  S  . (9.14)
 
Then F S  , −Sf  S, where −Sf : Y → X is the linear functional system defined
by
 
−Sf (y)  −Sf (y), (9.15)

for every y ∈ Y .
The proof is straightforward and omitted.
Theorem 9.4. For each linearsystem S ⊂ X × Y , there exists a linear functional
system Sf : Y → X such that F S, Sf  S; that is, S is a feedback system of itself.
Proof: The desired linear functional system Sf : Y → X can be defined by

Sf (y)  0x , (9.16)
 
for every y ∈ Y . Then, equation (9.13) says that S  F S, Sf . QED
Theorem 9.5 discusses the following control problem: A linear structure (say,
system) S ⊂ X × Y is fixed. According to some predefined targets, we need to
produce a new linear structure (say, a linear subspace)
 S  ⊂ X × Y . Can we design

a feedback process Sf : Y → X such that F S, Sf  S ?
Theorem 9.5. Assume the axiom of choice. For each linear system S ⊂ X × Y

and arbitrary linear subspace
 S ⊂ X × Y , there exists a linear functional system
Sf : Y → X such that F S, Sf  S if and only if R(S )  R(S) and N(S  )  N(S),
where N(W ) indicates the null space of the input–output system W ⊂ X × Y , defined
by
116 9 The Feedback Mechanism
   
N (W )  x ∈ X : x, 0y ∈ W . (9.17)

Proof: Necessity.
  Suppose that there exists a linear functional system Sf : Y → X
such that F S, Sf  S  . Then Theorem 9.2 implies that R(S )  R(S) and that
     
N S   x ∈ X : x, 0y ∈ S 
     
 x ∈ X : x + Sf 0y , 0y ∈ S
   
 x ∈ X : x + 0x , 0y ∈ S
 N (S).

Sufficiency. Suppose that S and S  have the same output space and the null space;
i.e., R(S  ) R(S) and
 N(S

)  N(S). We define a linear functional system Sf : Y → X

such that F S, Sf  S .
 
For each element y ∈ R S  , let
   
Xy  x ∈ X : x , y ∈ S  (9.18)

and

Xy  {x ∈ X : (x, y) ∈ S}. (9.19)

From the Axiom of Choice, it follows that there exist mappings

  
C  : Xy : y ∈ R S  → X (9.20)

and
 
C: Xy : y ∈ R(S) → X , (9.21)

    
such that C  Xy  C  (y) ∈ Xy , i.e., C  (y), y ∈ S  , and C Xy  C(y) ∈ Xy , i.e.,
(C(y), y) ∈ S. We are now ready to check that for each y ∈ R(S  )  R(S),

Xy  Xy . (9.22)

Thus, the mappings C and C  are well defined.

Let {yi : i ∈ I } be a basis in the linear space R(S  ), where I is a finite or infinite
index set. We now define a linear functional system

Sf : Y → X , (9.23)

satisfying Sf (yi )  C(yi ) − C  (yi ), for each i ∈ I. Then we make the following claim.
 
Claim. The systems S, S  , and Sf have the property that F S, Sf  S  .
9.2 Feedback Systems 117
 
In fact, for each element (x, yi ) ∈ S  , for an i ∈ I, we have x + Sf (yi ), yi ∈ S,
since
   
x + Sf (yi ), yi  x + C(yi ) − C  (yi ), yi
     
 x − C  (yi ), 0y + (C(yi ), yi ) ∈ N S  × 0y + S
 
 N (S) × 0y + S.
 S. (9.24)

Therefore, from equation (9.24), it follows that

 
(x, yi ) ∈ S  → x + Sf (yi ), yi ∈ S. (9.25)
 
Now, let (x, yi ) ∈ S; we then have x − Sf (yi ), yi ∈ S  , because
   
x − Sf (yi ), yi  x − C(yi ) + C  (yi ), yi
     
 x − C(yi ), 0y + C  (yi ), yi ∈ N (S) × 0y + S 
   
 N S  × 0y + S 
 S . (9.26)

Hence, we have from equation (9.26) the opposite implication of equation (9.25).
That is, for every i ∈ I,
 
(x, yi ) ∈ S  ↔ x + Sf (yi ), yi ∈ S. (9.27)

It now remains to show that for each pair (x, y) ∈ X × Y , the relation in equa-
tion (9.9) holds true. Let (x, y) be an arbitrary element in S  . Then y  i ci yi for
a finite number of yi ’s fromthe basis {yi : i ∈ I }, where each ci is a nonzero number.
Then we have x + Sf (y), y ∈ S. In fact,
 
  
x + Sf (y), y  x + Sf ci yi , y
i

   
 x+ ci C(yi ) − C  (yi ) , ci yi
i i
 
  

 x− ci C (yi ), 0y + ci C(yi ), ci yi
i i i
 
     
 (x, y) − ci C  (yi ), y + ci (C(yi ), yi ) ∈ N S  × 0y + S
i i
 
 N (S) × 0y + S  S. (9.28)
This gives us the implication from the right to left in equation
 (9.9). Theopposite
implication can be shown as follows. For each pair (x, y) ∈ S, x − Sf (y), y ∈ S  . In
118 9 The Feedback Mechanism

Figure 9.2 Structure of the

cascade connection of
systems


fact, let y  i ci yi , for some finite number of nonzero coefficients ci , where each
yi is from the basis {yi : i ∈ I }. Then, we have

  
x − Sf (y), y  x − ci Sf (yi ), y
i

   
 x− ci C(yi ) − C  (yi ) , ci yi
i i
 
    

 x− ci C(yi ), 0y + ci C (yi ), ci yi ∈ N (S) × 0y + S 
i i i
   
 N S  × 0y + S   S  . (9.29)
This completes the proof of Theorem 9.5. QED

We conclude this chapter with a discussion of a more general definition of the

concept of feedback systems. Let * be a system connecting operator, termed the
cascade (connecting) operation. The operation * is defined as follows: Let Si , i 
1, 2, 3, be input–output systems such that
 
S1 ⊂ X1 × Y1∗ × Z , (9.30)
 
S2 ⊂ X2∗ × Z × Y2 , (9.31)

and
   
S3 ⊂ X1 × X2∗ × Y1∗ × Y2 , (9.32)

where it is assumed that no algebraic properties are given on each set involved. The
three systems satisfy S1 ∗ S2  S3 if and only if

((x1 , x2 ), (y1 , y2 )) ∈ S3
↔ ∃z((x1 , (y1 , z)) ∈ S1 and ((x2 , z), y2 ) ∈ S2 ) (9.33)

Figure 9.2 shows the geometric meaning of the cascade connection of systems.
We now let S and Sf be two input–output systems such that
 
S ⊂ (X × Zx ) × Y × Zy , (9.34)

and
9.2 Feedback Systems 119

Figure 9.3 Construction of

general feedback systems

Sf ⊂ Zy × Zx , (9.35)

where Sf is called a feedback component of S. The relationship between the two

systems is displayed in Figure
 9.3.

The feedback system F S, Sf of S by Sf is defined by
       
(x, y) ∈ F S, Sf ↔ ∃ zy , zx ∈ Sf (x, zx ), y, zy ∈ S . (9.36)
 
When constructing F S, Sf step by step, we first cascade S and Sf (i.e., find
S ∗ Sf ) and then input one of the two outputs of the system S ∗ Sf .
Without special explanation in this book, by feedback system we always mean
the definition given by equation (9.9).
The concept of general linear systems was first studied in detail by Mesarovic and
Takahara (1975). Theorem 9.1 developed in this chapter is a combination of a result
in (Lin 1989) and a fundamental theorem given by Mesarovic and Takahara (1975).
The concept of feedback systems has been a major research topic in systems theory
and studied in many different ways by, for example, Mesarovic and Takahara (1975),
Bunge (1979), Rugh (1981), Deng (1985), Klir (1985), and Saito and Mesarovic
(1985). The definition of feedback systems, given in this chapter, was introduced by
Saito and Mesarovic (1985), and some results are from Lin and Ma (1990).

References

Bunge, B.: Treatise on Basic Philosophy. Vol. 4: A World of Systems. Reidel, Dordrecht, Holland
(1979)
Deng, J.-L.: Grey Control Systems. Hua-zhong Institute of Technology Press, Wuhan (1985)
Klir, G.: Architecture of Systems Problem Solving. Plenum Press, New York (1985)
Lin, Y.: Some properties of linear systems. Int. J. Syst. Sci. 20(6), 927–937 (1989)
Lin, Y., Ma, Y.H.: General feedback systems. Int. J. Gen. Syst. 18(2), 143–154 (1990)
Mesarovic, M.D., Takahara, Y.: General Systems Theory: Mathematical Foundations. Academic
Press, New York (1975)
Rugh, W.J.: Nonlinear systems theory: the Volterra/Wiener approach. Johns Hopkins University
Press, Baltimore (1981)
Saito, T., Mesarovic, M.D.: A meaning of the decoupling by feedback of linear functional time
systems. Int. J. Gen. Syst. 11(1), 47–61 (1985)
Chapter 10
Properties Invariant Under Feedback

To continue our investigation of single-relation systems, this chapter studies such

important concepts as MT-time systems, linear time systems, stationary systems, and
time invariably realizable systems, along with some feedback-invariant properties.

10.1 Linearity

Some properties of original systems may be kept by their feedback systems and
some may not. Those properties of original systems invariant under the feedback
transformation F are of special interest in research on feedback systems. Feedback
invariance has played a crucial role in characterizing homeostatic or explosive system
behaviors in various areas such as biology, engineering, social science, and so on.
A property of a linear system S ⊂ X × Y is called feedback invariant
 with respect
to Sf if the property also holds for the feedback system F S, Sf for each linear
functional system Sf ∈ Sf .

Corollary 10.1. Let S ⊂ X × Y and Sf : Y → X be a linear  system

 and a linear
functional system, respectively.
   Then the feedback
  system
 F S, Sf is a linear system
over X × Y and R F S, Sf  R(S) and N F S, Sf  N (S).
 
Proof: The linearity
  of the
 feedback system   F S,Sf follows from Theorem 9.2.
The equations R F S, Sf  R(S) and N F S, Sf  N (S) come from Theorem
9.5 here the Axiom of Choice is not used. QED

Theorem 10.1. Let S: X →  Sf : Y → X be linear functional

 Y and  systems
 such that
S is surjective. Then D F S, Sf  D(S) if and only if I −S ◦ Sf : Y → Y is
surjective, where I: Y → Y indicates the identity mapping; i.e., I(y)  y, for each
y ∈ Y.

© Springer Nature Switzerland AG 2018 121

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_10
122 10 Properties Invariant Under Feedback
  
Proof: Sufficiency. It is clear that D F S, Sf ⊂ D(S). Let
 x ∈ D(S) be arbitrary.
There then exists some y ∈ Y such that (x, y) ∈ S. Since I −S ◦ Sf : Y → Y is
surjective, there exists exactly one ŷ ∈ Y such that
  
I −S ◦ Sf ŷ  y  S(x). (10.1)

This implies that

  
S x + Sf ŷ  ŷ (10.2)
      
That is, x + Sf ŷ , ŷ ∈ S. Hence, x ∈ D F S, Sf .

Necessity. Pick an arbitrary y ∈ Y. From the hypothesis that S isa surjective

  system,
it follows that there exists some x ∈ X such that S(x)  y. Since
 D F
 S, Sf   D(S),
for this x there exists some ŷ ∈ Y such that (x, ŷ) ∈ F S, Sf ; i.e., x + Sf ŷ , ŷ ∈ S.
Hence,
  
ŷ  S x + Sf ŷ
 
 S(x) + S ◦ Sf ŷ
 
 y + S ◦ Sf ŷ . (10.3)
      
Thus, ŷ − S ◦ Sf ŷ  y, i.e., I −S ◦ Sf ŷ  y, This proves that I −S ◦ Sf :
Y → Y is surjective. QED

Theorem 10.2.  Let  S ⊂X × Y be a linear system and Sf : Y → X a linear functional

system. Then D F S, Sf  X if and only if for each x ∈ X there exists a y ∈ R(S)
such that (x + Sf (y), y) ∈ S.
The proof is clear from equation (9.9).

Theorem 10.3. Let S: X → Y be a linear

 functional
 system. Then for each arbitrarily
fixed Sf ∈ Sf , the feedback system F S, Sf is injective if and only if the original
system S is injective.

Proof: Sufficiency. Suppose that the original  S is injective, and let Sf ∈ Sf

 system
be arbitrary. Suppose
 the feedback
 system F S, Sf is not injective. Thus, there exist
distinct x1 , x2 ∈ D F S, Sf such that
 
(x1 , y), (x2 , y) ∈ F S, Sf , (10.4)
   
for some y ∈ R(S). By equation (9.9), we have x1 + Sf (y), y , x2 + Sf (y), y ∈ S.
Therefore, from the hypothesis that the system S: X → Y is injective, it follows that
 x1 x2 , contradiction.
x1 + Sf (y)  x2 + Sf (y); i.e.,
Necessity. Suppose F S, Sf is injective. From Theorem 10.3, we have
   
F F S, Sf , −Sf  S
10.1 Linearity 123

Applying the proof for sufficiency, we know that S is injective. QED

Combining what have been obtained in this chapter, the following theorem is
evident.

Theorem 10.4. Range space, null space, linearity, injectivity, surjectivity, and bijec-
tivity of original systems are feedback invariant with respect to Sf .
In the rest of this chapter, we study feedback-invariant properties of MT-time
systems. Let T be the time axis defined as the positive half of the real number line;
i.e., T  [0, +∞). Let A and B be two linear spaces over the same field A. We define

AT  {x : x is a mapping T → A} (10.5)

and

BT  {x : x is a mapping T → B}. (10.6)

Then the sets AT and BT can be made into linear spaces over A as follows: For
any elements f , g ∈ AT (respectively, ∈ BT ), and α ∈ A,

(f + g)(t)  f (t) + g(t) (10.7)

and

(αf )(t)  α · f (t), (10.8)

for each t ∈ T.

Assume the input space X and the output space Y in the preceding discussion are
linear subspaces of AT and BT , respectively; i.e., X ⊂ AT and Y ⊂ BT .

Proposition 10.1. For any linear system S ⊂ X × Y and any linear functional
system Sf : Y → X, the feedback system F S, Sf is a linear system over AT × BT .
The proof is straightforward and is omitted.

10.2 Mesarovic–Takahara Time Systems

Each input–output system S ⊂ BT AT × is termed an MT-time system, (Mesarovic—

Takahara time system). Therefore, Proposition 10.1 implies that the feedback trans-
formation defined in Chapter 9 is well defined on the class of all MT-time systems.
For any t, t  ∈ T with t < t  , let

T t  [ 0, t) , Tt  [ t, +∞) (10.9)

and
124 10 Properties Invariant Under Feedback

Figure 10.1 Graph of x in

the plane T × A

Ttt   [t, t  ), T̄ t  0, t
, T̄tt   [t, t  ]. (10.10)

Then the restrictions of a time function x ∈ X with respect to these time intervals
are denoted by

xt  xT t , xt  x|Tt , xtt   x|Ttt  (10.11)

and
 
x̄t  xT̄ t , x̄tt   xT̄tt  . (10.12)

For each y ∈ Y , or in general, for sets and vectors

 similar
 notation will be used;
for example, for each (x, y) ∈ X × Y , (x, y)t  xt , yt , (x, y)t  (xt , yt ), etc.
A subset S ⊂ X × Y ⊂ AT × BT is called a linear time system if S is a linear
subspace of X × Y and the input space D(S) satisfies the condition:
 
∀x, x ∈ X ∀t ∈ T x, x ∈ D(S) → xt ◦ xt ∈ D(S) , (10.13)

where xt ◦ xt is a time function ∈ AT , called the concatenation of xt and xt , and is
defined by the following: For any s ∈ T ,

 x(s), if s < t
x ◦ xt (s) 
t
. (10.14)
x (s), if s ≥ t

Without loss of generality, we will always assume that D(S)  X and R(S)  Y.
Let σ τ be the shift operator defined as follows: For each x ∈ X,

σ τ (x)(ξ )  x(ξ − τ ), (10.15)

for each ξ ∈ Tτ , where τ ∈ R, the set of all real numbers, can be any value if σ τ (x) is
meaningful. Figures 10.1, 10.2, and 10.3 show the geometric meaning of the concept
of shift operators. In Figures 10.2 and 10.3, the dotted A × T planes indicate the
locations of the A × T plane before the shift operation σ τ functions. Therefore, when
τ > 0, the application of σ τ implies that we move the graph of x τ units to the right;
when τ < 0, the application of σ τ tells us to consider the portion of the graph of x
on the right of the vertical line t  τ .
10.2 Mesarovic-Takahara Time Systems 125

Figure 10.2 Graph of σ τ (x)

with τ > 0

Figure 10.3 Graph of σ τ (x)

with τ < 0

A linear time system S ⊂ X × Y is called strongly stationary if S satisfies the

condition
 
∀t ∈ T σ −1 (S|Tt  S). (10.16)

Theorem 10.5. Let S ⊂ X × Y and Sf : Y → X be a linear time system and a

strongly stationary
  linear functional time system, respectively. Then the feedback
system F S, Sf is strongly stationary if and only if the original system S is strongly
stationary.
 
Proof: Necessity. Suppose that the feedback system F S, Sf is strongly stationary.
Then we have
   
S x + Sf (y), y : (x, y) ∈ F S, Sf . (10.17)

It suffices to show that for any (x, y) ∈ S, σ −1 ((x, y)|Tt ) ∈ S, for every t ∈ T. we
now pick an arbitrary (x, y) ∈ S and t ∈ T. Then we have
  
σ −1 ((x, y)|Tt )  σ −1 x − Sf (y) + Sf (y), y |Tt
       
 σ −1 x − Sf (y) Tt + σ −1 Sf (y)Tt , σ −1 (y|Tt ) . (10.18)
 
From the hypothesis that the systems F S, Sf and Sf are strongly stationary, it
follows that
 −1       
σ x − Sf (y) Tt , σ −1 (y|Tt ) ∈ F S, Sf , (10.19)

and
126 10 Properties Invariant Under Feedback
 −1   
σ (y|Tt ), σ −1 Sf (y)Tt ∈ Sf . (10.20)
  
Let x 
 σ −1 x − Sf (y) Tt and y  σ −1 (y|Tt ). Then, equation (10.20) shows
that Sf y  σ −1 Sf (y)|Tt . Therefore, we have
   
σ −1 ( (x, y)|Tt )  x + Sf y , y , (10.21)
   
where x , y ∈ F S, Sf so equation (10.21) implies that σ −1 ( (x, y)|Tt ) ∈ S.
Sufficiency. This follows from Theorem 10.3 and the necessity part of this proof.
QED
A linear time system S ⊂ X × Y is called precausal if it satisfies the condition:
   
(∀t ∈ T )(∀x ∈ X ) x̄t  0̄t → S(x)T̄ t  S(0)T̄ t , (10.22)

where S(x)  {y ∈ Y : (x, y) ∈ S}. A linear functional time system S: X → Y is

called causal if it is precausal.

Theorem 10.6. Let S ⊂ X ×  Y be a linear time

 system.Then
  S is precausal if and
only if (∀t ∈ T )(∀x, y ∈ X ) x̄t  ȳt → S(x)T̄ t  S(y)T̄ t .
t
Proof: Necessity. Suppose
 t S is precausal.
 t Then for any x, y ∈ X , (x − y)   0̄t
implies that S(x − y)T̄  S(0)T̄ . This shows that when x̄t  ȳt , S(x)T̄ t 

S(y)T̄ t because
t
 t (x − y)  0̄t is the same as x̄t − ȳt 
 0̄t . Hence, x̄ t  ȳ
t t t

and S(x − y)T̄   S(0)T̄ implies that S(x)T̄ − S(y)T̄  S(0)T̄ . So,
t t

S(x)T̄ t  S(y)T̄ t + S(0)T̄ t  S(y)T̄ t .  t  


Sufficiency. Suppose that (∀t ∈ T )(∀x, y ∈ X ) x̄t  ȳt → S(x) T̄  S(y)T̄ t .
 t  
This implies that (∀t ∈ T )(∀x ∈ X ) x̄  0̄t → S(x)T̄ t  S(0)T̄ t . QED

Theorem 10.7. Let S: X →Y andSf : Y → X be linear functional causal time systems.
Then the feedback
   F S,Sf is causal if and only if the composition time system
 system
S ◦ Sf ◦ F S, Sf : D F S, Sf → Y is causal.

Proof: The argument of the necessity part follows from the fact that each composi-
tion of causal functional time systems is causal.  
Sufficiency. Suppose that the composition system S ◦ Sf ◦ F S, Sf is causal. That
   t
is, for any t ∈ T and any x ∈ D F S, Sf , x̄  0̄t implies that
  
S ◦ Sf ◦ F S, Sf (x)T̄ t  0̄t . (10.23)
 
Let y  F S, Sf (x). We need to show that ȳt  0̄t . By equation (9.9), we have
 
S x + Sf (y)  y. (10.24)

Therefore,
 S(x) + S ◦ Sf (y) y, where   
S(x)T̄ t  0̄t and S ◦ Sf (y)T̄ t  S ◦ Sf ◦ F S, Sf (x)T̄ t  0̄t .
10.2 Mesarovic-Takahara Time Systems 127

Thus, we obtain
 
ȳt  S(x) + S ◦ Sf (y) T̄ t
 
 S(x)T̄ t + S ◦ Sf (y)T̄ t
 0̄t + 0̄t
 0̄t
 
This implies that the feedback system F S, Sf is causal. QED

10.3 Time-Invariable Realization

A linear functional time system S: X → Y is called time invariably realizable if it

satisfies the condition
 
∀t ∈ T ∀x ∈ X λt S(0t ◦ σ t (x)  S(x), (10.25)

where λt (·)  σ t (·|Tt ).

Theorem 10.8. A linear functional time system S: X → Y is time invariably realizable

if and only if for any t ∈ T and any x ∈ X
 
S 0t ◦ σ t (x) |Tt  σ t S(x). (10.26)

Proof: Necessity. Suppose S is time invariably realizable. Then S satisfies the rela-
tion in equation (10.25). It suffices to show that equation (10.26) is derivable from
the relation in equation (10.25). For any t ∈ T and any x ∈ X, let y  S(0t ◦ σ t (x).
Then, we have
    
λt S 0t ◦ σ t (x)  λt (y)  σ −t yT̄ t  S(x). (10.27)

The last equality of equation (10.27) shows that

  
σ t ◦ σ −t yT̄ t  y|Tt
 
 S 0t ◦ σ t (x) |Tt
 σ t S(x). (10.28)

Sufficiency. Suppose S satisfies equation (10.26) for any t ∈ T and any x ∈ X.

Then
 
σ −t S 0t ◦ σ t (x) T̄t  σ −t σ t S(x)  S(x). (10.29)
128 10 Properties Invariant Under Feedback

Comparing equation (10.29) and the relation in equation (10.25), we see that S is
time invariably realizable. QED

Theorem 10.9. Suppose that S: X → Y and Sf : Y → X are two linear functional

time systems
 such  that S and Sf are time invariably realizable. Then the feedback
system F S, Sf is time invariably realizable if and only if for any t ∈ T and any
x ∈ D F S, Sf ,
    
S ◦ Sf ◦ F S, Sf 0t ◦ σ t (x) |Tt  σ t S ◦ Sf ◦ F S, Sf (x), (10.30)
 
i.e., the system S ◦ Sf ◦ F S, Sf is time invariably realizable.
 
Proof: Sufficiency. Suppose the system  S ◦S f ◦F S, Sf is time invariably realizable.
Then for any t ∈ T and any x ∈ D F S, Sf , equation (10.30) holds true. From the
hypothesis that S is time invariably realizable, it follows from Theorem 10.8 that
equation (10.26) holds true. Therefore, we have
      
S 0t ◦ σ t (x) |Tt + S ◦ Sf F S, Sf 0t ◦ σ t (x) |Tt
 
 σ t S(x) + σ t S ◦ Sf ◦ F S, Sf (x). (10.31)

This implies that

     
S 0t ◦ σ t (x) + Sf F S, Sf 0t ◦ σ t (x) Tt
   
 σ t S x + Sf ◦ F S, Sf (x) . (10.32)

This equation is the same as

    
F S, Sf 0t ◦ σ t (x) |Tt  σ t F S, Sf (x). (10.33)

 Therefore,
 applying Theorem 10.8, we have showed that the feedback system
F S, Sf is time invariably
 realizable.

Necessity. Suppose F S,Sf  is time
 invariably realizable. Then Theorem 10.8
says that ∀t ∈ T and ∀x ∈ D F S, Sf ; we have
    
F S, Sf 0t ◦ σ t (x) |Tt  σ t F S, Sf (x). (10.34)
    
Let y  F S, Sf 0t ◦ σ t (x) and y  F S, Sf (x). Then from equation (10.34),
we have
    
S 0t ◦ σ t (x) + Sf (y) |Tt  σ t S x + Sf y , (10.35)

and
   
S 0t ◦ σ t (x) |Tt + S ◦ Sf (y)|Tt  σ t S(x) + σ t S ◦ Sf y . (10.36)
10.3 Time-Invariable Realization 129

From the hypothesis that S is time invariably realizable and Theorem 10.8, it
follows that equation (10.36) can be written as
  
S ◦ Sf (y)|Tt  σ t S Sf y . (10.37)

Equation (10.37) is the same as equation (10.30). This ends the proof. QED
We conclude this chapter with some discussion of chaos, attractor, and feedback
transformation.

Theorem 10.10. For each linear system S ⊂ X × X, there exists a feedback compo-
nent system Sf : X → X, such that D ⊂ X is a chaos, attractor, or strange attractor,
respectively,
  of S, if and only if D is a chaos, attractor, or strange attractor of
F S, Sf , respectively.
The proof follows from Theorem 9.4.

Theorem 10.11. Suppose that S ⊂ X × X and Sf : X → X are a linear system and a

functional linear system, respectively. Then a subset D ⊂ X is a chaos of S (if and
only if D is a chaos of the feedback system F(S, Sf )), if

(X − D) ± Sf (D) ⊂ X − D. (10.38)
 
Proof: Necessity. Suppose D ⊂ X is a chaos of S but not a chaos  of F S, Sf .
Pick an element (x, y) ∈ D2 ∩ F S, Sf . Then we have x + Sf (y), y ∈ S; so x 
/ D. Now x  x − Sf (y).
x + Sf (y) ∈  This contradicts relation in equation (10.38).
Therefore, D must be a chaos of F S, Sf .  
Sufficiency. Suppose D ⊂ X is a chaos
 of F S, Sf butnot achaos of S. Pick an
element (x, y) ∈ D2 ∩ S. Then we have x − Sf (y), y ∈ F S, Sf . Therefore, it must
be that x  x − Sf (y) ∈
/ D, Now, x  x + Sf (y), which contradicts equation (10.38).
So D must be a chaos of S. QED

Theorem 10.12. Suppose that S ⊂ X × X and Sf : X → X are a linear system and a

functional linear system, respectively. Then D ⊂ X is an attractor of S (if and only if
D is an attractor of the feedback system F(S, Sf )), if
 
X − D ⊂ x − Sf (y) : (x, y) ∈ S ∩ ((X − D) × D) , (10.39)

and
   
X − D ⊂ x + Sf (y) : (x, y) ∈ F S, Sf ∩ ((X − D) × D) . (10.40)

Proof: Necessity.
 Suppose D ⊂ X , satisfying equation (10.39), is an attractor of S
but not of F S, Sf . Then there exists an element x ∈ X − D such that
 
F S, Sf (x) ∩ D  ∅. (10.41)
130 10 Properties Invariant Under Feedback

From equation (10.39), it follows that there  (z, y) ∈ S ∩

 exists  ((X − D) × D)
such that z − Sf (y)  x. Thus, (x, y) ∈ F S, Sf ; i.e., y ∈ F S, Sf (x) ∩ D, a
contradiction.
 Sufficiency.
 Suppose D ⊂ X, satisfying equation (10.40), is an attractor of
F S, Sf , but not of the original system S. There then exists an element x ∈ X −
D such that

S(x) ∩ D  ∅. (10.42)
 
From equation (10.40), it follows that there exists (z, y) ∈ F S, Sf ∩
((X − D) × D) such that z + Sf (y)  x. Therefore, (x, y) ∈ S; y ∈ S(x) ∩ D, a
contradiction. QED
Feedback transformation has been a really hot topic in recent years. Because all
living and many nonliving systems are feedback systems, the method of studying
these systems is important.
The concept of feedback has been closely studied in control systems (Zadeh
1965; Bayliss 1966; Milsum 1966; Wonham 1979; Wu 1981; Henig 1983; Deng
1985; Negoita 1992; Saito and Mesarovic 1985; Takahara and Asahi 1985; Saito
1986, 1987) and economics (see, for example, (Forrest et al. 2018) and references
given there). Based on the definition of general feedback systems, first introduced
by Saito and Mesarovic (1985), some results presented in this chapter are from Lin
and Ma (1990).

References

Bayliss, L.E.: Living Control Systems. San Francisco (1966)

Deng, J.-L.: Grey Control Systems. Hua-zhong Institute of Technology Press, Wuhan (1985)
Forrest, J., Ying, Y.R., Gong, Z.W.: Currency Wars: Offense and Defense through Systemic Think-
ing. CRC Press, an imprint of Taylor and Francis, New York (2018)
Henig, M.I.: Vector-Valued Dynamic Programming. SIAM J. Control Optim 3, 237–245 (1983)
Lin, Y., Ma, Y.H.: General feedback systems. Int. J Gen. Syst. 18(2), 143–154 (1990)
Milsum, J.H.: Biological Control Systems Analysis. McGraw-Hill, New York (1966)
Negoita, C.V.: Pullback versus feedback. In: Negoita, C.V. (ed.) Cybernetics and Applied Systems,
pp. 239–247. Marcel Dekker, New York (1992)
Saito, T.: Feedback transformation of linear systems. Int. J. Syst. Sci. 17(9), 1305–1315 (1986)
Saito, T.: Structure of the class of linear functional time systems under output feedback transfor-
mation. Int. J. Syst. Sci. 18(3), 1017–1027 (1987)
Saito, T., Mesarovic, M.D.: A meaning of the decoupling by feedback of linear functional time
systems. Int. J. Gen. Syst. 11(1), 47–61 (1985)
Takahara, Y., Asahi, T.: Causality invariance of general systems under output feedback transforma-
tions. Int. J. Syst. Sci. 16(10), 1185–1206 (1985)
Wonham, W.M.: Linear Multi-Variable Control: A Geometric Approach. Springer, New York (1979)
Wu, X.M.: Pansystems analysis—a new exploration of interdisciplinary investigation (with appen-
dices on discussions on medicine, chemistry, logic and economics). Sci. Explor. 1(1), 125–164
(1981)
Zadeh, L.: Fuzzy sets. Inf. Control 8(3), 338–353 (1965)
Chapter 11
Decoupling of Single-Relation Systems

Because the investigations of systems are mostly about the relationship between the
“whole” and the “parts,” this chapter looks at how “whole” single-relation systems
can be decomposed into factor systems by feedback. And at the end, we use an
example to illustrate how the developed theory can be practically applied.

11.1 The Concept

Let S ⊂ X × Y be an input–output system. We assume that the input space X and

the output space Y can be represented in terms of the factor sets; i.e., there are sets
{Xi : i ∈ I } and {Yi : i ∈ I }, where I is a fixed index set, such that

X  {Xi : i ∈ I }, (11.1)

and

Y  {Yi : i ∈ I }. (11.2)

For each index i ∈ I , let

 
i  ix , iy : X × Y → Xi × Yi (11.3)

be the projection defined by letting ix  pi : X → Xi and iy  pi : Y → Yi such

that
     
ix xj j∈I  xi , iy yj j∈I  yi ,

© Springer Nature Switzerland AG 2018 131

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_11
132 11 Decoupling of Single-Relation Systems

and
           
i xj j∈I , yj j∈I  ix xj j∈I , iy yj j∈I  (xi , yi ).

We now can represent factor systems of S on the factor sets Xi and Yi ; i.e., for
each i ∈ I , the factor system Si of S can be defined by

Si  i (S) ⊂ Xi × Yi . (11.4)

The system S can now be viewed as decomposed into a family of factor systems
S  {Si : i ∈ I }, and S can be represented as a relation among its factor systems (i.e.,
S ⊂ Π {Si : i ∈ I } by identifying ((xi )i∈I , (yi )i∈I ) ∈ S ⊂ X ×Y with ((xi , yi )i∈I ) ∈ S ⊂
Π {Si : i ∈ I }. The systems S and Si are called the overall system and the factor (or
component) system, respectively, and S ⊂ Π {Si : i ∈ I } is called a complex system
representation over {Si : i ∈ I }.
Generally, there are many interactions among subsystems, so S is a proper subset
of the product of its factor systems Π {Si : i ∈ I }. If S  Π {Si : i ∈ I }, S is said to be
noninteracted.

11.2 Transferrable Properties

We now consider what properties are transferable from the overall system S to its
component systems Si or from the component systems Si to an overall system S.

Proposition 11.1. Let S ⊂ Π {Si : i ∈ I } be an input–output system which has been

decomposed into a family of factor systems S  {Si : i ∈ I }. If S is a linear system
over a field A, each component system Si is then also a linear system over the field A.

Proof: From the hypothesis that S is a linear system over the field A, it follows that,
for each i ∈ I , Si is a nonempty system on Xi × Yi . We need to show two things: (1)
For any (x1 , y1 ), (x2 , y2 ) ∈ Si , (x1 , y1 ) + (x2 , y2 ) ∈ Si , and (2) for any (x, y) ∈ Si and
any α ∈ A, α(x, y) ∈ Si .

The argument of (1). For any (ai , ci ) and (bi , di ) ∈ Si , there are (x1 , y1 ) and
(x2 , y2 ) ∈ S such that i (x1 , y1 )  (ai , ci ) and i (x2 , y2 )  (bi , di ). Therefore, we
have

(ai , ci ) + (bi , di )  (ai + bi , ci + di )

 i (x1 + x2 , y1 + y2 ) ∈ Si , (11.5)

because (x1 + x2 , y1 + y2 ) ∈ S.
The argument of (2). For any (ai , ci ) ∈ Si and any α ∈ A, there exists a pair
(x, y) ∈ S such that i (x, y)  (ai , ci ). Therefore,
11.2 Transferrable Properties 133

α(ai , ci )  (αai , αci )

 i (αx, αy)
 i (α(x, y))
∈ Si , (11.6)

because α(x, y) ∈ S. QED

Proposition 11.2. Suppose that S ⊂ X × Y is a linear time system that has

been decomposed into a family of component systems {Si ⊂ Xi × Yi : i ∈ I }; i.e.,
S ⊂ Π {Si : i ∈ I }. Then, each component system Si is also a linear time system.

Proof: It suffices from Proposition 11.1 to show that for any x, y ∈ D(Si ) and any
t ∈ T , the concatenation of x, y, xt ◦ yt ∈ D(Si ) for each i ∈ I . In fact, for any
x, y ∈ D(Si ) and any t ∈ T , there exist u, v ∈ D(S) such that
 
(u)  x and (v)  y. (11.7)
ix ix

Therefore, xt ◦ yt  Πix (ut ◦ vt ) ∈ D(Si ), because ut ◦ vt ∈ D(S). QED

Proposition 11.3. Under the same assumption as Proposition 11.2 and if the system
S is strongly stationary, then each component system Si is also strongly stationary.

Proof: Let (xt , yt ) ∈ Si |Tt . be arbitrary. There then exists some element (xi , yi ) ∈ Si
such that (xt , yt )  (xit , yit ). Furthermore, there exists some (x, y) ∈ S such that
i (x, y)  (xi , yi ). Since λt S ⊂ S, λt (x, y)  (λt x, λt y) ∈ S, where i (λt x, λt y) 
(λt xi , λt yi ), which implies that (λt xi , λt yi ) ∈ Si . Hence, (xt , yt )  (xit , yit ) ∈ σ t (Si ).
Consequently, Si |Tt . ⊂ σ t (Si ).

Conversely, let (xt , yt ) ∈ σ t (Si ) be arbitrary. There then exists some (xi , yi ) ∈ Si
such that (xt , yt )  σ t (xi , yi ). Furthermore, there exists some (x, y) ∈ S such that
i (x, y)  (xi , yi ). Since σ t (S) ⊂ S|Tt ., we have σ t (x, y)  (σ t (x), σ t (y)) ∈ σ t S ⊂
S|Tt .. That is, there exists some (x , y ) ∈ S such that (σ t (x), σ t (y))  (x , y )|Tt ..
Hence,

(xt , yt )  σ t (xi , yi )
 σ t i (x, y)
 
 i σ t (x), σ t (y)
  
 i x , y |Tt .

That is, σ t (Si ) ⊂ Si |Tt .

134 11 Decoupling of Single-Relation Systems

Combining these arguments, we obtain the following: For each index i ∈ I and
any t ∈ T , Si |Tt .  σ t (Si ); i.e., σ −t (Si |Tt .)  Si . From equation (10.15), we can
show that the component system Si is strongly stationary. QED

Proposition 11.4. Under the same assumption as in Proposition 11.2 and if the
system S is precausal with D(S)  X, each component system Si is also precausal.

Proof: Let i ∈ I be arbitrarily fixed. It can then be seen that D(Si )  Xi . In fact, if
D(Si )  Xi , we can pick an element xi ∈ Xi − D(Si ). This implies that D(S)  X ,
because no x ∈ D(S) satisfies ix (x)  xi , a contradiction.

Let xi , xi ∈ Xi , t ∈ T be arbitrary, and assume x̄it  x̄it . Since it is assumed that S
t t
is precausal, for any x, y ∈ X , x̄t  ȳt implies that S(x)|T  S(y)|T , from Theorem
10.6. Therefore, when the elements x, y ∈ X satisfy ix (x)  xi , ix (y)  xi , and
j (x)  j (y), for every j ∈ I with j  i, we have x̄t  ȳt . Thus,

t t
Si (xi ) T  (S(x)) T
iy

t
 (S(y)) T
iy
  t
 Si xi T . (11.8)

Applying Theorem 10.6, we have shown that each component system Si is pre-
causal. QED

Example 11.1. We will construct an example to show that even though an overall
system S is functional, none of the component systems Si is functional.

Suppose the index set I  {1, 2} and the sets X1 , X2 , Y1 , and Y2 are all the same,
namely the set Z of all integers. Let S be a functional linear system from X1 × X2 to
Y1 × Y2 defined by

S(n, m)  (m, n), (11.9)

for any (n, m) ∈ X1 × X2 . Then S1  {(n, m): (n, m) ∈ X1 × Y1 } and S2 

{(m, n): (m, n) ∈ X2 × Y2 }. Then it can be easily seen that neither S1 nor S2 is func-
tional.
11.2 Transferrable Properties 135

Proposition 11.5. Suppose that S ⊂ X × Y is a functional system that has

been decomposed into a family of component systems {Si ⊂ Xi × Yi : i ∈ I } with
S  Π {Si : i ∈ I }. Then each component system Si is also functional. If, moreover,
the overall system S is bijective, each component system Si is also bijective.

Proof: For any fixed i ∈ I , let (xi , yi ), (xi , zi ) ∈ Si be arbitrary. Since S 

Π {Si : i ∈ I } due to the assumption, there exist elements (x, y), (x, z) ∈ S such
that Πi (x, y)  (xi , yi ) and Πi (x, z)  (xi , zi ). Because S is functional, y  z,
so yi  Πiy (y)  Πiy (z)  zi . That is, the component system Si is functional.

Suppose S is a bijective function from X to Y. For any yi ∈ Yi , there exists an

element y ∈ Y such that Πiy (y)  yi . From the hypothesis that S is bijective, it follows
that there exists an element x ∈ X such that S(x)  y. Therefore, for xi  Πix (x),
Si (xi )  yi . That is, Si is surjective. Let (xi , yi ), (wi , yi ) ∈ Si be arbitrary. Then there
exist (x, y) and (w, y) ∈ S such that Πi (x, y)  (xi , yi ) and Πi (w, y)  (wi , yi ).
From the hypothesis that S is injective, it follows that x  w, that is, xi  Πix (x) 
Πix (w)  wi . Hence, Si is bijective. QED

Proposition 11.6. Suppose that S ⊂ X × Y is a functional linear time system that S

has been decomposed into a family of component systems {Si ⊂ Xi × Yi : i ∈ I } such
that S is noninteracted; i.e., S  Π {Si : i ∈ I }. If S is time invariably realizable, each
component system Si is then also time invariably realizable.

Proof: From the hypothesis that S is a functional system from X to Y , it follows that
D(S) ⊂ X . Thus, from the proof of Proposition 11.4, for each fixed i ∈ I , D(Si )  Xi .

We now let t ∈ T and xi ∈ Xi be arbitrary. There exists an element x0 ∈ X such

that Πix (x0 )  xi . Since S is time invariably realizable, from equation (10.25) it
follows that for t ∈ T , we have
 
λt S 0t ◦ σ t (x0 )  S(x0 ). (11.10)

Let A  {x ∈ X : Πix (x)  xi }. Then x0 ∈ A  ∅ and each x ∈ A satisfies equation

(11.10), which implies that

   

λt S 0t ◦ σ t (x) : x ∈ A  {S(x): x ∈ A}. (11.11)

Therefore, we have
136 11 Decoupling of Single-Relation Systems


Si (xi )  {S(x): x ∈ A}
ix

   
 λt S 0t ◦ σ t (y) : y ∈ A
ix


  t  
 λ S 0 ◦ σ (y) : y ∈ A
t t

ix


   
 λ t
S 0 ◦ σ (y) : y ∈ A
t t

ix
  

 
 λ Si
t
0 ◦ σ (y) : y ∈ A
t t

ix
  


 λ Si 0 ◦
t t
σ (y) : y ∈ A
t

ix
   


 λt Si 0t ◦ σ t (y) :y ∈ A
ix
 
 λt Si 0t ◦ σ t (xi ) . (11.12)

Applying equation (10.25), we complete the proof that Si is time invariably real-
izable. QED
The following result combines Propositions 11.1–11.6.

Theorem 11.1. Let S ⊂ X × Y be an input–output system which has been decom-

posed into a family of factor systems S  {Si : i ∈ I } such that S  Π {Si : i ∈ I }.
Then the following statements hold true:
(1) S is a linear system if and only if each Si is a linear system.
(2) S is a linear time system if and only if each Si is a linear time system.
(3) S is strongly stationary if and only if each Si is strongly stationary.
(4) S is causal if and only if each Si is causal.
(5) S is bijective if and only if each Si is bijective.
(6) The functional system S is time invariably realizable if and only if each Si is
functional and time invariably realizable.
The proof is not hard and is left to the reader.

Theorem 11.1 shows that the theoretical properties of many important systems
are either hereditary from the overall system S to each component system Si or able
to be lifted up from Si to S. In practice, we always hope to study S by studying each
component system Si .
11.3 Decoupling Through Feedback 137

11.3 Decoupling Through Feedback

Let S ⊂ X × Y be a linear system. We say that S can be decoupled by feedback if

there exists some functional system Sf ∈ Sf such that

F(S, Sf )  { (F(S, Sf )): i ∈ I }, (11.13)
i

where it is assumed that X  Π {Xi : i ∈ I }, Y  Π {Yi : i ∈ I }, and i : X × Y →

Xi × Yi is the projection defined by equation (11.3).
The concept of decoupling by feedback shows the following actions: The original
system S may not be noninteracted, so there might be some properties which cannot
be lifted up from component systems to the overall system S. If S is decoupled by
feedback, after applying appropriate feedback transformation, S is transformed into
a system which has a noninteracted complex systems representation. We now can
use that representation to analyze and control S.

Theorem 11.2. Suppose that S ⊂ X × Y and Sf : Y → X are a linear system and a

functional linear system, respectively, such that S is decomposed into S  {Si : i ∈ I }
and Sf is decomposed into S f  {Sfi : i ∈ I }. Then, for each i ∈ I , the following holds
true:
    
Π F S, Sf  F Si , Sfi . (11.14)
i

Proof: For each (xi , yi ) ∈ Πi (F(S, Sf )), there exists (x, y) ∈ F(S, Sf ) such that
Πi (x, y)  (xi , yi ). Since Πix (Sf (y))  Sfi (yi ), it follows that
⎛ ⎞
    
xi + Sfi (yi ), yi  ⎝ x + Sf (y) , (y)⎠
ix iy
 
 x + Sf (y), y
i

∈ (S)
i
 Si . (11.15)

Equation (11.15) implies that (xi , yi ) ∈ F(Si , Sfi ). That is, Πi (F(S, Sf )) ⊂
F(Si , Sfi ).
Let (xi , yi ) ∈ F(Si , Sfi ) be arbitrary. There then exists (x, y) ∈ S such that
  
(x, y)  xi + Sfi (yi ), yi . (11.16)
i

Therefore, (x − Sf (y), y) ∈ F(S, Sf ). From this, we have

138 11 Decoupling of Single-Relation Systems
 
(xi , yi )  xi + Sfi (yi ) − Sfi (yi ), yi
⎛ ⎞
  
⎝ x − Sf (y) , (y)⎠
ix iy
 
 x − Sf (y), y
i
  
∈ F S, Sf . (11.17)
i

Equation (11.17) gives F(Si , Sfi ) ⊂ Πi (F(S, Sf )). QED

Theorem 11.3. Assume the Axiom of Choice. Suppose that S ⊂ X × Y is a linear
system that can be decomposed into a family of component systems S  {Si : i ∈ I }.
Then S can be decoupled by feedback if and only if R(S)  {R(Si ): i ∈ I } and
N(S)  N (S)  {N (Si ): i ∈ I }.
Proof: Necessity. Suppose S can be decoupled by feedback. By the definition of
decoupling by feedback, there exists a functional linear system Sf ∈ Sf such that

 
F(S, Sf )  (F(S, Sf )) : i ∈ I . (11.18)
i

Therefore, by Theorem 9.5 we have

  
R(S)  R F S, Sf
  
   
 R F S, Sf :i ∈ I
i
    
 R F Si , Sfi : i ∈ I

 {R(Si ): i ∈ I }, (11.19)

and
  
N (S)  N F S, Sf
  
   
 N F S, Sf :i ∈ I
i
    
 N F Si , Sfi : i ∈ I

 {N (Si ): i ∈ I }. (11.20)

Sufficiency. Let S  ⊂ X × Y be the system defined by

S   {Si : i ∈ I }. (11.21)
11.3 Decoupling Through Feedback 139

Since R(S)  {R(Si ): i ∈ I } and N (S)  {N (Si ): i ∈ I } due to our assumption,
we see that R(S)  R(S  ) and N(S)  N(S  ). By applying Theorem 9.5, we find there
exists a functional linear system Sf ∈ Sf such that
 
F S, Sf  S 

 {Si : i ∈ I }

   
 F S, Sf : i ∈ I . (11.22)
i

That is, S can be decoupled by feedback. QED

11.4 An Example

In this section, we look at how to apply what has been established earlier to decouple
a system of differential equations. This example has seen practical applications in
the study of currency wars (Forrest et al. 2018). For some related discussion, see
Chapter 22 in this book.

Example 11.2. As an example of decoupling by feedback, consider the multivariable

system S defined by
⎧
⎪
⎨ ż  Az + Bx
y  Cz + Dx (11.23)
⎪
⎩ z(0)  0

where z is an m × 1 variable vector, A an m × m constant matrix, B an m × n constant

matrix, C an n × m constant matrix, D an n × n nonsingular constant matrix, and
  
X  x: 0, +∞) → Rn : x is piecewise continuous (11.24)

is the input space, where R is the set of all real numbers, and
  
Y  y: 0, +∞) → Rn : y is piecewise continuous (11.25)

is the output space. For each i = 1, 2, …, n, define the system Si by

⎧
⎨ ż  Az + Bi xi
⎪
yi  Ci z + Di xi (11.26)
⎪
⎩
z(0)  0
140 11 Decoupling of Single-Relation Systems

where Bi is an m × 1 constant matrix such that B  [B1 B2 . . . Bn ], Ci is a 1 × m

constant matrix such that
⎡⎤
C1
⎢C ⎥
⎢ 2⎥
C⎢ ⎥
⎢ .. ⎥, (11.27)
⎢. ⎥
⎢ ⎥
⎢C ⎥
n

Di is a nonzero constant, and

Xi  {xi : [ 0, +∞) → R: xi is piecewise continuous} (11.28)

is the input space and

Yi  {yi : [ 0, +∞) → R: yi is piecewise continuous} (11.29)

is the output space. Then the systems S and Si , i = 1, 2, …, n, satisfy

R(S)  {R(Si ): i ∈ I } and N (S)  {N (Si ): i ∈ I }. (11.30)

In fact, for any x ∈ X , the solution of the differential equation (11.23) in S is

given by
t
ϕ(t, 0, x)  ∫ e(t−s)A · B · x(s)ds. (11.31)
0

Thus, D(S)  X. At the same time, since the matrix D is nonsingular, the inverse
system S −1 of S is obtained as follows:
⎧  
⎪ −1 −1
⎨ ż  A − BD C z + BD y
⎪
x  −D−1 Cz + D−1 y . (11.32)
⎪
⎪
⎩ z(0)  0

Therefore, a similar argument shows that R(S)  D(S −1 )  Y . For the same
reason, we can show that for each i = 1, 2, …, n, R(S i )  Y i . That is, we have

R(S)  Y

 {Yi : i  1, 2, . . . , n}

 {R(Si ): i  1, 2, . . . , n}. (11.33)

For the null spaces of the systems S and Si , i = 1, 2, …, n, we have

11.4 An Example 141

N (S)  {−D−1 Cz: ż  (A − BD−1 C)z}, (11.34)

and
   
N (Si )  −Di−1 Ci z: ż  A − Bi Di−1 Ci z . (11.35)

Therefore, (S)  {N (Si ): i  1, 2, . . . , n}. By Theorem 11.3, S can be decou-
pled by feedback. In fact, define
⎡ ⎤
⎢ A 0 ⎥
⎢ ⎥
⎢ ⎥
⎢ A ⎥
⎢ ⎥
⎢ ⎥
α⎢ ⎢ ⎥
⎥
⎢ .. ⎥
⎢ . ⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
0 A

⎡ ⎤
⎢ B1 0 ⎥
⎢ ⎥
⎢ ⎥
⎢ B ⎥
⎢ 2 ⎥
⎢ ⎥
⎢
β⎢ ⎥
⎥
⎢ .. ⎥
⎢ . ⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
0 Bn

⎡ ⎤
⎢ C1 0 ⎥
⎢ ⎥
⎢ ⎥
⎢ C ⎥
⎢ 2 ⎥
⎢ ⎥
γ ⎢
⎢
⎥
⎥
⎢ .. ⎥
⎢ . ⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
0 Cn
142 11 Decoupling of Single-Relation Systems

⎡ ⎤
⎢ D1 0 ⎥
⎢ ⎥
⎢ ⎥
⎢ D ⎥
⎢ 2 ⎥
⎢ ⎥
δ⎢
⎢
⎥
⎥
⎢ .. ⎥
⎢ . ⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
0 Dn

Then the system Sd  S1 × S2 × . . . × Sn is represented by

⎧
⎨ ż  αz + βx
⎪
y  γ z + δx (11.36)
⎪
⎩
z(0)  0

Since δ is nonsingular, Sd−1 is obtained as follows:

⎧  
⎪
⎪ ż  α − βδ −1 γ z + βδ −1 y
⎨
x  −δ −1 γ z + δ −1 y . (11.37)
⎪
⎪
⎩ z(0)  0

If we define a functional linear system Sf : Y → X as follows:

⎧ ! ! " # !
⎪
⎪
⎪
⎪ ż A − BD−1 C 0 z BD−1
⎪
⎪  ·  + ·y
⎪ ż 
⎪ 0 α − βδ −1 γ z βδ −1
⎪
⎪
⎪
⎨ $ % "z#  
−1 −1 −1 −1 . (11.38)
⎪ x  −D C −δ γ · z  + D − δ
⎪
·y
⎪
⎪
⎪
⎪ " #
⎪
⎪ z
⎪
⎪
⎩ z  (0)  0
⎪

then the original system S can be transferred into Sd by this feedback component
system Sf . In fact, for any pair (x, y) ∈ X × Y , (x, y) ∈ Sd if and only if the input x
and the output y satisfy the systems representation in equation (11.37) if and only if
(x + Sf (y), y) ∈ S. We show the last “if and only if” condition as follows.
First, suppose that the input x and the output y satisfy the systems representation in
equation (11.37). Let z  be the solution of the differential equation in equation (11.37),
and let z be the solution of the differential equation in the systems representation in
equation (11.32). Then [z z  ]T is the solution of the differential equation in equation
(11.38). From the definition of Sf , it follows that
11.4 An Example 143

Sf (y)  −D−1 Cz + δ −1 γ z  + D−1 y − δ −1 y

   
 −D−1 Cz + D−1 y − −δ −1 γ z  + δ −1 y . (11.39)

Therefore, we have
   
x + Sf (y), y  −D−1 Cz + D−1 y, y ∈ S, (11.40)

where the membership relation comes from the systems representation in equation
(11.32) of the inverse system S −1 .
Second, suppose that a pair (x, y) ∈ X × Y satisfies (x + Sf (y), y) ∈ S. From the
systems representations in equations (11.32) and (11.38), we have

x + Sf (y)  −D−1 Cz + D−1 y. (11.41)

That is,
 
x  −D−1 Cz + D−1 y − Sf (y)
   
 −D−1 Cz + D−1 y − −D−1 Cz + δ −1 γ z  + D−1 y − δ −1 y
 −δ −1 γ z  + δ −1 y. (11.42)

Combining equation (11.42) with the systems representation in equation (11.37)

proves that the pair (x, y) satisfies the systems representation in equation (11.37).
All living systems can adjust themselves according to the changes in the environ-
ments. So, they are feedback systems, where the feedback mechanism is employed
to tell what adjustments are needed and by how much. Here, business organizations
are also seen as living systems, where those business entities that cannot make use
of the feedback mechanism generally go out of business sooner or later.
At the same time, the feedback mechanism also works with many nonliving sys-
tems, such as celestial bodies that orbiting around a planet or star. The interac-
tion between feedback transformation and feedback decoupling shed lights on the
“whole” system and the “parts” interrelate with each other. Jointly, these concepts tell
when the system under consideration can be divided into smaller systems. For related
researches, see Lin (1999), Takahara and Asahi (1985), and Saito (1986, 1987).

References

Forrest, J., Ying, Y.R., Gong, Z.W.: Currency Wars: Offense and Defense Through Systemic Think-
ing. CRC Press, An Imprint of Taylor and Francis, New York (2018)
Lin, Y.: General Systems Theory: A Mathematical Approach. Kluwer Academic and Plenum Pub-
lishers, New York (1999)
Saito, T.: Feedback transformation of linear systems. Int. J. Syst. Sci. 17(9), 1305–1315 (1986)
Saito, T.: Structure of the class of linear functional time systems under output feedback transfor-
mation. Int. J. Syst. Sci. 18(3), 1017–1027 (1987)
Takahara, Y., Asahi, T.: Causality invariance of general systems under output feedback transforma-
tions. Int. J. Syst. Sci. 16(10), 1185–1206 (1985)
Chapter 12
Decomposability Conditions

As we have learned in the previous chapters, the study of systemhood is more or less
about the relationship between the “whole” of the system of concern and the “parts”
of the system. In this chapter, we look at conditions under which the “whole” can be
decomposed into “parts.”

12.1 Global and Complex System Representations

Intuitively speaking, a complex system consists of several subsystems (see, for exam-
ple, Chapters 3 and 6). Let us first consider the following example. Suppose that

Si ⊂ Xi × Yi , (12.1)

for i  1, 2, 3, are three input–output systems such that X1  X2  X3  Y1  Y2 

Y3  R, the set of all real numbers, (xi , yi ) ∈ Si , i  1, 2, 3, if and only if y1  x1 ,
y2  2x2 , and y3  2x3 hold true. Suppose these three systems are connected in 

series; that is, x2  y1 and x3  y2 hold true; furthermore, suppose X  X1 and

 

Y  Y3 are the input space and the output space of a new system S , defined as

  

S  (x, y) ∈ X × Y : y  4x . (12.2)


Then S is the composition of S1 , S2 , and S3 . However, in equation (12.2), the

component systems S1 , S2 , and S3 completely disappeared.
Let us now consider another construction of a new system S. We let

X  X1 × X2 × X3 and Y  Y1 × Y2 × Y3 , (12.3)

and the new system S ⊂ X × Y is defined by

© Springer Nature Switzerland AG 2018 145

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_12
146 12 Decomposability Conditions

Figure 12.1 Structure of the



system S

Figure 12.2 Structure of the

system S

S  {((x1 , x2 , x3 ), (y1 , y2 , y3 )) ∈ X × Y :
(xi , yi ) ∈ Si and x2  y1 and x3  y2 }. (12.4)


Figures 12.1 and 12.2 show the geometric meanings of S and S. The system S is a
representation of the whole system, which explicitly reflects the fact that the whole 

system is composed of three component systems Si , i  1, 2, 3. The structure of S

is called a global system representation of the whole system, while S is a complex
system representation of the whole system. In fact, S is the product system of Si , i
 1, 2, 3.
The complex system S is defined over X and Y , but to emphasize the fact that S con-
sists of the component systems S1 , S2 , and S3 , we identify ((x1 , x2 , x3 ), (y1 , y2 , y3 )) ∈
X × Y with ((x1 , y1 ), (x2 , y2 ), (x3 , y3 )) ∈ S1 × S2 × S3 and denote the complex system


S by S ⊂ S1 × S2 × S3 . For the global system S , we emphasize the structure by



writing S  S3 ◦ S2 ◦ S1 .


The global system S and the complex system S are intimately related to each other.
 

In fact, if we define two mappings h1 : X → X and h2 : Y → Y by h1 (x1 , x2 , x3 )  x1

and h2 (y1 , y2 , y3 )  y3 , then we have
 

∀(x, y) ∈ X × Y (x, y) ∈ S → (h1 (x), h2 (y)) ∈ S , (12.5)

and

  

∀(x̂, ŷ) ∈ X × Y (x̂, ŷ) ∈ S →

  

∃(x, y) ∈ S (h1 (x), h2 (y))  (x̂, ŷ) ∈ S . (12.6)

 

If we define two mappings k1 : X → X and k2 : Y → Y by k1 (x̂)  (x̂, x̂, 2x̂) and

k2 (ŷ)  (ŷ/4, ŷ/2, ŷ), then k1 and k2 satisfy the following relations

 

∀(x̂, ŷ) ∈ X × Y (x̂, ŷ) ∈ S → (k1 (x̂), k2 (ŷ)) ∈ S , (12.7)

and
12.1 Global and Complex System Representations 147
 

∀(x̂, ŷ), (x̂ , ŷ ) ∈ X × Y ((k1 (x̂), k2 (ŷ)) 

 (k1 (x̂ ), k2 (ŷ )) → (x̂, ŷ)  (x̂ , ŷ )). (12.8)

Combining equations (12.5)–(12.8), we can see that for any (x, y) ∈ X × Y , we

have

(k1 , k2 ) ◦ (h1 , h2 )(x, y)  (k1 , k2 )(h1 (x), h2 (y))

 (k1 h1 (x), k2 h2 (y))
 (x, y), (12.9)
 

and for any (x̂, ŷ) ∈ X × Y , we have

(h1 , h2 ) ◦ (k1 , k2 )(x̂, ŷ)  (h1 , h2 )(k1 (x̂), k2 (ŷ))

 (h1 k1 (x̂), h2 k2 (ŷ))
 (x, y). (12.10)


Therefore, S and S are similar systems.

Generally, let S ⊂ X × Y and S ⊂ X × Y be two input–output systems, and hi ,
i  1, 2, be mappings such that

h1 : X → X and h2 : Y → Y , (12.11)

satisfying the condition

∀(x, y) ∈ X × Y ((x, y) ∈ S → (h1 (x), h2 (y)) ∈ S ). (12.12)

Then h  (h1 , h2 ) is called a modeling relation from S to S and is denoted by

h
(h1 , h2 ): S → S or S → S . Notice that h is actually a mapping from S to S .

Proposition 12.1. Suppose that h  (h1 , h2 ) and k  (k1 , k2 ) are modeling rela-
tions from system S1 to system S2 and from S2 to system S3 , respectively. Then the
composition k ◦ h  (k1 , k2 ) ◦ (h1 , h2 )  (k1 ◦ h1 , k2 ◦ h2 ) is a modeling relation
from S1 to S3 .

The proof is straightforward and is omitted.

Suppose that {Xi : i ∈ I } and {Yi : i ∈ I } are families of sets and that
{fi : Xi → Yi : i ∈ I } is a family of mappings. The product mapping f 
i∈I fi : {Xi : i ∈ I } → {Yi : i ∈ I } is defined by

f ((xi )i∈I )  (fi (xi ))i∈I , (12.13)

for any (xi )i∈I ∈ {Xi : i ∈ I }. When I  {1, 2, …, n} is finite, we denote the product
mapping f by f1 × f2 × · · · × fn . The following result shows the relation between the
concept of modeling relations and that of product mappings.
148 12 Decomposability Conditions

Proposition 12.2. Suppose that h  (h1 , h2 ) and k  (k1 , k2 ) are two modeling
relations from system S to system S . Then the restriction of the product mappings
h1 × h2 and k1 × k2 on S are mappings from S to S . Furthermore, if D(S)  X and
R(S)  Y , then h1 × h2 |S  k1 × k2 |S if and only if hi  ki , i  1, 2.

The proof is straightforward and is omitted.

Due to Proposition 12.2, we will still use h for the mapping h1 × h2 |S and the
modeling relation (h1 , h2 ): S → S , whenever no confusion occurs.
A modeling relation h  (h1 , h2 ): S → S is called an epimorphism from the
system S to the system S if for any fixed system S and modeling relations h and
h from S to S , h ◦ h  h ◦ h implies h  h .

Theorem 12.1. Let h  (h1 , h2 ) be a modeling relation from a system S to a system S

such that D(S)  X, R(S)  Y , D(S )  X , and R(S )  Y . Then h is an epimorphism
from S to S if and only if h1 and h2 are surjective.

Proof: Necessity. Suppose h is an epimorphism from S to S and, without loss of

generality, the mapping h1 is not surjective. That is, h1 (X )
 D(S )  X . There
exist mappings h 1 and h 1 such that

h1 |h1 (X ).  h 1 |h1 (X ). and h 1
 h 1 , (12.14)

where, for each arbitrarily fixed system S ⊂ X × Y , h 1 and h 1 are from X to X .

Pick mappings h 2 and h 2 : Y → Y such that h 2  h 2 . Then we h ◦ h  h ◦ h but
h
 h . This contradicts the fact that h is an epimorphism from S to S .
Sufficiency. Suppose h1 and h2 are surjective. Let S be an arbitrary system over
X × Y and h  (h 1 , h 2 ) and h  (h 1 , h 2 ) modeling relations from S to S such

that h ◦ h  h ◦ h. Then, we have

h ◦ h  h ◦ h → h ◦ h(x, y)  h ◦ h(x, y)
→ h 1 ◦ h1 (x)  h 1 ◦ h1 (x)
→ h 1  h 1 . (12.15)

The last implication follows because h1 : X → X is surjective and because D(S) 

X. A similar argument gives h 2  h 2 . Hence, applying Proposition 12.2, we have
h  h . That is, h is an epimorphism from S to S . QED

Theorem 12.2. Under the same assumption as in Theorem 12.1, if the modeling
relation h is also surjective as a mapping (that is, h1 × h2 |S is surjective), then h is
an epimorphism. The converse is not true.

Proof: It can be clearly seen from the assumptions that the mappings h1 and h2 are
surjective. Applying Theorem 12.1, it follows that h is an epimorphism.
We now see an example when the converse is not true. Let X  Y , S 
{(x, x): x ∈ X }, S  X × X , and h1  h2  idX . Then h  (h1 , h2 ) is an epi-
morphism but not surjective. QED
12.1 Global and Complex System Representations 149

Theorem 12.3. Under the same assumption as in Theorem 12.1, if the modeling
relation h is also a relation (that is, h has a right cancellation), then h is surjective
as a mapping. The converse is not true.
Proof: Let k: S → S be a right cancellation of h; i.e., h ◦ k  (idX , idY ). Let
(x , y ) ∈ S be arbitrary. Then h ◦ k(x , y )  h(k(x , y ))  (x , y ). That is, h is
surjective.
We now construct an example to show the converse may not be true. Suppose that
S ⊂ X × Y and S ⊂ X × Y are systems defined by

S  {(x1 , y1 ), (x2 , y2 ), (x3 , y2 )}, (12.16)

and

S  {(x1 , y1 ), (x1 , y2 ), (x2 , y2 )}. (12.16)

Let h1 : D(S) → D(S ) and h2 : R(S) → R(S ) be two mappings defined by h1 (x1 ) 
h1 (x2 )  x1 , h1 (x3 )  x2 , h2 (y1 )  y1 , and h2 (y2 )  y2 . If h  (h1 , h2 ) has a right
cancellation k  (k1 , k2 ), we have k1 (x1 )  x1 or x2 , k2 (y1 )  y1 , and k2 (y2 )  y2 .
Case 1: Suppose that k1 (x1 )  x1 . It then follows that (k1 , k2 )(x1 , y2 )  (x1 , y2 ) ∈/
S, which contradicts the hypothesis that k is a modeling relation from S to S.
Case 2: Suppose k1 (x1 )  x2 . We then have that (k1 , k2 )(x1 , y1 )  (x2 , y1 ) ∈ / S,
which contradicts the hypothesis that k is a modeling relation from S to S.
The contradictions imply that the modeling relation h does not have a right can-
cellation; i.e., h is not a retraction. QED

12.2 Isomorphisms and Monomorphisms

A modeling relation h from a system S to a system S is an isomorphism from S to

S if h has both right and left cancellation. In this case, S is isomorphic to S .
Theorem 12.4. A modeling relation h: S → S is an isomorphism from S to S if and
only if h is bijective from S to S as a mapping.
Proof: Necessity. Suppose h: S → S is an isomorphism. From Theorem 12.3, it
follows that h: S → S is surjective as a mapping. It remains for us to show that h
is injective. Let p: S → S be a left cancellation of h. That is, p ◦ h  (idX , idY ).
By contradiction, suppose h: S → S is not injective as a mapping. There then exist
distinct (x, y) and (u, v) ∈ S such that h(x, y)  h(u, v). So, p ◦ h(x, y)  p ◦ h(u, v),
which contradicts the assumption that p ◦ h  (idX , idY ) is the identity mapping on
S. QED
Example 12.1. We show that retraction does not guarantee an isomorphism. Let
S ⊂ X × Y and S ⊂ X × Y be two systems defined by
150 12 Decomposability Conditions

Figure 12.3 Properties of the mapping h from S to S with D(S)  X and R(S)  Y

S  {(x1 , y1 ), (x1 , y2 ), (x2 , y2 ), (x3 , y2 )}, (12.18)

and

S  {(x1 , y1 ), (x1 , y2 ), (x2 , y2 )}. (12.19)

Define a modeling relation h  (h1 , h2 ) from S to S by letting h1 (x1 )  h1 (x2 ) 

x1
and h1 (x3 )  x2 , and h2 (y1 )  y1 and h2 (y2 )  y2 . Let k  (k1 , k2 ) from
D(S ) × R(S ) → D(S) × R(S) be defined by

k1 (x1 )  x1 , k1 (x2 )  x3 , k2 (y1 )  y1 , k2 (y2 )  y2 . (12.20)

Then k is a modeling relation from S to S and h ◦ k  (idX , idY ). But, according

to Theorem 12.4, S is not isomorphic to S because |S|
 |S |.
Combining Theorems 12.2 and 12.3, we obtain Figure 12.3.
A modeling relation h  (h1 , h2 ) from S to S is termed as a monomorphism from
S to S if h is injective as a mapping from S to S . Then, similar to the previous results,
we have the following.

Theorem 12.5. Under the same assumption as in Theorem 12.1, the following hold
true:
(i) If both mappings h1 and h2 are injective, then h is a monomorphism, but the
converse is not true.
(ii) If the modeling relation h is a section (that is, h has a left cancellation), then
h1 and h2 are injective, but the converse is not true.
(iii) If h is an isomorphism, then h is a section, but the converse is not true.

Proof: (i) If h1 and h2 are injective, then the modeling relation h is injective as a
mapping from S to S . Therefore, h is a monomorphism. Conversely, define two
systems S ⊂ X × Y and S ⊂ X × Y such that

S  {(x1 , y1 ), (x2 , y3 ), (x3 , y2 ), (x3 , y3 )}, (12.21)

and

S  {(x1 , y1 ), (x1 , y2 ), (x2 , y1 ), (x2 , y2 )}, (12.22)

and define h  (h1 , h2 ): S → S by letting h1 (x1 )  h1 (x2 )  x1 , h1 (x3 )  x2 ,

h2 (y1 )  h2 (y2 )  y1 , and h2 (y3 )  y2 . Then h: S → S is an injective modeling
relation, but neither the mapping h1 nor h2 is injective.
12.2 Isomorphisms and Monomorphisms 151

(ii) Let k  (k1 , k): S → S be a left cancellation of h; i.e., k ◦ h 

(k1 ◦ h1 , k2 ◦ h2 )  (idX , idY ). Then h1 and h2 are injective. Conversely, define
systems S ⊂ X × Y and S ⊂ X × Y such that

S  {(x1 , y1 ), (x2 , y2 )}, (12.23)

and

S  {(x1 , y1 ), (x2 , y2 ), (x3 , y1 ), (x3 , y2 )}, (12.24)

and define mappings h1 and h2 by h1 (x1 )  x1 , h1 (x2 )  x2 , h2 (y1 )  y1 , and

h2 (y2 )  y2 . Then h1 and h2 are both injective. Now, if the modeling relation h has a
left cancellation k  (k1 , k2 ), then either k1 (x3 )  x1 or k1 (x3 )  x2 . If k1 (x3 )  x1 ,
then

k(x3 , y2 )  (x1 , y2 ) ∈
/ S, (12.25)

a contradiction. If k1 (x3 )  x2 , then

k(x3 , y1 )  (x2 , y1 ) ∈
/ S. (12.26)

This contradicts the hypothesis that k is a modeling relation from S to S.

(iii) We need only define two systems S ⊂ X × Y and S ⊂ X × Y and a section
h: S → S such that h is not an isomorphism. Let

S  {(x1 , y1 ), (x2 , y2 ), (x1 , y2 )}, (12.27)

and

S  {(x1 , y1 ), (x1 , y2 ), (x2 , y2 ), (x3 , y1 ), (x3 , y2 )}, (12.28)

and let the modeling relation h be defined as in the proof of (ii). Define a left cancel-
lation k  (k1 , k2 ) of h as follows: k1 (x1 )  k1 (x3 )  x1 , k1 (x2 )  x2 , k2 (y1 )  y1 ,
and k2 (y2 )  y2 . But h cannot be an isomorphism from S to S by Theorem 12.4
because |S|
 |S |. QED

Theorem 12.6. A modeling relation h  (h1 , h2 ): S → S is an isomorphism if and

only if h1 : D(S) → D(S ) and h2 : R(S) → R(S ) are bijective.

Proof: Necessity. Suppose h is an isomorphism. Then Theorems 12.3 and 12.5

imply that the mappings h1 : D(S) → D(S ) and h2 : R(S) → R(S ) are surjective and
injective.
Sufficiency follows from Theorem 12.4. QED
152 12 Decomposability Conditions

Theorem 12.7. Let h1 : X → X and h2 : Y → Y be two mappings and S ⊂ X × Y

and S ⊂ X × Y be two systems. Then the product mapping h1 × h2 reduces a
modeling relation from S to S ; i.e., h1 × h2 |S is a mapping from S to S if and only
if for any x ∈ D(S), h2 [S(x)] ⊂ S (h1 (x)).
Proof: Necessity. Suppose that h1 × h2 |S is a mapping from S to S . Then for any
x ∈ D(S) and any y ∈ S(x), we have h1 × h2 |S(x, y)  (h1 (x), h2 (y)) ∈ S . That is,
h2 (y) ∈ S (h1 (x)). This gives h2 [S(x)] ⊂ S (h1 (x)).
Sufficiency. Suppose that for any x ∈ D(S), h2 [S(x)] ⊂ S (h1 (x)). Then, for each
(x, y) ∈ S, h2 (y) ∈ S (h1 (x)); i.e., (h1 (x), h2 (y)) ∈ S . Therefore, h  (h1 , h2 ) is a
modeling relation from S to S . QED
From Theorem 12.6, it can be seen that if a system S is isomorphic to a system
S , then S is also isomorphic to S. In fact, let h  (h1 , h2 ) be an isomorphism from
S to S . Then there are mappings k1 : X → X and k2 : Y → Y such that

k1 |D(S )  h−1 −1
1 and k2 |R(S )  h2 . (12.29)

Applying Theorem 12.6, it follows that k  (k1 , k2 ) is an isomorphism from S to

S. Therefore, in the future we can speak about isomorphic systems.

12.3 Input–Output System Covering

Suppose that S ⊂ X × Y is an input–output system such that D(S)  X and R(S) 

Y , and S  {Sα ⊂ S: Sα
 ∅, α ∈ I } where I is an index set, is a covering of S; i.e.,
S satisfies ∪S  ∪{Sα : α ∈ I }  S. The collection S is referred to as an input–output
system covering of S if S satisfies the following two conditions:
(i) There are coverings X  {Xα : α ∈ Ix } and Y  {Yα : α ∈ Iy } of the input space
X and the output space Y , respectively.
(ii) There is a relation IS ⊂ Ix × Iy with D(IS )  Ix and R(IS )  Iy such that S
and {Xi × Yj : (i, j) ∈ IS } are related by a one-to-one correspondence ψ: IS → I
satisfying

(Xi × Yj ) ∩ S  Sψ(i,j) . (12.30)

In particular, if the collections S, X , and Y are partition s of the sets S, X, and Y ,

respectively, the corresponding input–output system covering of S will be referred
to as an input–output system partition. The system covering will be denoted by
{S, X , Y , IS }.
A system covering will be used when viewing a system macroscopically by
ignoring its detailed structures. Generally, not every covering is compatible with
an input–output model structure. When it is compatible, it can generate an approxi-
mation of the original input–output system. Here, the relation IS is a representation
12.3 Input–Output System Covering 153

of the approximation and will be called an associated input–output system of the

covering.

Theorem 12.8. Each input–output system S ⊂ X × Y has an input–output system

covering, which is also an input–output system partition.

Proof: We define the collections S, X , and Y as follows:

S  {S}, X  {X }, Y  {Y }. (12.31)

The rest of the proof is clear. QED

Example 12.2. An input–output system covering (respectively, partition)

{S, X , Y , IS }

is termed a trivial covering (respectively, partition) if |Ix | 1 or |Iy | 1. We now

construct an input–output system S ⊂ X × Y which does not have a nontrivial
partition. Let S ⊂ X × Y be the system defined by

S  {(x1 , y1 ), (x2 , y2 )}, (12.32)

where x1 , x2 ∈ X and y1 , y2 ∈ Y are distinct elements in X and Y , respectively.

Then S may not have a nontrivial partition. In fact, by contradiction, suppose that
S has a nontrivial input–output system partition {S, X , Y , IS }. Then the associated
input–output system IS ⊂ Ix × Iy must satisfy |Ix | 2  |Iy |. In fact, if |Ix |> 2, there
is an i ∈ Ix such that Xi ∈ X contains neither x1 nor x2 . That is, for each j ∈ Iy ,

(Xi × Yj ) ∩ S  ∅. (12.33)

Equation (12.33) contradicts the condition that (Xi × Yj )∩S ∈ S and each element
in S is nonempty. A similar argument shows that |Iy | has to be 2.
Suppose X  {X1 , X2 } and Y  {Y1 , Y2 }. Then for each i  1, 2, we have

Xi ∩ D(S)
 ∅
 Yi ∩ R(S). (12.34)

To guarantee that for each i  1, 2 and j  1, 2,

(Xi × Yj ) ∩ S  Sψ(i,j) ∈ S, (12.35)

the sets Xi and Yj must contain {x1 , x2 } and {y1 , y2 }, respectively. This fact implies
that if the input space X  {x1 , x2 } and the output space Y  {y1 , y2 }, then the
system covering {S, X , Y , IS } would be a trivial covering, contradiction.
When a system covering is a partition, the associated input–output system IS is
characterized by the following results.
154 12 Decomposability Conditions

Theorem 12.9. Let S  {Si ⊂ S: i ∈ I } be an input–output system partition of a

system S ⊂ X × Y . Then the following holds true:

(i, j) ∈ IS ↔ (∃(x, y) ∈ S)(x ∈ Xi and y ∈ Yj )

↔ (Xi × Yj ) ∩ S
 ∅, (12.36)

where IS is the associated input–output system of the partition, Xi ∈ X and Yj ∈ Y .

In particular, if the system IS exists, IS is unique for the given sets S, X , and Y .

Proof: The following is clear:

(∃(x, y) ∈ S)(x ∈ Xi and y ∈ Yj ) ↔ (Xi × Yj ) ∩ S

 ∅. (12.37)

Suppose (i, j) ∈ IS . The definition of a system covering implies that (Xi × Yj ) ∩

S  Sψ(i,j)
 ∅; so there is (x, y) ∈ Sψ(i,j)  (Xi × Yj ) ∩ S. Consequently,
(∃(x, y) ∈ S)(x ∈ Xi and y ∈ Yj ). Conversely, suppose there exists (x, y) ∈ S such
that x ∈ Xi and y ∈ Yj . Since (x, y) ∈ (Xi × Yj ) ∩ S ⊂ S  ∪α∈I Sα , we have

(x, y) ∈ Sk for some k ∈ I . (12.38)

Let ψ −1 (k)  (i , j ). Since (Xi × Yj ) ∩ S  Sψ(i ,j )  Sk and (x, y) ∈ Sk hold,

we have x ∈ Xi and y ∈ Yj . Since X and Y are partitions, we must have Xi  Xi
and Yj  Yj . The uniqueness of existence of the associated input–output system IS
comes from the fact that (i, j) ∈ IS ↔ (Xi × Yj ) ∩ S
 ∅. QED
Let f : X → Y be a mapping from a set X to a set Y . An equivalence relation Ef
can then be defined on X as follows:

xEf y ↔ f (x)  f (y). (12.39)

The quotient set X /Ef will be denoted by X /f .

Theorem 12.10. Suppose that h  (h1 , h2 ): S → S is a modeling relation with

D(S)  X and R(S)  Y . There then exists an input–output system partition
{S, X , Y , IS } of S such that X  X /h1 , Y  Y /h2 , and IS ⊂ (X /h1 ) × (Y /h2 )
is defined by

([x], [y]) ∈ IS ↔ ([x] × [y]) ∩ S

 ∅. (12.40)

A one-to-one correspondence ψ: IS → S/h1 × h2 is given by

ψ([x], [y])  [(x, y)]. (12.41)

Proof: We first show that the correspondence ψ is well defined. Let ([x], [y]) ∈ IS
and (x̂, ŷ) ∈ X × Y be arbitrary such that [x]  [x̂] and [y]  [ŷ]. It suffices to show
that
12.3 Input–Output System Covering 155

Figure 12.4 Epi–mono

decomposition of a modeling
relation h

[(x, y)]  [(x̂, ŷ)]. (12.42)

From the definition of equivalence class, it follows that [(x, y)] 

{(x̂, ŷ) ∈ X × Y : h(x̂, ŷ)  (x, y)}  [(x̂, ŷ)]; that is, ψ is a mapping from IS to
X × Y /h1 × h2 . We now need to show that for each ([x], [y]) ∈ IS , ψ([x], [y]) ∈
S/h1 × h2 ; i.e., there exists (x̂, ŷ) ∈ S such that [(x, y)]  [(x̂, ŷ)]. The last equation
follows from the fact that ([x] × [y]) ∩ S
 ∅.
Second, we show that ψ: IS → S/h1 × h2 is a one-to-one correspondence. Let
[(x̂, ŷ)] ∈ S/h1 × h2 be arbitrary. Since ψ([x̂], [ŷ])  [(x̂, ŷ)] holds, ψ is surjective.
Suppose ψ([x], [y])  ψ([x̂], [ŷ]). Then [x]  [x̂] and [y]  [ŷ]; i.e., the mapping
ψ is injective. QED

Theorem 12.11. Suppose that S ⊂ X ×Y and S ⊂ X ×Y are input–output systems

such that D(S)  X, R(S)  Y , D(S )  X , and R(S )  Y , and h  (h1 , h2 ): S → S
is a modeling relation. Let μ1 : X → X /h1 and μ2 : Y → Y /h2 be the canonical
mappings. Then the following statements hold:
(i) μ  (μ1 , μ2 ) is an epimorphism from S to S/h. In particular, μ is surjective.
(ii) A modeling relation σ  (σ1 , σ2 ): S/h → S can be defined by σ ([x], [y]) 
(h1 (x), h2 (y)) such that σi , i  1, 2, are injective.
(iii) The diagram in Figure 12.4 is commutative; that is, there is an epi–mono
decomposition of h.
(iv) If h1 and h2 are injective, then the modeling relation μ is an isomorphism.
(v) If h is surjective, then σ is an isomorphism.

Proof: (i) It is clear that the canonical mappings μ1 : X → X /h1 and μ2 : Y →

Y /h2 are surjective when defined by μi (m)  [m], for i  1, 2. If (x, y) ∈ S,
and (x, y) ∈ ([x] × [y]) ∩ S, thus μ(x, y)  (μ1 (x), μ2 (y))  ([x], [y]) ∈ S/h.
By Theorem 12.1, the modeling relation μ is an epimorphism. Since there is
(x , y ) ∈ S for any ([x], [y]) ∈ S/h such that [x ]  [x] and [y ]  [y], μ is
surjective.
(ii) Since [x]  [x ] implies h1 (x)  h1 (x ), σ1 ([x])  h1 (x) is well defined. The
mapping σ2 is also well defined. Suppose ([x], [y]) ∈ S/h. Then the definition
of S/h implies there exists (x̂, ŷ) such that (x̂, ŷ) ∈ ([x] × [y])∩S. Consequently,
h1 (x)  h1 (x̂), h2 (y)  h2 (ŷ), and (x̂, ŷ) ∈ S; i.e., σ ([x], [y])  (h1 (x), h2 (y)) ∈
S . Hence, σ is a modeling relation, and σ1 and σ2 are clearly injective.
(iii) The commutativity of the diagram in Figure 12.4 is clear from the definitions
of μ and σ .
156 12 Decomposability Conditions

(iv) If h1 and h2 are injective, h1  σ1 ◦ μ1 says that μ1 is injective. Similarly, the

mapping μ2 is also injective. From statement (i) and Theorem 12.6, we know
that μ is an isomorphism.
(v) The desired result can be shown as in the proof of (iv). QED

12.4 Surjective Complex System Representation of Order n

  

Suppose that S ⊂ X × Y is a system, {Si ⊂ Xi × Yi : i  1, 2, . . . , n} is a collection

of systems, and S ⊂ X × Y satisfies the conditions

X ⊂ X1 × X2 × · · · × Xn and Y ⊂ Y1 × Y2 × · · · × Yn , (12.43)

and

S ⊂ S1 × S2 × · · · × Sn , (12.44)

where we identify each element ((x1 , x2 , . . . , xn ), (y1 , y2 , . . . , yn )) ∈ S with

((x1 , y1 ), (x2 , y2 ), . . . , (xn , yn )) ∈ S1 × S2 × · · · × Sn . 

The system S is called a surjective complex system representation of S with order

  

n if there exist mappings h1 : X → X and h2 : Y → Y such that h1 × h2 (S)  S . S



is called an injective complex system representation of S with order n if there exist

  

mappings k1 : X → X and k2 : Y → Y such that k1 × k2 (S ) ⊂ S and k1 and k2 are 

injective. S is called an isomorphic complex system representation of S with order

 

n if there exist mappings h1 : X → X and h2 : Y → Y such that h  (h1 , h2 ) is an

isomorphism from S to S .
   

Proposition 12.3. For each system S ⊂ X × Y and each natural number n > 0, S
has a surjective, an injective, and an isomorphic complex system representation with
order n.

n 
n
Proof: Let X ⊂ X and Y ⊂ Y be the diagonals of the product sets, respectively.
That is,


X  {(x1 , x2 , . . . , xn ): x ∈ X (xi  x, i  1, 2, . . . , n)} (12.45)

and


Y  {(y1 , y2 , . . . , yn ): y ∈ Y (yi  y, i  1, 2, . . . , n)}. (12.46)

We now define S ⊂ X × Y by letting



((x, x, . . . , x), (y, y, . . . , y)) ∈ S ↔ (x, y) ∈ S , (12.47)

12.4 Surjective Complex System Representation of Order n 157
 

and define mappings h1 : X → X and h2 : Y → Y by h1 (x, x, . . . , x)  x and

h2 (y, y, . . . , y)  y. Then we have the desired results with respect to (h1 , h2 ). QED
  

Theorem 12.12. A system S ⊂ X ×Y has an injective complex system representation

S with order n over a class of systems {Si ⊂ Xi × Yi : i  1, 2, . . . , n} if and only if
  

there exist n partitions Pi  {S ij ⊂ S : j ∈ Ii } of S , for each i  1, 2, …, n, such that

the following conditions hold:


(i) Pi is an input–output system partition of S , for each i  1, 2, …, n.

 

(ii) There exists an embedding σi  (σ1i , σ2i ): S i → Si , where S i is an associated

input–output system of Pi with σ1i and σ2i injective.
 

(iii) Let Xi  {[xj ]i : xj ∈ X } and Yi  {[yj ]i : yj ∈ Y } be the input and output spaces


of S i , respectively, where Xi and Yi are partitions by definition and [xj ]i and

[yj ]i are equivalence classes of Xi and Yi , respectively. Then, for arbitrary
[xi ]i ∈ Xi and [yi ]i ∈ Yi , for each i  1, 2, …, n,
n


n   
[xi ]i
 ∅ → ∃x̂ ∈ X̂ [xi ]i  {x̂} (12.48)
i1 i1

and
n


n   
[yi ]i
 ∅ → ∃ŷ ∈ Ŷ [yi ]i  {ŷ} . (12.49)
i1 i1

Proof: Necessity. Let S ⊂ X × Y be an injective complex system representation

with order n, where X ⊂ X1 × X2 × · · · × Xn and Y ⊂ Y1 × Y2 × · · · × Yn . Let
i  (ix , iy ): X ×Y → Xi ×Yi be the projection defined by equation (11.3). Then


i : S → Si is a modeling relation. Let e  (e1 , e2 ): S → S be an injective modeling



relation. Since i ◦ e  (ix ◦ e1 , iy ◦ e2 ): S → Si is a modeling relation, we can



define a partition Pi of S by


Pi  S /(ix ◦ e1 ) × (iy ◦ e2 ). (12.50)

We now check the conditions (i)–(iii) for these partitions Pi , i  1, 2, …, n. Since

Theorem 12.10 implies that Pi is an input–output system partition, condition (i) holds
  

true. Since the associated input–output system S i is given by S i  S /i ◦ e, there is



an embedding mapping σ : S i → Si due to Theorem 12.11(ii). Hence, condition (ii)

also holds true. We now check condition (iii). In the present case, we have
 

Xi  X /ix ◦ e1 and Yi  Y /iy ◦ e2 . (12.51)

Choose an equivalence class [xi ]i from each Xi , for i  1,2, …, n. The definition
of Xi implies ix ◦ e1 ([xi ]i )  {x̂i } for some x̂i ∈ Xi , so
158 12 Decomposability Conditions

e1 ([xi ]i ) ⊂ −1
ix ({x̂i }). (12.52)

Suppose x̂ ∈ ∩ni1 −1 −1

ix ({x̂i }). Since x̂ ∈ ix ({x̂i }), for each i, we have

(∀i  1, 2, . . . , n)(ix (x̂)  x̂i ). (12.53)

Hence, x̂  (x̂1 , x̂2 , . . . , x̂n ). Consequently, we have

 −1
n  
n  −1
n 
({x̂i })
 ∅ → ({x̂i })  {(x̂1 , x̂2 , . . . , x̂n )}. (12.54)
i1 ix i1 i1 ix

Since, in general, e1 (−1 −1

ix [xi ]i ) ⊆ ∩i1 e1 ([xi ]i ), e1 ([xi ]i ) ⊂ ix ({x̂i }) implies
n


n 
n
e1 ( [xi ]i )
 ∅ → e1 ( [xi ]i )  {(x̂1 , x̂2 , . . . , x̂n )}. (12.55)
i1 i1

Since e1 is injective, ∩ni1 [xi ]i is empty or singleton. A similar argument works

for the collection Yi . Hence, we have showed condition (iii).
  

Sufficiency. Suppose that S satisfies conditions (i)–(iii). Let be S ⊂ S 1 × S n ×



· · · × S n defined by

(([x1 ]1 , [x2 ]2 , . . . , [xn ]n ), ([y1 ]1 , [y2 ]2 , . . . , [yn ]n )) ∈ S ↔

n

   
n
(∀i  1, 2, . . . , n) ([xi ]i , [yi ]i ) ∈ Ŝi and [xi ]i , [yi ]i ∈ Ŝ, (12.56)
i1 i1

where we identify singleton sets with the elements. Let X  D(S ) and Y  R(S )
 

and mappings k1 : X → X and k2 : Y → Y be defined by

k1 (x̂)  ([x̂]1 , [x̂]2 , . . . , [x̂]n ) and k2 (ŷ)  ([ŷ]1 , [ŷ]2 , . . . , [ŷ]n ), (12.57)
 

for each x̂ ∈ X and ŷ ∈ Y . 

Claim. The mapping k  (k1 , k2 ) is an isomorphism from S to S .

 

In fact, let (x̂, ŷ) ∈ S be arbitrary. Theorem 12.9 implies that ([x̂]i , [ŷ]i ) ∈ S i .
Since condition (iii) implies that
n

 
n
[x̂]i , [ŷ]i  (x̂, ŷ) ∈ Ŝ, (12.58)
i1 i1

we have k1 × k2 (x̂, ŷ) ∈ S . That is, k is a modeling relation from S to S . To

show k is one-to-one, we let k1 (x̂)  k1 (x̂ ). It follows from the definition of k1 that
∀i([x̂]i  [x̂ ]i ). The condition (iii) implies that
12.4 Surjective Complex System Representation of Order n 159


n 
n
{x̂}  [x̂]i  [x̂ ]i  {x̂ }. (12.59)
i1 i1

Hence, k1 is injective. The same argument can be used to show that k2 is also
injective. 

To show that k is surjective from S to S , let

(([x1 ]1 , [x2 ]2 , . . . , [xn ]n ), ([y1 ]1 , [y2 ]2 , . . . , [yn ]n )) ∈ S



be arbitrary. Then the definition of S implies that there exists (x̂, ŷ) ∈ S such that


n 
n
{x̂}  [xi ]i and {ŷ}  [yi ]i . (12.60)
i1 i1

Consequently, for any i, x̂ ∈ [xi ]i and ŷ ∈ [yi ]i ; i.e., [x̂]i  [xi ]i and [ŷ]i  [yi ]i .
Therefore,

k1 × k2 (x̂, ŷ)  (([x1 ]1 , [x2 ]2 , . . . , [xn ]n ), ([y1 ]1 , [y2 ]2 , . . . , [yn ]n )). (12.61)


This completes the proof that k is an isomorphism from S to S . 

We now let  S1 × S2 × · · · × Sn . Since there is an embedding σi from S i to Si ,

an injective modeling relation from to S is

((σ11 × σ12 × · · · × σ1n ) ◦ k1 , (σ21 × σ22 × · · · × σ2n ) ◦ k2 ). (12.62)

This ends the proof of Theorem 12.12. QED

  

Theorem 12.13. If a system S ⊂ X × Y has a surjective complex system represen-

tation S with order n over a class {Si ⊂ Xi × Yi : i  1, 2, . . . , n}, then the following
conditions hold:


(i) S has n input–output coverings


i 

Ci  {S j ⊂ S : j ∈ Ii }, i  1, 2, . . . , n; (12.63)


(ii) There is a surjective mapping from Si to an associated input–output system S i

of Ci .


Proof: Let h  (h1 , h2 ): S → S be a surjective modeling relation and i 

(ix , iy ): S → Si the projection. For each i, let
⎧⎛
⎛ ⎞⎞ ⎫
⎨ −1
 −1 ⎬
Ci  ⎝h1 (j) × h2 ⎝ (k)⎠⎠ ∩ Ŝ : (j, k) ∈ Si . (12.64)
⎩ ⎭
ix iy
160 12 Decomposability Conditions
 

Then Ci is a covering of S . In fact, let (x̂, ŷ) ∈ S be arbitrary. Since h is surjective,

(x̂, ŷ)  h1 × h2 (x, y), for some (x, y) ∈ Si . If i (x, y)  (j, k) ∈ Si , then
−1
 −1

(x, y) ∈ (j) × (k). (12.65)
ix iy

That is,
−1
 −1

(x̂, ŷ)  h(x, y) ∈ h( (j) × (k))
ix iy
−1
 −1

 h1 ◦ (j) × h2 ◦ (k). (12.66)
ix iy

Next, let Ci be the subset of Ci defined by deleting empty sets and duplicated sets
and write
−1
 −1
 

Ci  {(h1 ◦ (j) × h2 ◦ (k)) ∩ S :(j, k) ∈ ISi ⊂ Si }. (12.67)

ix iy

Then Ci is an input–output system covering. In fact, let Ixi  D(ISi ) and Iyi  R(ISi )
and
−1

Xi  {h1 ◦ (j): j ∈ Ixi }, (12.68)
ix

and
−1

Yi  {h2 ◦ (k): k ∈ Iyi }. (12.69)
iy

 

Then Xi and Yi are coverings of the sets X and Y , respectively. Next, let the desired
relation between Ixi and Iyi be ISi , and let ψ be the identity mapping on ISi . Then (ii)
holds true. QED
  

Theorem 12.14. Suppose that S ⊂ X × Y is a system and

{Si ⊂ Xi × Yi : i  1, 2, . . . , n} is a family of systems. If the following condi-


tions are true, then S has a surjective complex system representation with order n,
S ⊂ X × Y , where

X  X1 × X2 × · · · × Xn and Y  Y1 × Y2 × · · · × Yn , (12.70)
12.4 Surjective Complex System Representation of Order n 161

and

S ⊂ S1 × S2 × · · · × Sn . (12.71)


(i) S has n input–output coverings


i 

Ci  {S j ⊂ S : j ∈ I i }, i  1, 2, . . . , n. (12.72)

(ii) There is a surjective modeling relation σi  (σ1i , σ2i ) from Si to an associated



input–output system S i of Ci .
i
 
i  

(iii) Let Xi  {X j ⊂ X : j ∈ Ixi } and Yi  {Y k ⊂ Y : k ∈ Iyi } be the input and output

 
i 
i
spaces of S i , respectively. Then for any X i ∈ Xi and Y i ∈ Yi ,

 n 

n  
 i
X̂ii
 ∅ →  X̂i   1, (12.73)
 
i1 i1

and
 n 

n  
 
Ŷii
 ∅ →  Ŷii   1. (12.74)
 
i1 i1

Proof: Suppose (i)–(iii) hold true and ISi is an associated input–output system of Ci .
Define a system S ⊂ IS1 × IS2 × · · · × ISn as follows:
n

 
n

(j1 , . . . , jn , k1 , . . . , kn ) ∈ S ↔ ∀i((ji , ki ) ∈ IS ) and
i
X̂ji ,
i
Ŷji ∈ Ŝ, (12.75)
i

i1 i1

where each singleton set is identified with its element. Let X  D(S ) and Y 
 

R(S ) and h1 : X → X and h2 : Y → Y be the mappings defined by


n 
i
h1 (j1 , . . . , jn )  X ji , (12.76)
i1

and


n 
i
h2 (k1 , . . . , kn )  Y ki . (12.77)
i1

Condition (iii) and the definition of S ensure that h1 and h2 are well defined.
We now show that h  (h1 , h2 ) is a surjective modeling relation from S to S . If
162 12 Decomposability Conditions


(x, y) ∈ S , then the definitions of S and h imply that (h1 (x), h2 (y)) ∈ S and hence h


is a modeling relation. Suppose that (x̂, ŷ) ∈ S is arbitrary. For each i,

  
i 
i
S⊂ S j → ∃r ∈ I i ((x̂, ŷ) ∈ S r ). (12.78)
j∈I i

The definition of an input–output system covering implies that for some (ji , ki ) ∈
ISi ,

i 
i 
i


(X ji × Y ki ) ∩ S  S r . (12.79)

i 
i
Consequently, x̂ ∈ X ji and ŷ ∈ Y ki ; that is,

h1 × h2 (j1 , . . . , jn , k1 , . . . , kn )  (x̂, ŷ). (12.80)

The modeling relation h is surjective. We now define S ⊂ S1 × S2 × · · · × Sn and

σ : S → S by

(x, y) ∈ S ↔ ∀i((xi , yi ) ∈ Si ) and

σ (x, y)  (σ11 (x1 ), . . . , σ1n (xn ), σ21 (y1 ), . . . , σ2n (yn )) ∈ S . (12.81)


Then h ◦ σ is a surjective modeling relation from S to S . QED

  

Theorem 12.15. A system S ⊂ X × Y has an isomorphic complex system repre-

sentation S of order n over a family {Si ⊂ Xi × Yi : i  1, 2, . . . , n} of systems if and


only if there are partitions Pi , i  1, 2, …, n, of S such that the following conditions

hold true:
(i) Pi is an input–output system partition. 

(ii) There is an isomorphism σi from an associated input–output system S i of Pi

to Si , where Si , ⊂ Si and S ⊂ S1 × · · · × Sn .
(iii) Condition (iii) in Theorem 12.14 is true.
The proof is similar to those of Theorems 12.12 and 12.14 and is left to the reader.

The study of decomposability conditions enlightens the research of relations

between “parts” and “whole.” The results in this chapter are mainly based on the
excellent article (Takahara 1982). For additional related results, see Lin (1999). The
concept of decomposition of systems shows its power in Chapter 11.

References

Lin, Y.: General Systems Theory: A Mathematical Approach. Kluwer Academic and Plenum Pub-
lishers, New York (1999)
Takahara, Y.: Decomposability conditions for general systems. Large Scale Syst. 3(1), 57–65 (1982)
 Part II
The Systemic Intuition
Chapter 13
How Systems Could Be Intuitively Seen

Starting with this chapter, this book turns its attention and focus to applications of
the abstract systems theory established in the previous chapters in areas of business
decision making by introducing the so-called systemic yoyo model. This model plays
the role of intuition and playground for developing appropriate analytic models to
attack various issues encountered in the business world. In other words, this yoyo
model provides an actual stage for various investigations of business scenarios to take
place at the height of holistic thinking so that realistically beneficial conclusions can
be derived rigorously.
The previous 11 chapters concentrated on the theoretical development of the foun-
dation of systems science in order to unveil some of the most elementary properties
of the general system. For that reason, all the logic thinking and reasoning were done
by using the abstract language of set theory and symbolic logic. From this chapter on,
because this book will focus on applications in business decision making, all the rel-
evant thinking and reasoning will be presented in a language and fashion appropriate
for business professionals. Other than deriving conclusions definitely and rigorously,
we will also emphasize how conclusions are intuitively seen first before relevant rig-
orous proofs are provided. This emphasis is practically important for the business
professionals who are actually working in the business world, because for most of
these professionals time is money and time is an opportunity. They cannot sit around
for a long time to first come up with some delicate technical proofs to support their
professional intuitions; they do not have the necessary time to wait for data collection
and vigilant analysis of the data before they have to take actions in order to capture
the fleeting business opportunities that could disappear in front of their eyes at any
time moment. This urgency in making business decisions has been especially true for
the current world of business dominated mostly by waves of transient and short-lived
competitive advantages (McGrath 2013).
Because of how this book is divided into three parts: the foundation, the systemic
intuition, and applications in business decision making, if the reader is mathemat-
ically inclined, then she will be able to fully enjoy reading all the chapters in this
book and see how the results from the first part are actually employed either directly
© Springer Nature Switzerland AG 2018 165
J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_13
166 13 How Systems Could Be Intuitively Seen

or indirectly in the other two parts. In particular, studies on large-scale systems in

various chapters of the first part will be practically employed to design strategies
for protecting national economic security against attacks of international speculative
capital in Chapter 22. If the reader is good at quantitative reasoning, but not willing
to spend additional time to study the fundamental properties of general systems, then
she can jump right into the second part and finish up the book through the third part
in order to acquire some of the most up-to-date discoveries and conclusions in the
relevant research endeavors. And if the reader is only interested in learning how the
systemic thinking can be directly used to make business decisions in a timely fashion,
then she can go right to the third part of this book, where each rigorously established
conclusion is accompanied by discussions on what the result means in real business
life. In other words, in this case, each and every one of the rigorous proofs in the third
part of this book can be skipped over without interrupting the learning of some of
the most intriguing conclusions in business decision making developed on the basis
of systemic thinking and reasoning.
This chapter consists of five sections. Section 13.1 introduces the systemic yoyo
model. Section 13.2 looks at some of the fundamental properties of systemic yoyo
fields. Section 13.3 studies how systemic fields interact with each other. Section 13.4
develops four laws on state of motion for the purpose applying systemic logic of
thinking beyond natural science into social sciences. Section 13.5 concludes this
chapter by addressing four questions: Historically, how the development of calculus
becomes possible? What does this history imply to the research of systems science?
Why is intuition very important in terms of the exploration of knowledge and the
development of scientific theories? And how valid is figurative analysis in decision
making?

13.1 The Systemic Yoyo Model

In the area of business decision making, most, if not all, scenarios under consideration
are such systems that are open, meaning that they interact with their environments,
are complex, meaning that they involve mutually influencing factors making it diffi-
cult or impossible to really tell which affects what, and are giant, meaning that many
variables are involved making any conventional scientific study difficult. Specifically,
the great many variables involved in the study alone generally make such investiga-
tions hard to manage. And, the large-scale complexity involved can make the already
tedious and difficult problems even more drastically challenging.
In the traditional science, each and every number is seen as a point on the real
number line or a point in a high-dimensional Euclidean space. Each and every quanti-
tative variable is treated as a moving point on the real number line or in the Euclidean
space. Therefore, when great masses of such moving points are jointly considered,
the problem is most likely beyond the limited capability of human imagination.
Among others, this challenge is really the core reason for the extreme difficulty
faced in dealing with systems that are open, complex, and giant involved in most
13.1 The Systemic Yoyo Model 167

business decision-making scenarios. Thus, the following natural question arises at

this particular junction:
Can we introduce a systemic intuitive background different from that of Euclidean spaces
so that such a great number of moving particles and masses of various combinations of these
particles can be more easily managed?

The reason why the desired systemic intuition needs to be different from Euclidean
spaces is because many recent and age-old challenges, such as nonlinearity, chaos,
etc., facing the traditional science are really consequences of how Euclidean spaces
are composed of: linear (straight) real number lines. In other words, the traditional
science attempts to investigate curvatures using tools developed linearly in fictitious
spaces [for more in-depth discussion, see Lin and OuYang (2010)]. And the intu-
ition underneath the previous question is that because systems science, as the second
dimension of science, studies systemhood, while the traditional science studies thing-
hood [for details, see Klir (1985)], the difficulty one faces when dealing with open,
complex, giant systems is really a difficulty one experienced in the first dimension of
science. Hence, there should be a relatively more manageable means in the second
dimension for one to carry out such large-scale tasks of decision making.
To geometrically comprehend this logic of reasoning and why the intuitive feeling
might work out in terms of dealing with open, complex, giant systems, one can simply
imagine a two-dimensional city that is surrounded by a circular steady wall (or an
enclosed bounded area in the two-dimensional plane). If there is no gap in the wall,
which is a closed curve in the plane, then it will be difficult or impossible for any
army to break into the city from within the two-dimensional space. Now, if one smart
engineer designs an air strike by making use of a third dimension, which was not
known before because everyone was only thinking within the two-dimensional plane,
then the forces of his or her side can be easily parachuted into the city along the third
dimension. That is, difficulties one faces when making decisions in business that
involve open, complex, giant systems are partially and mainly due to the reason that
he/she maintains himself or herself in the first dimension of science without truly
taking advantage of systems science, the newly found second dimension of science.
When we study the business world and treat every scenario of decision making
as system problem solving (Klir 1985), we can theoretically confirm the observa-
tion that many seemingly unrelated business decisions are actually related one way
or another and different areas of the business landscape—large or small, global or
regional—evolve and develop in concert. When one decision is changed, many other
seemingly unrelated decisions and consequent human behaviors also change accord-
ingly. That means that changes in the business world should be seen as a whole, and
the whole evolution of each and every system of concern should be considered in
order to understand how business entities or systems evolve as wholes and how they
are related to each other.
In whole evolutions, what is extremely important is the commonly existing dis-
continuity with which transitional change s (or blow-ups), such as the appearance of
disruptive technologies, occur. These blow-ups represent how the gradual and con-
tinuous changes of old structures, such as business models, management techniques,
168 13 How Systems Could Be Intuitively Seen

Figure 13.1 Relationship between planar infinity and three-dimensional North Pole

competitive advantages, etc., are being replaced by new structures. If the system,
abstracted out of the relevant scenario of business decision making, is described
truthfully by a mathematical model, then the model is generally nonlinear (Wu and
Lin 2002), and the blow-ups of the model reflect the destruction of old structures and
establishment of new ones.
By borrowing the formality of calculus, the concept of blown-ups can be written as
follows: For a given (mathematical or symbolic) model that truthfully describes the
system abstracted out of our business decision making, if its solution u  u(t; t0 , u0 ),
where t stands for time and u0 is the initial state of the system at the initial time
moment t0 , satisfies

lim |u|  +∞, (13.1)

t→t0

and at the same time moment when t → t0 , the underlying business system also goes
through a transitional change, then the solution u  u(t; t0 , u0 ) is called a blown-up
solution, and the relevant business development expresses a blown-up. Thousands
of real-life evolutionary systems (Lin 2009) indicate that disastrous events appear
at the moments of blown-ups in the whole evolutions of systems and magnificent
opportunities appear with the occurrence of disruptive technologies (Christensen
2016).
For nonlinear models in independent variables of time (t) and space (x; x and y;
or x, y, and z), the concept of blown-ups is similar, where blow-ups in the model and
the underlying business decision-making system can appear in time or in space or in
both.
One of the key features of the concept of blown-ups is the quantitative ∞, a
mathematical indeterminacy. This indeterminacy in applications has caused troubles,
such as instabilities and calculation spills that interrupt each and every working
computer (Lin and OuYang 2010). To systemically understand this symbol ∞, let
us look at the Riemann ball (Figure 13.1), representing a curvature space, which is
well studied in the complex analysis in mathematics.
The modeling used in problem solving in the traditional science is to project a
moving point xi of the curvature surface of the ball onto a point xi of the plane, a
13.1 The Systemic Yoyo Model 169

simplified model for a multi-dimensional Euclidean space. By using this projection,

both −∞ and +∞ are connected through a blow-up. In particular, when the dynamic
point xi travels through the North Pole N on the sphere, the corresponding image
xi on the plane shows up as a reversal change from −∞ to +∞ through a blow-
up. That is, the planar points ±∞ stand implicitly for direction changes of one
dynamic point on the sphere at the polar point N instead of any indeterminacy.
Speaking differently, the phenomenon of being directionless, as shown by blow-ups
of a lower-dimensional space, represents a direction change of movement in a higher-
dimensional curvature or nonlinear space. Therefore, blow-ups and brown-ups can
specifically represent implicit transformations of spatial dynamics, which is the same
with disruptive technologies that represent transformations of the landscape of the
business world. Through blown-ups, problems of indeterminacy of a distorted space,
as developed in the traditional science, are transformed into determinant situations
of a more general system in a curvature and nonlinear space.
Corresponding to the previous implicit transformation between the plane and the
Riemann ball, one can also relate quantitative ±∞ in one-dimensional space to a
dynamic movement on a circle, a curvature space of a higher dimension through the
point of view of modeling using differential equations.
This general discussion on how problems are modeled in the traditional sci-
ence indicates that nonlinearity mathematically stands (mostly) for singularities in
Euclidean spaces and realistically stands for eddy motions or movements in curva-
ture and nonlinear spaces. Such motions are about structural evolutions, a natural
consequence of uneven evolutions of materials, resources, information, investments,
etc.
Next, let us look at the general dynamic system and how it is related to eddy
motions. Newton’s second law of motion is that the acting force is equal to the product
of the mass of the object being acted on its acceleration of motion. Symbolically, we
have

d v 
m  F. (13.2)
dt
In the business world, we know that the elapse of time is different from one
company to another, and the availability of resources, talents, information, etc., is
different from one firm to another. That is, time is uneven across the landscape of the
business world and the distribution of all other economic variables is also uneven,
which is the same as Einstein’s concept of uneven time and space of materials’ evo-
lutions. So, we can assume that F  −∇S(t; x, y, z), where S  S(t; x, y, z) stands
for the time–space distribution of the external acting object. Let ρ  ρ(t; x, y, z) be
the density or the internal structure of the business entity being acted upon. Then,
equation (13.2) for a unit structure of the business entity being acted upon can be
written as

d u 1
− ∇S(t, x, y, z), (13.3)
dt ρ(t, x, y, z)
170 13 How Systems Could Be Intuitively Seen

where u is used to replace the original v in order to represent the meaning that each
change of some economic factor is a consequence of mutual reactions of the internal
structures of various economic entities. Now, if ρ is not a constant, as expected in
the business world, then equation (13.3) becomes
 
d (∇x × u ) 1
 −∇x × ∇S  0. (13.4)
dt ρ

where ∇x represents a change in the x-direction. So, equation (13.4) stands for
an eddy motion because of the cross products. In other words, a nonlinear mutual
reaction between materials’ or business entities’ uneven internal structure s and the
unevenness of the external forcing objects and entities will definitely produce eddy
motions.
At this junction, it is important to note that the commonality of eddy motions
is confirmed not only by daily observations of surrounding natural phenomena and
events of the business world but also by laboratory studies from as small as atomic
structures to as huge as nebular structures of the universe. In fact, what is shown in
the previous paragraphs implies that eddies appear out of the unevenness of inter-
nal structures of the entity of concern and in the distribution of various resources.
Thus, if the business world is seen at the height of structural evolutions, then the
business world is simple with only two forms of motions: clockwise rotations and
counterclockwise rotations. Also, the vertical vectority in structures has been very
practically implemented in the common form of motion of the business world and
the universe.
Now, a natural question would arise for the reader: Why are all structures and
organizations of the universe in rotational movements? According to what is just
discussed earlier about the uneven distribution in internal structures or organization
of various resources, such as investments, talents, information, etc. (or Einstein’s
uneven space and time), we can assume that all materials or organizations have
uneven structures. Out of these uneven structures, there naturally exist gradient s,
a concept of calculus. With gradients, there appear forces. Combined with uneven
arms of forces, the carrying materials or organization will have to rotate in the form
of moments of forces.
Based on the previous paragraphs, we can now imagine that the landscape of the
world economy is entirely composed of eddy currents—smaller regional economies
and then in even much smaller scales, companies, and individuals. The existing
eddies exist in different sizes and scales and interact with each other. That is, the
landscape of the world business is a huge ocean of eddies, which change and evolve
constantly. One of the most important characteristics of spinning fluids, including
spinning solids, is the difference between the structural properties of inwardly (con-
verging) and outwardly (diverging) spinning pools and the discontinuity between
these pools. Due to the stirs in the form of moments of forces, in the discontinuous
zones, there exist sub-eddies and sub-sub-eddies (see Figure 13.2, where sub-eddies
are created naturally by the large eddies M and N). Their twist-ups (the sub-eddies)
contain highly condensed amounts of materials and energies, representing where
13.1 The Systemic Yoyo Model 171

Figure 13.2 Appearance of

sub-eddies

(a) How two harmonic yoyo fields create subeddies

(b) How two inharmonic yoyo fields create subeddies

business opportunities emerge. In other words, the traditional frontal lines and sur-
faces (in meteorology or, in other words, business trends) are not simply expansions
of particles or smaller scale phenomena without any structure. Instead, they represent
twist-up zones or new opportunities concentrated with irregularly or unconventional
structured entities and energies and investments. That is where the so-called small
probability events or disruptive technologies appear and small probability infor-
mation and small probability chance of success are observed and acted upon by
entrepreneurs.
Based on what is discussed above, the concepts of inputs, outputs, and converging
and diverging eddy motions are coined together (Wu and Lin 2002) in the model
shown in Figure 13.3 for each object and every system imaginable. In particular,
each system or object considered in a study or business decision making is a multi-
dimensional entity that spins about its axis. If we fathom such a spinning entity in our
three-dimensional space, we will have a structure as shown in Figure 13.3a. The side
of the inputs sucks in all things, such as materials, information, talents, investments,
etc. After funneling through the neck, all things, such as products, services, etc.,
are spit out in the form of outputs. Some of the materials, spit out from the end of
outputs, never return to the other side, and some will (Figure 13.3b). Such a structure,
as shown in Figure 13.3a, is called a (systemic) yoyo due to its general shape.
More specifically, what this model says is that each entity in the universe, be it
a tangible or intangible, a business entity, an organization, a culture, a civilization,
etc., can all be seen as a kind of realization of a certain multi-dimensional spinning
yoyo with a spin field around it, either visible or invisible. It stays in a constant
spinning motion as depicted in Figure 13.3a. If it does stop its spinning, it will
no longer exist as an identifiable system. What Figure 13.3c shows is that due to
the interactions between the eddy field, which spins perpendicularly to the axis of
spin, of the model and the meridian field, which rotates parallel to the axis of spin
(Figure 13.3b), all the materials returning to the input side travel along a spiral
trajectory. As for why materials in the universe and organizations and individuals
172 13 How Systems Could Be Intuitively Seen

(a) The yoyo model in our 3dimensional space

(b) The side view (c) The spiral trajectory of meridian field

Figure 13.3 Eddy motion model of the general system

in the business world evolve rotationally in the first place and continue to do so,
according to the uneven distributions of international structures, as discussed earlier,
and the concept of uneven space and time of Einstein’s theory of relativity (1997),
we know that all materials and economic entities possess uneven internal structures
or uneven distribution of resources. Out of the unevenness, there naturally exist
gradients. With gradients, there will appear forces. Combined with uneven arms of
forces, the carrying materials and organizations will have to rotate in the form of
moments of forces (Lin 2009, p. 31).

13.2 Properties of Systemic Yoyo Fields

Because each yoyo spins as in Figure 13.3a, the spin field consists of an eddy field that
is perpendicular to the axis of rotation of the yoyo structure as shown in Figure 13.3b,
and a meridian field that moves parallel to the axis (see Figure 13.3b, c). The meridians
go into the center of the input side, through the neck, and then out the output side.
They travel through the space and return to the center of the input side. Somehow,
we can imagine that these meridians help to hold different layers of the eddy field
of the yoyo structure together. For the convenience of communication, for any given
yoyo structure, the input side will be referred to as the South Pole of the structure
and the output side the North Pole.
13.2 Properties of Systemic Yoyo Fields 173

Here, the word “spin” is used to capture the “angular momentum” or the presence
of “angular momentum” intrinsic to a body as opposed to orbital angular momentum
of angular momentum that is the movement of the object about an external point. For
example, the spin of the earth stands for the earth’s daily rotation about its polar axis.
The orbital angular momentum of the earth is about the earth’s annual movement
around the sun. Generally, a two-dimensional object spins around a center (or a
point), while a three-dimensional object rotates around a line called an axis, where
the center and the axis must be within the body of the object.
The concept of spin has been, respectively, studied in mathematics, astronomy,
quantum mechanics, and social science areas. For example, the theory and practice
of public relations heavily involve the concept of spin, where a person, such as a
politician, or an organization, such as a publicly traded company, signifies his often
biased favor of an event or situation. Although traditional public relations rely on
creative presentation of the underlying facts, “spin” tends to imply disingenuous,
deceptive, and/or highly manipulative tactics used to influence public attitudes and
opinions (Stoykov and Pacheva 2005; Bernays 1945).
In theory, we can think of the totality of some interrelated business entities and
economic agents as a systemic yoyo if this yoyo is situated in isolation from other
yoyo structures. Since business systems are of various kinds and scales, the business
world can be seen as an ocean of eddy pools of different sizes, where each pool spins
about its visible or invisible center or axis. For example, the spin field of air in a
tornado is a yoyo field in our three-dimensional physical space. In the solenoidal
structure, at the same time when the air within the tornado spins about the eye in
the center, the systemic yoyo structure continuously sucks in and spits out air. The
tornado takes in things, such as water on the bottom, and lifts up everything it took
in; then it gives out the things from the top of the spinning field. Simultaneously, the
tornado also breathes in and out with air in all horizontal directions and elevations.
When the tornado takes in more than it gives out, the tornado will grow larger with
increasing effect on everything along its path. If the opposite holds true, then the
tornado is in its process of dying out. If the intake and output of the tornado are
in equilibrium, then the tornado can last for at least a while. Generally, a systemic
yoyo experiences a period of stable existence after its initial formation and before
its disappearance.
For the convenience of our discussion in this book, we assume that the spinning
of yoyo structures follows the following left-hand rule 1:

Left-Hand Rule: When holding our left hand, the four fingers represents the spin-
ning direction of the eddy plane and the thumb points to the North Pole direction
along which the yoyo structure sucks in and spits out things at its center (the neck).
(Note: It can be seen that in the business world, systemic yoyos do not have to comply
with this left-hand rule.)

As influenced by the eddy spin, the meridian directional movement of things in the
yoyo structure is actually slanted instead of being perfectly vertical. In Figure 13.3c,
the horizontal vector stands for the direction of spin on the yoyo surface toward the
174 13 How Systems Could Be Intuitively Seen

Figure 13.4 Three-quark

yoyos

reader and the vertical vector the direction of the meridian field, which is opposite of
that in which the yoyo structure sucks in and spits out materials. Other than breathing
in and out things from the input (the South Pole) and output (the North Pole) sides,
the yoyo structure also takes in and gives out things in all horizontal directions and
elevations, just as in the case of tornadoes discussed earlier.
In the process of taking in and giving out things, formed is an outside surface of
the yoyo body that is mostly imaginary of our human mind; this surface holds most
of the contents of the spinning yoyo. The density of things of this surface decreases
as one moves away from the yoyo structure. The maximum density is reached at
the center of the eddy field. As the spin field, which is the combination of the eddy
and meridian fields, constantly takes in and gives out things, there does not exist any
clear boundary between the yoyo structure and its environment, which is analogous
to the circumstance of a tornado that does not have a clear-cut separation between
itself and its surroundings.
To study the general structure of systemic yoyos, let us borrow the term of quark
structure from Chen (2007). Following the notations of Chen (2007), each spinning
yoyo, as shown in Figure 13.3a, is seen as a two-quark structure, where we imagine
the yoyo is cut through its waist horizontally in the middle, and then the top half is
defined as an absorbing quark and the bottom half a spurting quark. A three-quark
yoyo can look like one of those in Figure 13.4a or b. Here, in Figure 13.4a, the
absorbing u-quarks stand for local spinning pools, while together they also travel
along in the larger eddy field P in which they are part of. Similarly in Figure 13.4b,
the spurting d-quarks are regional spinning pools. At the same time when they spin
individually, they also travel along in the large yoyo structure of N. In all these cases,
the spinning directions of these u- and d-quarks are the same except that each u-quark
spins convergently (inwardly) and each d-quark divergently (outwardly).
Different yoyo structures have different numbers of absorbing u-quarks and d-
quarks. And, the u-quarks and d-quarks in different yoyos are different due to vari-
ations in their mass, size, spinning speed and direction, and the speed of absorbing
and spurting materials. This end is well supported by discoveries in particle physics
and organizational structures of business entities, where the number of divisions
related to attracting investments, securing resources, talents, etc., and the number of
13.2 Properties of Systemic Yoyo Fields 175

(a) An observed uniform flow (b) An observed chaotic pattern

Figure 13.5 Patterns observed in a dishpan experiment

divisions related to selling are generally not equal to each other. All related details
are omitted here.
To better understand what is discussed next, let us look at the well-known dishpan
experiment. In the late 1950s, Fultz et al. (1959) constructed the following exper-
iment: They partially filled a cylindrical vessel with water, placed it on a rotating
turntable, and subjected it to heating near the periphery and cooling near the center.
The bottom of the container is intended to simulate one hemisphere of the earth’s sur-
face; the water, the air above this hemisphere; the rotation of the turntable, the earth’s
rotation; and the heating and cooling, the excess external heating of the atmosphere
in low latitudes and the excess cooling in high latitudes.
To observe the pattern of flows at the upper surface of the water, which was
intended to simulate atmospheric motion at high elevations, these authors sprinkled
some aluminum powder. A special camera that effectively rotated with the turntable
took time exposures so that a moving aluminum particle would appear as a streak,
and sometimes each exposure ended with a flash, which could add an arrowhead to
the forward and end of each streak. The turntable generally rotated counterclockwise,
as does the earth when viewed from above the North Pole.
Even though everything in the experiment was arranged with perfect symmetry
about the axis of rotation, such as no impurities added in the water, and the bottom of
the container was flat, Fultz and his colleagues observed more than they bargained
for. First, both expected flow patterns appeared, as shown in Figure 13.5a, b, and the
choice depended on the speed of the turntable’s rotation and the intensity of the heat-
ing. Briefly, with fixed heating, a transition from circular symmetry (Figure 13.5a) to
chaotic flow (Figure 13.5b) would take place as the rotation increased past a critical
value. With the sufficiently rapid but fixed rate of rotation, a similar transition would
occur when the heating reached a critical strength, while another transition back to
symmetry would occur when the heating reached a still higher critical strength.
Now, let us fit this dishpan experiment to the discussion above by imagining either
the top or the bottom of each yoyo as a spinning dish of fluids. Then the patterns
as observed in the dishpan experiment suggest that in theory there could exist such
a yoyo structure that it has n u-quarks and m d-quarks, where n ≥ 1 and m ≥ 1 are
arbitrary natural numbers, and each of these quarks spins individually and along with
each other in the overall spinning pool of the yoyo structure.
176 13 How Systems Could Be Intuitively Seen

13.3 Interactions of Systemic Fields

When more than two business entities are concerned with, the interactions between
them, a multi-body problem, become impossible to describe analytically. This end is
witnessed by the difficulties scholars have faced in the investigation of the three-body
problem since about three hundred years ago when Newton introduced his laws of
motion regarding two-body systems. So, to achieve anything meaningful in terms of
the interactions between business organizations (or systemic yoyos), in this section,
we will focus on figurative expressions.
First, let us look at how yoyo fields are classified.
Just as electric or magnetic fields, where like fields repel each other and opposite
fields attract each other, for yoyo fields the same principle holds true, where the like
ends (the North Poles or South Poles) repel and the opposite attract. In particular,
when the South Poles of two yoyos face each other, Figure 13.6a, where S stands for
the input side and N the output side, although the two South Poles have the tendency
to attract each other, the meridian fields X and Y actually repel each other apart, where
mX1 repels against mY1 and mX2 against mY2 . So, in this case, there does not exist any
attraction between the two input sides, since these S sides cannot feed anything to
each other directly, and both have to get input materials from their meridian fields.
When the North Poles of yoyos X and Y face off (head on), Figure 13.6b, the yoyos
are pushed apart, where in comparison the meridian fields of X and Y only come
into place with attraction when the edges of the North Pole sides of X and Y directly
face each other. When the South Poles of a yoyo face the North Pole of another
yoyo (Figure 13.7a), where what is spurted out of yoyo Y can go and do go directly
into yoyo X. At the same time, the meridian fields of both yoyos X and Y have the
tendency to combine into one field too (Figure 13.7b).
The yoyo’s (vertical) absorbing/spurting and (all directional) taking in/giving out
along the meridian field direction create an active field that can be visibly seen as a

Figure 13.6 Repulsion of like yoyo fields

Figure 13.7 How two yoyos can potentially become one yoyo
13.3 Interactions of Systemic Fields 177

Figure 13.8 Tendency for yoyos to line up

field with source (or active field). In particular, in a given yoyo, if its eddy field is
identified as its magnetic field, then the active (the meridian) field will correspond
to the electric field. In this case, the so-called N and S poles of the magnetic field
correspond, respectively, to the diverging and the converging sides of the yoyo struc-
ture, while the positive and negative electric fields to the converging and diverging
eddy fields. Similarly, the eddy field can be identified as an electric field and so cor-
respondingly the meridian field as the magnetic field. Now, if the eddy field is seen
as a gravitational field, then the meridian field will be the corresponding magnetic
field. In this case, we can see that the gravitational field can be both attraction and
repulsion that exist side by side, the former corresponds to the converging side of
the eddy field, and the later corresponds to the diverging side.
Next, let us look at the relative positioning of two yoyo structures X and Y in
Figure 13.8. Then the meridian field A of X fights against C of Y so that both X and
Y have the tendency to realign themselves in order to reduce the conflicts along the
meridian directions. Similarly in Figure 13.8b, the meridian field A1 of yoyo X fights
against B1 of Y. So, the yoyos X and Y also have the tendency to realign themselves
as in the previous case.
Of the two yoyos X and Y above, let us assume that one of them is mighty and
huge and the other so small that it can be seen as a tiny particle. Then, the tiny particle
yoyo m will be forced to line up with the much mightier and larger spinning yoyo
M in such a way that the axis of spin of the tiny yoyo m is parallel to that of M
and that the polarities of m and M face the same direction. For example, Figure 13.9
shows how the particle yoyo m has to rotate and reposition itself under the powerful
influence of the meridian field of the much mightier and larger yoyo structure M.
In particular, if the two yoyos M and m are positioned as in Figure 13.9a, then the
meridian field A of M fights against C of m so that m is forced to realign itself by
rotating clockwisely in order to reduce the conflicts with the meridian direction A
of M. If the yoyos M and m are positioned as in Figure 13.9b, the meridian field A1
of yoyo M fights against B1 of m so that the particle yoyo m is inclined to readjust
itself by rotating once again clockwisely. If the yoyos M and m are positioned as in
Figure 13.9c, then the meridian field A2 of yoyo M fights against B2 of m so that the
tiny particle yoyo m has no choice but to reorient itself clockwisely to the position
as in Figure 13.9b. As what has been just analyzed, in this case, yoyo m will further
be rotated until its axis of spin is parallel to that of M and its polarities face the same
directions as M.
178 13 How Systems Could Be Intuitively Seen

Figure 13.9 How mighty spinning yoyo M bullies particle yoyo m

Figure 13.10 Manifestation

of an induced yoyo field

Now, let us imagine a tiny particle yoyo m, say a unit in the supply chain of a
business activity, where the supply chain is theoretically seen as the circular loop in
Figure 13.10a, where the X’s stand for an overriding yoyo field of the environment
going into the page. Then, the axis of spin and its polarities of the tiny yoyo m have
to live up with those of an abstract mighty spinning yoyo, part of whose eddy field
is symbolized by the X’s in Figure 13.10a.
Based on the left-hand rule 1 and the north direction of the yoyo field outside
the circuit or supply chain (the reason why we use the direction outside the circuit
instead of inside is because in the eddy field of the mighty spinning yoyo, the circuit
would be only the size of a dot), we can see how the particle yoyos, such as different
units in the supply chain, line up inside the circuit as shown in Figure 13.11. This
line-up of particle yoyos inside the circuit also determines the north direction of the
superpositioned eddy field along the circuit, which fights against the increase in the
intensity of the original yoyo field. This end in fact provides a systems theory support
for the validity of the following generalized Lenz’s law.

Generalized Lenz’s Law: There is an induced yoyo current in any closed, conduct-
ing loop if and only if the yoyo flux that is a yoyo field through the loop is changing.
The direction of the induced current is such that its induced yoyo field opposes the
change in the yoyo flux.

If we ponder over the discussion above in more detail, it is not hard to see that
the imaginary return circuit is not necessary. As long as the yoyo field (an eddy or
meridian field), as indicated by the X’s, changes, there will be a manifested yoyo
13.3 Interactions of Systemic Fields 179

Figure 13.11 How particle yoyos line up inside the circuit

Figure 13.12 Formation of yoyo dipoles

field in the space (Figure 13.10b). This kind of manifested yoyo field is a special
kind of yoyo field.
To better understand the generalized Lenz’s law, let us imagine a person, living
in an established environment. If one day he suddenly plans to change his own
behaviors completely, what will happen to him is that he will find that everyone else
in the environment instantly becomes united in the fight against his planned changes.
The more he is determined to carry out his planned changes, the more resistance he
will experience from the united environmental front.
Second, let us look at the phenomenon of the so-called yoyo dipoles. Similar to
the concept of electric charge, if an organization, when seen as a high-dimensional
spinning yoyo, possesses an overall attraction pull or repulsion push, we say that the
organization has a yoyo charge. When an organization with neutral yoyo charge is
affected by an external yoyo charge, the organization can experience the phenomenon
of polarization, just as the case when scraps of paper are affected by an electric
charge. In particular, when a positive yoyo charge is placed near the organization,
the negative component yoyos inside the organization have the potential to face
the positive charge, while positive component yoyos face away from the external
charge. Similarly, if the external charge is negative, the charged component yoyos
inside the organization will face in directions opposite to those described above. That
is, under the influence of an external powerful yoyo charge, a yoyo dipole within the
organization can be formed, Figure 13.12. Because opposite yoyo poles attract, the
external yoyo charge attracts the created yoyo dipole within the organization.
When a yoyo dipole is situated in a uniform yoyo field E,  its two poles of the
dipole will be acted upon by the yoyo field so that the yoyo dipole experiences a
torque, Figure 13.13.
180 13 How Systems Could Be Intuitively Seen

Figure 13.13 Yoyo dipole

experiences a torque

Figure 13.14 Yoyo dipole

in an uneven yoyo field

Figure 13.15 Systemic

yoyos combine through their
meridian fields

When a yoyo dipole is placed inside an uneven yoyo field E,  where in different
locations the field intensities and directions are different. For instance, in the field
of a particle yoyo, the yoyo dipole will spin until it is parallel to the direction of the
field. If the particle yoyo carries a positive charge, then the positive direction of the
yoyo dipole agrees with that of the field. Because the yoyo field intensity is different
from one location to another, there is a difference between the forces experienced
by the two poles of the yoyo dipole. The end closer to the yoyo charge experiences
a pull greater than the push at the end pointing to the positive direction. Therefore,
the net force acting on the yoyo dipole is not zero and points to the yoyo charge.
If the yoyo charge is negative, similar results hold true (Figure 13.14). That is, the
following result holds true: The force experienced by a yoyo dipole placed in an
uneven yoyo field always points to the direction of greater field intensity.
When systemic yoyos combine through their meridian fields, Figure 13.15, there
exists a field potential pit (trap) over which the yoyos attract to each other. At the
same time, between the same poles of the meridian fields, a field potential rampart
exists to repel the poles from each other. No matter whether it is the field potential
pit or rampart, the field intensity is different from one yoyo structure to another. For
yoyos of relatively stronger field strengths, their field potential pits are relatively
deeper and their potential ramparts are relatively higher, meaning further away from
the centers of the yoyo structures; for yoyos of relatively weaker field strengths, their
field potential pits are relatively shallower and their potential ramparts relatively
lower.
When systemic yoyos link to each other through their meridian fields so that
the input sides (the South Poles) face with their output sides (the North Poles),
13.3 Interactions of Systemic Fields 181

Figure 13.16 Combined

yoyo structure of four
component yoyos

their eddy fields also link and interact with each other, Figure 13.16. The overall
combined yoyo structure looks like a large ring along which the component yoyos
are connected through their quarks. In Figure 13.16a, the oval arrows stand for the
eddy fields’ directions of the component yoyos and the line arrows the meridian
fields’ directions. Because the component yoyos are connected through their quarks
and their eddy and meridian fields’ directions satisfy the left-hand rule, the resultant
field structure is well depicted in Figure 13.16b. In this figure, the horizontal ellipse
stands for the loop made up of the component yoyos connected through meridian
fields and the vertical ellipses the eddy fields of the components.
When spinning yoyos link through their opposite meridian poles, their relative
positions will be fixed. So, between these yoyos, a kind of meridian field bond is
formed, where the spinning centerlines are connected to each other with the meridian
fields pointing to the same direction. As indicated by Figure 13.16, each field bond
involves such parameters as bond strength, bond angle, bond force, bond length,
bond energy, etc.
All different kinds of interactions between systemic fields, some of which are
discussed in this section, can be readily employed to support the investigation on
how business organizations interact and associate with each other.

13.4 Laws on State of Motion

Based on the discussions in the previous sections, we know that spin is the funda-
mental evolutionary feature and characteristic of organizations. In this section, we
will study the figurative analysis method of the systemic yoyo model and see how to
apply it to establish laws on state of motion that have been useful in studying eco-
nomic growth and development. Specifically, after introducing the figurative analysis
method, we will learn four laws on state of motion that generalize Newton’s laws
of motion while no external forces are required for these laws to work. As what’s
well known, these Newton’s laws are one of the reasons why physics is an “exact”
science. And, the rest of this book will show that these new laws of state of motion
are readily applicable to social sciences and humanity areas. This section is based
on Lin (2007).
The First Law on State of Motion: Each imaginable and existing entity in the uni-
verse and the business world is a spinning yoyo of a certain dimension. Located on
182 13 How Systems Could Be Intuitively Seen

(a) Object m is located in a diverging eddy (b) Object m is located in a diverging eddy and
and pulled by a converging eddy M pulled or pushed by a diverging eddy M

(c) Object m is located in a converging eddy (d) Object m is located in a converging eddy and
and pulled by a converging eddy M pulled or pushed by a diverging eddy M

Figure 13.17 Acting and reacting models with harmonic yoyo structures

the outskirt of the yoyo is a spin field. Without being affected by another yoyo struc-
ture, each particle in the said entity’s yoyo structure continues its movement in its
orbital state of motion.

This law implies that when a force impresses on an object or organization, the
force must be from the outside of the entity being affected and that the external force
comes from the spin field of the yoyo structure of another entity. In other words, the
acting force is in fact the attraction or gravitation of the acting yoyo field, created
within the acting field by the unevenness of its internal structure.
By combining this first law on state of motion with the yoyo model, we derive
models on how a small entity m is acted upon by another entity M. In particular,
Figure 13.17a depicts the scenario that m is originally situated in a diverging eddy
before the converging eddy M pulls it over into the spin field of M. Figure 13.17b
describes how m in a diverging eddy can be either pushed (direction (a)) or pulled
(direction (b)) by M. When m is pulled along the direction (b), it will be captured
by the diverging eddy M. In this case, if the spin field M is not powerful enough,
m will not be pulled to the spin field M out of its original field. Figure 13.17c, d
is models for the scenarios on how m is pulled away from its original converging
eddy by another eddy of the harmonic spinning pattern. Here, two spinning patterns
are said to be harmonic if other than the spinning directions, the spin fields look the
same.
To be more specific, yoyo M, exerting a gravitational pull (Figure 13.17a, c) or
a gravitational push (Figure 13.17b, d) on yoyo m, must be much greater than m in
order to have an obvious effect on the state of motion of m and to capture m. That
is, the yoyos M and m are from different levels and have different scales. Otherwise,
if they are of the same scale or level, instead of a dot, m will be the entire eddy N
on the left-hand side in Figure 13.17a–d. Then, m will not be pulled over to M in
Figure 13.17a. Instead, m will be a supplier of resources for the converging eddy
M. In Figure 13.17b, yoyos M and N push each other away. In Figure 13.17c, the
13.4 Laws on State of Motion 183

(a) Object m is located in a diverging eddy (b) Object m is located in a diverging eddy
and pulled by a converging eddy M and pushed or pulled by a diverging eddy M

(c) Object m is located in a converging eddy (d) Object m is located in a converging eddy
and pulled by a converging eddy M and pushed or pulled by a diverging eddy M

Figure 13.18 Acting and reacting models with inharmonic yoyo structures

converging eddies N and M attract each other and might become one depending on
whether or not their spinning directions and angles agree with each other. And in
Figure 13.17d, M serves as a supplier of the converging eddy N.
As for yoyos N and M with their yoyo structures spinning in inharmonic directions
(or patterns), the acting and reacting are modeled in Figure 13.18. From this analysis,
we have the following:
The Second Law on State of Motion: When a constantly spinning yoyo structure
M does affect a tiny yoyo m, which is located in the spin field of another yoyo field
N, the velocity of m will change. More specifically, m experiences an acceleration a
toward the center of M such that the magnitude of a is given by

v2
a
r
where r is the distance between m and the center of M and v the moving speed of
the field of M about distance r away from the center of M. And, the magnitude of the
net pulling force F net that M exerts on m is given by

v2
Fnet  ma  m
r
Similar to what has been done earlier, let us now analyze the situation of two eddy
motions acting on each other in two different angles: (1) They act and react to each
other’s spin field, and (2) one of them is acted upon by another that is of a higher
level and scale. For the first situation, the interaction of the two eddy fields can be
modeled systemically as in Figure 13.19.
By analyzing each spinning direction, it can be seen that in Figure 13.19a, the
acting and reacting pair is more of N exerting a force on M, and N serves as a
supplier of things, energies, etc., to the converging need of M. When both N and M
are relatively stable, M also exerts an opposite but reacting force on N. Otherwise,
184 13 How Systems Could Be Intuitively Seen

(a) N diverges and M converges (b) Both N and M diverges

(c) Both N and M converge (d) N converges and M diverges

Figure 13.19 Same scale acting and reacting spinning yoyos of the harmonic pattern

without such a reacting force, M would collapse and N becomes diluted. That is, the
moment both N and M would have to realign their balances and territories or they
both would simply dissolve and disappear together.
When both fields N and M represent yoyo structures in Figure 13.19b, they truly
exert equal, but opposite, action and reaction forces on each other. Because their spin
fields fight against each other, they will push each other further apart. This is one
situation Newton’s third law is really talking about.
When N and M both converge harmonically as shown in Figure 13.19c, they
exert an attraction or pull on each other. If they are relatively stable, their spin fields
achieve a temporary (roughly) balanced acting and reacting forces. However, sooner
or later, these two spin fields will try to combine and have a tendency to become one
spin field.
As for the situation in Figure 13.19d, it is the same as that in Figure 13.19a except
that the roles of N and M are shifted.
For objects N and M with inharmonic spin fields, we have the modeling diagrams
in Figure 13.20. In particular, for the yoyo structures N and M in Figure 13.20a,
even though the situation is similar to that of Figure 13.19a where N diverges and M
converges, due to the inharmonic spinning directions, N and M act and react to each
other differently of that of Figure 13.19a. In this current case, N exerts a pushing force
on M and does not easily serve as a supplier of things, energies, etc., for M. When
compared to Figure 13.19b, the yoyo structures N and M in Figure 13.20b can work
together more peacefully, because their spin fields do not fight against each other,
even though they also push each other apart. For the situations in Figure 13.20c, the
spin fields N and M attract each other. But from their inharmonic spinning directions,
it can be seen that they will not combine and will never become one. Situation in
Figure 13.20d is exactly the same as that in Figure 13.20a with the roles of N and M
switched.
Based on the analysis above, we have the following:
13.4 Laws on State of Motion 185

(a) N diverges and M converges (b) Both N and M diverge

(c) Both N and M converge (d) N converges and M diverges

Figure 13.20 Same scale acting and reacting spinning yoyos of inharmonic patterns

The Third Law on State of Motion: When the spin fields of two yoyo structures N
and M act and react on each other, their interaction falls in one of the six scenarios
in Figure 13.19a–c and Figure 13.20a–c. And, the following hold true:
(1) For the cases in (a) of Figures 13.19 and 13.20, if both N and M are relatively
stable temporarily, then their action and reaction are roughly equal but in oppo-
site directions during the temporary stability. In terms of the whole evolution
involved, the divergent spin field (N) exerts more force on the convergent field
(M) than M’s reaction peacefully in the case of Figure 13.19a and violently in
the case of Figure 13.20a.
(2) For the cases (b) in Figures 13.19 and 13.20, there are permanent equal, but
opposite, actions and reactions with the interaction more violent in the case of
Figure 13.19b than in the case of Figure 13.20b.
(3) For the cases in (c) of Figures 13.19 and 13.20, there is a permanent mutual
attraction. However, for the former case, the violent attraction may pull the
two spin fields together and have the tendency to become one field. For the
latter case, the peaceful attraction is balanced off by their opposite spinning
directions. And, the spin fields will coexist permanently.

Now, for the second situation, where one of the acting and reacting yoyo fields is
acted upon by another that is of a higher level and scale. Assume the former is m,
and the latter is M. This situation can be modeled systemically by the diagrams in
Figures 13.17 and 13.18, where m is a particle in a higher-level eddy flow N before
it is acted upon on by M.
If the spinning yoyo m is located in a spin field N and affected by the converging
spin field M (Figure 13.17a), due to the harmonic spinning directions of N and M, m
experiences an action from M pulling it away from its current orbit and a reaction from
N to keep m on its current orbit. Because the influence of M is external, depending
on the structure of the spin field of m, m might continue its rotation in the spin field
N with its orbit altered, or it might be on its way to depart from N. If m does leave
186 13 How Systems Could Be Intuitively Seen

(a) Small m is situated as in Figure 13.17(b) (b) Small m is situated as in Figure 13.18(a)

(c) Small m is situated as in Figure 13.17(c) (d) Small m is situated as in Figure 13.18(d)

Figure 13.21 Yoyo field m is thrown into a sub-eddy created jointly by N and M

the spin field of N, it has two possibilities: It might fall into the spin field of M, and
it might simply be shot out to an area unrelated to either N or M. When m takes the
second possibility, it experiences an equal but opposite action from M and reaction
from N.
In terms of the situation in Figure 13.17b, due to the violent nature facing each
other between the spin fields of N and M, either m will be kept in the spin field of
N longer than it would be without the effect of M, or it will be captured by the spin
field of M, or it will fall in a sub-eddy created by both N and M; see Figure 13.21a.
In terms of Figure 13.17c, due to their conflicting spinning directions at the adja-
cent area between N and M, yoyo field m experiences both a push and a pull from the
spin field of M. The act of pushing comes from the spinning direction of M opposite
to that in which m travels. The experience of pull is caused by the fact that M is a
converging eddy motion. In this case, the acts of push and pull experienced by m are
neither equal nor opposite to each other, Figure 13.21c.
When m is located in a converging field N and acted upon by a diverging M,
Figure 13.17d, what m experiences is (1) more of a push toward the center of N
without much of any reaction or (2) thrown out of N and M along the spinning
directions of N and M into their adjacent area.
The situation in Figure 13.18a is similar to that of Figure 13.17b. The scenario in
Figure 13.18b is similar to that of Figure 13.17a except that m can only be thrown
out by both N and M along their spinning directions to areas not related to either
N or M. This is when m experiences equal but opposite action and reaction from
N and M individually. The experience m in Figure 13.18c is similar to that of
Figure 13.17d except that both N and M apply a gravitational pull on m. Depending
on the characteristics of m’s own spin field, it may stay in the field of N, or travel to the
field of M, or be thrown out by both N and M along their common spinning direction
in the adjacent area. When m is thrown out, it experiences an equal, but opposite,
action and reaction from N and M individually. Now, the situation in Figure 13.18d
13.4 Laws on State of Motion 187

is similar to that of Figure 13.17c except that M helps to push m to converge to the
center of N faster. See Figure 13.21 for more details.
Note: The sub-eddies in Figure 13.21a are both converging, because N and M
are suppliers for them and sources of forces for their spins. The sub-eddies in
Figure 13.21b are only spinning currents. They serve as middle stop before supplying
to the spin field of M. The sub-eddies in Figure 13.21c are diverging. And, the sub-
eddies in Figure 13.21d are only spinning currents similar to those in Figure 13.21b.
This round of analysis leads to the following:
The Fourth Law on State of Motion: When the spin field M acts on a tiny parti-
cle m, rotating in the spin field N, m experiences equal, but opposite, action and
reaction, if it is either thrown out of the spin field N and not accepted by that of M
(Figures 13.17a, d and 13.18b, c) or trapped in a sub-eddy motion created jointly
by N and M (Figures 13.17b, c and 13.18a, d). In all other possibilities, m does not
experience equal and opposite action and reaction from N and M.
When organizations and business firms are modeled as yoyo fields, these laws on
state of motion vividly describe how these socioeconomic entities interact with each
other.

13.5 Remarks

This section consists of three parts. The first part discusses the importance of devel-
oping the systemic yoyo model as an intuition and playground for systemic reasoning
and thinking. The second part demonstrates by using particular examples the use-
fulness of intuitions played by the well-known Cartesian products. And the third
part addresses the validity of analysis and reasoning based on shapes and dynamic
graphs.

13.5.1 How Calculus, a Pinnacle Achievement of Science,

Was Made Possible?

The seventeenth century is very conspicuous in the history of mathematics. In partic-

ular, early in the century, John Napier revealed his study of logarithms, Thomas Har-
riot and William Oughtred contributed to the codification of algebra, Galileo Galilei
started the science of dynamics, and Johann Kepler announced his laws of planetary
motion. And later in the century, Gerard Desargues and Blaise Pascal opened a new
field of pure geometry. Rene Descartes launched modern analytic geometry, Pierre
de Fermat laid the foundation of modern number theory, and Christiaan Huygens
contributed majorly to the theory of probability and other fields. Then toward the
end of the century, Isaac Newton and Gottfried Wilhelm Leibniz created calculus,
which immediately caught the attention of the scientific world.
188 13 How Systems Could Be Intuitively Seen

Figure 13.22 Two-

dimensional Cartesian
coordinate system

The gateway toward the beginning of calculus was the modern analytic geometry,
where Descartes used a simple way, known as Cartesian coordinate system, to bridge
geometry and algebra so that algebra can be readily employed to investigate problems
of geometry. In particular, the two-dimensional Cartesian coordinate system is the
well-known plane consisting of two real number lines, Figure 13.22, where all plane
geometric figures studied in Euclidean geometry can be expressed by using numbers
and algebraic equations.
What is extremely important about such Cartesian coordinate systems is the result
that any n-dimensional linear space, as defined in abstract algebra, is isomorphic to
the n-dimensional Cartesian coordinate system that is made up of n real number
lines, where n is a natural number greater than or equal to 1.
Other than providing a whole box of tools for solving practical problems, the
importance of calculus is also demonstrated by what age-old problems it can help
resolve. For example, integration can be employed to solve such problems as com-
puting areas, volumes, and lengths of arcs, the challenge of which can be traced back
to at least the time of the Eudoxus (ca. 370 B.C.), involving throughout history such
famous scholars as Archimedes, Kepler, Cavalieri, Torricelli, Fermat, Pascal, Saint-
Vincent, Barrow, and others. On the other hand, differentiation has originated in the
problem of drawing tangents to curves and in finding maximum and minimum values
of functions on the ideas set forth by Fermat in 1629 with Isaac Barrow (1630–1677)
making important contributions to the theory of differentiation.

13.5.2 Usefulness of Intuition

Graphic intuitions have been widely employed throughout history in developing ini-
tial understandings, theories, and applications. For example, the original Euclidean
geometry from more than two thousand years ago is entirely built on visual imagina-
tions of the fictitious plane. Even so, this theory has been successfully employed for
practical uses since the ancient times. And the theory of ancient medicine, such as the
one originated in China more than 5 thousand years ago, is partially constructed on
13.5 Remarks 189

Figure 13.23 Intuition used

in calculus

the basis of the visual abstraction of graphic presentations. For example, the theory
of acupuncture is developed by connecting known acupuncture points on the general
graphic representation of human bodies (Forrest 2013).
As for our modern times, calculus and its related theories, which jointly constitute
the foundation of the entire modern science, are also constructed on the graphic
intuition of the Cartesian coordinate system. For example, to show the limit of the
function sin x/x is 1 as the variable x approaches 0, and the intuition in Figure 13.23
is employed. So, one can establish the following inequality through comparing the
shaded areas of triangle OAB (Figure 13.23a), sector OAC (Figure 13.23b), and
triangle ODC (Figure 13.23v):
sin x cos x
1< < ,
x 1
from which follows the result: lim sin x/x  1.
x→0
As a matter of fact, permeating the entire spectrum of modern science, including
such fundamental methodologies as calculus and all related theories, as mentioned
above, probability theories, statistics, and all other theories on uncertainty, similar
intuitions based on Cartesian coordinate systems are widely employed. For details,
see Kline (1972).

13.5.3 Validity of Figurative Analysis

In this chapter, our discussions have been developed heavily on the analysis of shapes
and dynamic graphs. So, to any educated mind out of the current formal school
system, he/she will naturally question the validity of such a method. To address this
concern, let us start with the concept of equal quantitative effects; for details, see Wu
and Lin (2002; Lin 1998a).
By an equal quantitative effect, it means the eddy effects with nonuniform vor-
tical vectorities existing naturally in systems of equal quantitative movements due
to the unevenness of materials. Here, equal quantitative movements stand for such
movements that quasi-equal acting and reacting entities, be they objects or organi-
zations, are involved or two or more quasi-equal mutual constraints are concerned
with. What’s significant about equal quantitative effects is that they can easily throw
calculations of equations into computational uncertainty (Lin and OuYang 2010).
For example, if two quantities x and y are roughly equal, then x − y becomes a
190 13 How Systems Could Be Intuitively Seen

computational uncertainty involving large quantities with infinitesimal increments.

This end is closely related to the second crisis in the foundations of mathematics
(Forrest 2013).
Based on studies of chaos (Lorenz 1993), Wu and Lin (2002) show that nonlinear
equation systems represent equal quantitative movements, where minor changes in
the initial condition s lead to dramatic changes in the corresponding solutions. Such
extreme volatility existing in the solutions can be easily caused by changes of a digit
many places after the decimal point. Such a digit place far away from the decimal
point in general is no longer practically meaningful. That is, when equal quantitative
effects are involved, one faces with either the situation where no equation can be
reasonably established or the situation that the established equation cannot be solved
with valid and meaningful solution(s); for details, see Lin and OuYang (2010).
In other words, the concept of equal quantitative effects computationally declares
that equations are not eternal and that there is not any equation under equal quanti-
tative effects. That is the reason why in practical applications one needs to convert
numbers (quantities) back into shapes (figurative structures) (Lin 1998a), because
she wants to describe and make use of the formality of eddy irregularities, such as
blown-ups, discontinuities, etc., caused by interactions of two or more eddy fields.
These irregularities represent some of the very difficult challenges facing the transi-
tional science.
Because each scientific theory is developed on a particular set of axioms or pos-
tulates, there are the problems of disagreement between scientific theories and irreg-
ularities existing in objective evolutions of matters and business organizations and
that distinct properties are quantified and abstracted into indistinguishable numbers
or numerical variables. Such incapability of the traditional science, especially the
modern mathematics, has been shown time and again in areas of practical prediction
s. For example, because theoretical studies cannot yield any meaningful and effec-
tive method to foretell drastic weather changes, especially about small or near-zero
probability disastrous weather systems [In fact, the study of chaos theory indicates
that weather patterns are chaotic and unpredictable. A little butterfly fluttering its
tiny wings in Australia can drastically change the weather patterns in North America
(Gleick 1987)], the technique of live report has been widely employed. However,
in the area of financial market prediction s, it has not been so lucky that the tech-
nique of live report can possibly applied effectively. Due to equal quantitative effects,
the movements of prices in the financial marketplace have been truly chaotic when
viewed from the contemporary scientific point of view.
In other words, the concept of equal quantitative effects helps to make the episte-
mology of natural sciences gone from solids to fluids and complete the unification of
natural and social sciences. More specifically, after the previous four laws on state of
motion are established, they can be readily employed to study social systems, such as
military conflicts, political struggles, economic competitions, etc.; for details, see the
following chapters in this book. The situation here is expected to be parallel to that
when Newton’s laws of motion were introduced physics became an “exact” science.
13.5 Remarks 191

After having discussed the importance of equal quantitative effects and inevitable
failures of traditional science, including modern mathematics, a natural question
arises on figurative analysis: How can such methods also suffer from the same doom?
As a matter of fact, the usage of graphs in human exploration of the nature goes
back as far as our recorded history can go. For example, any written language consists
of an array of graphic figures. In terms of figurative analysis, one early work is the
book, named Yi Ching [or the Book of Changes, as known in English (Wilhalm and
Baynes 1967)]. For now, no one knows exactly when this book was written and who
wrote it. All known is that the book has been around since about three thousand
years ago. In that book, the concept of Yin and Yang was first introduced and graphic
figures are used to describe supposedly the development and evolution of all matters
and events in the world. When Leibniz (a founder of calculus) had a hand on that
book, he introduced the binary number system and base p number system in modern
mathematics (Kline 1972). Later on, Bool furthered this work and laid down the
foundation for the modern computer technology to appear.
In modern days, figures and figurative analysis are readily seen and employed in
many areas of human endeavors. One good example is π, the number closely related
to circles. Because this number cannot be written in the traditional fashion in either
the decimal form or the fraction form, we simply use a figure π to indicate it. The same
idea is employed to write all irrational numbers. In the area of weather forecasting,
figurative analysis is used each and every day in terms of weather maps. In terms
of studies of financial markets, a big part of the technical analysis is developed on
graphs.

References

Bernays, E.: Public Relations. Bellman Publishing Co., Boston, MA (1945)

Chen, G.R.: The Original State of the World: The Theory of Ether Whirltrons. Tianma Books
Limited, Hong Kong (2007)
Christensen, C.M.: The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail,
Reprint edn. Harvard Business Review Press, Boston (2016)
Einstein, A.: Collected Papers of Albert Einstein. Princeton University Press, Princeton (1997)
Forrest, J.Y.L.: A Systemic Perspective on Cognition and Mathematics. CRC Press, an imprint of
Taylor and Francis, Balkema, The Netherlands (2013)
Fultz, D., Long, R.R., Owens, G.V., Bohan, W., Kaylor, R., Weil, J.: Studies of thermal convection
in a rotating cylinder with some implications for large-scale atmospheric motion. Meteorological
Monographs, vol. 21, no. 4. American Meteorological Society (1959)
Gleick, J.: Chaos: Making a New Science. Viking Press, New York (1987)
Kline, M.: Mathematical Thought from Ancient to Modern Times. Oxford University Press, Oxford
(1972)
Klir, G.: Architecture of Systems Problem Solving. Plenum Press, New York (1985)
Lin, Y. (guest editor): Mystery of Nonlinearity and Lorenz’s Chaos. A special double issue. Kyber-
netes: Int. J. Cybern. Syst. Manage. Sci. 27(6–7), 605–854 (1998a)
Lin, Y.: Systemic yoyo model and applications in Newton’s, Kepler’s laws, etc. Kybernetes: Int. J.
Cybern. Syst. Manage. Sci. 36(3–4), 484–516 (2007)
192 13 How Systems Could Be Intuitively Seen

Lin, Yi: Systemic Yoyos: Some Impacts of the Second Dimension. Auerbach Publications, an
imprint of Taylor and Francis, New York (2009)
Lin, Y., OuYang, S.C.: Irregularities and Prediction of Major Disasters. Auerbach Publications, an
imprint of Taylor and Francis, New York (2010)
Lorenz, E.N.: The Essence of Chaos. Washington University Press, Seattle (1993)
McGrath, R.G.: The End of Competitive Advantage: How to Keep Your Strategy Moving As Fast
as Your Business. Harvard Business Review Press, Boston (2013)
Stoykov, L., Pacheva, V.: Public Relations and Business Communications. Ot Igla Do Konetz, Sofia
(2005)
Wilhalm, R., Baynes, C.: The I Ching or Book of Changes, 3rd edn. Princeton University Press,
Princeton, NJ (1967)
Wu, Y., Lin, Y.: Beyond Nonstructural Quantitative Analysis: Blown-Ups, Spinning Currents and
the Modern Science. World Scientific, River Edge, New Jersey (2002)
Chapter 14
Whole Evolutions, Where Systemic Yoyos
Come from

After introducing the brief history of the concept of systems and the systemic yoyo
model in the previous chapter, this chapter looks at the theoretical foundation on why
such an intuitive model of general systems holds for each and every system that is
either tangible or imaginable.
This chapter is organized as follows: Section 14.1 defines the concept of blown-
ups and provides its mathematical and realistic meanings. Section 14.2 looks at the
properties of blown-ups by using quadratic models, nth degree polynomial models,
n ≥ 3, and nonlinear time–space evolution equations. Section 14.3 considers issues
facing the quantitative infinity ∞ by carefully analyzing an implicit transformation
between Euclidean spaces and curvature spaces, motions represented by general
dynamic system, and by comparing wave motions and eddy motions. Section 14.4
addresses the concepts of equal quantitative movements and effects. The presentation
of this chapter is concluded in Sect. 14.5.

14.1 Blown-Ups: Moments of Transition in Evolutions

When we study nature and treat everything we see as a system (Klir 2001), one fact
we can easily observe is that many systems in nature evolve in concert. When one
thing changes, many other seemingly unrelated things alter their states of existence
accordingly. If we attempt to make predictions regarding an evolving event, such as
the price of a stock in the financial market, where the market behavior is dependent on
the behaviors of the market participants, or what is forthcoming in the weather system,
where nature goes through its normal course of evolution without being affected by
how we think it would come out to be, we may very well find ourselves in a situation
that is somewhere in between where we do not have enough information and where
we have too much information. It has been shown (Soros 2003; Liu and Lin 2006;
Lorenz 1993) that no matter which situation we are in, either too much information or
too little, we face with uncertainties. That is why we propose (OuYang et al. 2009) to
© Springer Nature Switzerland AG 2018 193
J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_14
194 14 Whole Evolutions, Where Systemic Yoyos Come from

look at the evolution of the system or event whose future outlook we need to predict
as a whole. That is, when developments and changes naturally existing in the natural
environment are seen as a whole, we have the concept of whole evolutions. And,
in whole evolutions, other than continuities, widely assumed and studied in modern
mathematics and science, what seems to be more important and more common is
discontinuity, with which transitional changes (or blown-ups) occur. These blown-
ups reflect not only the singular transitional characteristics of the whole evolutions
of nonlinear equations, but also the changes in old structures being replaced by new
structures.
Thousands of case studies have shown the fact (Lin 2009) that reversal and tran-
sitional changes are the central issue and an extremely difficult open problem of
prediction science, because the well-developed method of linearity, which tries to
extend the past rise-and-fall into the future, does not have the ability to predict
forthcoming transitional changes and what will occur after the changes. In terms
of nonlinear evolution models, blown-ups reflect destructions of old structures and
establishments of new ones. Although studies of these nonlinear models reveal the
limits and weaknesses of calculus-based theories, we can still make use of their ana-
lytic forms to describe to a certain degree the realisticity of discontinuous transitional
changes of objective events, materials, and organizations. So, the concept of blown-
ups is not purely theoretical; it is also realistic. By borrowing the form of calculus,
we can write the concept of blown-ups as follows: For a given (mathematical or
symbolic) model that truthfully describes the real-life situation of our concern, if its
solution u  u(t; t0 , u0 ), where t stands for time and u0 the initial state of the system,
satisfies

lim |u|  +∞, (14.1)

t→t0

and at the same time moment when t → t0 ; the underlying real-life system also goes
through a transitional change, then the solution u  u(t; t0 , u0 ) is called a blown-up
solution and the relevant real-life movement expresses a blown-up. Our analysis of
thousands of real-life cases of various evolutionary systems (Lin 2009) indicates
that disastrous events appear at the moments of blown-ups in the whole evolutions
of systems.
For nonlinear models in independent variables of time (t) and space (x, or x and
y, or x, y, and z), the concept of blown-ups is defined similarly, where blow-ups in
the model and the underlying real-life system can appear in time or in space or in
both. For instance, if the time–space evolution equation is written as follows:

∂t u  g(t; u, ∂x u), (14.2)

where u is an n×1 matrix of the state variables, g(t; u, ∂x u) n×1 matrix of functions in
t, u, and ∂x u, ∂x and ∂t stand for the differential operations ∂/∂x and ∂/∂t, respectively.
Assume that the initial (or boundary) condition is
14.1 Blown-Ups: Moments of Transition in Evolutions 195

Figure 14.1 A transitional

blown-up

u(t0 , x)  u0 (x). (14.3)

Then, when the solution u  u(t, x; t0 , u0 ) or ux  ux (t, x; t0 , u0 ) changes, for t ∈

[t0 , tb ), continuously, and when t → tb , the following holds true

lim |u|  +∞, (14.4)

t→t0

or

lim |ux |  +∞, (14.5)

t→t0

then either u or ux is called a blown-up solution, if the boundary value problem,

consisting of equations (14.2) and (14.3), truly describes a real-life system and at the
time moment t  tb , this real-life system goes through a transitional change. In the
rest of this chapter, we assume that this assumption holds true.
Based on the evolutionary behaviors of the real-life system before and after the
transition, we can classify blown-ups into two categories: transitional and nontran-
sitional blown-ups. A blown-up is transitional if the development of the real-life
system after the special blow-up moment in time or space is completely opposite to
that before the blow-up. For example, if the evolution of the system grows drastically
before the blow-up and the development of the system after the blow-up starts from
nearly ground “zero,” then such blown-up is transitional (Figure 14.1). Otherwise, it
is called nontransitional (Figure 14.2).

14.2 Properties of Blown-Ups

To understand the mathematical characteristics of blown-ups, let us look at the fol-

lowing constant-coefficient equation:
196 14 Whole Evolutions, Where Systemic Yoyos Come from

Figure 14.2 A
nontransitional blown-up

u̇  a0 + a1 u + · · · + an−1 un−1 + un  F(u), (14.6)

where u is the state variable and a0 , a1 , . . . , an−1 are constants.

14.2.1 Blown-Ups of Quadratic Models

When n  2, equation (14.6) becomes

u̇  a0 + a1 u + u2 , (14.7)

which is one of the simplest general nonlinear evolution models and is also a spe-
cial Riccati equation [for all omitted technical details in this chapter, please consult
with (Wu and Lin 2002)]. This equation has found a wide range of applications in
the science and applications of prediction. Let us now first discuss this equation’s
properties of blown-ups, which are important in terms of both theoretical and prac-
tical implications. Here, we will not apply Riccati variable transformations to solve
equation (14.7). Instead, we will present our analysis with different values of a0 and
a1 .
Case 1. When   a12 − 4a0  0, equation (14.7) becomes
 2
1
u̇  u + a1 . (14.8)
2

Now, integrating by separating variables produces

1
u  − a1 − (A0 + t)−1 , (14.9)
2
where A0 is the integrating constant, determined by the given initial value. Evidently,
the solution equation (14.9) contains a blown-up when t  −A0 . To this end, let us
analyze the details as follows:
14.2 Properties of Blown-Ups 197

Figure 14.3 Curve #2

stands for the integral curve
when A0 > 0, while curves #1
and #3 for A0 < 0, where
u-axis is located at t  −A0

(a) If A0 > 0, then u gradually stabilizes at the equilibrium state − 21 a1 as time t

approaches +∞. So, in the interval t ∈ [0, +∞), the evolution of u is continuous.
(b) If A0 < 0, then when t → t0  −A0 , the state u → +∞. At this moment, u
experiences a discontinuous singularity and a transitional change after the sin-
gularity. This kind of discontinuity has nothing to do with the smoothness of the
given initial field, showing the essential characteristics of nonlinear evolutions.
See Figure (14.3).
Case 2. When   a12 − 4a0 > 0, equation (14.7) becomes
 2
1 1 
u̇  u + a1 − a12 − 4a0 . (14.10)
2 4

By separating the variables and integrating the previous equation, we obtain

 
 U (u) − X 

ln   Xt + A0 , (14.11)
U (u) + X 

where U (u)  u + a1 /2, X  a12 − 4a0 /2, and A0 is the integration constant. For
this scenario, we can analyze the blown-up situation as follows:
(a) If the movement of our interest is limited to |U (u)| < X , which is a bounded
movement, then the solution of equation (14.10) is given as follows:
 
A0 a1
u  X · tanh Xt + − . (14.12)
2 2

In this case, the development of the evolution is continuous without experiencing

any blown-up, Figure 14.4.
(b) If the range of movement satisfies |U (u)| > X , which stands for a unbounded
motion, then we have
198 14 Whole Evolutions, Where Systemic Yoyos Come from

Figure 14.4 Continuous

evolution without
experiencing any blown-up
for the case of a1 < 0

Figure 14.5 A blown-up

solution for the case of a1 < 0

 
A0 a1
u  X · coth −Xt − − . (14.13)
2 2

If A0 < 0 and when

A0
t → tb  − , (14.14)
2X
the solution in equation (14.13) experiences a blown-up, Figure 14.5.
What is analyzed indicates that under a fixed set of conditions, whether or not
a model experiences discontinuities is dependent on the given initial field. So, it is
difficult for numerical solutions of nonlinear evolution equations to be bounded in
each iteration step, no matter which smoothing scheme is employed. In this sense, it
can be said that various integration schemes, designed to solve nonlinear evolution
equations, cannot really avoid experiencing “explosive” growth or being trapped in
“error-value spiral” computations due to the evolutionary singularities. This is where
many of the paradoxes and “fascinating” conclusions in the Lorenz’s chaos theory
are originated.
Case 3. When   a12 − 4a0 < 0, equation (14.7) becomes

u̇  U 2 (u) − X 2 . (14.15)
14.2 Properties of Blown-Ups 199

Integrating this equation provides the following:

a1
u  Y · tan(Yt + A0 ) − . (14.16)
2

where Y  4a0 − a12 /2 and A0 is the integration constant.
So, when
π
t → tb  Y −1 + n − A0 ,
2
equation (14.16) contains periodic transitional blown-ups.
From this detailed analysis, it can be seen that under different conditions, the
solution of the same nonlinear model can either be continuous and smooth or expe-
rience either blown-ups or periodic transitional blown-ups. So, even for the simplest
nonlinear evolution equations, the well-posedness of their evolutions in terms of dif-
ferential mathematics is conditional. Here, the requirements for well-posedness are:
the solution exists; it is unique, and it is stable (or continuous and differentiable).
Note: The condition of well-posedness is very important for the majority of the
applied mathematics to work.

14.2.2 Blown-Up Properties of nth Degree Polynomial

Models, n ≥ 3

When n  3, other than the situation that F(u)  0 has two real solutions, one of
which is of multiplicity 2, equation (14.6) experiences blown-ups. For details, please
consult with (Wu and Lin 2002). For the general case n, even though the analytical
solution of equation (14.6) cannot be found exactly, the blown-up properties of the
solution can still be studied through qualitative means. To this end, based on the
fundamental theorem of algebra, let us assume that equation (14.6) can be written as
 q1  qm
u̇  F(u)  (u − u1 )p1 . . . (u − ur )pr u2 + b1 u + c1 . . . u2 + bm u + cm ,
(14.17)

where pi and qj , i  1, 2, …, r and j  1, 2, …, m, are positive whole numbers, such

that

m
n pi + 2 qj ,   b2j − 4cj < 0
i1 j1

for j  1, 2, …, m. Without loss of generality, assume that u1 ≥ u2 ≥ . . . ≥ ur , then

the blown-up properties of the solution of equation (14.17) are given in the following
theorem.
200 14 Whole Evolutions, Where Systemic Yoyos Come from

Theorem 14.1. If the following conditions hold true, then the solution of an initial
value problem in the form of equation (14.3) contains blown-ups,
1. When ui , i  1, 2, …, r, does not exist, that is, F(u)  0 does not have any real
solution; and
2. If F(u)  0 does have real solutions ui , i  1, 2, …, r, satisfying u1 ≥ u2 ≥
. . . ≥ ur ,
(a) When n is an even number, if u > u1 , then u contains blown-up(s);
(b) When n is an odd number, no matter whether u > u1 or u < ur , there always
exist blown-ups.

A detailed proof of this theorem can be found in (Wu and Lin 2002, pp. 65–66)
and is omitted here.
For higher-order nonlinear evolution systems, in general, each of them can be
simplified into systems of nonlinear evolution equation s in a form similar to that
of equation (14.6). Since the general case is difficult to analyze, let us look at the
following second order nonlinear evolution equation:

ü  a0 + a1 u + a2 u2 + u3 , (14.18)

where ü  ∂ 2 u/dt 2 and ai (i  0, 1, 2) are constants. Now, this equation can be

reduced to the following system of first order equations in two variables:

u̇  v
. (14.19)
v̇  a0 + a1 u + a2 u2 + u3

Multiplying the left-hand side of the second equation by v and the right-hand side
by u and then integrating the resultant equation provides the following:
 
1 1 1
v2  E(u)  2 u4 + a2 u3 + a1 u2 + a0 u + h0 , (14.20)
4 3 2

where h0 is the integration constant. Now, the system in equation (14.20) can be
reduced into
√
u̇  ± E. (14.21)

Now, we study the blown-up properties of equation (14.21) as follows:

1. If E  0 has a real solution of multiplicity four, then based on equation (14.8), it
can be shown that u contains blown-ups.
2. If E  0 contains two distinct real solutions of multiplicity two, then
1 2 2
E u + b1 u + c1
4
14.2 Properties of Blown-Ups 201

and b21 − 4c1  0. Now, the analysis is done in two cases:

(a) When b21 − 4c1 > 0, based on equation (14.10) it follows that u contains a
continuous bounded solution and a blown-up solution.
(b) When b21 − 4c1 < 0, based on equation (14.14) it follows that u contains a
periodic blown-up solution.

3. If E  0 contains only one real solution, which has multiplicity two, and two
conjugate complex solutions, then E  u2 (u2 + b2 u + c2 ), where we assumed
that the real solution of multiplicity two is zero, and b22 − 4c2 < 0. In this case,
equation (14.21) becomes

u̇  ±u u2 + b2 u + c2 . (14.22)

Let u  1/v. Then the previous equation can be reduced into

dv
  ±dt. (14.23)
c2 V 2 (v) + X

where V (v)  v + b2 /2c2 and X  (4c2 − b22 )/4c2 . Now, let us analyze the
scenario in two cases.
(a) When c2 < 0, we have b22 − 4c2 < 0 and an integration of equation (14.23)
can produce
V (v) √
−arcsin √  −c2 (±t + A0 ), (14.24)
−X

where A0 is a constant. When t takes on a certain special value, v  0 and

u  1/v → ∞. So, the solution of equation (14.22) experiences blown-ups.
(b) When c2 > 0 and either b22 − 4c2 > 0 or b22 − 4c2 < 0, integrating
equation (14.23) provides
√  √
−ln c2 V (v) + c2 V 2 (v) + X  c2 (±t + A0 ), (14.25)

which evidently contains blown-ups.

4. If E  0 has four distinct real solutions, let us assume u1 > u2 > u3 > u4 , then
equation (14.21) becomes

u̇  ± (u − u1 )(u − u2 )(u − u3 )(u − u4 ). (14.26)

So, the continuous bounded movements of equation (14.26) can be written by

using an elliptic function. As for the unbounded movement, we will have to apply
the method of estimation to prove that it contains blown-ups. As a matter of fact,
202 14 Whole Evolutions, Where Systemic Yoyos Come from

when u > u1 and take the “+” on the right-hand side of equation (14.26), we can
obtain the following estimate:

u̇ ≥ (u − u1 )2 . (14.27)

It is not hard to prove that in this case, u contains blown-ups. Similarly, it can be
shown that when u < u4 , u also contains blown-ups.

Example 14.1 To see how the previous procedure actually works out, let us look at
the following example of the nonlinear elasticity model

Ẍ  −fX + X 3 , (14.28)

where X stands for displacement of position, and f the linear elasticity coefficient.
As what was done to equation (14.18), we have
1 4 1 2
E 2  Ẋ 2  X − fX + h0 . (14.29)
4 2
Now, equation (14.28) can be reduced to the following first-order nonlinear evo-
lution equation
√
Ẋ  ± E. (14.30)

Our earlier discussion indicates that under certain conditions, the evolution of
X contains blown-ups. Therefore, the nonlinear elasticity change is fundamentally
different from that of linear elasticity change. That is, within the range of elasticity,
linear elasticity models evolve continuously and smoothly. And, the evolution of
nonlinear elasticity models contains discontinuous singular blown-ups.

14.2.3 Blown-up of Nonlinear Time–Space Evolution

Equations

As for evolution equations involving changes in space, they can directly and intu-
itively reflect the real-life meanings of blown-ups. The one-dimensional advection
equation is one of the simplest nonlinear evolution equations involving changes in
space. Its Cauchy problem can be written as follows:

ut + uux  0
, (14.31)
u|t0  u0
14.2 Properties of Blown-Ups 203

where u stands for the speed of a flow, ut  ∂u/∂t and ux  ∂u/∂x with t representing
time and x the one-dimensional spatial location. By using the method of separating
variables without expansion, we let

u  A(t)v(x) and u0 (x)  A(0)v(x). (14.32)

So, equation (14.31) can be reduced to the following:

Ȧ
 −vx  −λ, (14.33)
A2
where λ is a constant. Then we have

Ȧ + λA2  0 and vx  λ. (14.34)

Integrating equation (14.34) by separating the variables and taking A(0)  A0 pro-
vides
A0
A . (14.35)
1 + A0 vx t

Multiplying both sides of this equation by ux and taking u0x  A0 vx produce

u0x
ux  . (14.36)
1 + u0x t

Based on the definition of the degree of divergence, ux is the first-dimensional

degree of divergence. When u0x > 0, that is when the initial field is divergent, ux
declines continuously with time t until the diverging motion disappears. If u0x <
0, that is when the initial field is convergent, then the solution of ux → ∞ and
experiences a discontinuous singularity with time t → tb  −1/u0x . Evidently,
t < tb , ux < 0 evolves continuously.
When t > tb , ux > 0. So, the convergent movement of the initial field (u0x < 0)
can be transformed into a divergent movement (u0x > 0) through a blown-up. The
characteristics of this kind of movement cannot be truly and faithfully described by
linear analysis or statistical analysis. They reflect the fundamental characteristics of
nonlinear evolution.

14.3 Issues Facing Quantitative Infinity

One of the features of blown-ups is the quantitative infinity ∞ , which stands for
indeterminacy mathematically. So, a natural question is how to comprehend this
mathematical symbol ∞, which in applications causes instabilities and calculation
spills that have stopped each and every working computer.
204 14 Whole Evolutions, Where Systemic Yoyos Come from

Figure 14.6 Riemann

ball—relationship between
planer infinity and
three-dimensional North
Pole

14.3.1 An Implicit Transformation Between Euclidean

Spaces and Curvature Spaces

To address the previous question, let us look at the mapping relation of the Rie-
mann ball, which is well studied in complex functions (Figure 14.6). This so-called
Riemann ball, a curved or curvature space, illustrates the relationship between the
infinity on the plane and the North Pole N of the ball. Such a mapping relation con-
nects −∞ and +∞ through a blown-up. Or in other words, when a dynamic point
xi travels through the North Pole N on the sphere, the corresponding image xi on
the plane of the point xi shows up as a reversal change from −∞ and +∞ through a
blown-up. So, treating the planar points ±∞ as indeterminacy can only be a product
of the thinking logic of a narrow or low-dimensional observe-control, since, speaking
generally, these points stand implicitly for direction changes of one dynamic point
on the sphere at the polar point N. Or speaking differently, the phenomenon of direc-
tionless, as shown by blown-ups of a lower-dimensional space, represents exactly a
direction change of movement in a higher-dimensional curvature space.
Therefore, the concept of blown-ups can specifically represent implicit transfor-
mations of spatial dynamics. That is, through blown-ups, problems of indeterminacy
of a narrow observe-control in a distorted space are transformed into determinant
situations of a more general observe-control system in a curvature space. This discus-
sion shows that the traditional view of singularities as meaningless indeterminacies
has not only revealed the obstacles of the thinking logic of the narrow observe-control
(in this case, the Euclidean space), but also the careless omissions of spatial proper-
ties of dynamic implicit transformations (bridging the Euclidean space to a general
curvature space).
Corresponding to the previous implicit transformation between the imaginary
plane and the Riemann ball, let us now look at how we can relate quantitative ±∞
(symbols of indeterminacy in one-dimensional space) to a dynamic movement on a
circle (a curved space of a higher dimension) through the modeling of differential
equations.
In the beginning of Section 14.2, we studied the blown-ups of quadratic models.
Now, let us look at the essence of singularities by investigating spatial dynamic
transformations. To make this presentation easier to follow, let us first cite some
14.3 Issues Facing Quantitative Infinity 205

relevant results from Section 14.2 on quadratic evolution models. For the evolution
described by

u̇  a0 + a1 u + u2 , (14.37)

we have three possibilities:

1. When a12 − 4a0  0, we have
 
1 1
u− − a0 . (14.38)
t + A0 2

2. When a12 − 4a0 > 0, there exist the following two possibilities:

(a) If |a1 /2 + u| < a12 − 4a0 /2, then we have
 
1√ 1√ 1 a1
u− X · tanh X t + A0 − , (14.39)
2 2 2 2

where X  a12 −
4a0 .
(b) If |a1 /2 + u| > a12 − 4a0 /2, then we have
√ √ 
X X A0 a1
u coth − − − . (14.40)
2 2 2 2

3. When a12 − 4a0 < 0, we then have

√ √ 
−X −X t A0 a1
u tan + − . (14.41)
2 2 2 2

These results imply that other than equation (14.40), which represents a local
smooth continuity, all other cases, as shown in equations (14.38), (14.39), and (14.41),
experience discontinuous blown-ups in their corresponding whole evolutions. Now,
let us explain how these results on evolutions of quadratic evolutions correspond to
the dynamic implicit transformations of the projection mapping between a planar
circle and a straight line.
Case 1. Transform implicitly a planar circle to a line tangent to the circle.
For this situation, Figure 14.7 depicts the dynamic relationship between point
pi on the circle and its projection point pi on the horizontal tangent line. Here, the
set of all points pi is one-to-one corresponding to the set of all points pi on the
tangent line. Now, let the point N on the circle corresponds to the singular point +∞
on the tangent line. So, when the point pi travels directly from the singular point
+∞ to the other singular point −∞ on the straight line, it simply reflects how the
point pi traveled on the circle through the polar point N with a change in direction.
206 14 Whole Evolutions, Where Systemic Yoyos Come from

Figure 14.7 Implicit

transformation between a
circle and a tangent line

Figure 14.8 Implicit

transformation between a
circle and a secant line

Combining this explanation with the situation of two equal real roots of the quadratic
form in equation (14.38), the movement of the point pi is limited by u0 (a real root
or an equilibrium state). When the point pi travels upward on the right-hand side of
the circle, the corresponding point pi on the tangent line travels to +∞. When the
point pi goes across the polar point N, the corresponding point pi leaps from +∞
to −∞ directly. That is, a direction change on the circle stands for a discontinuous
singularity on the line, which is shown as a blown-up in equation (14.38).
Case 2. Transform implicitly a planar circle to a line secant to the circle.
If we focus on locally bounded movements, such as the point pi1 in Figure 14.8,
whose movement is limited by the distinct real roots u1 and u2 , then the point pi1
cannot go across the boundaries u1 and u2 and would not be able to travel over the

polar point N. So, the corresponding projection point pi1 on the secant line will not

approach either +∞ or −∞. That is, the point pi1 does not experience any singularity.
This fact corresponds exactly to the smooth continuous solution in equation (14.39).
If we do not limit ourselves to bounded movements, such as the movement of the
point pi2 in Figure 14.8, which can go across the polar point N, then the corresponding

projection point pi2 can approach +∞ and then be transformed into −∞ when the
point pi2 goes across point N. This situation corresponds to the blown-up solution in
equation (14.40).
Case 3. Implicitly transform implicitly a planar circle onto a line disjoint from the
circle.
This situation corresponds to the periodic blown-ups of equation (14.41) with
14.3 Issues Facing Quantitative Infinity 207

Figure 14.9 Implicit

transformation between a
circle and a disjoint line

a12 − 4a0 < 0

and
  
A0 1
4a0 − a12 t +  + 2k π, k  0, ±1, ±2, . . .
2 2

Since the movement of the point pi3 is not limited by any condition, it can go

across the polar point N repeatedly, and the corresponding projection point pi3 on
the line detached from the circle experiences periodic transformations from +∞ to
−∞ or from −∞ to +∞ as shown in Figure 14.9.
Our discussion here indicates that the traditional view of singularities as mean-
ingless indeterminacies has not only revealed the obstacles of the thinking logic of
the narrow observe-control, but also the careless omissions of spatial or dynamic
implicit transformations.
Summarizing what has been discussed in this chapter, we can see that nonlinear-
ity, speaking mathematically, stands (mostly) for singularities in Euclidean spaces,
the imaginary plane or the straight line discussed above. In terms of the world’s
reality, nonlinearity represents eddy motions, movements on a curvature space, the
Riemann ball, or the circle above. Such motions represent a problem about struc-
tural evolutions, which are a natural consequence of uneven evolutions of materials.
So, nonlinearity accidentally describes discontinuous singular evolutionary charac-
teristics of eddy motions (in curvature spaces) from the angle of a special, narrow
observe-control system, the Euclidean spaces. To support this end, please go to
Section 15.1: Bjerknes’ Circulation Theorem, for details.

14.3.2 Motions Represented by General Dynamic System

This subsection looks at the general dynamic system and how it is related to eddy
motions. The following is Newton’s second law of motion
208 14 Whole Evolutions, Where Systemic Yoyos Come from

d v

m  F. (14.42)
dt
So, Einstein’s concept of uneven time and space of evolutions leads to the follow-
ing assumption

F
 −∇S(t; x, y, z), (14.43)

where S  S(t; x, y, z) stands for the time–space distribution of the external acting
object. If ρ  ρ(t; x, y, z) stands for the density of the object or organization being
acted upon. Then, the kinematic equation (14.42) for a unit mass or structure of the
object or organization being acted upon can be written as

d u
1
− ∇S(t; x, y, z), (14.44)
dt ρ(t; x, y, z)

where u
is used to replace the original v
in order to represent the meaning that each
movement of some materials or evolution of organizations is a consequence of mutual
reactions of materials’ or organizations’ structures. Evidently, if ρ is not a constant,
then equation (14.44) becomes
 
d (∇x × u
) 1
 −∇x × ∇S  0, (14.45)
dt ρ

which stands for an eddy motion because of the cross product involved. In other
words, a nonlinear mutual reaction between materials’ or organizations’ uneven
structures and the unevenness of the external forcing object or organization will
definitely produce eddy motions.
Conversely, because nonlinearity stands for eddy sources, it represents a problem
about structural evolutions. This end has essentially resolved the problem of how
to understand nonlinearity and also ended the particle assumption of Newtonian
mechanics, where all acting and reacting entities are assumed to be particles without
any volume and size, and the methodological point of view of melting shapes into
numbers, which has been formed since the time of Newton in natural sciences.
More important is that the concept of uneven eddy evolutions reveals the fact that
forces exist in the structures of evolving objects or organizations, and do not exist
independently outside of these objects or organizations, as what Aristotle and Newton
believed so that the movement of all things, be they physical objects or organizations
of people, had to be pushed first by God.
Based on this logic of reasoning, the concept of second stir of materials’ or orga-
nizations’ movements is introduced naturally. At this junction, we need to point out
that as early as over 2,500 years ago, Lao Tzu of China once said: “Tao is about
physical materials. Even though nothing can be seen clearly, there exist figurative
structures in the fuzziness.” [Chapter 21, Tao Te Ching, (English and Feng 1972)].
It is no doubt a scientific and epistemological progress that in our modern time,
14.3 Issues Facing Quantitative Infinity 209

Einstein proposes that gravitations are originated from the unevenness of time and
space, essentially ending the era of Aristotelian concept of forces existing indepen-
dently outside of the acting and reacting things. However, the fact that Einstein did
not notice the unevenness of time and space of the entity being acted upon implies
that he did not successfully reveal the essence of nonlinear mutual reactions.
For example, in terms of the phenomenon of the universe’s evolution, Newton
needs the hands of God, the first push, which can be vividly seen in the second law of
motion. Einstein realized the materialism of forces. However, based on his general
relativity theory, it is concluded that the universe is originated from a big bang out of a
singular point. Obviously, the singular point represents the transition of an evolution.
This understanding possesses more realisticity of things than Newton’s hands of God.
Therefore, the scientific community has accepted such a big bang theory. However,
the acceptance does not mean that such a big bang theory has revealed the essence of
evolution of the universe, because the theory of big bang is established on the basis
of the universe’s background radiation (3 K, where K is the absolute temperature
index), which is assumed to be uniform in all directions. Next, should there be
only one singular point? Since the so-called background radiation should also be
originated from the unevenness of things, can the universe’s background radiation
be calm? Even though a calm state could be reached temporarily, there should be a
presingular point universe. Since unevenness is the fundamental property of spinning
things and the origin of all multiple levels of materials’ and organizations’ eddies,
the corresponding singular point explosions should also be multiple. So, the big bang
theory, as a scientific theory, is still not plausible.
As a matter of fact, the concept of second stir can also be extended into a theory
about the evolution of the universe. Since the concept of the second stir assigns forces
to evolving materials’ or organizations’ structures, it naturally ends Newton’s God.
The duality of rotations—rotations appear in pairs—leads to differences in spin-
ning directions. Such differences surely create singular points within singular zones.
Through sub-eddies and sub-sub-eddies, breakages (or big bangs) are represented so
that evolutionary transitions are accomplished. Evidently, according to the concept
of second stir, the number of singular points would be greater than one, and the
big bang explosions would also have their individual multiplicities. Therefore, the
concept of second stir stands for a thought left behind by the scientific world from
the twentieth century, concerning the real-life essence of materials’ or organizations’
evolutions, worthy of further and deepened thinking and investigation.

14.3.3 Wave Motions Versus Eddy Motions

Wave motions, originated in the morphological changes of things in motion, such

as developing business entities, vibrating solids, water, and sound waves caused
by local content distributions in fluids, etc., were initially assumed to be caused
by reciprocating movements. Later, this concept was extended to studies of other
physical phenomena; and the effort went to such an extent that to suit mathematical
210 14 Whole Evolutions, Where Systemic Yoyos Come from

Figure 14.10 A wave

motion

Figure 14.11 An eddy

motion

needs all absolute concepts about disturbances are treated as wave motions. That
is the reason how this concept of wave motions has been so widely employed in a
great many branches of natural and social sciences. Because this concept, current
employed in a wide array of studies, is no longer the same as that initially introduced
in physics, we will refer to this greatly generalized concept of wave motions as
extended wave motions for the convenience of communication.
Extended wave motions, on the other hand, define eddy and wave motions on
the basis of physics of the root causes and properties of the relevant disturbances
instead of the need to suit some detailed mathematical requirements. If a distur-
bance is the morphological change caused by reciprocating (linear) movements of
things, it is called a wave motion. If the disturbance is the morphological changes
caused by spinning (curvilinear) movements of things, it would be referred to as an
eddy motion. Therefore, wave motions and eddy motions are two different concepts
introduced in the study of evolutions. To this end, each wave motion is at least two
directional—vertical and horizontal, while each eddy motion is unidirectional. Uni-
directionality is the characteristic of flows. So, eddy motions can also be named eddy
currents or spinning currents (Figures 14.10 and 14.11). In terms of mathematics,
wave motions represent linear problems, while eddy motions nonlinear issues. The
current widely employed concept of extended wave motions is essentially developed
by generalizing the original concept of wave motions to satisfy the requirements of
relevant mathematical theories in order to establish the desired outcomes. As ana-
lyzed here, such generalization has caused confusion of concepts in theoretical and
applied investigations. Typical examples of such confusion are extrapolations of the
thinking logic of particle’s continuity, widely used in decision makings and predic-
tions. Even if we assume such a generalization is OK, the assumed continuity of the
generalized wave motion s is still conditional.
14.3 Issues Facing Quantitative Infinity 211

Speaking differently, even based on the traditional formal analysis, one can see
the phenomenon of “break-offs” appearing in wave motions. To confirm this end,
we will in the following use nonlinear dispersive “wave motions” as our platform to
show this phenomenon.
Let the vibrating wave of an uneven material be written as follows:

ϕ(t, x)  Acosξ, (14.46)

where A  A(t, x) and ξ  ξ (t, x) with t being the time variable and x the one-
dimensional space. Then, the local wave number and frequency are given by
∂ξ ∂ξ
R(t, x)  and ω(t, x)  − . (14.47)
∂x ∂t
Now, by eliminating ξ from equation (14.47), we produce

Rt + ωx  0. (14.48)

Let us define the group speed as

∂ω
C(R)  , (14.49)
∂R
so that equation (14.48) becomes
∂ω
Rt + Rx  0. (14.50)
∂R
From equation (14.49), it follows that the local wave number R(t, x) satisfies the
following quasi-linear hyperbolic equation

Rt + C(R)Rx  0. (14.51)

Now, equation (14.51) stands for a typical blown-up problem. If tb is a point

of discontinuity in the solution of equation (14.51), then when t < tb , R evolves
continuously and the corresponding wave motion is continuous. When t  tb , R
approaches ∞ or L  2π/R → 0 experiences a blown-up with “broken” wavelength.
So, nonlinear dispersive wave motions can evolve through blown-up(s) so that long
waves are transformed into “short waves” or “broken waves.”
If when t → tb , Rt  ∂R/∂x → ∞, that is, the first-order derivative of the
wave number with respect to space experiences a blown-up, then from R  2π/L, it
follows that
   
2π Lx L2 2π
Rx   −2π 2 or − ,
L x L Lx Rx
212 14 Whole Evolutions, Where Systemic Yoyos Come from

where Rx → ∞ and L → 0. Therefore, uneven materials’ nonlinear dispersive

waves are a blown-up problem. This is fundamentally different from the continuous
movement of linear dispersive waves of even materials.
Of course, the discussion above is about nonlinear “waves” in their mathemati-
cal forms. So, what has been said is only about the difference in the mathematical
properties of linearity and nonlinearity. That difference should not and could not
be employed to illustrate the corresponding real-life morphologies and properties.
So, nonlinear “waves” of the mathematical form are not equivalent to any real-life
existence of mutual reacting waves. Because wave motions are originated from the
diverging or converging currents of things, and because both diverging and converg-
ing currents of things in objective and realistic processes always contain unevenness
that causes twisting forces so that sub-eddies are formed and rotating eddy motions
are resulted. The morphology of broken waves of things must be spinning “spindrifts”
instead of waves. Thus, wave motions are only local and relatively small-scale real-
life phenomena, while eddy motions, on the other hand, are the commonly existing
real-life morphology. Because the mutual reaction of uneven structures within the
underlying materials and organizations constitutes twisting forces, which is “twisting
motions,” spinning motions are inevitable results of mutual reactions of the uneven-
ness existing in materials or organizations, forming the origin from which eddy
currents are from.

14.4 Equal Quantitative Movements and Effects

Another important concept studied in the blown-up theory is that of equal quantitative
effects. Although this concept was initially proposed in the study of fluid motions, it
essentially represents the fundamental and universal characteristics of all movements
of materials or all evolutions of organizations. More importantly, this concept reveals
the fact that nonlinearity is originated in the figurative structures of materials or
organizations instead of nonstructural quantities of the materials or organizations.
The so-called equal quantitative effects stand for the eddy effects with nonuniform
vortical vectorities existing naturally in systems of equal quantitative movements due
to the structural unevenness of materials or organizations. And, by equal quantitative
movements, it means such movements caused by quasi-equal acting and reacting
things or under two or more quasi-equal mutual constraints. For example, the relative
movements of two or more planets of approximately equal masses are considered
equal quantitative movements; the developments of two or more business firms of
approximately equal characteristics are considered equal quantitative evolutions. In
the microcosmic world, an often seen equal quantitative movement is the mutual
interference between the particles to be measured and the equipment used to make
the measurement. In the business world, a frequently appearing equal quantitative
movement is the mutual interference between the economic variables that are to be
measured and the instruments designed and used to actually collect the needed data.
And many phenomena in daily lives, such as wars, politics, economies, chess games,
14.4 Equal Quantitative Movements and Effects 213

races, plays, etc., can all be considered as equal quantitative movements with equal
quantitative effects.
Comparing to the concept of equal quantitative effects, Aristotelian and Newto-
nian framework of separate objects and forces is about unequal quantitative move-
ments established on the assumption of particles—all things considered have no size
and no volume. On the other hand, equal quantitative movements are mainly char-
acterized by the kind of measurement uncertainty that when I observe an object, the
observation of the object is constrained by me. When an object or an organization is
observed, the object or the organization cannot really be separated from the observer.
At this juncture, it can be seen that the Su-Shi Principle of Xuemou Wu’s panrel-
ativity theory (1990), Bohr (Bohr 1885–1962) principle and the relativity principle
about microcosmic motions, von Neumann’s Principle of Program Storage, etc., all
fall into the uncertainty model of equal quantitative movements with separate objects
and forces.
What is practically important and theoretically significant is that eddy motions are
confirmed not only by daily observations of surrounding natural phenomena, but also
by laboratory studies from as small as atomic structures to as huge as nebular struc-
tures of the universe. At the same time, eddy motions show up in mathematics—the
very foundation of modern science—as nonlinear evolutions. The corresponding lin-
ear models can only describe straight-line-like spraying currents and wave motions
of the morphological changes of reciprocating currents. What is interesting here is
that wave motions and spraying currents are local characteristics of eddy move-
ments. This fact is very well shown by the fact that linearity is a special case of some
nonlinearity instead of an approximation of the nonlinearity.
Since 99% of all materials in the universe are fluids, and since under certain con-
ditions solids can be converted to fluids, OuYang (1998a) pointed out that “when
fluids are not truly known, the amount of human knowledge is nearly zero. And, the
epistemology of the Western civilization, developed in the past 300 plus years, is
still under the constraints of solids.” This statement no doubt points to the central
weakness of the current theoretical studies and also located the reason why “math-
ematics met difficulties in the studies of fluids,” as said by Engels (1939). It also
well represents Lao Tzu’s teaching (English and Feng 1972): “If crooked, it will be
straight. Only when it curves, it‘ will be complete.” As a matter of fact, mankind
exists in a spinning universe with small eddies nested in bigger eddies or multiple
eddies. Through sub-eddies and sub-sub-eddies, heat-kinetic energy transformations
are completed. These nested eddies and energy transformations vividly represent the
forever generation changes of blown-up evolutions of all things in the universe.
The birth–death exchanges and the nonuniformity of vortical vectorities of eddy
evolutions provide natural explanations for where and how quantitative irregularities,
complexities, and multiplicities of materials’ and organizations’ evolutions, when
seen from the traditional narrow observe-control system, come from. Evidently, if
the irregularity of eddies comes from the internal structural unevenness of materials
or organizations, and if the world is seen at the height of structural evolutions, then the
world is simple. And, it is so simple that there are only two forms of motions. One is
clockwise rotation, and the other counterclockwise rotation. The vortical vectority in
214 14 Whole Evolutions, Where Systemic Yoyos Come from

the structures of materials and organizations has very intelligently resolved the Tao of
Yin and Yang of the “Book of Changes” of the eastern mystery (Wilhalm and Baynes
1967) and has been very practically implemented in the common form of motion of
all things in the universe. That is where the concept of invisible organizations of the
blown-up system comes from.
The concept of equal quantitative effects not only possesses a wide range of
applications, but also represents an important omission of the traditional science,
developed in the past 300 plus years. Evidently, not only are equal quantitative effects
more general than the mechanic system of particles with far-reaching significance,
but also have they directly pointed to some of the fundamental problems existing in
the traditional science. For instance,
1. Equal quantitative effects generally throw equations into computational uncer-
tainty. For example, if x ≈ y, then the difference x − y becomes a mathematical
problem of computational uncertainty, involving large quantities with infinitesi-
mal increments. For details, see Kline (1972) and Lin (2008) for the second crisis
in the foundations of mathematics. Although this end has been well known, prac-
tical applications are still often misguided into such uncertainties unconsciously.
For example, in meteorological science, one situation involves
1

× V
h ≈ − ∇ph
2Ω
ρ

where the left-hand side stands for the deviation force caused by the earth’s
rotation and the right-hand side the stirring force of the atmospheric density
pressure. However, scholars have tried for many decades to compute d V
h /dt
under the influence of such quasi-equal computational uncertainty. As a matter
of fact, the concept of equal quantitative effects has computationally declared that
equations are not eternal, or there is not any equation under equal quantities. That
is why OuYang introduced the methodology of abstracting numbers (quantities)
back into shapes (figurative structures). The purpose of abstracting numbers
back into shapes is to describe the formalization of eddy irregularities, which is
different from formalized quantification of structures. That is, we should very
well see that if 300 years ago it was human wisdom to abstract numbers out of
things, then it would be human stupidity to continue to do so today.
2. Because the current variable mathematics is entirely about “computational”
schemes designed on some commonly accepted axioms, there must be the prob-
lem of disagreement between the conclusions of variable mathematics and irreg-
ularities existing in the objective evolutions of materials and organizations. So,
the corresponding quantified comparability can only be relative. At the same
time, there exists a problem with quantification where distinct properties cannot
be distinguished because the relevant quantifications produce the same indis-
tinguishable numbers. Because of this, it is both incomplete and inaccurate to
employ quantitative comparability as the only standard for judging the scien-
tificality of works. For example, in terms of weather forecasting or studies of
organizational behaviors, it should be clear that difficulties we have been facing
14.4 Equal Quantitative Movements and Effects 215

Figure 14.12 How forces naturally exist within uneven structures

are consequences of the incapability of the existing theories in handling equal

quantitative effects. In other words, the quantitative science, developed in the
past 300 plus years, is incomplete and incapable of resolving problems about
figurative structures.
3. The introduction of the concept of equal quantitative effects has not only made the
epistemology of natural sciences gone from solids to fluids, but also completed
the unification of natural and social sciences. That is because many social phe-
nomena, such as military conflicts, political struggles, economic competitions,
chess games, races, plays, etc., can all be analyzed on the basis of figurative struc-
tures, such as the systemic yoyo fields, of equal quantitative effects. However,
what needs to be pointed out is that the current system of the traditional science
is basically extensions of the research under unequal quantitative effects.
In order for us to intuitively see why equal quantitative effects are so difficult
for the traditional science to handle by using the theories established in the past
300 plus years, let us first look at why all things in the universe are in rotational
movements. According to Einstein’s uneven space and time, we can assume that
all materials and organizations have uneven internal structures. Out of these uneven
structures, there naturally exist gradients. With gradients, there will appear forces.
Combined with uneven arms of forces, the carrying materials will have to rotate in
the form of moments of forces. See Figure 14.12 for details, where we look at two
points P1  (x1 , y1 , z1 ) and P2  (x2 , y2 , z2 ) within an entity of density (or structural
homogeneity) ρ  ρ(x, y, z) that are a certain distance away from each other. Then
each of these points experiences the act of a gradient force
 
−
→ ∂ρ(xi , yi , zi ) ∂ρ(xi , yi , zi ) ∂ρ(xi , yi , zi )
Pi  , , , (14.52)
∂x ∂y ∂z

for i  1, 2. Therefore, as long as the density (or structural homogeneity) ρ 

ρ(x, y, z) is not constant, a twisting force between the points P1  (x1 , y1 , z1 ) and
216 14 Whole Evolutions, Where Systemic Yoyos Come from

P2  (x2 , y2 , z2 ) is created, leading to a spinning motion. That also explains why

solids do not exhibit any clear feature of spin, because the internal density of a solid
is constant so that at each chosen point within the solid, the gradient force is 0.

That is exactly what the ancient Chinese Lao Tzu, (English and Feng 1972)
said: “Under the heaven, there is nothing more than the Tao of images,” instead
of Newtonian doctrine of particles—under the heaven, there is such a Tao that is
not about images but sizeless and volumeless particles. The former stands for an
evolution problem of rotational movements under stirring forces. Since structural
unevenness is an innate character of materials and organizations, that is why it is
named second stir, considering that the phrase of first push was used first in history
(OuYang et al. 2000). What needs to be noted is that the phrases of first push and
second stir do not mean that the first push is prior to the second stir.
Now, it can be seen that the world, be it natural or artificial, physical or economic,
and/or the universe, is composed of entirely eddy currents, where eddies exist in
different sizes and scales and interact with each other. That is, the universe is a
huge ocean of eddies, which change and evolve constantly. One of the most impor-
tant characteristics of spinning fluids, including spinning solids, is the difference
between the structural properties of inwardly and outwardly spinning pools and the
discontinuity between these pools. Due to the stirs in the form of moments of forces,
in the discontinuous zones there appear sub-eddies and sub-sub-eddies (Figure 14.13,
where sub-eddies are created naturally by the large eddies M and N). Their twist-ups
(the sub-eddies) contain highly condensed amounts of materialistic and structural
contents and energies. In other words, the traditional frontal lines and surfaces (con-
sidered in meteorology) are not simply expansions of particles without any internal
structure. Instead, they represent twist-up zones concentrated with irregularly struc-
tured contents of materials, organizations, and energies. That is where the so-called
small probability events appear and small probability information is observed and
collected. So such information (event) should also be called irregular information
(and event). In terms of basic energies, these twist-up zones cannot be formed by only
the pushes of external forces and cannot be adequately described by using symbolic
forms of separate objects and forces. Since evolution, development and growth are
about changes in the internal structures of materials and organizations, it cannot be
simply represented by different speeds of movements. Instead, they are mainly about
transformations of rotations in the form of moments of forces ignited by structural
irregularities. The enclosed areas in Figure 14.14 stand for the potential places for
equal quantitative effects to appear, where the combined pushing or pulling is small
in absolute terms. However, it is generally difficult to predict what will come out of
the power struggles. In general, what comes out of the power struggle tends to be
drastic and unpredictable by using the theories and methodologies of the traditional
science.
14.5 Remarks 217

Figure 14.13 Appearance of sub-eddies

Figure 14.14 Structural

representative of equal
quantitative effects

14.5 Remarks

When a mathematical model, which truthfully and adequately describes the real-life
situation of concern, blows up at a specific time moment or a specific spatial location
or both and the underlying real-life system also goes through a transitional change,
then the solution of the model is called a blown-up solution and the relevant real-life
movement expresses a blown-up. Blown-up phenomena appear in life and business
all the time, especially in studies involving evolutions, predictions, and decision
making.
The main points of the discussion in this chapter can be described as follows. For
the situation that blown-ups appear only with time, by analyzing nonlinear equa-
tions and more general nonlinear models, it is found that in most cases, nonlinearity
implies blown-ups and the requirements for well-posedness of the traditional sci-
ence (existence, uniqueness, and stability) do not hold true, meaning that most of
the methods and thinking logic in the traditional science cannot be employed to
resolve the relevant problems involving nonlinearity. Only under very special local
conditions, there will not be any blown-up. That is when the available methods and
theories in the traditional science can be applied. When blown-ups appear in both
time and space, it is shown that if a system, when seen as a rotational entity, is initially
divergent, then over time the whole evolution of the system is continuous and the
divergent development of the system will eventually disappear smoothly and quietly.
However, if the initial state of the system is convergent, then it is guaranteed that a
moment of blown-up will appear at a definite time moment in the foreseeable future.
After the system goes through a transitional change (blown-up), it will restart as a
divergent system.
218 14 Whole Evolutions, Where Systemic Yoyos Come from

By introducing the concept of implicit transformations between a Euclidean space

and a curvature space, it is shown that blown-ups in the Euclidean space are sim-
ply some transitional changes in the curvature space. And, periodic blown-ups of
the Euclidean space and rotational movements in the curvature space correspond
very well, illustrating a method on how to resolve the problems of the quantitative
infinity ∞, numerical instability, and computational spills, by reconsidering the cor-
responding situations in curvature spaces. By doing so, all these problems, which
seem unsolvable in the traditional science, are avoided.
It is concluded that the real-life characteristics of blown-ups are spinning currents,
which will be shown theoretically and empirically in the next chapter. One of the
very significant outcomes of this discovery is that discontinuities and singularities
existing in calculus-based models can actually be applied as a tool to predict forth-
coming transitional changes. By looking at the general dynamic system, it is shown
that Newton’s second law of motion actually indicates that as long as the acting and
reacting entities have uneven internal structures, their mutual reaction will be non-
linear and a rotational movement. So, the concept of second stir is introduced. On
the basis of this new concept, new explanations and a generalization of the big bang
theory are provided in terms of a new theory on the evolution of the universe.
With the theoretical background established for rotations to be the common form
of movements in the universe, the concepts of equal quantitative effects and equal
quantitative movements are introduced. Through using these concepts, it is analyzed
that mathematical nonlinearity is created by uneven internal structures of the under-
lying materials and organizations instead of the nonstructural quantities abstracted
out of the materials and organizations. It is found that equal quantitative effects lead
to computational uncertainties and have the ability to unify natural and social sci-
ences, since in both of these sciences, figurative structural analyses can be employed.
By looking at the geometry of equal quantitative effects, one can easily locate where
small probability events would appear and where small probability information can
be observed.
As for the references on which this chapter is based, other than at specific loca-
tions we have directly pointed to the source of information, the presentation of this
chapter is mainly based on Glassey (1977), Guo (1995), Hess (1959), Holton (1979),
Kuchemann (1965), Lin (2000), Lin and Fan (1997), Lin and OuYang (1998), Lin
and Wu (1998), Lorenz (1993), OuYang (1994, 1995, 1998a, b), OuYang et al. (1998,
2000), Saltzman (1962), Wu (1990, 1998), and Wu and Lin (2002). For more details
and references, please consult with these works.

References

English, J., Feng, G.F.: Tao De Ching. Vintage Books, New York (1972)
Glassey, R.T.: On the blowing up of solution to Cauchy problems for nonlinear Schrodinger equation.
J. Math. Phy. 18, 1794–1797 (1977)
Guo, B.L.: Nonlinear Evolution Equations. Shanghai Press of Scientific and Technological Educa-
tion, Shanghai (1995)
Hess, S.L.: Introduction to Theoretical Meteorology. Holt, Rinehart and Winston, New York (1959)
References 219

Holton, J.R.: An Introduction to Dynamic Meteorology, 2nd edn. Academic Press, New York (1979)
Kline, M.: Mathematical Thought from Ancient to Modern Times. Oxford University Press, Oxford
(1972)
Klir, G.: Facets of Systems Science. Springer, New York (2001)
Kuchemann, D.F.: Report on the I.U.T.A.M. symposium on concentrated vortex motions in fluids.
J. Fluid Mech. 21, 1–20 (1965)
Lin, Y.: Some impacts of the second dimension. In: Proceedings of the 15th European Meeting in
Cybernetics and Systems Research, pp. 3–8, Vienna, Austria, 25–28 April 2000
Lin, Y. (guest editor): Systematic studies: the infinity problem in modern mathematics. Kybern. Int.
J. Cybern. Syst. Manage. Sci. 37(3–4), 385–542 (2008)
Lin, Y.: Systemic Yoyos: Some Impacts of the Second Dimension. Auerbach Publications, an imprint
of Taylor and Francis, New York (2009)
Lin, Y., Fan, T.H.: The fundamental structure of general systems and its relation to knowability of
the physical world. Kybern. Int. J. Syst. Cybern. 26, 275–85 (1997)
Lin, Y., OuYang, S.C.: Invisible Tao and realistic nonlinearity propositions. Kybern. Int. J. Syst.
Cybern. 27, 809–822 (1998)
Lin, Y., Wu, Y.: Blown-ups and the concept of whole evolutions in systems science. Probl. Nonlinear
Anal. Eng. Syst. 4, 16–31 (1998)
Liu, S.F., Lin, Y.: Grey Information: Theory and Practical Applications. Springer, London (2006)
Lorenz, E.N.: The Essence of Chaos. Washington University Press, Seattle (1993)
OuYang, S.C.: Break-Offs of Moving Fluids and Several Problems of Weather Forecasting. Press
of Chengdu University of Science and Technology, Chengdu (1994)
OuYang, S.C.: Pansystems point of view of prediction and explosive reverses of fluids. Appl. Math.
Mech. 16, 255–262 (1995)
OuYang, S.C.: Weather Evolutions and Structural Forecasting (in Chinese). Meteorological Press,
Beijing (1998a)
OuYang, S.C.: Some remarks on “explosive” evolutions of general nonlinear evolution equations
and related problems. Appl. Math. Mech. 19, 165–173 (1998b)
OuYang, S.C., Wu, Y., Lin, Y., Li, C.: The discontinuity problem and “chaos” of Lorenz’s model.
Kybern. Int. J. Syst. Cybern. 27, 621–635 (1998)
OuYang, S.C., Miao, J.H., Wu, Y., Lin, Y., Peng, T.Y., Xiao, T.G.: Second stir and incompleteness
of quantitative analysis. Kybern. Int. J. Cybern., Syst. Manage. Sci. 29, 53–70 (2000)
OuYang, S.C., Chen, Y.G., Lin, Y.: Digitization of Information and Prediction. China Meteorological
Press, Beijing (2009)
Saltzman, B.: Finite amplitude free convection as an initial value problem. I. J. Atmos. Sci. 19,
329–341 (1962)
Soros, G.: The Alchemy of Finance. Wiley, New York (2003)
Wilhalm, R., Baynes, C.: The I Ching or Book of Changes, 3rd edn. Princeton University Press,
Princeton, NJ (1967)
Wu, X.M.: The Pansystems View of the World. People’s University of China Press, Beijing (1990)
Wu, Y.: Blown-ups: a leap from specified local confusions to a general whole evolution. Kybern.
Int. J. Cybern. Syst. 27, 647–655 (1998)
Wu, Y., Lin, Y.: Beyond Nonstructural Quantitative Analysis: Blown-Ups, Spinning Currents and
the Modern Science. World Scientific, River Edge, New Jersey (2002)
Chapter 15
Some Empirical Justifications

Based on the discussions in Chapter 14, this chapter will look at empirical evidences
and observations that underline the existence of the yoyo structure behind each and
every system, be tangible or intellectually imaginable.
Historically, the plane or the three-dimensional spaces naturally used in Euclidean
geometry, calculus, statistics, and other fundamental areas underneath the entire
spectrum of modern science are just introduced by describing what they are with-
out any justification. However, because the systemic yoyo model is different from
and counterintuitive to our localized intuition—the plane and the three-dimensional
space—let us simply present a few empirical evidences to show why such a systemic
model actually exists in nature. As a matter of fact, when a model is employed sim-
ply as an intuition underlying rigorous reasoning of logic, there is no need for us to
scientifically show its theoretical existence.

15.1 Eddy Effects Caused by Changes in Density of Media

of Movement

Based on rigorous theoretical reasoning, it is concluded in Section 14.3 that nonlin-

earity stands mostly for singularities in Euclidean spaces, the imaginary real number
line, or the plane described by the crossing of two real numbers lines, or the n-
dimensional Cartesian coordinate system consisting of n crossing real number lines.
In terms of real-life applications, nonlinearity represents eddy motions that take place
in curvature spaces, such as the Riemann ball or the circle discussed earlier. From
each such a motion arises a problem regarding structural evolution as natural conse-
quences of uneven evolutions of the underlying things, such as investment, resource
availability, information flow, etc. So, in the traditional science nonlinearity acciden-
tally describes discontinuous singular evolutionary characteristics of eddy motions
(in curvature spaces) from the angle of a special, linearized local observe-control

© Springer Nature Switzerland AG 2018 221

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_15
222 15 Some Empirical Justifications

Figure 15.1 Definition of a

closed circulation

system, the Euclidean spaces (the imaginary line or plane or n-dimensional Carte-
sian coordinate system). To further support this theoretical realization, let us in this
chapter look at the well-known Bjerknes’ Circulation Theorem (1898); for details,
see (Hess 1959).
At the end of the nineteenth century, Bjerknes (1898) discovered the eddy effects
due to changes in the density of the media in each and every movement of the
atmosphere and ocean. He consequently established the circulation theorem, which
was later named after him. For the completeness of this book, let us look at this
theorem briefly.
By a circulation, it is meant to be a closed contour in a fluid, such as water, the
atmosphere, or any media within which the thing of interest flows. For example, in
business scenarios, the abstract, but realistically existing, media for the money to flow
is a market where money changes hands of ownership from one person to another,
from one person to a business entity, from a general economic agent to another.
Although such a media cannot be visualized exactly, it indeed physically exists.
Theoretically, each circulation, denoted as , is defined as the line integral about the
contour of the component of the velocity vector locally tangent to the contour. In
symbols, if V stands for the speed of a moving fluid, S an arbitrary closed curve, δr
the vector difference of two neighboring points of the curve S (Figure 15.1), then
a circulation  located within the two-dimensional Cartesian coordinate system, is
defined as follows:

  V δ r . (15.1)
S

Its rate of change with respect to time is given by the corresponding derivative

d d
 V δr . (15.2)
dt dt S

The right-hand side of this equation can be written as follows:

  
d d V  d
V δr  · δ r + V· δr . (15.3)
dt S S dt S dt
15.1 Eddy Effects Caused by Changes in Density of Media of Movement 223

Based on the assumption of continuity needed for computing the derivative, we

have
 
d dr
δ r  δ  δ V . (15.4)
dt dt

By substituting equation (15.4) into the second term on the right-hand side of
equation (15.3), we obtain
   2 
 d V
V · δ r  δ  0. (15.5)
S dt S 2

Now, equation (15.2) becomes


d d V
 δr , (15.6)
dt S dt

which is known as an accelerating circulation. The Euler equation in fluid mechanics

is given by

d V 1 −
→
 − ∇p − 2  × V + g , (15.7)
dt ρ

where p stands for the pressure of the media, ρ the density of the media, g the
acceleration of the gravitational pull of the system on which the media exists, and
−
→
 the system’s rotational angular speed. To help visualize the relevant concepts, the
−
→
symbols p, ρ, g , and  can be, respectively, imagined as the atmospheric pressure,
the density of the atmosphere, the gravity of the earth, and the earth’s rotational
angular speed. To imagine what these symbols represent in terms of business decision
making, we first treat the business entity as a spinning yoyo field with a rotational
−
→
angular speed  and a gravitational pull g on things within the field. Then, we treat
p as the pressure of the yoyo field on the particular economic agent of concern within
the yoyo field, and ρ the local density of the yoyo field.
The first term on the right-hand side of equation (15.7) is referred to as a pressure-
gradient force, the second term the Coriolis force. By substituting equation (15.7)
into equation (15.6), we produce
      
d 1 −→
 − ∇p δ r − 2  × V · δ r + g · δ r . (15.8)
dt S ρ S S

Assume that the area enclosed by the closed curve S is A (Figure 15.1). Then, the
second term on the right-hand side of equation (15.8) is
    
−
→ −
→ −
→
2  × V · δ r  2  · V · δ r  2  · n Vr · δ r, (15.9)
S S S
224 15 Some Empirical Justifications

where n is the unit vector perpendicular to the plane generated by V and δr , V r the
component of V in the direction perpendicular to δ r . That is,


V sin θ δ r  Vr δ r.

So, we have
 −  
→ −
→ dσ
2  × V · δ r  2  · n Vr · δ r  2 , (15.10)
S S dt

where σ is the projection of the area A on the equator plane.

If g is taken to be the gradient of the potential function ϕ, that is, g  −∇ϕ, then
we have
  
g δ r  − ∇S · δ r  δϕ  0. (15.11)
S S S

Now, substituting Equations (15.10) and (15.11) into equation (15.8) leads the
following well-known Bjerknes’ Circulation Theorem:
¨  
d V 1 dσ
 ∇ × (−∇p) · δσ − 2 , (15.12)
dt σ ρ dt

where σ is the projection area on the equator plane of the area enclosed by the closed
curve S, p the atmospheric pressure, ρ the density of the atmosphere, and  the
earth’s rotational angular speed.
The left hand side of equation (15.12) represents the acceleration of the moving
fluid. According to Newton’s second law of motion, this acceleration is equivalent to
the force acting on the fluid. On the right-hand side of the equation, the first term is a
solenoid term, as known in meteorology. It is originated from the interaction of the p-
and ρ-planes (or the interaction between density and pressure) due to the unevenness
existing in the density ρ so that a twisting force is created. Consequently, the move-
ment of the fluid must be a rotation with the rotating direction determined by the
equal p- and ρ-plane distributions (Figure 15.2). The second term in equation (15.12)
comes from the rotation of the carrying system—the earth.
In college textbooks, this Bjerknes’ Circulation Theorem is employed to explain
the formation of land-ocean breezes. However, historically, some people, such as
Saltzman (1962), see this circulation theorem as a major betrayal of the classical
mechanics, while S. C. OuYang assigned this theorem a very high value, believing
that in revealing the commonly existing and practically significant eddy effects of
fluid motions, Bjerknes’ Circulation Theorem has and will play an increasingly
important role. What is important for us in this book is that this theoretical result
helps to address the mystery of nonlinearity (Lin 1998), which has been puzzling the
mankind since the time of Isaac Newton, and reveals the universal laws on state of
motion.
15.1 Eddy Effects Caused by Changes in Density of Media of Movement 225

Figure 15.2 A diagram for solenoid circulations

The significance of this theorem also includes the following:

(1) Dynamically speaking, an uneven density ρ implies eddy effects instead of the
so-called effects of elastic pressure as commonly referred to. The pressure-
gradient force (−∇p/ρ) should also be referred to as a stirring gradient force of
the density pressure. Therefore, it is an eddy source. Speaking with mathematical
formality, this gradient force is nonlinear, which thus implies an eddy source that
leads to eddy motions. Beyond the observation that uneven eddy motions are
the most common form of movements observed in the universe, this Bjerknes’
Circulation Theorem clearly and factually shows this observation as a fact.
(2) In the universe, the common form of movements is that of rotations, such as the
eddy motions of the solar system, galaxies, nebula, etc. In the cosmic level, there
are polar eddies, cyclones, anticyclones, etc. In the realm of abstract thoughts
and beliefs, noncompromising thoughts and different and conflicting opinions
always exist. In the business world, there always exist inconsistent and con-
tradictory scenarios, each of which leads to profitable opportunities. And in
the meso-scale level on earth, there are various rotational movements found in
atomic structures. That is why Kuchemann (1965) once claimed that “the tendon
of moving fluids is eddies,” and that each field of eddies represents both a place of
high concentration of different forms of energies and a place of efficient transfor-
mations of energies from one form into another (Lin 2009). In other words, each
business world consisting of many, many competing firms—eddy pools—rep-
resents a high concentration of opportunities and a place where opportunities
are efficiently transformed into innovative products and consequently profits.
Through concentration of energies/opportunities and through transformations
of energies/opportunities into other forms, eddy pools actually consume some
of the energies. Such a rise-and-fall in the magnitudes of different forms of
energies determines the equilibrium of various acting and reacting forces and
economic agents, and reflects the close relationship between the quantities of
eddy pools and transformations of energies and opportunities.
(3) Mr. Shi-jia Lu, a well-known Chinese scholar, once pointed out that “the essence
of fluids is eddies, because fluids cannot stand twisting forces and each twist
creates eddies,” (unpublished lecture notes). Therefore, as long as competition
226 15 Some Empirical Justifications

is allowed to appear, each open-market economy will naturally invite innova-

tion that signals where twisting forces exist. Because uneven densities, such
as investments, talents, resources, etc., create twisting forces, fields of spin-
ning currents (or competing business firms) are naturally created. Such fields
do not have uniformity in terms of types of currents. Clockwise and counter-
clockwise eddies always coexist, leading to destruction of the initial smooth,
if any, fields of currents, as demonstrated by the dishpan experiments. When
such evolutions are reflected and modeled by using the analytical methodology
of any calculus-based local observe-control system, discontinuous singularities
in solutions naturally appear. On the other hand, discontinuities of nonlinear
models offer us an analytic method for possibly predicting major transitional
changes that exist in evolutions and developments and for potentially predicting
the appearance of fields of eddy currents.

15.2 Conservation of Informational Infrastructure

Some branches of the traditional science, such as physics, chemistry, etc., are made
“exact” by establishing various laws of conservation, even though, at the times when
these laws were initially proposed, there might not have been any “theoretical or
mathematical” proofs for them. Walking along the similar lines, after Lin (1995)
provides a systemic explanation for several laws of conservation by employing results
of general systems theory, Lin and Fan (1997) show that human understanding of
nature is very much limited by human sensing organs, even though the mankind
constantly attempts to comprehend the true state of the nature.
In this section, based on an intuitive understanding of the concept of general sys-
tems, one can naturally see that there should be a law of conservation that emphasizes
on the uniformity of “spatial structures” of systems. On top of this intuition, several
areas of scientific research are searched and similarities among these areas are found.
Based on these similarities that exist in diversely different areas of learning, a law of
conservation of informational infrastructure is proposed and some possible impacts
and consequences of this law are explored. Based on what has been discussed in the
earlier chapters, this section provides an empirical evidence for the validity of the
systemic yoyo model for each and every system, be it general or particular, physically
tangible or intellectually imaginable.
Now, let us look at an intuitive understanding of the concept of general systems,
which forms the heuristic foundation of this section on the development of the law
of conservation of informational infrastructure.
From a practical point of view, a system is what is distinguished as a system (Klir
1985). From a mathematical point of view, a system is defined as follows (Lin 1987):
S is a (general) system, provided that S is an ordered pair (M, R) of sets, where M is
the set of objects of the system S and R a set of some relations on the set M. The sets
M and R are called the object set and the relation set of the system S, respectively
(For those readers who are mathematically inclined, for each relation r in R, there
15.2 Conservation of Informational Infrastructure 227

exists an ordinal number n  n(r), a function of r, such that r is a subset of the

Cartesian product of n copies of the set M.). The idea of using an ordered pair of
sets to define a general system is to create the convenience of comparing systems
and uncovering properties of systems. In particular, when two systems S1 and S2 are
given, by writing each of them as an ordered pair (Mi , Ri ), i  1, 2, we can make use
of the known facts of mathematics to show that S1  S2 , if and only if M1  M2
and R1  R2 . When two systems S1 and S2 are not equal, with their ordered pair
structures we can readily investigate their comparisons, such as how and when they
are similar, congruent, or one is structurally less than the other, and other properties
between systems. See Chapters 2–12, for more details.
By combining these two understandings of general systems, one can intuitively
see the following: Each thing that can be imagined in human minds is a system
according to Klir’s definition so that this thing would look the same as that of an
ordered pair (M, R) according to Lin’s definition. Furthermore, relations in R can be
about some information of the system, its spatial structure, its theoretical structure,
etc. That is, there should be a law of conservation that reflects the uniformity of all
tangible and imaginable systems with respect to:
(1) The content of information;
(2) Spatial structures;
(3) Various forms of movements, etc.
In the rest of this section, to support this intuition of (general) systems let us
look at examples from several different disciplines of the traditional science. To this
end, owning to the wide range of disciplines covered in this section, many well-
established scholars in the relevant fields have been consulted, including Profes-
sors Zhang Guodong, Xiao Xinghua, Huang Yongchang, Zheng Jiyan, Song Zhang-
hai, and Wang Wei. Their time, advice, and beneficial discussions that have greatly
improved the presentation of this section are appreciated. Roughly speaking, what
is presented in this section summarizes the collective wisdom of more 300 years of
scientific training, teaching, and discovery. The presentation in this section is mainly
from (Ren et al. 1998).

15.2.1 Physical Essence of Dirac’s Large Number Hypothesis

Dirac (1937), the founder of quantum mechanics, proposes the following well-known
large number hypothesis: The ratio of the static electrical force and the universal
gravitation in an H2 atom is given as follows:

e2
 2 × 1039 , (15.13)
Gmp me

where e2 is the static electric force in an H2 atom, G the gravitational constant, mp the
mass of the proton, and me mass of the electron. In terms of the atomic unit, the age
228 15 Some Empirical Justifications

of the universe is 2 × 1039 . Dirac considers these two nondimensional large numbers
being very close and believes that their closeness could not be a coincidence and that
it must stand for something fundamental.
One important contribution Dirac makes here is that he formally establishes a
connection between the universe and the microscopic world. However, regarding
the most essential physical meaning of such a connection of the two worlds of
different scales, Dirac conjectures that the large number hypothesis implies that the
gravitational “constant” G decreases as time advances and causes various matters to
be created. Now, let us explore this large number hypothesis from a different angle.
According to Allen (1976), in an H2 atom, the static electrical force is

e2  2.307113 × 10−19 (static electricity unit),

the gravitational constant is

G  6.672 × 10−8 dyne × cm2 /g,

the mass of the proton is

mp  1.6726485 × 10−24 g,

the mass of the electron is

me  9.109534 × 10−28 g,

the speed of light is

c  2.99792458 × 1010 cm/s,

and the radius of a classical electron is


re  e2 / me c2  2.817938 × 10−13 cm.

Based on Pan (1980), the Hubble constant is

H  55 km s−1 · Ma par s−1  1.782428367 × 10−18 s−1 .

From all these data values, the following can be computed:

(1) In an H2 atom, the ratio of the static electrical force and the universal gravitation
is

e2
 2.269 × 1039 . (15.14)
Gmp me
15.2 Conservation of Informational Infrastructure 229

(2) The ratioof the age  of the universe (=1/H, approximately a Hubble age) and
e2 /(me c2 )
the time c
needed for a light beam to travel through the distance equal
to the radius of an electron is given as follows:
1/H
 5.96866 × 1040 , (15.15)
e2 /(me c2 )/c

which can be rewritten as the ratio of the “radius of the universe” (= c/H, the
2
Hubble distance) and the radius of an electron ( mee c2 ) as follows:
c/H
 5.96866 × 1040 , (15.16)
e2 /(me c2 )

From Equations (15.14) and (15.15) or Equations (15.14) and (15.16), it can be
seen that these two large numbers differ by only a multiple of ten. Therefore, for
numbers of such a magnitude, they can be seen as approximately equal. At this
junction, we can tell that some undiscovered important physical essence might be
very well implied by such uniformity in the structural information of the microcosm
and the macrocosm. If these two quantities are seen as roughly equal, we have

e2 1/H c/H
≈ 2 ≈ 2 . (15.17)
Gmp me e /(me c )/c
2 e /(me c2 )

Cross-multiplying equation (15.17) gives

1 e2 /(me c2 )
Gmp me × ≈ e2 × , (15.18)
H c
or

c e2
Gmp me × ≈ e2 × . (15.19)
H me c2

The meaning of Equations (15.18) and (15.19) is that: The product of the universal
gravitation and the age of the universe (or the “radius of the universe”) approximately
equal the product of the static electrical force (in an H2 atom) and the time for a light
beam to travel the distance of the radius of an electron (or “the radius of an electron”).
Because measures for the universe are connected by the universal gravitation, and
measures for atoms in the microcosm are connected by electromagnetic force, the
physical meaning of Equations (15.18) and (15.19) becomes clear: The product of
physical quantities of the universal scale approximately equals the product of relevant
physical quantities of the microcosm. That is to say that there is uniformity between
the universe and the atomic world. Does this physical meaning of the large number
hypothesis imply the following: Besides the potential unification of the four basic
physical forces in the physical world, electromagnetic interaction, weak interaction,
strong interaction, and gravitational interaction, there might be a grand unification
230 15 Some Empirical Justifications

in which the products of some relevant physical quantities in the universal scale and
in the micro-scale approximately equal a fixed constant?
To this end, we have the following facts: According to Xian and Wang (1987),
the radius of an H2 atom (Bohr radius) is given by

ry  0.529 × 10−8 cm,

the course velocity (reaction rate) of electromagnetic interactions is

Vce  1016 − 1019 s−1 .

The forcing distance of strong interactions satisfies

rq < 10−13 cm,

∼

the course velocity of strong interactions is given by

Vcq  1021 −1023 s−1 .

So, the following is obtained:

ry × Vce  0.529 × 108 − 0.529 × 1011 cm s−1 (15.20)

and

rq × Vcq ≈ 108 − 1010 cm s−1 . (15.21)

That is, the following has been obtained

ry × Vce ≈ rq × Vcq . (15.22)

The meaning of equation (15.22) is that the product of some physical quantities
of electromagnetic interactions approximately equals that of relevant physical quan-
tities of strong interactions. This end implies that there exists uniformity between
electromagnetic interactions and strong interactions.

15.2.2 The Angular Momentum Mystery of the Solar System

This mystery of our solar system is the following: In the solar system, the total mass
of all planets accounts for 0.135% of the total mass of the solar system, while the total
angular momentum of all planets accounts for 99.421% of the total angular momen-
tum of the solar system. In other words, the mass of the sun amounts to 99.865% of
15.2 Conservation of Informational Infrastructure 231

the mass of the solar system, while the angular momentum of the sun only amounts
to 0.579% of the total angular momentum of the solar system. Correspondingly, two
natural questions arise here: How did this mystery form? And, are the masses and
the relevant angular momentums related?
As in the previous subsection, this subsection will treat the masses and the relevant
angular momentums of the solar system similarly: Multiply them together; see (Ren
and Hu 1989) for more details.
In (Dai 1979), the masses and relevant angular momentums with respect to the
center mass of the solar system of the sun and the planets in the solar system are given.
So, the sum of all the products of the masses (mp ) and their angular momentums (Jp )
of all the planets can be obtained as follows:

mp Jp  4.144 × 1088 g2 cm2 s−1 , (15.23)
p

and the product of the mass m⊕ and the angular momentum J⊕ with respect to the
center of mass of the solar system of the sun can be obtained as follows:

m⊕ × J⊕  4.143 × 1088 g2 cm2 s−1 . (15.24)

The numerical values in Equations (15.23) and (15.24) are very close and can be
seen as the same. In other words, we have seen the following law of conservation:
The product of the mass and the angular momentum of the sun are equal to the sum
of the products of the masses and the relevant angular momentums of all the planets.
That is, the mystery of the solar system’s angular momentum can be resolved as
follows: Under the conditions of time and space of the solar system, there exists the
conservation of phenomenon of equal mass-angular-momentum products between
the sun and the planets. Hence, it is very possible that the stability of the solar system
is realized through the law of conservation that mass-angular-momentum products
between the sun and the planets are equal.
In terms of celestial mechanics, Ren and Hu (1989) show that in a system, consist-
ing of two celestial bodies—a two-body problem—the motion of the smaller-mass
body can be obtained by solving the dynamic equation that describes how the smaller-
mass body (m) circles around the larger-mass body (M ). From
dσ
J  mr 2  m G(M + m)a(1 − e2 ), (15.25)
dt
where J stands for the orbital angular momentum, r the orbital vector, d σ/dt the
angular velocity, a the orbital radius, and e the orbital eccentric rate, establishing a
rectangular coordinate system with the origin located at the center of mass of the
two-body system leads to the angular momentum J 1 of the center of larger-mass
celestial body and the angular momentum J 2 of the smaller-mass celestial body as
follows:
232 15 Some Empirical Justifications

Mm
J1  J (15.26)
(M + m)2

and

M2
J2  J. (15.27)
(M + m)2

Since the center mass of the two-body system is always located on the line segment
connecting the two celestial bodies, it now follows obviously from Equations (15.26)
and (15.27) that

MJ1  mJ2 . (15.28)

That is, the predicted law of conservation has been obtained: The product of the
mass and the angular momentum with respect to the center of mass of the two-body
system of the larger-mass celestial body (center body) is the same as that of the
smaller-mass celestial body (circling body).
As an example, let us look at the two-body system of the earth and the moon. Based
on the numerical values of the masses and the angular momentums with respect to
the center of mass of the earth–moon system (Dai 1979), we can calculate and obtain
that

J2  2.786 × 1041 g cm2 s−1 ,

J1  3.4266 × 1038 g cm2 s−1 ,

and

MJ1  2.0477 × 1067 g2 cm2 s−1  mJ2 . (15.29)

Equation (15.29) implies that in our earth–moon system, the law of conservation
of identical mass-angular-momentum products holds true.
For a system of a large-mass celestial body and many small-mass circling celestial
bodies, Xian and Wang (1987) show that the related multiplicative relationships of
mass-angular-momentums approximately hold true in the form of additive compo-
nents. So, it can be seen that in a celestial body system of a simple mechanics, there
is a law of conservation of mass-angular-momentum products between the center
celestial body and the circling celestial bodies.

15.2.3 Movements of the Atmosphere on Earth

In meteorology, weather systems, representing movements of the earth’s atmosphere,

are classified as large-(lar), medium-(mid), small-(li), and micro-(mic) scale systems.
15.2 Conservation of Informational Infrastructure 233

These systems of different classifications have their scales of spatial level measure-
ments (L) set at approximately 108 , 107 , 106 , 105 cm, their measures of vertical
velocity (W) set at approximately 100 , 101 , 102 , 103 cm s−1 , and the measures of
their life spans (τ ) at approximately at 106 , 105 , 104 , 103 s, respectively.
Similar to what has been done previously, let see the products of relevant quantities
for weather systems of the same classification as follows:
(1) For the products of spatial level measures and relevant vertical velocities, we
have

Llar × Wlar ≈ 108 cm2 s−1

≈ Lmid × Wmid
≈ Lli × Wli
≈ Lmic × Wmic . (15.30)

(2) For the products of life spans and relevant vertical velocities, we have

τlar × Wlar ≈ 106 cm2

≈ τmid × Wmid
≈ τli × Wli
≈ τmic × Wmic . (15.31)

Equations (15.30) and (15.31) obviously mean that no matter which classification
of a weather system is in, the products of its spatial level measure and its vertical
velocity are always approximately equal to a fixed constant; and the same holds true
for the product of the system’s life span and its vertical velocity. That is to say, in
the earth atmosphere, there also exists a law of conservation of products between
different spatial measures and relevant vertical velocities or between different life
spans and relevant velocities.

15.2.4 Conjecture of a Law of Conservation of Informational

Infrastructure

As is well known, laws of conservations are the most important laws of science. For
example, in physics, there are many established laws of conservation dealing with
many important concepts, such as mass, energy, momentum, moment of momentum,
electric charge, etc. The research of modern physics indicates that each moving object
in an even and isotropic space and time, no matter whether the space is microscopic, or
mesoscopic, or macroscopic, a particle or a field must follow the laws of conservation
of energy, momentum, and moment of momentum. That is, a unification of space has
been achieved, which is called a Minkowski space. In Einstein’s relativity theory, the
234 15 Some Empirical Justifications

concepts of mass, time, and space have been closely connected; and the mass–energy
relation realizes the unification of mass and energy.
In the discussion above, it has empirically shown that between the macrocosm
and the microcosm, between the electromagnetic interactions of atomic scale and
the strong interactions of Quark’s scale, between the central celestial body and the
circling celestial bodies of celestial systems, between and among the large, medium,
small, and micro-scales of the earth’s atmosphere, there also exist laws of conserva-
tion of products of spatial physical quantities. Based on this empirical fact, we have
the following conjecture: There might exist a more general law of conservation in
terms of structure, in which the informational infrastructure, including time, space,
mass, energy, etc., is approximately equal to a constant. Symbolically, this conjecture
can be written as follows:

AT × BS × CM × DE  a (15.32)

or more generally,

AT α × BS β × CM γ × DE ε  a, (15.33)

where α, β, γ , ε and a are constants, T , S, M, E and A, B, C, D are, respectively,

time, space, mass, energy, and their coefficients. These two formulas can be applied to
various conservative systems of the universal, macroscopic, and microscopic levels.
The constants α, β, γ , ε, and a are determined by the initial state and properties of
the natural system of interest.
In equation (15.32), when two (or one) terms of choice are fixed, the other two (or
three) terms will vary inversely. For example, under the conditions of low speed and
the macrocosm, all the coefficients A, B, C, and D equal 1. In this case, when two
terms are fixed, the other two terms will be inversely proportional. This end satisfies
all principles and various laws of conservation in the classical mechanics, including
the laws of conservation of mass, momentum, energy, moment of momentum, etc.
So, the varieties of mass and energy in this case are reflected mainly in changes in
mass density and energy density. In the classical mechanics, when time and mass are
fixed, the effect of a force of a fixed magnitude becomes the effect of an awl when
the cross section of the force is getting smaller. When the space and mass are kept
unchanged, the same force of a fixed magnitude can have an impulsive effect, since
the shorter the time the force acts the greater density the energy release will be. When
time and energy are kept the same, the size of working space and the mass density
are inversely proportional. When the mass is kept fixed, shrinking acting time and
working space at the same time can cause the released energy density reaching a
very high level.
Under the conditions of relativity theory, that is, under the conditions of high
speeds and great amounts of masses, the coefficients in equation (15.32) are no longer
equal to 1, and equation (15.33) becomes more appropriate, and the constants A, B,
C, D, and a and the exponents α, β, γ , and  satisfy relevant equations in relativity
15.2 Conservation of Informational Infrastructure 235

theory. When time and space are fixed, the mass and energy can be transformed back
and forth according to the well-known mass–energy relation:

E  mc2 .

When traveling at a speed close to that of light, the length of a pole will shrink
when the pole is traveling in the direction of the pole and any clock in motion will
become slower. When the mass is sufficiently great, light and gravitation deflection
can be caused. When a celestial system evolves to its old age, gravitation collapse
will appear and a black hole will be formed. Based on equation (15.32), it can be
imagined that when our earth evolves sufficiently long, say a billion or trillion years,
the relativity effects would also appear. More specifically speaking, in such a great
time measurement, the creep deformation of rocks could increase and solids and
fluids would have almost no difference so that solids could be treated as fluids.
When a universe shrinks to a single point with the mass density infinitely high, a
universe explosion of extremely high energy density could appear in a very short
time period. So, a new universe is created!

15.2.5 Impacts of the New Law of Conservation

If the perceived law of conservation of informational infrastructure holds true, which

seems to be so as suggested by all the empirical data presented earlier, its theoretical
and practical significance is obvious.
The hypothesis of this law of conservation contains the following facts:
(1) Multiplications of relevant physical quantities in either the universal scale or
the microscopic scale approximately equal a fixed constant; and
(2) Multiplications of either electromagnetic interactions or strong interactions
approximately equal a fixed constant.
In the widest domain of human knowledge of our modern time, this law of con-
servation deeply reveals the structural unification of different layers of the universe.
It can be a new evidence for the big bang theory and supports the model for an infi-
nite universe with border and the oscillation model for the evolution of the universe,
where the universe evolves as follows:

. . . → explosion → shrinking → explosion → shrinking → . . .

This law of conservation also supports the hypothesis that there exist universes
“outside of our universe.” The truthfulness of this proposed law of conservation is
limited to the range of “our universe” with its conservation constant being determined
by the structural states of the initial moment of “our universe.”
All examples employed earlier show that to a certain degree, the proposed law
of conservation holds true. That is, there indeed exists some kind of uniformity in
236 15 Some Empirical Justifications

terms of time, space, mass, and energy among different natural systems of various
scales under either macroscopic or microscopic conditions or relativity conditions.
Therefore, there might be a need to reconsider some classical theoretical systems so
that our understanding about nature can be deepened. For example, under the time
and space conditions of the earth’s atmosphere, the traditional view in atmospheric
dynamics is that since the vertical velocity of each atmospheric huge scale system is
much smaller than its horizontal velocity, the vertical velocity is ignored. As a matter
of fact (Ren and Nio 1994), since the atmospheric density difference in the vertical
direction is far greater than that in the horizontal direction, and since the gradient
force of atmospheric pressure to move the atmospheric system 10 m vertically is
equivalent to that of moving the system 200 km horizontally, the vertical velocity
should not be ignored. The law of conservation of informational infrastructure, if
it holds true for all scales used in the earth’s atmosphere, might provide conditions
for a unified atmospheric dynamics applicable to all atmospheric systems of various
scales. As a second example, in the situation of our earth where time and mass do
not change, in terms of geological time measurements (sufficiently long time), can
we imagine the force, which causes the earth’s crust movements? Does it have to be
as great as what is believed currently?
As for applications of science and technology, tremendous successes have been
made in the macroscopic and microscopic levels, such as shrinking working spatial
sectors, shortening the time length for energy releasing, and sacrificing partial masses
(say, the usage of nuclear energy). However, the law of conservation of informational
infrastructure might very well further the width and depth of applications of science
and technology. For example, this law of conservation can provide a theory and
thinking logic for us to study the movement evolution of the earth’s structure, the
source of forces or structural information which leads to successful predictions of
major natural and social events, and to find the mechanisms for the formation of
torrential rains, for the arrival of earthquakes, and for the appearance of economic
disasters (Ren 1996).
Philosophically, the law of conservation of informational infrastructure indicates
that in the same way as mass, energy is also a characteristic of physical entities.
Time and space can be, respectively, seen as forms of physical existence of entities
with motion as their basic characteristics. This law of conservation connects time,
space, mass, and motion closely into an inseparable whole. So, time, space, mass, and
energy can be seen as attributes of systems, be they natural or consisting of people.
With this understanding, the concept of mass is generalized and the wholeness of
the world is further proved and the thoughts of never diminishing mass and never
diminishing universes are evidenced.
15.3 Bodily Communications that Can Be Felt Undoubtedly 237

15.3 Bodily Communications that Can Be Felt Undoubtedly

This section looks at how the systemic yoyo model is manifested in different areas of
life by briefly visiting relevant experimental and clinical evidences. All the omitted
details can be found in the relevant references.
Each person is a system, which is made up of various organs, body parts, thoughts,
cells, etc. So, the yoyo model of each and every system suggests that each human
being is a three-dimensional realization of a spinning yoyo structure of a certain
dimension higher than three. At the same time, each human being possesses a spin-
ning field that interests with the environment. To illustrate this end, we consider two
simple and easy-to-repeat experiments.

15.3.1 Visual Stares

It is well known that if we stare at a person briefly with certain intensity, that person
and his or her partner, if there is one such partner at the time moment of the stare, will
sense it almost instantly and they can locate the origin of stare without any trouble.
To further illustrate this scenario, let us imagine that we go to a sport event, say a high
school swim meet, where three regional high school teams are competing. Assume
that the area of competition satisfies the following conditions: (1) There is a pool
of the Olympic size; (2) along one long side of the pool there are about 200 seats
available for spectators, such as the parents and other relatives of the swimmers, to
watch the events of the meet; and (3) the pool area is enclosed with a roof and walls
all around.
Now, we physically enter the pool area for a forthcoming event, where 6 swimmers
will soon compete with butterfly strokes. As soon as we enter the enclosed area of
competition, we immediately fall into a steaming pack of screaming and jumping
spectators, cheering for their favorite swimmers currently competing in the pool.
After finding our seats in the bleacher area with a distance away from the pool
deck, still waiting for the current event to finish, let us purposelessly pick a voluntary
helper standing or walking on the pool deck for whatever reason, be it her beauty or
his strange look or body posture. Next, let us stare at this person intensively for a
moment. Then, here is what will happen almost instantaneously:
Magically enough, our stare will be felt by the person from quite a good distance.
She/he will turn around and locate us in no time out of the reasonably sized audience
filled with screaming and jumping spectators.
By using the systemic yoyo model, one explanation for why this happens and
how the silent communication takes place can be provided. In particular, each side,
the person being stared at and us, is a high-dimensional spinning yoyo. Even though
we are separated by space and by informational noise, the stare of one party on the
other has directed that party’s spin field of the yoyo structure into the spin field of
the yoyo structure of the other party. Even though the latter party initially did not
238 15 Some Empirical Justifications

know the forthcoming stare, when original state of her/his spin field is interrupted
by the sudden intrusion of another unexpected spin field, the person surely senses
the exact direction and location from which the intrusion is coming from. That is the
underlying mechanism for the silent communication to be established.
When this experiment is done in a large auditorium where the person being stared
at is on the stage, the afore-described phenomenon does not occur. It is because when
many spin fields interferes with the field of a same person, these interfering fields
actually interrupt and destroy their originally organized flows of materials and energy
so that the person who is being stared at can only feel the overwhelming pressure
from the entire audience instead of from individual persons. If a person is not good at
public speaking, then the person can more easily feel the said overwhelming pressure.
This easily repeatable experiment in fact has been conducted by some of the local
high school students. For example, when these students eat out in a restaurant and
after they run out of topics to gossip about, they play the game they call “feel the
vibe.” What they do to play this game is to stare as a group at a randomly chosen guest
of the restaurant to see how long it takes the guest to feel their stares. As described in
the situation of swim meet above, the chosen guest can almost always feel the stares
immediately and can locate the intruders in no time.

15.3.2 Reciprocal Impressions of People

In this subsection, let us look at an interesting phenomenon that widely exists in sit-
uations of human relationship—the reciprocal impression of people.
When a person A has a good impression about another person B, magically, person
B also has a similar and almost identical impression about A. In particular, if A does
not like B and describes B as a dishonest person with various undesirable traits, it has
been clinically proven in psychology that what A describes about B is exactly who
A is himself. On the other hand, A’s liking of B is identically reciprocated with B’s
liking of A (Hendrix 2001). Such quiet and unspoken evaluations of one another can
be seen in any working environment or family establishments. For instance, in a work
situation, quality is not and cannot be quantitatively measured, such as teaching at
such an institution in the U.S.A. that emphasizes on teaching only. When one teacher
does not perform well in his line of work, he generally uses the concept of quality
loudly in day-to-day settings in order to cover up his own deficiency in quality. When
one does not have honesty, he tends to use the term honesty all the time. It is exactly
as what Lao Tzu (exact time unknown, Chapter 1) said over 2,000 years ago: “The
one who speaks of integrity all the time does not have integrity.”
So, a natural question that arises here is: What is the underlying mechanism for
such a quiet and unspoken evaluation of each other to take place in real life? Being
able to address this question appropriately and apply the corresponding techniques
adequately is very important for professionals who are still climbing the corporate
ladder in their careers and those who are still struggling to achieve higher levels of
accomplishments.
15.3 Bodily Communications that Can Be Felt Undoubtedly 239

As a matter of fact, the underlying mechanism for such a quiet and unspoken eval-
uation of each other is based on that each human being stands for a high-dimensional
spinning yoyo and its rotational field. Our feelings toward each other are formed
through the interactions of our either visible or invisible yoyo structures and their
spin fields. So, when person A feels good about another person B, it generally means
that their underlying yoyo structures possess the same or very similar attributes, such
as spinning in the same direction, both being either divergent or convergent at the
same time, both having similar intensities of spin, etc. When person A does not like
B and lists many undesirable traits B possesses, it fundamentally stands for the sit-
uation that the underlying yoyo fields of A and B are fighting against each other in
terms of either opposite spinning directions, or different degrees of convergence, or
in terms of other attributes. For a more in-depth analysis along a similar line, please
consult with (Lin and Forrest 2011).
When an experiment is designed and conducted to demonstrate the wide-ranging
existence of this phenomenon with local high school students as participants, it is
found that after their friendship between two students A and B turned sour, they have
both been feeling sadly that the other party no longer likes her. In particular, when
we ask A why she does not like B anymore, the answer is exactly what we expect:
“Because she does not like me anymore!”

15.4 Remarks

By revisiting Bjerknes’ Circulation Theorem, it is empirically shown that any force

acting on a fluid must be a twisting force, creating spinning currents in the fluid. Other
than employing this theorem to explain the formation of land–ocean breezes, this
theorem also implies that each (mathematical) nonlinearity stands for an eddy source
that creates consequently eddy motions. This theorem in fact provides a theoretical
explanation for why the mostly seen form of motion in the universe is rotation,
because everything in the universe behaves like a body of fluid, when seen from
a land distance, and uneven densities lead directly to eddy sources and then eddy
motions. And, because when one eddy rotates in one direction, its environment, when
seen as another eddy, always rotates in the opposite direction. Such forever existing
duality of eddies in rotational directions destroys the initial smooth fields, if any. And
that explains why there must be situations, processes, and phenomena that theories
developed on the assumption of smoothness, such as calculus-based theories and
statistics-based methods, cannot explain.
Based on how the concept of (general) systems is understood, the law of con-
servation of informational infrastructure is introduced empirically. That provides
another vivid evidence for the physical existence of systemic yoyo structures behind
systems. And, in the last section of this chapter, easily repeatable observations in set-
tings of different social events and human interactions are analyzed in the name of
the systemic yoyo model. These experiments clearly demonstrate the validity of the
systemic yoyo model in areas of social sciences, such as sociology and humanities.
240 15 Some Empirical Justifications

References

Allen, C.W.: Astrophysical Quantities, 3rd edn. The Athlone Press, New York, NY (1976)
Dai, W.S.: The Evolution of the Solar System, vol. 1. Shanghai Science and Technology Press,
Shanghai (1979)
Dirac, P.A.M.: The cosmological constants. Nature 139, 323 (1937)
Hendrix, H.: Getting the Love you Want: A Guide for Couples. Owl Books, New York (2001)
Hess, S.L.: Introduction to Theoretical Meteorology. Holt, Rinehart and Winston, New York (1959)
Klir, G.: Architecture of Systems Problem Solving. Plenum Press, New York, NY (1985)
Kuchemann, D.F.: Report on the I.U.T.A.M. symposium on concentrated vortex motions in fluids.
J. Fluid Mech. 21, 1–20 (1965)
Lin, Y.: A model of general systems. Math. Model. Int. J. 9, 95–104 (1987)
Lin, Y.: Developing a theoretical foundation for the laws of conservation. Kybernetes: Int. J. Syst.
Cybern. 24, 52–60 (1995)
Lin, Y.: Mystery of nonlinearity and Lorenz’s chaos. Kybernetes: Int. J. Cybern. Syst. Manag. Sci.
27(6–7), 605–854 (1998)
Lin, Y.: Systemic Yoyos: Some Impacts of the Second Dimension. Auerbach Publications, an imprint
of Taylor and Francis, New York (2009)
Lin, Y., Fan, T.H.: The fundamental structure of general systems and its relation to knowability of
the physical world. Kybernetes: Int. J. Syst. Cybern. 26, 275–285 (1997)
Lin, Y., Forrest, B.: Systemic Structure behind Human Organizations: From Civilizations to Indi-
viduals. Springer, New York (2011)
Pan, N.B.: Hubble constant. In: Large encyclopedia of China: astronomy, p 108. Chinese Large
Encyclopedia Press, Beijing (1980)
Ren, Z.Q.: The resolution of the difficulties in the research of earth science. In: Wu, F.H., He, X.J.
(eds.) Earth Science and Development, pp. 298–304. Earthquake Press, Beijing (1996)
Ren, Z.Q., Hu, Z.W.: The product relation of mass-angular momentum in celestial body systems.
Advances of Comprehensive Study of Mutual Relations in Comos: Earth and Life, pp. 269–274.
Chinese Science and Technology Press, Beijing (1989)
Ren, Z.Q., Lin, Y., OuYang, S.C.: Conjecture on law of conservation of informational infrastructures.
Kybernetes: Int. J. Syst. Cybern. 27, 543–552 (1998)
Ren, Z.Q., Nio, T.: Discussions on several problems of atmospheric vertical motion equations.
Plateau Meteorol. 13, 102–105 (1994)
Saltzman, B.: Finite amplitude free convection as an initial value problem. I. J. Atmos. Sci. 19,
329–341 (1962)
Xian, D.C., Wang, Z.X.: Particle physics. In: Large Encyclopedia of China: Physics, pp. 737.
Chinese Large Encyclopedia Press, Beijing (1987)
 Part III
Applications in Business
Decision Making
Chapter 16
Economic Events and Processes

In this chapter, we learn how the systemic yoyo model and the new figurative analysis
method (Chapter 13) can be applied to the study of economics. Because humans are
involved in each economic situation and their desires and beliefs alter the evolution
of the outcome, leading to unpredictable, chaotic consequences (Soros 1998), this
chapter shows that when each economic entity is seen as a rotating yoyo with a spin
field around it, the seemingly unpredictable, chaotic consequences of human desires,
beliefs, and corresponding behaviors no longer look unpredictable and chaotic.
What is new in the rest of the book from the mainstream literature in economics is
that we learn how various problems considered in economics can be seen as problems
of (general) systems and spinning yoyos in light of whole evolutions of these systems
(Lin 1999). Thus, it is shown that in an open market of free competition, the concept
of demand and supply is in fact about how different economic forces mutually restrict
each other and mutually react on each other. Because of this reason, we lay down
the theoretical foundation for the introduction of the systemic yoyo model and its
methodology into the research of economic entities, organizations, relationships, and
evolutions.
As an example of application, we apply such a yoyo structure to model the evolu-
tion of competitions between economic sectors and individual enterprises from the
moment when a sector or an enterprise is born to the time when it is gone. By looking
at economic competitions as interactions between spinning yoyos, we are ready to
employ the methods and results discovered and well developed in fluid dynamics to
the research of economics. And to prepare for the rest of this book, in the third section
of this chapter, we rigorously establish two theoretical results to confirm what is fig-
uratively seen from the yoyo model indeed holds true generally. In particular, the first
theorem shows how market-wide competition stimulates and invites the entrance of
new competitors, while the second result explains the reason why each organization
with at least two employees has to experience internal competitions.

© Springer Nature Switzerland AG 2018 243

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_16
244 16 Economic Events and Processes

16.1 How Market Demand and Supply Interact with Each

Other

Let us consider an open market of free competition. The price P of a consumer good
is closely related to the demand D and the supply S. In fact, changes in the price P
are directly proportional to the difference of the demand and supply if we only look
at the current momentary relationship between these variables, that is, the general
relationship between these variables is linearized for the current time moment:
dP
 k(D − S), k > 0 (16.1)
dt
where t stands for time and k is a constant. With all other conditions fixed, it is clear
that both the demand D and the supply S are functions of the price P, written as:

D  D(P) and S  S(P). (16.2)

The relationship between the demand D and the price P is generally linear if once
again we focus only the current time moment. This linear relationship can be written
as follows:

D(P)  −λP + β, λ, β > 0, (16.3)

where λ is the rate of change of the demand with respect to the price P and β is the
saturation constant of the demand.
On the other hand, the relationship between the supply S and the price P is
generally nonlinear, because initially when the price P moves higher, the supply S
also increases; however, when the supply S reaches the level of market saturation,
no matter how the price can still potentially go higher, the overproduction will force
the supply to drop. Let us write this nonlinear relationship between the supply S and
the price P symbolically as (at least by means of local approximation)

S(P)  δ + αP + γ P 2 (16.4)

where δ > 0 is a constant and α and γ are, respectively, the linear and nonlinear
intensities of the supply, satisfying α > 0 and γ < 0.
Substituting equations (16.3) and (16.4) into equation (16.1) produces
dP
 AP 2 + BP + C, (16.5)
dt
where A  −kγ > 0, B  −k(λ + α) < 0, and C  k(β − δ) > 0 (because β is the
market demand when the price P = 0 and δ the corresponding market supply).
For equation (16.1) or (16.5), the majority of the relevant literature focuses on
the study of the price stability at the demand–supply equilibrium, while ignoring the
16.1 How Market Demand and Supply Interact with Each Other 245

whole evolutionary characteristics of the price. Because the price evolution model
in equation (16.5) is quadratic, the price in general changes discontinuously and has
the characteristic of singular reversal transitions. In particular, the discriminant of
equation (16.5) can be computed as follows:
 
  k 2 (λ + α)2 + 4γ (β − α) .

So, we have three possibilities:  = 0,  > 0, and  < 0. In the following, let us
analyze the evolution model in each of these cases.
Case 1:  = 0. At the equilibrium there is only one equilibrium price, P 
−B/2A > 0. The whole evolution of the price is characterized by
−1 B
P − , (16.6)
At + P0 2A

where P0 is the integration constant that is determined by the given initial condition.
• If P0 > 0, the price P increases continuously with time and approaches the equi-
librium state (−B/2A).
• If P0 < 0, then when t  tb  −P0 /A, a discontinuity occurs to the price change.
– When t < tb , the greater the demand, the higher the price.
– When t  tb , the price falls abruptly, indicating that either the demand reaches
its level of saturation or the supply increases drastically, causing the abrupt drop
in the price, and
– When t > tb , the price starts to rebound continuously with time and eventually
approaches the demand–supply equilibrium P1 .
This process indicates that through market adjustments, the price is no longer
growing as blindly as during the first period of time. Instead, by taking the market
situation into consideration, to keep a reasonable equilibrium between the supply
and demand, the price eventually stabilizes within a certain range.
This whole evolutionary analysis of the price P is not only more complete than the
stability analysis at the equilibrium, but also more practically realistic than studies
of continuous evolutions when compared to the objective situations existing in the
marketplace of free competitions, where sudden and disruptive developments appear
quite regularly.
Case 2:  > 0. Solving equation (16.5) leads to
 
 2P + X − √Y   √ 
 
 √   exp A Y t + P20 (16.7)
 2P + X + Y 

where X  B/A
√ < 0, Y  (B/A) − (4C/A), and P20 is the integration constant. If
2

|2P + X |  Y , then equation (16.7) indicates that changes in price are continuous.
Otherwise, equation (16.7) indicates that the price evolution contains blown-ups,
246 16 Economic Events and Processes

because in this case, the left-hand side of equation (16.7) is equal to 0, while the
right-hand side is not.
Case 3:  < 0. Solving equation (16.5) produces
 
Z AZt X
P  tan + P30 − , (16.8)
2 2 2

where Z  (4C/A) − (B/A)2 and P30 is the integration constant. This equation
implies that at
2
π 
t  tb  + nπ − P30 ,
Z 2
for n = 0, ±1, ±2, …, periodic blown-ups in the price occur. That is, the price of the
consumer good shows the behavior of periodic rise-and-fall.
This analytic reasoning and symbolic computation shows that the concept of
demand and supply is in fact about mutual restrictions or mutual reactions of quasi-
equal mutual constraints of different forces under equal quantitative effects. In par-
ticular, the so-called demand and the so-called supply actually represent a pair of
quasi-equal acting and reacting forces of the marketplace. When they exert their
quasi-equal quantitative effects on each other, a system of equal quantitative move-
ments, where the demand and supply evolve around each other and reinforced one
another, creating the so-called equal quantitative effects, which are the eddy effects
with nonuniform vortical vectorities existing naturally in the system of movements
due to the unevenness existing in the interactions between the demand and supply.
For example, in a principal–agent problem, the principal and the agent can be
treated as a market demand and market supply, respectively, so that they represent
a pair of interacting and mutually constraining parties and forces. Each interaction
between the parties is a real-life equal quantitative movement. The effect of such a
movement can be seen as an equal quantitative effect. It is shown that if the inter-
actions between these parties are investigated with the concepts of time and space,
then the interacting forces between the parties actually indicate that they attract and
reinforce the parties. In other words, market demands and supplies in fact reinforce
each other in one way or another so that they grow stronger or weaker together. Their
interacting forces are created within each of themselves by the unevenness of the
internal structures. So, from Chapters 13 and 14, it follows that as soon as calculus-
based symbolic computation is employed to the study of such a relationship between
demand and supply, one has to face the problem of nonlinearity, an unsettled math-
ematics problem. From the blown-up theory, a general theory of development and
evolution, it follows that nonlinearity is about structures of the entities involved. So,
to understand nonlinearity for theoretical and practical reasons, the systemic yoyo
model comes into play.
For our specific purpose, let us see in the next section how such a systemic yoyo
structure can be employed to model the evolution of competitions between economic
sectors and individual enterprises.
16.2 Model Economic Cycles by Using Systemic Yoyos 247

16.2 Model Economic Cycles by Using Systemic Yoyos

For the sake of convenience of presentation, we will only look at the general situation
for a new economic sector to appear and evolve. To do so, let us model each existing
economic sector and/or an individual company as a spinning yoyo, where the sector’s
products and/or services are the outputs spit out of the yoyo, and the profits and other
economic benefits the inputs sucked into the yoyo.
When several yoyo fields coexist side by side (Figure 16.1), their mutual inter-
actions lead to the appearance of new products and new services, modeled by the
sub-eddies between the larger yoyo pools N and M in Figure 16.1. To cash in on the
profit opportunities coming along with the new products and services, new companies
are first established, followed by a new economic sector. At this stage of growth and
evolution, the technology of producing the new products is generally quite inefficient,
because it is a preliminary combination of the technology and equipment from vari-
ous existing sectors, where the idea of the new products was initially based. So, the
cost of production is quite high. Additionally, because these new products/services
are relatively unknown in the marketplace and have no loyal followings, the demand
is very small. Consequently, the need for mass production is slow. Even though the
spinning of the new yoyo seems to be slow, the gradually increased market awareness
and demand indicate that a new technology and an appropriate management routine
need to be introduced. And the increasing spinning strength of the yoyo provides the
adequate capital for such needed technological and managerial innovations.
Specifically, Figure 16.1a models the scenario that some component m in an
existing economic sector N is pushed outward by the yoyo field N, while it is also
pushed outward by another economic sector M. Here, the component m can be a
division of an existing company, the idea of a new product which the market does not
see highly valuable, etc. Because neither of the existing economic sectors is willing
to host the component m, an initially embryonic field, the small eddy fields located
between N and M, started to form. Similar analyses are applicable to Figure 16.1b–d,
each of which is different from the scenario in Figure 16.1a that the yoyo field of
one of the existing economic sector is altered from diverging spin to converging spin
or from harmonic spin pattern to inharmonic pattern. That is, the cases depicted in
Figure 16.1a–d represent all possible cases on how the yoyo fields of two existing
economic sectors would interact with each other.
In the second stage of development and evolution, additional companies, some of
which are trend followers, start to pop up at a high speed to occupy market shares
in the newly emerging market. The situation is similar to that when a calm pond
of water is initially disturbed, with the creation of a big whirlpool all areas of the
originally static water start to move. As a consequence, the relevant investments of
capital, human resource, technology, etc., are drastically increased so that the level
of production is brought up to a higher level.
To gain an upper hand in the market competition in terms of market shares (the
diverging side or output side of the yoyo body) and cash flows (the converging side
or input side of the yoyo body), the technology for mass production is improved
248 16 Economic Events and Processes

(a) Object m is located in a diverging eddy N and pushed by another diverging eddy M

(b) Object m is in a diverging eddy N and pulled or pushed by another converging eddy M

(c) Object m is located in a converging eddy N and pulled by another converging eddy M

(d) Object m is in a converging eddy N and pulled or pushed by another diverging eddy M

Figure 16.1 Some ways small yoyo structures can possibly be created

and becomes more efficient with lowered costs (the spinning strength of some yoyo
bodies becomes strong). So, the market territory of this new economic sector, led by
a few strong representative companies, opens up quickly.
After entering the third stage of evolution, the marketplace is gradually saturated
with a more-than-sufficient supply. As the revenue increases, the profit elasticity of
the products and services starts to erode. So, due to the weakening profit margins
(fewer materials can be inputted from the input side), further advances become slower
and more difficult.
At the end, the entire sector stops any attempt for further advancement. And the
production stabilizes at the level of replacing aged and broken products purchased
16.2 Model Economic Cycles by Using Systemic Yoyos 249

earlier. As a yoyo’s spin becomes weaker, meaning that fewer materials are given
off (from the output side) and sucked in (from the input side), caused by appear-
ances of other different but relevant new yoyos (economic sectors or companies), the
yoyo might be absorbed by another faster and stronger spinning structure, or simply
destroyed by several adjacent, more energetic yoyos.
As a matter of fact, each real-life economic entity goes through such a three-
staged life cycle with some minor variations and different specifics. Comparing to
the literature, the major contribution of this yoyo model is to see how the problem
of concern can be seen from a different angle as mutual interactions between several
spinning yoyos, that is, from the angle of overall development and whole evolution.
It stands for a dynamic model instead of a static one. Because of this characteristic,
one can sensibly expect that such a model can help bring forward new insights into
various existing and new researches.

16.3 Emergence of Market Competitions

The discussion in the previous section demonstrates intuitively how a new economic
sector appears out of the interaction or competition of the existing economic sectors.
In this section, we establish two theoretical results to confirm what is figuratively
seen earlier by using the rigor of game theory. In particular, the first result shows how
market-wide competition invites the entrance of new competitors, while the second
result details why each organization with at least two employees witnesses internal
competitions.
To develop the first result, let us look at market competitions that are external to
any incumbent firm of an established market. To this end, assume that the oligopoly
market of our concern consists of m firms, m  1, 2, …, providing consumers with
mutually substitutable products, with their respective shares of loyal consumers, who
only purchase products from their companies as long as the selling price is not more
than their reservation price, which is set to 1 without loss of generality. To protect
their turfs while potentially increase their consumer bases, they compete over the
switchers with adjustable prices charged to their customers in order to deter the
potential entrance of new competitions. So, assume that these firms produce their
horizontally differentiated products at constant marginal costs, which is set to zero
without loss of generality. Assume consequently the managements of these m firms
are well aware of the pricing strategies of the other firms and have established their
best responses by playing the Nash equilibrium through pure self-analyses.
Speaking differently, the market is in a state of mutual forbearance, where incum-
bent firms mitigate rivalry by dividing markets in proportion to firm strength. They
cede dominance to their stronger competitors in those market segments where they
are less efficient, while in exchange the latter do the same in segments where the
former are more efficient. The firms’ codependence gradually motivates them to de-
escalate rivalry. Eventually, the rates of entry and exit in the market decrease, and
inter-firm hostility declines. So, without loss of generality, let us consider the aggre-
250 16 Economic Events and Processes

Figure 16.2 Systemic

bird’s-eye view of our
marketplace of a previously
prosperous market

gate of the incumbent firms as one firm, whose share of occupancy of the market is
α so that β  1 − α represents the size of the market segment of switchers who base
their purchase decision on which price is lower. Then, the following result holds true:

Theorem 16.1. In the given oligopoly market with m incumbent firms, m  1, 2,

…, at least one firm profitably enters the market to compete with these m incumbent
firms, if, and only if, the size of the market segment of switchers is greater than zero.

Generally speaking, the previously assumed market condition means that the
technology involved and the relevant business operations have been standardized.
So, for a new small firm to enter such a market with profit potential, it is reasonable to
assume that this firm has come up with a more efficient technology and/or operation
that can greatly reduce the overall business expenditure. Further assume that this
small entrant uniformly randomizes its price P over the interval [0, 1] as long as the
firm could make profits on the average. Again, the constant marginal costs of this
small entrant are set to zero without loss of generality.
Before we see the detailed proof for this theorem, let us first see intuitively why
such a result holds true. To this end, let us fathom the marketplace is as an abstract
yoyo field, and we look at the multi-dimensional yoyo body at a distance from above
either the convergent side or the divergent side, while imagine that everything here
takes place in our three-dimensional space. That is, we are looking at a pool of
spinning fluid, where the word “fluid” is an abstraction of movement of all kinds
of media, such as goods, information, money, credit, etc., that appear and exist in
business activities. In other words, graphically one is looking at the market of concern
as the pool of spinning fluid shown in Figure 16.2.
Associated with this end of intuition, the well-known dishpan experiment shows
that when the movement of the fluid within the rotational dish is under enough
pressure created by either the sufficient speed of rotation or sufficient difference
in the temperature between the center and the periphery of the dish, the pattern of
uniform movement, as shown in Figure 16.2, will develop into the chaos, as shown in
Figure 16.3. The number of local eddy leaves is determined either by the rotational
speed or by the temperature difference or both and increases with the speed and the
temperature difference.
Now, the systemic modeling and laboratory experiment suggest that the fluid
nowhere within this spinning dish could avoid being disturbed by the flows, either
16.3 Emergence of Market Competitions 251

Figure 16.3 Asymmetric

flow observed in Fultz’s
dishpan experiment

orderly or chaotically, of the pan. And being disturbed regionally means that a local
flow pattern will appear inevitably.
Proof of Theorem 16.1. (⇒) Suppose that by randomizing its price over the
interval [0, 1] a small firm enters into the oligopoly market of m firms, which are
collectively seen as one aggregate firm, because these m firms are in a state of mutual
forbearance. So, the consumer surplus must satisfy β  1 − α > 0.
(⇐) Assume that the consumer surplus satisfies β  1 − α > 0. Firstly, let α0 be
a real number so that β  1 − α > α0 > 0, and α  α0 , where is a large natural
number, indicating that the market has been largely taken by the incumbent firms.
Secondly, let us imagine that the aggregate firm is divided into many identical
“firms,” named i, i  1, 2, …, . Each of them provides consumers with identical
products and enjoys the market share α0  α/ of loyal consumers. These imaginary
firms compete over the switchers with adjustable prices. Because these imaginary
firms are really equal partitions of the same aggregate firm, they have the same
constant marginal cost, which is set to zero without loss of generality, the manage-
ments of these firms are fully aware of the pricing strategies used by all the firms
(because the firms are managed by the same administrative unit), and they establish
their best, identical responses by playing the Nash equilibrium through their unified
self-analyses.
Thirdly, these imaginary firms do not have any symmetric pure strategy Nash
equilibrium (For the setup here, there is no need to consider asymmetric pure strategy
Nash equilibrium, because all these imaginary firms take identical actions.). In fact,
for any symmetric pure strategy portfolio (x1 , x2 , . . . , x ), where xi  xj , for i, j  1,
2, …, , a randomly chosen Firm j (∈ {1, 2, . . . , }) can slightly lower its price from
xj to xj to produce additional profits for all the firms as long as xj β > (xj − xj )α,
which is possible to do by adjusting xj sufficiently close to xj . So, (x1 , x2 , . . . , x ) is
not a Nash equilibrium. Even so, Forrest et al. (2017) show that these firms do have
a symmetric mixed-strategy Nash equilibrium.
For the rest of this proof, it suffices to show that there is one small firm that will be
expected to profit by entering this market through uniformly randomizing its price
strategy over the interval [0, 1].
Let F(P) be the price distribution of Firm j, one of the imaginary firms of the
aggregate firm. The aggregate firm or equivalently each of the imaginary forms
sets its price after taking into account the price of the new firm and those of all
252 16 Economic Events and Processes

other imaginary firms. Hence, the profits for Firm jfrom its loyal consumers are
α0 P and those from its share of the switchers are ij (1 − P)[1 − F(P)]βP 
βP(1 − P)[1 − F(P)] −1 . Hence, the profits
Firm j generates when the firm sells
its product at price P are

α0 P + βP(1 − P)[1 − F(P)] −1

and the objective function of Firm j is

+∞
 
maxF(P) E(
)  α0 P + βP(1 − P)[1 − F(P)] −1 dF(P)
−∞
1
 
 α0 P + βP(1 − P)[1 − F(P)] −1 dF(P)
0

where E(
) stands for Firm j’s expected profits for all possible prices, and the objec-
tive for Firm j is to maximize its expected profits by choosing its price distribution
F(P). The reason why the upper and lower limits of the integral are changed, respec-
tively, from +∞ and −∞ to 1 and 0 is because when P < 0 or when P > 1, the profits
are zero.
The equilibrium indifference condition of Firm j is

α0 × P + β × P(1 − P)[1 − F(P)] −1  α0 × 1 (16.9)

So, for the imaginary firms, solving equation (16.9) leads to their symmetric equi-
librium pricing strategy as follows:
  −1
1
α0
F(P)  1 − (16.10)
βP

From β > α0 , it follows that α0 /β < 1. So, for any price P, satisfying 1 ≥ P ≥ α0 /β,
equation (16.10) is a well-defined probability distribution. This end implies that for
the imaginary firms, or equivalently, the aggregate firm, the lowest allowed price
is α0 /β.
This proof is complete from the fact that the profits of the small entrant

0 /β
α +∞
E(
)  βPdP + βP[1 − F(P)] dP (16.11)
0 α0 /β

0 /β
α 1   /( −1)
α0
 βPdP + βP[1 − F(P)] dP + β (16.12)
β
0 α0 /β
16.3 Emergence of Market Competitions 253

are positive, where the first term in the right-hand side of equation (16.11) stands
for the expected profits of the small entrant when it charges the lowest price in the
marketplace and captures the entire segment of the switchers, the second term is the
small entrant’s expected profits when it is in direct competition with the incumbent
firms, and the third term in equation (16.12) comes from the mass point of size
(α0 /β)1/( −1) of F(P) at the reservation price P  1. This end implies that if the
consumer surplus β  1 − α > 0, there will be at least one small entrant that will
enter the market to compete with the incumbent firms. QED
The result in Theorem 16.1 indicates that although the m incumbent firms are
risk neutral and want to continuously reap in their respective profits by securely
defending their established turfs, they still have to fight over the switchers of the
marketplace in order to eliminate the switcher segment. Otherwise, new competi-
tion(s) will inevitably enter the market with potential of making good profits. On the
other hand, the proof of Theorem 16.1 provides a practically useful pricing method
for a firm to enter a profitable market occupied by powerful incumbent firms: Ran-
domize the selling price between the cost of the product and the reservation price the
incumbents charge their loyal customers.
Next, let us look at competitions that materially exist within any organizational
entity with at least two employees as reflected by the individual value systems of
employees.
Generally, no matter which business entity is concerned with, there are com-
petitions in terms of how the organization should be managed, how the detailed
operations should be carried out, and how employees’ efforts and devotions should
be directed. And each stakeholder of the organization always seems to have ideas
about how things could improve.
One reason why we discover abundant competitive situations is because each
person, as a living being that is severely limited by its sensing organs, looks at the
world with a pair of colored eyes. The word “color” in the literature is also known
by such terms as personal values and/or philosophical assumptions about the world.
In other words, because philosophical assumptions and value systems vary from one
person to another, from one people to another, from one culture to another, …, the
same physical world becomes extremely beautiful and multi-colored when people
individually try to describe what they see and what the world is really about.
By underlying assumptions and values of philosophy, we mean the value system
of a person that consists of his/her beliefs about how the world functions and his/her
moral codes with which he/she is recognized with his/her particular identity and
integrity. From the systemic yoyo model, it follows that each human being lives in
a vast ocean of spinning yoyo fields, which consists of the fields of other people,
physical objects, abstract thoughts, and myriad of other things and matters. Soon
after a person is born, he/she starts to interact with the world. It is these interactions
with people, physical objects, abstract thoughts, and the myriad of other things and
matters that shape the person’s philosophical assumptions and values, similar to how
a civilization formulates its value system. Because of the subtle differences between
the interactions experienced by one person from those by another person, each per-
son has his/her own set of very specific philosophical assumptions and values, which
254 16 Economic Events and Processes

dictate the behaviors and decision making of the person for the rest of his/her life.
Although the differences might be “subtle” when seen from the angle of the magnifi-
cent scale of the entire ocean of spin fields, they are generally major to the individuals
involved, causing important differences in the relevant personal value systems. That
actually explains why children who grow up in the same household may have quite
different personalities, characteristics, and thinking processes. And that explains why
different people have different underlying philosophical assumptions and values (the
value systems), because firstly no two people grow up within a perfectly identical
environment, and secondly with age people’s philosophical assumptions and values
evolve according to their respectively changing environments.
Proposition 16.1. Competitions always exist within any organizational system that
has at least two employees.
In fact, because no two employees share the same system of values, any two
employees will look at many aspects of the organizational system differently. That
difference between their value systems leads to competitive consequences of the two
employees.
Jointly, what is established in this section explains why for an economic entity to
survive and then to succeed within its environment, it has to balance the interactions
between internal and external competitions while make use of these competitions to
advance itself.

16.4 Concluding Remarks

Many scholars over time have looked at the concepts of demand and supply from
different angles. The discussion on these concepts from the angle of whole evolution,
as presented in Section 16.1, is based on Wu and Lin (2002). These authors were the
first to reveal how the interaction of market demand and supply is actually nonlinear
and experiences blown-ups. Section 16.2 is mainly based on Lin (2009). It provides a
nonlinear view on how economic cycles emerge so that the section provides a brand
new angle to look at and method to study these cycles. As for the emergence of market
competition, both internal and external to a firm, as discussed in Section 16.3, it is
based on the work by Forrest and Nightingale (2017).
It is OuYang (1994), see also Lin (1998), that the concept of quasi-equal quanti-
tative effects was initially introduced and applied to the forecast of the occurrence of
nearly-zero probability disastrous weather conditions. Lin (2007) shows that within
time and space the interacting forces between two entities actually reinforce each
other in one way or another so that they grow stronger or weaker together. The
blown-up theory was first systemically presented by Wu and Lin (2002), where they
show that nonlinearity implies brown-ups, or the break-up of the old structure and
emergence of new structures.
The well-known dishpan experiment was initially conducted successfully by Hide
(1953) of Cambridge University, England, and then by Dave Fultz and his colleagues
16.4 Concluding Remarks 255

of University of Chicago (1959) independently. The results of this experiment have

been employed in the study of meteorology.
The theoretical need for fighting off potential new competitions is rigorously
established by Forrest et al. (2017). These authors demonstrate that only through
adjusting their selling prices, the incumbent firms, no matter how powerful they
are, could possibly maintain their loyal customers while attracting a few switchers
who make purchase decisions based on whose price is lower. The special market
state, known as a state of mutual forbearance, has been well investigated by various
scholars, such as Bernheim and Whinston (1990), Li and Greenwood (2004), Yu and
Cannella (2012), Fuentelsaz and Gómez (2006), Haveman and Nonnemaker (2000),
and others.
Villalobos and Vargas (2015) and Terán et al. (2015) systematically considered
the concepts of personal values and philosophical assumptions about the world. Lin
and Forrest (2011) investigated how the value system of a person dictates the per-
son’s beliefs about how the world functions and her moral codes with which she is
recognized with her particular identity and integrity. Forrest and Orvis (2016) pro-
vide an explanation for why different people have different underlying philosophical
assumptions and values (the value systems).
Dr. Jennifer Nightingale, School of Business, Slippery Rock University (Slippery
Rock, PA 16057, USA; e-mail: Jennifer.nightingale@sru.edu), also contributed to
the work presented in this chapter.

References

Bernheim, B.D., Whinston, M.D.: Multimarket contact and collusive behavior. RAND J. Econ.
21(1), 1–26 (1990)
Forrest, Y.L.J., Nightingale, J.: Successfully transition into the era of transient competitive advan-
tages. In: Proceedings of 2017 NABET Annual Conference, 26–27 Oct 2017, State College, PA.,
pp. 57–79 (2017)
Forrest, Y.L.J., Buttermore, J., Wajda, T.A.: At Nash equilibrium when new market competitions
appear? Kybern. Int. Cybern. Syst. Manag. Sci. 46(2), 256–271 (2017)
Forrest, Y.L.J., Orvis, B.: Principles of management efficiency and organizational inefficiency.
Kybern. Int. J. Cybern. Syst. Manag. Sci. 45(8), 1308–1322 (2016)
Fuentelsaz, L., Gómez, J.: Multipoint competition, strategic similarity and entry into geographic
markets. Strateg. Manag. J. 27(5), 477–499 (2006)
Fultz, D., Long, R.R., Owens, G.V., Bohan, W., Kaylor, R., Weil, J.: Studies of Thermal Convection
in a Rotating Cylinder with Some Implications for Large-Scale Atmospheric Motion, Meteorol.
Monographs (American Meteorological Society) vol. 21, no. 4 (1959)
Haveman, H.A., Nonnemaker, L.: Competition in multiple geographic markets: the impact on
growth and market entry. Adm. Sci. Q. 45(2), 232–267 (2000)
Hide, R.: Some experiments on thermal convection in a rotating liquid. Q. J. R. Meteorol. Soc. 79,
161 (1953)
Li, S.X., Greenwood, R.: The effect of within industry diversification on firm performance: synergy
creation, multi-market contact and market structuration. Strateg. Manag. J. 25(12), 1131–1153
(2004)
256 16 Economic Events and Processes

Lin, Y. (guest ed.): Mystery of nonlinearity and Lorenz’s chaos. Kybern. Int. J. Syst. Cybern. 27,
605–854 (1998)
Lin, Y.: General Systems Theory: A Mathematical Approach. Plenum and Kluwer Academic Pub-
lishers, New York (1999)
Lin, Y.: Systemic yoyo model and applications in Newton’s, Kepler’s laws, etc. Kybern. Int. J.
Cybern. Syst. Manag. Sci. 36(3–4), 484–516 (2007)
Lin, Y.: Systemic Yoyos: Some Impacts of the Second Dimension. Auerbach Publication, an imprint
of Taylor and Francis, New York (2009)
Lin, Y., Forrest, B.: Systemic Structure Behind Human Organizations: From Civilizations to Indi-
viduals. Springer, New York (2011)
OuYang, S.C.: Break-Offs of Moving Fluids and Several Problems of Weather Forecasting. Press
of Chengdu University of Science and Technology, Chengdu (1994)
Soros, G.: The Crisis of Global Capitalism: Open Society Endangered. Public Affairs, New York
(1998)
Terán, O., Sibertin-Blanc, C., Gaudou, B.: The influence of moral sensitivity on organizational
cooperation. Kybernetes 44(6/7), 1067–1081 (2015)
Villalobos, J., Vargas, M.: Towards a simple dynamical model of citizens’ perception. Kybernetes
44(6/7), 1134–1145 (2015)
Wu, Y., Lin, Y.: Beyond Nonstructural Quantitative Analysis: Blown-Ups, Spinning Currents and
Modern Science. World Scientific, River Edge (2002)
Yu, T., Cannella Jr., A.A.: A comprehensive review of multimarket competition research. J. Manag.
39, 76–109 (2012)
Chapter 17
Issues of Competition

By employing the concept of Nash equilibria of game theory and relevant methods of
reasoning on the intuition of systems research, this chapter investigates the dynamics
of a coordinately monopolized market with m incumbent risk-neutral firms: How
these firms compete for customers who switch from the product of one company to
another by adjusting prices; how their expected profits stay stagnant; why their bases
of loyal customers deteriorate while customers become less patient than ever before.
The main conclusions include: (1) In a developed marketplace, risk neutrality
would lead to stagnation in profits and irrational decision on pricing; (2) when the
competition of the market grows with more new firms entering the market, the loyal
customer base of the incumbent firms deteriorates over time; and (3) systemic mod-
els are developed to explain why competitive advantages of the past seem to be
sustainable, while present competitive advantages transient.
What is established in this chapter indicates that for the minimum objective of
business survival, existing firms, no matter how established, have to without any
choice participate in market competition.
The rest of this chapter is organized as follows. Section 17.1 provides a literature
review. Sections 17.2 and 17.3, based on Forrest et al. (2017), respectively, introduce
all the conditions of the market to be considered for the rest of the presentation and
establish the relationship between coordinated monopoly of the market and stag-
nation in expected profits. Sections 17.4 and 17.5, based on Forrest and Tallapally
(2017), explain, respectively, why markets have been changing faster and customers
becoming less patient than ever before and why once sustainable competitive advan-
tages become transient. Section 17.6 concludes the presentation of this chapter along
with several open problems to be considered for future research.

© Springer Nature Switzerland AG 2018 257

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_17
258 17 Issues of Competition

17.1 Some Background Information

Guo et al. (2012) address the role of downstream market competition under sym-
biotic production and demonstrates that incorporating different types of competi-
tion in product markets will partially eliminate the inefficiency as caused by double
marginalization. The study suggests that introducing callback services or internet
telephones creates such an environment that is similar to downward market com-
petition in the output market so that international tariffs are significantly reduced.
Chang et al. (2015) consider the effect of market competition on the relationship that
exists between corporate governance and capital structure dynamics. Based upon an
examination of this relationship, these authors find that product market competition
increases the incentives for firms with weak governance structures to maximize the
wealth of shareholders, thereby increasing the adjustment speed toward target lever-
age. Furthermore, the difference in such adjustment speed between firms with weak
and strong governance structures is found to be smaller than expected.
Debruyne and Reibstein (2005) investigate when incumbents should enter into
new market niches created by technological innovation. They argue that market con-
ditions and company-specific characteristics are not sufficient to explain the entry
timing of the incumbents; instead that entry is a contagious process. Their results
demonstrate that incumbents are more likely to respond to innovations in their indus-
try when their counterparts do so.
Because a new product (attacker) entering into a competitive market is likely to
provoke responses from some or all of the existing products (defenders), Kumar
and Sudharshan (1988) investigate the development of optimal defensive strategies
based on an understanding of the possible reactions of all the defenders to an optimal
attack. Similar to Lane (1980), these authors also assume that each of N products
sequentially enters with perfect foresight on the subsequent entry, and then, based on
new technology, an unanticipated attacker enters. The N defenders respond in price
without positioning themselves according to Lane’s model, and once this equilib-
rium is obtained, advertising and distribution response functions scale sales. Then
it is shown that under these decoupled response function models of advertising and
distribution, uniformly-distributed tastes, and nonincreasing market size, the optimal
defense for all existing brands is to decrease their respective prices, advertising, and
distribution. Similar results are also obtained by Hauser and Shugan (1983), where
different consumer response models and a different equilibrium assumption are used.
Because organizations evolve through periods of both incremental and evolution-
ary changes punctuated by discontinuous or revolutionary changes, Tushman and
O’Reilly III (1996) consider it a challenge for managers to adapt the culture and
strategy of their organizations to its current environment, while not undermining
the companies’ ability to adjust to radical changes in that environment. Day (1994)
studies how market-driven orientation of businesses can be achieved and sustained,
finding that the emerging capabilities approach to strategic management, coupled
with total quality management, offers a rich array of ways to design change pro-
grams that will enhance a market orientation with increased capabilities in market
17.1 Some Background Information 259

sensing and customer linking. By contrasting two distinct business cases, Day and
Schoemaker (2016) assess the managerial implications of sensing, seizing, and trans-
forming critical capabilities for successful organizational adaptation.
Kaharuddin et al. (2017) investigate and measure the transient competitive advan-
tage readiness among hotels, cafes, and fashion retail industry in Bandung, Indonesia,
and find based on a total of 60 managers that most managers of hotels, cafes, and
fashion retails in Bandung do not have enough readiness and proper strategy to create
the transient competitive advantage. Contradicting Porter’s views, Bashir and Verma
(2017) highlight how business model innovation can serve as a competitive advan-
tage by reviewing and analyzing the literature on competitive advantage, because
imitating an entire novel system is quite difficult compared to imitating a product or
a service. Leavy (2016) considers what dynamic capabilities will be needed to com-
pete in a world of transient advantage, like modern China, with its complicated and
quickly changing demand pattern, hyper-competition, shifting industry boundaries,
and discontinuities in the regulatory context. By reviewing the literature on adapt-
ability, Koller (2016) introduces the concept of adaptive advantage and addresses the
problem of its implementation in an organization by looking at innovation culture,
decision-making style, and accumulated experience of a sample of “old” and innova-
tive firms. Purkayastha and Sharma (2016) inductively analyze the unique decisions
of three firms that develop a competitive advantage by shaping their business model.
The authors emphasize on the criticality of the business model as a higher-level
construct formed from multiple structural and strategic decisions that, eventually,
become a source of competitive advantage.
So, comparing what is developed in this chapter with the literature, it is readily
clear that this work enriches the relevant knowledge at the height of theoretical
abstraction with a much wider range of practical applicability.

17.2 The Market Conditions

Continuing the setup in Section 16.3, let us consider an oligopoly marketplace with
m incumbent firms, named 1, 2, …, m. These firms offer the market with mutually
substitutable products, although the products are different in terms of quality and
follow-up services. Assume that these firms have firmly controlled their respective
shares of the market with their individually solid bases of loyal consumers; conse-
quently, these firms have become risk neutral and plan to continuously reap in their
respective profits by securely defending their established turfs. As in a real-life mar-
ket, assume that there are customers in the marketplace who make their purchases
based on whose price is the lowest. So, these customers will be called switchers; they
are loyal to any of the incumbent firms. To attract the business of these switchers, the
incumbent firms adjust their selling prices. In other words, assume that the incumbent
firms produce their products that are horizontally differentiated at constant marginal
costs, which is set to zero without loss of generality.
260 17 Issues of Competition

Figure 17.1 Eddy motion model of the general system

Based on the assumptions above, it is natural to assume consequently that the

managements of the m incumbent firms are well aware of the pricing strategies of the
other firms and have established their best responses by playing the Nash equilibrium
through pure self-analyses. Because of all these assumptions, in this chapter, such a
market will be referred to as coordinately monopolized (by these existing firms).
First, let us look at the situation with m  2. Assume that the market share of the
loyal consumers of Firm k is αk , k  1, 2, such that these consumers only purchase
the product of Firm k provided that the price is no more than their reservation value,
which is set to 1. Let β  population size and β − α1 − α2 be the size of the market
segment of switcher who base their purchase decisions on which price is lower. By
dividing the market share partition by the population size, the parameters α1 , α2 , and
β are normalized. Without loss of generality, these same symbols will still be used
for the relevant population proportions. That is, the following holds true

α1 + α2 + β  1.

A natural question that arises at this junction is how the market shares α1 and α2
are related to each other, because these two firms are situated in a perfectly identical,
symmetrical conditions? To this end, let us look at the systemic intuition provided
by the yoyo model.
Firstly, let us think of the marketplace of our concern as a closed system so that
the marketplace has to have the fundamental characteristics of the general system
(Lin 1999, 1987). In particular, the dynamics of the marketplace can be seen as an
abstract yoyo field (Lin 2009, 2007) as shown in Figure 13.3, which is reproduced
here as Figure 17.1 to make this chapter self-contained.
Specifically, on the basis of the blown-up theory (Wu and Lin 2002), a general
theory of development, and the discussion on whether or not the world can be seen
from the viewpoint of systems (Lin 1988; Lin et al. 1990), the concepts of inputs,
outputs, and converging and diverging eddy motions are coined together in the model
shown in Figure 17.1 for each object and every system imaginable, be it physical or
organizational, where a marketplace is seen as an organization composed of people
connected by flows of goods, services, money, information, etc. In other words, each
business entity can be modeled by a multi-dimensional entity that spins about its
axis. If such a spinning entity is fathomed in our three-dimensional space, one has
17.2 The Market Conditions 261

Figure 17.2 Systemic

bird’s-eye view of our
marketplace

a structure as shown in Figure 17.1a. The side of inputs sucks in all things, such as
resources, information, investments, profits, etc. After funneling through the neck,
all things are spit out as outputs. Some of the outputs never return to the other side
and some will (Figure 17.1b). For the sake of convenience of communication, such
a structure as shown in Figure 17.1a is referred to as a (Chinese) yoyo due to its
general shape.
What this systemic model says is that each physical or intellectual entity in the
universe, be it a tangible or intangible object, a living being, an organization, a
culture, a civilization, etc., can all be seen as a kind of realization of a certain multi-
dimensional spinning yoyo with an eddy field around. It stays in a constant spinning
motion as depicted in Figure 17.1a. If it does stop its spinning, it will no longer exist
as an identifiable system. What Figure 17.1c shows is that due to the interactions
between the eddy field, which spins perpendicularly to the axis of spin, of the model,
and the meridian field, which rotates parallel to axis of spin, all the materials that
actually return as inputs travel along a spiral trajectory.
As expected, this yoyo model has successfully played the role of intuition and
playground for scholars who investigate the world and explore new knowledge holis-
tically (Lin 2009; Lin and Forrest 2011; Forrest 2013, 2014; Forrest and Tao 2014;
Ying and Forrest 2015). In particular, this yoyo model of general systems has been
successfully applied in the investigation of Newtonian physics of motion, the concept
of energy, economics, finance, history, foundations of mathematics, small probabil-
ity disastrous weather forecasting, civilization, business organizations, the mind,
among others. Along this same line of logic, let us now see how the question just
raised earlier can be successfully addressed.
Secondly, when the marketplace of our concern, as described above, is fathomed as
an abstract yoyo field, one can look at the multi-dimensional yoyo body at a distance
from above either the convergent side or the divergent side, while imagining that
everything here takes place in our three-dimensional space. That is, one is looking
at a pool of spinning fluid, where the word “fluid” is an abstraction of movement of
all kinds of media, such as goods, information, money, credit, etc., that appear and
exist in business activities. In other words, graphically one is looking at the market
of concern as the pool of spinning fluid shown in Figure 17.2.
Now, the well-known dishpan experiment, which was initially conducted success-
fully by Hide (1953) of Cambridge University, England, and then by Dave Fultz and
his colleagues of University of Chicago (1959) independently, shows that when the
262 17 Issues of Competition

Figure 17.3 Asymmetric

flow observed in Fultz’s
dishpan experiment

movement of the fluid within the rotational dish is under enough pressure created
by either the sufficient speed of rotation or sufficient difference in the temperature
between the center and the periphery of the dish, the pattern of uniform movement,
as shown in Figure 17.2, will develop into the chaos, as shown in Figure 17.3. The
number of local eddy leaves is determined either by the rotational speed or by the
temperature difference or both and increases with the speed and the temperature
difference.
Now, the perfect symmetry that exists in the setup of the dishpan experiment, the
same also holds true for our afore-described marketplace, does not really help to stop
the occurrence of the local eddy leaves in the spinning fluid. That implies that if there
are only two local pools that appear within the dishpan, it is impossible for the two
local eddy pools to be of different sizes. If this end is related to the question about
the market shares α1 and α2 of our firms, then this systemic intuition would suggest
that α1  α2 . As a matter of fact, symbolically, the following result can be proven.

Theorem 17.1. In the mixed strategy Nash equilibrium, the two firms’ market shares
in this coordinately monopolized market are the same. That is, α1  α2 .

Proof. Let Fk (P) be the price distribution of Firm k, k  1, 2. Then the profits for
Firm 1 from its loyal consumers are α1 P, and those from its share of the switchers
are [1 − F2 (P)]βP. So, Firm 1’s objective function is
+∞
maxF1 (P) E(1 )  {α1 P + [1 − F2 (P)]βP}dF1 (P) (17.1)
−∞

where 1 is the profits of Firm 1, E(1 ) the expected profits, and Firm 1 likes to
maximize its expected profits by selecting its appropriate price distribution.
Because Firm 1 likes to attract as many switchers as possible in order to potentially
increase its profits from the guaranteed level α1 from its loyal consumers by charging
them the reservation value 1, the following holds true

α1 P + βP ≥ α1 .

So, P ≥ α1 /(α1 + β). Because P  1 is the reservation value of the loyal con-
sumers, the objective function of Firm i in equation (17.1) becomes
17.2 The Market Conditions 263

1
maxF1 (P) E(1 )  {α1 × P + [1 − F2 (P)]β × P}dF1 (P) (17.2)
α1
α1 +β

The equilibrium indifference condition for Firm 1 is

α1 × P + [1 − F2 (P)]β × P  α1 × 1,

which leads to
α1 (1 − P)
F2 (P)  1 − , (17.3)
βP

satisfying
 
α1
F2  0 and F2 (1)  1. (17.4)
α1 + β

Similarly, one has

α2 (1 − P)
F1 (P)  1 − , (17.5)
βP

satisfying
 
α2
F1  0 and F1 (1)  1. (17.6)
α2 + β

So, Firm 1’s expected profits are

1
E(1 )  {α1 P + [1 − F2 (P)]βP}dF1 (P)
α1
α1 +β

1
 α1 dF1 (P)
α1
α1 +β

 α1 F1 (P)|1 α1
α1 +β

 α2 . (17.7)

Similarly, the expected profits of Firm 2 are α1 .

If α1  α2 , without loss of generality assume α2 > α1 , then from the assumptions
about the loyal consumers, one has α1  E(2 ) ≥ α2 > α1 , a contradiction. So,
that means α1  α2 . QED
264 17 Issues of Competition

When seeing the result in Theorem 17.1 intuitively from the marketing perspec-
tive, because everything in this study is set up symmetrically, this result makes perfect
sense. Otherwise, the firm with smaller market share would try to adjust its price in
an attempt to expand its market weight so that this firm is no longer risk neutral
and the market is no more coordinately monopolized. On the other hand, in practice
this result is apparently incorrect due to the fact that each firm is constrained by
its own particular conditions. In other words, in the real business world, even in a
nearly ideal market of coordinated monopoly, the market shares of loyal consumers
of the existing firms would generally be different from one firm to another. However,
even so, the existing firms could still behave as what is assumed earlier due to their
respective realization of their individual restrictions and limitations.

17.3 Coordinated Monopoly and Stagnation in Expected

Profits

This section looks at the general case when the oligopoly and coordinately monop-
olized marketplace has m incumbent firms, as described in the previous section.
Based on what is established in Theorem 17.1 and the systemic intuition by using
the dishpan experiment, assume that the size of the loyal consumer base of each
firm is a constant α such that these consumers only purchase the product of their
respective firms provided that the price is not more than their reservation value,
which is again set to 1. Let the segment of all switchers, who switch from one firm
to another totally depending on the price, in the marketplace be of the size β. Then
the following equation holds true

mα + β  1 (17.8)

where both α and β are normalized as described earlier.

Theorem 17.2. In the symmetric Nash equilibrium of this coordinately monopolized

market, each of these m firms’ expected profits stays constant and is equal to α
although each of them tries to attract as many switchers as possible, while each of
these firms exploits its loyal consumers by charging them an as high price as possible.
Before proving this result, let us first look at the systemic intuition on why this
result should hold true. To this end, let us imagine the entire market of our concern as
the entire spinning dish in Figure 17.3, while each of the m firms is one of the local
eddy leaves so that there are a total of m local eddies in the spinning dish. Hence,
no matter how hard each of the local eddy spins in order to pull more of the fluid
particles into its field, some of the fluid particles always tend to wonder around the
spin field without actually becoming part of any of the local eddies. In other words,
as soon as such a pattern of fluid movement as the one shown in Figure 17.3 is formed
and becomes stabilized, it will be difficult or impossible for the local eddies to grow
any larger.
17.3 Coordinated Monopoly and Stagnation in Expected Profits 265

Proof of Theorem 17.2. When these symmetric firms compete on price, there is no
pure strategy Nash equilibrium, for the case that no symmetric pure strategy equilib-
rium exists, please consult with Narasimhan (1988), Varian (1980). In fact, for any
pure strategy portfolio (x1 , x2 , . . . , xm ), if there is a unique index i ∈ {1, 2, . . . , m}
such that

xi < xj , for j ∈ {1, 2, . . . , m} and j  i, (17.9)

then Firm i has successfully attracted all the price switchers and can therefore slightly
raise its price xi to bring in additional profits as long as the new price still satisfies the
condition in equation (17.9), while all other firms maintain their respective strategy
unchanged.  So, the pure strategy portfolio(x1,x2 , . . . , xm ) is not a Nash equilibrium.
If I  i ∈ {1, 2, . . . , m}: xi  minm
j1 xj has cardinality |I | great than 1, then
Firms k, k ∈ I , have absorbed all the price switchers. Due to the fact that everything
in this scenario is set up symmetrically, it is ready to see that each of these firms
would have taken in β/|I | portion of the switcher segment. So, Firm k’s profits, k ∈ I ,
are αxk + (β/|I |)xk . So, one of these firms, say Firm k, can lower its price slightly to
xk , satisfying

xk α + β/|I |
> ,
xk α+β

to bring in additional profits by attracting all the switchers. That is, the pure strategy
portfolio (x1 , x2 , . . . , xm ) is not a Nash equilibrium. Similar argument can be used
to show that even when not all firms k ∈ I share the same portion of the price
switchers, then the firm with the fewest price switchers could slightly lower its price
to increasing its profits by attracting all the price switchers. That is, once again, the
following has been shown that the pure strategy portfolio (x1 , x2 , . . . , xm ) is not a
Nash equilibrium.
Lastly, if for any i ∈ {1, 2, . . . , m}, xi  1, then Firm j, for any chosen j ∈
{1, 2, . . . , m}, can slightly lower its price from the reservation value 1 to anywhere
in the interval (α/(α + β), 1) to produce more profits by taking in the entire segment
of the switchers. So, (1, 1, …, 1) is not a Nash equilibrium, either. That is, the game
of concern does not have any pure strategy Nash equilibrium.
However, this game does have a symmetric mixed strategy Nash equilibrium
(Zhou et al. 2015). To this end, let Fi (P) stand for the price distribution of Firm i,
i ∈ {1, 2, . . . , m}.
First assume that there are only two competing Firms i and j, that is, m  2. At
price P, Firm i’s profits
from its loyal consumers are αP, and its profits from the
switchers are 1 − Fj (P) βP. So, the objective function of Firm i is
266 17 Issues of Competition

+∞


maxFi (P) E(i )  αP + 1 − Fj (P) βP dFi (P)
−∞
1


 αP + 1 − Fj (P) βP dFi (P) (17.10)
0

where i  i (P) stands for Firm i’s profits at price P, E(i ) the expected profits
for all possible prices, and Firm i’s objective is to maximize its expected profits by
choosing its particular price distribution Fi (P). The reason why the upper and lower
limits of the integral are changed, respectively, from +∞ and −∞ to 1 and 0 is
because when P < 0 or when P > 1, the profits are zero.
Assume that there are 3 firms, named i, j, k, involved in the price competition.
Then, at price P, Firm i’s revenue from its loyal
consumers is αP. The portion of
switchers Firm j does not get is γ  1 − Fj (P) β, which is still

available for Firms
i and k to take. Now, [1 − Fk (P)]γ  [1 − Fk (P)] 1 − Fj (P) β is the portion of the
switchers whom are taken up by neither Firm j nor Firm k. So, they are the switchers
left for Firm i to take in. So, the objective function of Firm i is

1


maxFi (P) E(i )  αP + 1 − Fj (P) [1 − Fk (P)]βP dFi (P) (17.11)
0

So, from mathematical induction it follows that when there are m price-competing
firms, the portion of switchers Firm i takes in is

1 − Fj (P) β.
ji

Hence, the profits i Firm i generates when the firm sells its product at price P
are

αP + 1 − Fj (P) βP,
ji

and the objective function of Firm i is

⎧ ⎫
1 ⎨ 
m ⎬
maxFi (P) E(i )  αP + [1 − Fk (P)]βP dFi (P) (17.12)
⎩ ⎭
0 ji

Firm i can earn α by charging the reservation value 1, because all the firm’s loyal
consumers will purchase its product at that price. However, to potentially maximize
profits, each firm likes to adjust the price P in order to take in as many of the
17.3 Coordinated Monopoly and Stagnation in Expected Profits 267

switchers as possible. At the same time, each firm does not have any incentive to
price its product below the price of α/(α + β), because any price below α/(α + β)
will yield profits less than α despite of attracting all the switchers, where
α
αP + βP ≥ α → P ≥
α+β

The equilibrium indifference condition for Firm i is


m
α
α×P+ [1 − Fk (P)] β × P  α × 1, ≤ P ≤ 1, (17.13)
ji
α+β

for i, j  1, 2, …, m, and i  j. So, the symmetric equilibrium price distribution is

the following continuous function
  m−1
1
(1 − P)α α
F(P)  Fi (P)  Fj (P)  1 − , ≤ P ≤ 1, (17.14)
P(1 − mα) α+β

satisfying the boundary conditions:

 
α
F  0 and F(1)  1. (17.15)
α+β

In this unique mixed strategy Nash equilibrium, there is no mass point of prices
that the firms charge with positive probability. Hence, each firm’s expected profits is
⎧ ⎫
+∞⎨ m ⎬
E()  αP + [1 − Fk (P)]βP dF(P)
⎩ ⎭
−∞ ji

1
 αdF(P)
α
α+β

 αF(P)|1 α  α. (17.16)
α+β

That is, in the symmetric mixed strategy Nash equilibrium, each firm’s expected
profits does not change although each of the firms tries to attract as many switchers
as possible, while each of these firms exploits its loyal consumers by charging them
as high price as possible. QED
268 17 Issues of Competition

17.4 Markets Change Faster and Customers Less Patient

Because of the advent of the internet and the knowledge-based economy in general,
decreases in protective trade regulations, and advances in technology, consumers
and markets are changing faster than ever, and customers are less patient with busi-
nesses than ever before. Consequently, the advantages that made many commercial
companies iconic, such as the 163-year-old Alliance Boots (UK), have ended. So,
many companies now seek to be quicker, more decisive, and more candid so that any
negative news is immediately addressed. They spend more time thinking about the
future than ever before. To capture opportunities and act as an organic whole, many
companies seek to break down internal structural solos, while creating courageous
leaders who are fully engaged in capturing opportunities and moving away from
strategies and practices that no longer represent the future (McGrath 2013, p. ix). So,
the following natural questions arise:

Question 17.1: What makes markets change faster and customers less patient than
ever?

Question 17.2: What makes past sustainable competitive advantages presently tran-
sient?

The answers to these two questions are closely related. Let us address Ques-
tion 17.1 first. To this end, let us continue to look at the oligopoly market of m
incumbent firms as described in the previous sections. Then we have the following
result.

Theorem 17.3. In the Nash equilibrium, when the competition of the market grows
with more new firms entering the market, the loyal customer base of the incumbent
firms deteriorates over time.

Proof. This market does not have any pure strategy Nash equilibrium and nonsym-
metrical mixed strategy Nash equilibrium (Forrest et al. 2017). Let Fi (P) stand for
the price distribution of Firm i, ∈ {1, 2, . . . , m}, which compete with each to attract
switchers. The assumption that an increasing number of firms enter the market implies
that the consumer surplus satisfies β  1 − α > 0. Assume that there are n new
firms that enter the market by uniformly randomizing its price P over the interval
[0,1], where their cost basis is also assumed to be constant and set to zero. Then, the
profits of incumbent Firm i are given by

αP + βP(1 − P) 1 − Fj (P)
n

ji

and the objective function of incumbent Firm i is

⎧ ⎫
1 ⎨  ⎬
m

maxFi (P) E(i )  αP + βP(1 − P)n 1 − Fj (P) dFi (P)

⎩ ⎭
0 ji
17.4 Markets Change Faster and Customers Less Patient 269

The equilibrium indifference condition for incumbent Firm i is

α × P + β × P(1 − P) 1 − Fj (P)  α × 1.
n

ji

So, the symmetric equilibrium price strategy of each incumbent Firm i, i  1, 2,

…, m, is
  m−1
1
−1
− m−1 α −1
 P m−1 (1 − P)− m−1 h(P)
n−1 n−1
F(P)  Fi (P)  1 − P m−1 (1 − P)
β
  m−1
1

where h(P)  P m−1 (1 − P) m−1 − βα

1 n−1
.
For this strategy F(P) to be valid, we must have F(P)  0, for P ≤ P, F(P)  1,
for P ≥ P, and F(P) ≥ 0, for P ≤ P ≤ P, where P and P are some fixed price levels
such that 0 ≤ P < P ≤ 1.
For such price levels P and P to exist, h(P) must satisfy the following: h(P) ≥ 0,
for P ≤ P ≤ P, where P and P are some fixed price levels such that 0 ≤ P < P ≤ 1,
and that h(0) < 0 and h(1) < 0. Since
1 n−1

h (P)  P m−1 (1 − P) m−1 P −1 − (n − 1)(1 − P)−1

1

m−1

it can be shown that h(P) reaches its maximum at P  1/n. That is, in order  for
the previously mentioned price levels P and P to exist, h(P) must satisfy h 1n > 0,
which means
     
α 1 1 n−1 1 n 1 1
< 1−  1− 1− .
β n n n n n

So, when n → ∞, α/β → (1/e) × 1 × 0  0. This end means that the base of
loyal customers for each incumbent firm gradually diminishes when an increasing
number of new firms enter the market. QED
The result in Theorem 17.3 provides an answer to Question 17.1: What has made
markets change faster and customers less patient than ever? In particular, because of
the growing globalization of the world commerce, consumer markets, which were
once regional and tightly controlled by local firms, are becoming international with
increasing number of firms from different corners of the world competing for the
same customers. In terms of the systemic yoyo model, initially, the world consisted
exclusively of local markets, as indicated by the small eddy leaves in Figure 17.4a,
where the overall circular pool stands for the world market. Because the interna-
tional commerce did not exist or the scale was very small and could be ignored,
the overall circular dish in Figure 17.4a does not spin. That is, the businesses of the
world are not seen as a system. Next, with the advancement of modern transportation
and communication technology, the world becomes more connected through inter-
270 17 Issues of Competition

(a) Regional markets (b) Globalizing world markets

Figure 17.4 World full of regional markets

national transfers of goods, information, and knowledge; the originally disconnected

regional markets become connected as shown in Figure 17.4b. The counterclockwise
spin of the overall dish models the fact that goods, information, and knowledge are
now traveling throughout the world, while originally local firms, the local eddies that
also rotate around with the overall dish, are also serving the entire world market. Such
much-increased sources of suppliers of goods and services make customers spoiled
so that they become less patient than ever before with anything unsatisfactory.

17.5 Once Sustainable Competitive Advantages Become

Transient

This section answers Question 17.2 given in the previous section in two parts: (1)
Why do competitive advantages of the past seem sustainable? And (2) what makes
present competitive advantages feel like transient?
To address the first part of Question 17.2, let us start by looking at Bjerknes’ Cir-
culation Theorem established in the year of 1898 over one hundred years ago (Hess
1959); for details, see Section 15.1. This theorem shows that nonlinearity mathemat-
ically stands (mostly) for singularities, and in terms of physics, it represents eddy
motions. Such motions are a problem of structural evolutions, a natural consequence
of uneven evolutions of materials. In particular, at the end of the nineteenth century,
V. Bjerknes discovered the eddy effects due to changes in the density of the media
in the movements of the atmosphere and ocean. By a circulation, it is meant to be
a closed contour in a fluid. Mathematically, each circulation  is defined as the line
integral about the contour of the component of the velocity vector locally tangent
to the contour. In symbols, if V
stands for the speed of a moving fluid, S an arbi-
trary closed curve, δ
r the vector difference of two neighboring points of the curve S
(Figure 17.5), then a circulation  is defined as follows:

  V
δ
r . (17.17)
S
17.5 Once Sustainable Competitive Advantages Become Transient 271

Figure 17.5 Definition of a closed circulation

Figure 17.6 Solenoid circulations

Through some ingenious manipulations (Wu and Lin 2002), the following well-
known Bjerknes’ Circulation Theorem is obtained:
   
d V
1 dσ
 ∇ × (−∇p) · δσ − 2
, (17.18)
dt ρ dt
σ

where σ is the projection area on the equator plane of the area enclosed by the closed
curve S, p the atmospheric pressure, ρ the density of the atmosphere, and
the
earth’s rotational angular speed.
The left-hand side of equation (17.18) represents the acceleration of the moving
fluid, which according to Newton’s second law of motion is equivalent to the force
acting on the fluid. On the right-hand side, the first term is called a solenoid term in
meteorology. It is originated from the interaction of the p- and ρ-planes due to uneven
density ρ so that a twisting force is created. Consequently, materials’ movements
must be rotational with the rotating direction determined by the equal p- and ρ-plane
distributions (Figure 17.6). The second term in equation (17.18) comes from the
rotation of the earth.
This theorem reveals the commonly existing and practically significant eddy
effects of fluid motions and implies that uneven eddy motions are the most com-
mon form of movements observed in the universe. Because uneven densities create
twisting forces, fields of spinning currents are naturally created. Such fields do not
have uniformity in terms of types of currents. Clockwise and counterclockwise eddies
272 17 Issues of Competition

always coexist, leading to destructions of the initial smooth, if any, fields of currents.
What’s important is that the concept of uneven eddy evolutions reveals that forces
exist in the structures of evolving objects and do not exist independently outside of
the objects.
Now, let us look at the question we are addressing: Why do competitive advantages
of companies in the past seem sustainable?
At the early times, people lived in more primitive conditions when compared to
the present time. Due to the existing natural conditions and available resources within
the environment and societies, some people and households started to exchange their
surplus of goods with others for the purpose of improving the quality of their lives.
That led to the formation of very beginnings of business transactions. Because the
population density was low, tools available for production, transportation, and com-
munication were limited and inefficient, and minor obstacles of the environment in
today’s standard easily divided the flow of goods and services into small-scale shops.
Because other than some luxury goods and services, these individual and separated
shops mostly provided the basic necessities for human survival, they naturally deliv-
ered mostly identical set of goods and services with some minor differences. As time
went on, better tools for production and transportation and better practices of man-
agement were designed and employed in various individual shops. The natural desire
for better living conditions paved the way for inventions of new tools, discovery of
new methods of production, and introduction of more efficient ways of management
to pass around the land through word of mouth and improving technology of com-
munication. So, a circulation of information and people with special abilities and
resources started to form. As the circulation started to appear, Bjerknes’ Circulation
Theorem guarantees the appearance of abstract eddy motions over the land consisting
of migration of people, spread of knowledge and information, and transportation of
goods. That explains how local and then regional markets were initially formed and
why all markets have conditions and barriers against new entrants. That is, over time
incumbent firms have naturally established their respective competitive advantages
that have been part of the conditions and barriers that help the firms to exploit their
market shares while preventing competitions from new entrants. That explains why
competitive advantages of the past seem to be sustainable, and indeed, they were
for the most part sustainable. As a consequence, most of strategy frameworks and
tools in use were based on the single dominant notion: The purpose of designing and
adopting strategies is to achieve a sustainable competitive advantage.
At this junction, there is a natural need to justify the scientific validity for us to
employ the Bjerknes’ Circulation Theorem as in the previous paragraph, because in
theory this theorem holds true only for fluids. Firstly, when the systemic yoyo model
is initially introduced earlier, we have given a relevant explanation for how and why
each human organization is a spinning pool of fluid, consisting of flows of such fluids
as energy, information, materials, etc., that circulate within the inside of, go into, and
are given off from the organization. Secondly, the systemic yoyo model of systems
naturally leads to the realization that the universe is a huge ocean of eddies, which
changes and evolves constantly. That is, the totality of the physically existing world
can be studied as fluids. Thirdly, as described in the previous paragraph, people in
17.5 Once Sustainable Competitive Advantages Become Transient 273

the land helped to circulate information, knowledge, goods, etc., all of which are
studied using continuous or differentiable functions in social sciences in general
and economics in particular. When these aspects of a market are modeled by such
functions, they are generally seen in physics and mathematics as flows of fluids and
are widely known as flow functions. Specifically, in the formation of an economy,
it is these commonly shared aspects (or fluids) that make the land to have living
markets, where individual persons are simply local “impurities” of the fluids, and
each of the “impurities” carries some concentrated amount of “energy,” information,
knowledge, etc.
Secondly, let us see what makes present competitive advantages feel like tran-
sient. Recent frequently occurring breakthroughs and fast advances in communica-
tion technology, management methodologies, and manufacturing capabilities have
made important know-how information more widely available and helped remove
many insurmountable barriers of the past for the general public. In other words,
such detailed information as for how to start up a competitive company, how to run
a firm efficiently, and how to improve and market product beautifully have helped
companies, either incumbent or new coming, to compete with each other on an
unprecedented, equally levelled ground due to the fast advancement of communi-
cation technology. And, with the development of modern engineering technology,
manufacturing an imagined product, which combines all the best-selling functional-
ities of relevant products, has become doable and much less costly than before. So,
the secrets behind most of the so-called sustainable competitive advantages of the
past have become common knowledge, while as soon as a new advantage leads to
success, the underlying ideas become public information in no time. That is exactly
the reason why newer competitive advantages have to be introduced frequently and
evolved into fresh formats quickly.
In terms of the systemic yoyo model, the evolution of an economy, just as described
in the previous paragraphs, can be shown most closely and vividly by using the dish-
pan experiment. To this end, let us first look at the basics of this experiment. As
initially conducted successfully by Hide (1953) of Cambridge University, England,
and then by Dave Fultz and his colleagues of University of Chicago (1959) indepen-
dently, this experiment shows that when the movement of the fluid within a rotational
dish is under enough pressure created by either the sufficient speed of rotation or suf-
ficient difference in temperature between the center and the periphery of the dish, the
pattern of uniform movement, as shown in Figure 17.2, will develop into the chaos,
as shown in Figure 17.3. The number of local eddy leaves is determined either by
the rotational speed or by the temperature difference or both and increases with the
speed and the temperature difference.
Of course, for our situation in hands, the pattern in Figure 17.2 should be replaced
by that in Figure 17.4a, because different from the perfect symmetry in the setup
of the dishpan experiment, the initial distribution of local markets (and economies)
is not even or symmetric. Instead, it is dictated by natural conditions and availabil-
ity of resources. Additionally, the symmetrical distribution of local eddy leaves in
Figure 17.4b would not be so symmetrical for our model of the world economy, either.
Now, if we look at this model closely, we can see that as the pressure increases, caused
274 17 Issues of Competition

by either the spinning speed or the “temperature” difference between the center and
periphery of the dish, the number of eddy leaves will grow larger. So, competitive
advantages have to become temporary and transient in order for the incumbent eddy
leaves to maintain their quality of livelihood. That is, to win in volatile and fast-
moving environments, business leaders need to learn and master the skills of how
to discover and exploit short-lived opportunities with speed and decisiveness, while
realizing the fact that the deeply ingrained structures and systems that these leaders
used to reply on to extract maximum value from a competitive advantage are actu-
ally liabilities, which are outdated and even dangerous, in the current fast-moving
competitive environment.

17.6 Conclusions

Based on systems research in general and the systemic yoyo model in particular, this
chapter
(1) Investigates the dynamics of a coordinately monopolized market, where m estab-
lished and risk-neutral firms plan to continuously reap in their respective profits
by securely defending their bases of loyal consumers while competing over the
switchers by adjusting their prices, and
(2) Develops a general theory on what makes markets change faster and customers
less patient than ever before.
What is theoretically significant is that when such a coordinated monopoly as
described in this chapter is formed, the expected profits of each firm will become
stagnant, that as long as there are switchers, who make purchase decisions only
based on prices, in the marketplace, the existing firms, even with their coordinated
monopoly of the market, have to realistically try to procure as many of the switchers
as possible; otherwise new competitions would enter into the market (Theorem 16.1).
In terms of market entry timing, this work clears up a few indecisive conclusions
established in empirical studies of the literature. For example, we theoretically show
that (1) as long as the market contains sufficient consumer surplus, new competition
will naturally appear no matter what particular contingencies and antecedents are
prevalent, and (2) established business entities should not be early movers of any
new market until some first movers have demonstrated the existence and depth of
the market.
All the results established in this chapter are developed on the intuition of the
systemic yoyo model by looking at the market as a spinning yoyo field and by
observing the dynamics of the internal movement of the “fluid” in Figure 17.3.
However, the well-known dishpan experiment suggests that when either the rotational
speed of the dish or the temperature difference between the periphery and the center
of the dish increases, the number of the local eddy pools will grow and the asymmetric
pattern of flow will eventually return to that shown in Figure 17.2 with time. So, what
17.6 Conclusions 275

could be potentially more thought provoking than what has been established in this
chapter is to investigate the following open questions and related issues:
Question 17.3: Symbolically prove when there are m firms in the coordinately
monopolized market, the size α of the set of all loyal consumers of each firm is
the same, as what is shown in Theorem 17.1 for the special case when m  2. This
result is very important because it underlies the foundation for all the other results
in this chapter and related results in the literature to hold true.
Question 17.4: How would the existing firms adjust themselves if the established
market starts to shrink and eventually disappear, as the flow pattern in Figure 17.3
starts to transform back into the one shown in Figure 17.2 with time?
In short, all the results, presented in this chapter, attempt to capture momen-
tary “photographic” shots of the dynamics of an evolutionary process, where a new
market gradually emerges, grows larger and becomes mature with time, and then
dwindles and eventually disappears. In comparison, the most significant results will
be those that not only portrait momentary shots but also reveal the laws underlying
the evolution of the development process: germination–growth–maturity–disappear-
ance. Results of this nature would fit the description of post-modern science given
in Lin (2009).

Acknowledgements Drs. John Buttermore, Theresa Wajda, and Pavani Tallapally of School of
Business, Slippery Rock University, Slippery Rock, PA 16057, U.S.A., also participated in the
research presented in this chapter.

References

Bashir, M., Verma, R.: Why business model innovation is the new competitive advantage. IUP J.
Bus. Strategy 14(1), 7–17 (2017). (Hyderabad)
Chang, Y.K., Chen, Y.L., Chou, R.K., Huang, T.H.: Corporate governance, product market compe-
tition and dynamic capital structure. Int. Rev. Econ. Finance 38, 44–55 (2015)
Day, G.S.: The capabilities of market-driven organizations. J. Mark. 58(4), 37–52 (1994)
Day, G.S., Schoemaker, P.J.H.: Adapting to fast-changing markets and technologies. Calif. Manag.
Rev. 58(4), 59–77 (2016)
Debruyne, M., Reibstein, D.J.: Competitor see, competitor do: incumbent entry in new market
niches. Mark. Sci. 24(1), 55–66 (2005)
Forrest, J.Y.L.: A Systemic Perspective on Cognition and Mathematics. CRC Press, an imprint of
Taylor and Francis, Balkema, The Netherlands (2013)
Forrest, J.Y.L.: A Systems Perspective on Financial Systems. CRC Press, an imprint of Taylor and
Francis, Balkema, The Netherlands (2014)
Forrest, J.Y.L., Tallapally, P.: Fast changing markets, transient competitive advantages and key
organizational element for firms. In: Proceedings of 2017 NABET Annual Conference, October
26–27, 2017, State College, PA, pp. 80–93 (2017)
Forrest, J.Y.L, Tao, L.X.: Investment and Employment Opportunities in China. CRC Press, an
imprint of Taylor and Francis, New York (2014)
Forrest, J.Y.L., Buttermore, J., Wajda, T.: At Nash equilibrium when new market competitions
appear? Kybern. Int. J. Cybern. Syst. Manage. Sci. 46(2), 256–271 (2017)
276 17 Issues of Competition

Fultz, D., Long, R.R., Owens, G.V., Bohan, W., Kaylor, R., Weil, J.: Studies of Thermal Con-
vection in a Rotating Cylinder with Some Implications for Large-Scale Atmospheric Motion.
Meteorological Monographs (American Meteorological Society), vol. 21, no. 4 (1959)
Guo, W.C., Lai, F.C., Liu, C.J., Mai, C.C.: Symbiotic production and downstream market compe-
tition. Atlantic Econ. J. 40(3), 329–340 (2012)
Hauser, J.R., Shugan, S.M.: Defensive marketing strategies. Mark. Sci. 2(4), 319–360 (1983)
Hess, S.L.: Introduction to Theoretical Meteorology. Holt, Rinehart and Winston, New York (1959)
Hide, R.: Some experiments on thermal convection in a rotating liquid. Quart. J. R. Meteorol. Soc.
79, 161 (1953)
Kaharuddin, A., Handaru, A.W., Sardan, W., Ahmad Mohammed, H.A.A.: Transient competitive
advantage readiness: findings from hotels, cafés, and fashion retails in Bandung, Indonesia. Int.
J. Bus. Glob. 18(3) (2017). https://doi.org/10.1504/ijbg.2017.083240
Koller, M.R.T.: Exploring adaptability in organizations: where adaptive advantage comes from and
what it is based upon. J. Organ. Change Manage. 29(6), 837–854 (2016)
Kumar, K.R., Sudharshan, D.: Defensive marketing strategies: an equilibrium analysis based on
decoupled response function models. Manage. Sci. 34(7), 805–815 (1988)
Lane, W.J.: Product differentiation in a market with endogenous sequential entry. Bell J. Econ.
11(1), 237–260 (1980)
Leavy, B.: The next wave of global disruption and the role of China’s entrepreneurs. Strategy
Leadership 44(3), 27–37 (2016)
Lin, Y.: A model of general systems. Math. Modell. Int. J. 9(2), 95–104 (1987)
Lin, Y.: Can the world be studied in the viewpoint of systems? Math. Comput. Model. 11, 738–742
(1988)
Lin, Y.: General Systems Theory: A Mathematical Approach. Plenum and Kluwer Academic Pub-
lishers, New York (1999)
Lin, Y.: Systemic Yoyos: Some Impacts of the Second Dimension. Auerbach Publication, an imprint
of Taylor and Francis, New York (2009)
Lin, Y., Forrest, B.: Systemic Structure behind Human Organizations: From Civilizations to Indi-
viduals. Springer, New York (2011)
Lin, Y.: Systemic yoyo model and applications in Newton’s, Kepler’s laws, etc.”. Kybern. Int. J.
Cybern. Syst. Manage. Sci. 36(3–4), 484–516 (2007)
Lin, Y., Ma, Y., Port, R.: Several epistemological problems related to the concept of systems. Math.
Comput. Model. 14, 52–57 (1990)
McGrath, R.G.: The End of Competitive Advantage: How to Keep Your Strategy Moving As Fast
as Your Business. Harvard Business Review Press, Boston (2013)
Narasimhan, C.: Competitive promotional strategies. J. Bus. 61(4), 427–429 (1988)
Purkayastha, A., Sharma, S.: Gaining competitive advantage through the right business model:
analysis based on case studies. J. Strategy Manage. 9(2), 138–155 (2016)
Tushman, M.L., O’Reilly III, C.A.: The ambidextrous organizations: managing evolutionary and
revolutionary change. Calif. Manag. Rev. 38(4), 8–30 (1996)
Varian, H.R.: A model of sales. Am. Econ. Rev. 70(4), 651–659 (1980)
Wu, Y., Lin, Y.: Beyond Nonstructural Quantitative Analysis: Blown-ups, Spinning Currents and
Modern Science. World Scientific, River Edge, NJ (2002)
Ying, Y.R., Forrest, Y.L.J.: Capital Account Liberation: Methods and Applications. CRC Press, an
imprint of Taylor and Francis, New York (2015)
Zhou, B., Mela, C.F., Amaldoss, W.: Do firms endowed with greater strategic capability earn higher
profits? J. Mark. Res. 52, 325–336 (2015)
Chapter 18
Some Issues of the Family

This chapter is devoted to the discovery of necessary and sufficient conditions under
which the well-known rotten kid theorem, established in 1974 by Gary Becker,
a Nobel Laureate in economics, holds true generally by employing the systemic
yoyo model. Beyond that, this chapter also uses the yoyo model as the background
intuition to establish the following two results: (1) the Theorem of Never-Perfect
Value Systems, which shows how any developed value system would be flawed in
practice, and (2) such a theorem that if a tender loving parent exists, then selfish kids
would take advantage of his tenderness and love. At the end, some open, unsettled
problems are posted for the future research.
This chapter’s main contribution to the relevant literature is that it shows how
the systemic yoyo model can be employed to lay down the foundation for many
empirical findings, as a starting point and direction for one to construct traditional
models. Beyond that, this chapter shows how insightful understandings and rigorous
proofs can be mapped out analytically and symbolically on the basis of the figurative
analysis.
This chapter is organized as follows. Section 18.1 addresses the problem of how
to make selfish people behave unselfishly. Section 18.2 investigates two mysterious
phenomena of the family. Section 18.3 explains why any established value system
is imperfect. Then the chapter is concluded in Section 18.4.

18.1 How Could Selfish People Behave Unselfishly

The following is a very important result in household economics, established by

Becker (1974), a Nobel Laureate of 1992 in economics. It has found practical appli-
cation in a wide range of problems and issues.

Theorem 18.1 (Rotten Kid Theorem, the 1974 version) If a family has a head who
cares about all other members so much that he transfers his resources to them auto-

© Springer Nature Switzerland AG 2018 277

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_18
278 18 Some Issues of the Family

Figure 18.1 Time moments

of transfers

matically, then any redistribution of the head’s income among members of the house-
hold would not affect the consumption of any member, as long as the head continues
to contribute to all. Additionally, other members are also motivated to maximize the
family income and consumption, even if their welfare depends on their own con-
sumption alone.

A restricted version of this result is later given in (Becker 1991). Due to its
importance while it is not generally true, Bergstrom (1989) established his version
of this result, which was generalized by Dijkstra (2007) to a different level. Not only
so, Dijkstra also establishes a condition under which in the Samaritan’s dilemma,
the altruist reaches her first best when she moves after the parasite. Later, Lin and
Forrest (2008b) established an if and only if condition under which Theorem 18.1
holds true.
In practical situations, if the head transfers his resources to other family mem-
bers periodically at fixed time intervals, the implied evolution of the family can be
depicted in Figure 18.1, where scale marks on the timeline represent the moments
of individual asset distributions. Mark 0 stands for the beginning of our discussion
or focus; negative marks the moments of distribution of the past; and the positive
marks the future moments.
Assume that at each moment of distribution, the head transferred $Rj to selfish
member j. If another member k suffers from a financial setback in time period (1, 0),
and because of his consideration of others the head bails out k by providing k extra
monetary support at t  0 and maybe a few following pay periods, then Theorem 18.1
indicates that the head can adjust his amount of distribution to everyone in order to
finance his increased support for k; such adjustment will not affect the consumption
of any member, as long as the head continues to contribute to all. Intuitively, it is
because of such potential adjustments in transfers that all selfish members feel the
existence of the family and are willing to help maximize the family income and
consumption, since each of them might one day need such an increased support.
In contrast, Bergstrom (1989) focuses on time period (k, k + 1), and Dijkstra
(2007) looks at time periods (k, k + 1) and [k, k + 1], respectively and individually.
Specifically, Bergstrom’s result focuses on the case that the head’s transfer is made
after whatever the selfish kids do beforehand, while Dijkstra establishes conditions
under which the head reaches his first best when he makes transfers, respectively,
either before or after however the selfish kids behave. However, if we look at the
payment cycles of a work environment, where employees get paid periodically at
fixed time intervals as depicted in Figure 18.1, then Becker’s original version of the
rotten kid theorem applies more appropriately. In other words, because of the potential
future transfers, both the benevolent head and the selfish kids will have to adjust their
expectations and behaviors from one time period to the next in order to reach their
individual Pareto optimums over time. That is, in real life there are fundamental
18.1 How Could Selfish People Behave Unselfishly 279

differences between one-shot deals and repeated businesses. It is because of this

reason that there is a need to look at Becker’s 1974 result and see under what general
conditions it holds true.
Before we develop several important observations, let us first look at Bergstrom’s
(1989) lazy rotten kids with different sets of details added and slightly different
outcomes produced.

Example 18.1 (Bergstrom 1989, The Lazy Rotten Kids) Let Yj be an index, 0 ≤
Yj ≤ 1, j  1, . . . , n, designed to measure how hard kid j works to generate his own
income Ij . Symbolically, we assume that Ij is a function of Yj :
 
Ij  Ij Yj .

Assume that kid j’s utility Uj is a function

 of his consumption Xj of goods and
leisure, the latter of which is measured by 1 − Yj . Suppose that the benevolent
head h of the family decides on his distribution of his own resources to the kids by
maximizing his utility function

Uh  Uh (Xh , U1 , . . . , Un ),

where Xh represents the amount of consumption of h subject to the budget constraint


n 
n
Xj  Ij .
j1 j1

If h is unable to observe the

 levels of individual efforts Yj the kids put in to produce
their own incomes Ij  Ij Yj , then a lazy and selfish child will not have enough
incentive to work, because his extra effort will not provide him with additional pay.
Hence, h runs into the problem of incentives with the lazy child. In other words, if
the kid’s work is to help h make additional income, he would not automatically work
extra to maximize the family income.
Assume that benevolent h has a system to measure each kid’s income Ij and his
level of effort Yj placed in earning his income and that h’s utility depends on the
utilities and consumptions of the kids. If benevolent h can give only gifts of money
to the kids but not leisure, then, in general, Becker’s rotten kid theorem does not hold
true. Specifically, assume that the family consists of the head h and two lazy kids,
named 1 and 2. The utility function of kid j is
 
Uj  Xj 1 − Yj , j  1, 2,

and the utility function of the head h is

   
Uh  Xh + U1 + U2  Xh + X1 (1 − Y1 ) + X2 (1 − Y2 ), (18.1)
280 18 Some Issues of the Family

where Xk stands for the consumption of person k, k  1, 2, h. The scheme of resource

transfer from h to each kid is given by

hj  wYj , j  1, 2,

where, to be fair, w is a fixed positive constant and h is able to observe Yj , j  1, 2.

Now, the head h’s scheme of transfer of his resources to the kids is carried out
so that their consumptions will maximize Equation (18.1) subject to the following
constraints:

Xh + X1 + X2  Xh + I1 + I2 + h1 + h2  Xh + I1 (Y1 ) + I2 (Y2 ) + wY1 + wY2 (18.2)

and

Xj ≥ Ij , j  1, 2. (18.3)

Following the assumption of Theorem 18.1, assume that the head h’s income
Ih  Xh + h1 + h2 is sufficiently large so that he always chooses to make positive
monetary gifts to each kid. So, the constraints in equation (18.3) are not binding and
can be ignored. And, because the kids are lazy, assume that their incomes produced
by their own efforts are zero, that is, I1  0  I2 .
The first-order conditions for maximizing equation (18.1) subject to equa-
tion (18.2) are
⎡ ⎤ ⎡ ⎤
∂Uh 1 −1/2
⎢ ∂X1 ⎥ ⎢ 2 1 U (1 − Y )
1 ⎥ ⎡ ⎤
⎢ ∂Uh ⎥ ⎢ 1 −1/2 ⎥ −1
⎢ ∂X ⎥ ⎢ U ⎥
⎢ 2 ⎥ ⎢ 2 2 (1 − Y2 ) ⎥ ⎢ −1 ⎥
⎢ ∂U ⎥  ⎢ ⎥  λ⎢ ⎥
⎣ w ⎦,
⎢ h ⎥ ⎢ − 1 U −1/2 X ⎥
⎢ ∂Y1 ⎥ ⎢ 2 1 1 ⎥
⎣ ⎦ ⎣ ⎦ w
∂Uh 1 −1/2
∂Y2 − U
2 2
X2

where λ is the Lagrange multiplier. By dividing the third and the fourth row entry
equations by the first and the second, respectively, we obtain

w  X1 /(1 − Y1 ), (18.4)

and

w  X2 /(1 − Y2 ), (18.5)

Dividing equations (18.4) and (18.5) provides

X1 /X2  (1 − Y1 )/(1 − Y2 ). (18.6)

18.1 How Could Selfish People Behave Unselfishly 281

Solving equation (18.6) for X1 and X2 individually and inserting them into equa-
tion (18.2) leads to
  
hi  C + (1 − Yi )/ 1 − Yj + 1 Xj − hj , i, j  1, 2, i  j, (18.7)

where C  Xh − Ih − I1 − I2 is a constant.
What’s surprising in equation (18.7) is that each kid’s share hi of the benevolent
h’s income is a decreasing function of the effort Yi he puts in his work. That is, the
incentive problem in the present case is worse than that when h does not have a way
to observe the kids’ efforts. Once again, if the kids’ work is to help h make additional
income, the lazy kids might not automatically put in additional effort. QED
This example vividly shows that Theorem 18.1 does not automatically solve
household incentive problems by having a benevolent head. As a matter of fact,
what it shows is that by introducing a value system into the decision-making process
on how to distribute his own income to the kids, the head actually redirects all poten-
tial anger and dissatisfaction of the kids toward himself. In terms of maintaining
peace and harmony of the family, that is not good.
In Theorem 18.1, the only thing the benevolent head does is his voluntary and
forceful transfer of his resources to the selfish kids. In real life, it is very possible
that some of the kids just do not care about receiving the gifts, and some kids might
even refuse to accept anything from the head. In these cases, we say that the head’s
distribution is in conflict with the consumption preferences of the receiving members.
To this end, Bergstrom (1989) proves the rotten kid theorem of Becker’s 1991 version
with a short argument, which is correct only under the assumption, which is not
explicitly mentioned, that each selfish kid’s utility is an increasing function of how
much he receives from the head. The presentation in this chapter shows that this is
exactly where the situation could go wrong.
Based on Example 18.1, if Becker’s rotten kid theorem is seen in light of the
systemic yoyo model, then the correctness of this result must be conditional, not as
simply stated as in the theorem. In particular, the family can be seen in two different
fashions:
(1) The entire family, as a system, is a spinning yoyo with the family consumption
being the outputs and the family income the inputs.
(2) The family consists of the head and a collection of individual kids, each of
whom can be modeled as spinning yoyos that interact with each other, where
each individual’s consumption is the outputs and his income the inputs.
For case (1), if we stand above the yoyo from a distance and look at the output
side, then the flows of the outputs can be simulated by the fluid patterns in the
dishpan experiment. (For the convenience of our reader, we highlight the relevant
points of this experiment here.) In particular, Raymond Hide (1953) of Cambridge
University, England, and Dave Fultz and his colleagues of University of Chicago
(1959) conducted this experiment with different setups. For Fultz’s version of this
experiment, he partially filled a cylindrical vessel with water, placed it on a rotating
turntable, and subjected to heating near the periphery and cooling near the center. The
282 18 Some Issues of the Family

(a) Uniform pattern of flow (b) Chaotic flows with local eddies

Figure 18.2 Fultz’s dishpan experiment

bottom of the container simulates one hemisphere of the earth’s surface, the water the
air above this hemisphere, the rotation of the turntable simulates the earth’s rotation,
and the heating and cooling simulate the excess external heating of the atmosphere
in low latitudes and the excess cooling in high latitudes.
To observe the pattern of flows at the upper surface of the water, which was
intended to simulate atmospheric motion at high elevations, Fultz sprinkled some
aluminum powder. A camera that rotated with the turntable took time exposures
so that a moving aluminum particle would appear as a streak and sometimes each
exposure ended with a flash, which could add an arrowhead to the forwarded end
of each streak. The turntable generally rotated counterclockwise, as does the earth
when viewed from above the North Pole.
Although everything in the experiment was arranged with perfect symmetry about
the axis of rotation, the bottom of the container was flat, and Fultz and his colleagues
observed more than they initially expected. First, both expected flow patterns, as
shown in Figure 18.2, appeared, where the choice between the pattern in Figure 18.2a
and that in Figure 18.2b depended on the speed of the turntable’s rotation and the
intensity of the heating and cooling. Briefly, with fixed heating, a transition from the
circular symmetry in Figure 18.2a to that in Figure 18.2b took place as the rotation
increased past a critical value. With the sufficiently rapid, but fixed rate of rotation, a
similar transition occurred when the heating reached a critical strength, while another
transition back to the symmetry would occur when the heating reached a still higher
critical strength.
What is important about this experiment is that structures, such as jet streams,
traveling vortices, and fronts, appear to be basic features in rotating heated fluids and
are not peculiar to atmospheres only. With this understanding in place, it is safe for
us to conclude that for case (1) where the entire family is seen as a spinning yoyo,
sooner or later the spin field of the family yoyo will have local eddies, which in some
locations may not necessarily spin in the same direction of the overall family yoyo.
That is, case (1) turns into the situation described in case (2).
For case (2), to make the situation easier to analyze, assume that the family consists
of only two members, the benevolent head H and a selfish kid K. In this case, the
interactions between the yoyos can be depicted in Figure 18.3, where the benevolence
18.1 How Could Selfish People Behave Unselfishly 283

Figure 18.3 Gift giving

from H to K

(a) The pools of H and K spin harmonically

(b) The pools of H and K are inharmonic

of H is represented as a divergent whirlpool and the selfishness of K as a convergent

whirlpool. If both m1 and m2 stand for H’s voluntary money transfers, the spin field
of K in Figure 18.3 will accept the gift m1 happily and the transfer m2 unwillingly
(Figure 18.3b) or even reject such a monetary gift (Figure 18.3a). Here, m1 is given to
K without violating his preference of consumption, while m2 is forced on K against
his preferences. That is, we have seen the following result.
Theorem 18.2 (Lin and Forrest 2008b). Becker’s rotten kid theorem holds true, if
and only if the distribution of the benevolent head’s resources is not in conflict with
the consumption preference of any selfish member.
Proof. (⇒) Assume that Theorem 18.1 holds true. We need to show that the
distribution of the benevolent h’s resources is not in conflict with the consumption
preference of any selfish member. By contradiction, assume that Theorem 18.1 holds
true and that the way h distributes his resources to the kids is in conflict with the
consumption preference of member k.
Let Y be a variable representing one aspect of member k’s total consumption
preference so that k’s utility is an increasing function of Y and his share hk of h’s
distribution is a decreasing function of Y . To simplify the argument, assume that the
family consists of only two members: the benevolent head h and the selfish member
k. So, their utility functions can be written, respectively, as follows:

Uk  Uk (Xk , Y ), Uh  Uh (Xh , Uk ), (18.8)

where Xi is the total consumption of member i on various commodities, i  k, h,

and Y the special variable as given above satisfying

∂Uk /∂Y > 0. (18.9)

So, jointly equations (18.8) and (18.9) mean that the distribution hk  hk (Y ) of
h’s resources to k satisfies ∂hk /dY < 0. (The differentiability of hk is assumed here
284 18 Some Issues of the Family

to take advantage of the methods of calculus.) Without causing confusion, assume

that no other variables enter into the function hk .
To h, his distribution of resources maximizes his utility Uh in equation (18.8)
subject to the following budgetary constraint:

Xh + Xk  Ih + Ik  Xh + (Ik + hk ), (18.10)

where Ii is the personal income of i, i  k, h. So, the first-order conditions for this
maximization problem are
⎡ ⎤ ⎡ ⎤
∂Uh ∂Uh ⎡ ⎤
∂Xh ∂Xh 1
⎢ ⎥ ⎢ ⎥
⎢ ∂Uh ⎥ ⎢ ∂Uh ∂Uk ⎥ ⎢ ⎥
⎢ ∂Xk ⎥  ⎢ ∂Uk · ∂Xk ⎥  λ⎣ 1 ⎦. (18.11)
⎣ ⎦ ⎣ ⎦
∂Uh ∂Uh ∂Uk − dhk

∂Y ∂Uk
· ∂Y
dY

Dividing the second row by the third row in equation (18.11) leads to

∂Xk /∂Y  −dhk /dY . (18.12)

That is, equation (18.12) and the assumption of hk imply that ∂Xk /∂Y > 0. And,
from equation (18.10), it follows that Xk  Ik + hk and that ∂Xk /∂Y  dhk /dY < 0,
a contradiction. This contradiction implies that the assumption that the way h dis-
tributes his resources to other members is in conflict with the consumption prefer-
ences of member k is incorrect.
(⇐) Assume that the distribution of the benevolent h’s resources is not in conflict
with the consumption preference of any selfish member. We need to prove that
Theorem 18.1 holds true. Once again, we do this by contradiction by assuming
that Becker’s rotten kid theorem 18.1 does not hold true. It means that (1) there is a
redistribution of the head’s income among members of the household that does affect
the consumption of at least one member, say member , even though h continues to
contribute to all, and (2) there is at least one selfish member k who is not motivated
to maximize the head’s (the family’s) income.
Situation (1) implies that for that one-time redistribution of h’s income, the way
of distribution does not agree with how member  prefers to consume his income and
monetary gifts. That means that the way h distributes his income at least this time
is in conflict with the consumption preference of . This end contradicts the initial
assumption that no such conflict exists. That is, situation (1) would not occur.
Now for situation (2), let Y be a variable concerning one aspect of the consump-
tion preference of member k, representing the reason why k does not care about
maximizing h’s income. If k’s share of h’s distribution is hk , then it must mean that
∂hk /∂Y |Y Y0 < 0, where Y  Y0 stands for the special Y -value of the moment of
our concern. That is, locally at around Y  Y0 , hk is a decreasing function of Y . In
fact, if hk is not a decreasing function of Y at any local region of Y  Y0 , then hk
must satisfy either ∂hk /∂Y |Y Y0  0 or ∂hk /∂Y |Y Y0 > 0. If hk satisfies the second
18.1 How Could Selfish People Behave Unselfishly 285

condition, then it must be a local increasing function of Y at Y  Y0 . In this case,

the greater the Y -value (at least locally), the greater share k receives from h. Since
k’s utility is an increasing function of his consumption, he would naturally like to
help maximize the head’s income. This end contradicts the assumption of Y .
If the first condition above, ∂hk /∂Y |Y Y0  0, holds true, from the fact that k
is not motivated to maximize h’s income, it follows that hk must be in one of the
following three possibilities:
(1) It is a local constant function of Y .
(2) It reaches a local minimum at Y  Y0 .
(3) It reaches a local maximum at Y  Y0 .
Firstly, the function hk cannot be a local constant or have reached a local minimum
at Y  Y0 ; otherwise, this end would not agree with the fact that k’s utility is an
increasing function of his consumption, since if h’s income drops or stays the same,
k’s share hk will have to decrease or stay constant. So, hk cannot be a local constant
function of Y or have reached a local minimum at Y  Y0 .
Secondly, hk cannot have reached a local maximum at Y  Y0 , either; otherwise,
it would mean that k has been helping out h to maximize his income until Y  Y0 .
This end contradicts the definition of the variable Y in the first place.
That is, the distribution hk of the benevolent head’s income to member k is in
conflict with the consumption preference of the selfish k. This end contradicts the
initial assumption that no such conflict exists. Therefore, the assumption that Becker’s
rotten kid theorem 18.1 does not hold true is incorrect. QED

18.2 Two Mysterious Phenomena of the Family

First of all, it can be readily seen that when two selfish members of the family have
conflicting consumption preferences, the benevolent head might still get along with
both of them in terms of consumption preferences. Specifically, assume that the
utility functions of the selfish members i and j are given by

Ui  Xi − Y and Uj  Xj + Y . (18.13)

Then, the benevolent head’s utility could be chosen to be

   
Uh  Xh Ui + Uj  Xh Xi + Xj . (18.14)

In equations (18.13) and (18.14), Xk is the consumption of k, k  i, j, h, and Y is

a variable of consumption member i dislikes while member j likes. Equation (18.14)
indicates that in the face of conflict in terms of consumption preferences of i and j,
the head h is completely neutral.
Also, when the benevolent family head h and a selfish member i have conflict in
their consumption preferences, it does not necessarily mean that in the overall utility
286 18 Some Issues of the Family

function of h, he must have a direct conflict with i in terms of their consumption

preferences. For example, assume that Y stands for a specific consumption of goods
which h dislikes while i likes. Let the utility functions of h and i from consumptions
be given by

Ui  Xi Y and Uh  Xh /Y ,

where Xk stands for the total consumption of k besides Y , k  i, h. If the overall

utility function of h is given by

U  U (Uh , Ui )  Ui · Uh  Xi Xh ,

then it implies that although individually h and i have conflicting consumption pref-
erences, in his overall utility, the benevolent head h cares about member i as much
as if they did not have any conflict in their consumption preferences.
What is discussed above implies that no matter whether or not there exists a
conflict in terms of individual consumption preferences between family members,
the benevolent head could care about all members just the same as if there is no
conflict at all. That is, in general, the distribution of the head’s resources to all
members, as seen in Theorem 18.2, is not affected by his personal consumption
preferences.
Next, let us look at the relationship between the altruism of parents and the “merit
goods” of children, assuming that there is no direct bargaining between the parents
and children. By “merit goods” of children, it means those particular traits or behavior
of children that parents care about, such as whether they are lazy, studying hard at
school, visiting often, drinking excessively, marrying well, or being mean to siblings
(Becker 1991, p. 10).
As shown in Example 18.1, children’s laziness can cause a breakdown of Becker’s
rotten kid theorem as soon as a variable about children’s laziness is employed in the
distribution of parents’ income to the children.
Regarding the desire for children to study hard at school, the situation of estab-
lishing an index to make the relevant measurement appears. Differences in making
measurement can lead to significant differences in children’s academic achievements.
And no matter whether the developed index is measurable or not, Theorem 18.2 indi-
cates that Becker’s rotten kid Theorem 18.1 is no longer true as soon as this index is
used in the determination on how much money each child receives from his parents.
For example, some parents are happy with their children’s schoolwork as long as
their grades are B’s and A’s, especially when these kids are first-generation college
bound. However, if the parents’ educational status is relatively high, the index they
define to measure how well their children study at school tends to be content-rich.
This end could be used to explain why parental education and occupational status
have a positive and significant effect of their children’s educational achievements
(Hill and Duncan 1987; Norberg-Schonfeldt 2004).
When a child has grown and is no longer under his parents’ roof, the parents will
naturally like to see the adult child coming home to visit often. If the parents are
18.2 Two Mysterious Phenomena of the Family 287

financially able, they will likely employ monetary means to have the child visit them
as often as possible (Bernheim et al. 1986). However, due to various reasons, obliga-
tions, and commitments, the adult child might be too busy to visit his parents as often
as they desire. Now, if the parents employ financial means to have their child visit
them often, they will face the possible situation as described in Theorem 18.2, where
the distribution of the parents’ money is in conflict with the child’s consumption
preference. For example, the adult child might be
• Working for a promotion he has craved about for a long time;
• Dating the woman of his dreams; or
• Occupied with a second job that brings him considerably more than what his
parents could afford to pay him for his visits.
In real life, there are millions of possibilities that prevent the child from frequently
paying visits to his parents. On the other hand, for the case when the parents are
extremely wealthy and the child has no desire to achieve high in life for himself, then
Theorem 18.2 implies that Becker’s rotten kid theorem works, and this adult child
will pay his parents visits as often as they desire. And very possibly, the adult child
might simply move back into the parents’ house together with his family, treating the
opportunity of taking care of the aging parents as his full-time job. These possible
scenarios have been well studied and documented by Stanley and Danko (1996) in
their over twenty years of research.
For any other merit goods parents want to see their children “consume,” The-
orem 18.2 will apply as long as a variable is introduced to measure how well the
children act up to their parents’ expectations and the parents will distribute their mon-
etary awards to the children according to that specific variable. As long as there is a
conflict between the way how the parents give to the children and the children’s con-
sumption preferences, at least one of the children will not be motivated to maximize
the parent’s income.
Now, let us look at Becker’s Nightlight Controversial where the conflicting inter-
ests between the benevolent head and a selfish member of the family can be seen
clearly and vividly. After we add some necessary details, it is found that the con-
clusion of Bergstrom (1989) in this situation is not completely correct. And, our
modification seems to be more in line with the reality.
Example 18.2 [The Nightlight Controversial, (Becker 1974)]. This example was
initially constructed to show how the rotten kid Theorem 18.1 can be employed in
a family setting where the husband has to read at night. The reading light, however,
bothers his wife when she tries to sleep.
To compensate for her possible loss of sleep or inconvenience, the husband trans-
fers additional monetary contribution to his wife from his increased income, a result
of his nightly reading. If Becker’s rotten kid theorem holds true for this scenario,
then how much and how long both of them will be better off when the selfish wife
allows her altruistic husband decides to read in bed. The reason is that the utility of
the husband would go higher and he would increase his transfer to his wife, more
than compensating for the initial harm done by his nightly reading.
288 18 Some Issues of the Family

Bergstrom (1989) used the nightlight usage as the public good to construct an
explicit example, on which he concluded that for the wife and husband to be on the
same page about the nightlight usage, they have to have an explicit bargain struck.
On the basis of Bergstrom’s example, in the following we will show that such a
rational bargain cannot be reached in a long period of time because the husband is
not assured that his nighttime reading will lead to an increased income. Specifically,
let

UH  UH (Uh (Xh , Xw , Y ), Uw (Xw , Y ))  Uh Uwa , 0 < a < 1, (18.15)

be the benefit function of the household, consisting of the husband and wife, in two
variables Uh and Uw , the utility functions of the husband and wife, respectively,
defined by

Uh  Xh Xw (Y + 1) (18.16)

and

Uw  Xw e−Y , (18.17)

where Xi is the total consumption of goods of family member i, i = h, w, and Y the

amount of nighttime reading of the husband.
By substituting equations (18.16) and (18.17) into equation (18.15), we obtain

UH  Xh Xw1+a (Y + 1)e−aY .

The household wants to maximize UH subject to the budgetary constraint:

Xh + Xw  Ih + Iw ,

where Ii is the personal income of i, i  h, w. The first-order conditions for this

optimization problem are
⎡ ⎤ ⎡ ⎤
∂UH ⎡ ⎤
∂Xh Xw1+a (Y + 1)−aY 1
⎢ ⎥ ⎢ ⎥
⎢ ∂UH ⎥ ⎢ ⎢ 1 ⎥
⎢ ∂Xw ⎥  ⎣ (1 + a)Xwa Xh (Y + 1)e−aY ⎥
⎦  λ⎣ ⎦, (18.18)
⎣ ⎦
∂UH 1+a −aY
Xh Xw e [1 − a(Y + 1)] − dIh
dY
∂Y

where λ > 0 is the Lagrange multiplier. The (3.1) entries in equation (18.18) shows
that only when

Y > 1/a − 1, (18.19)

18.2 Two Mysterious Phenomena of the Family 289

the husband’s income Ih has an uptrend with an increase of Y . Because the value of
a in equation (18.19) is really not known to the couple and the wife’s consumption

Xw  [(1 + a)/(2 + a)](Ih + Iw )

does not have a clear connection with Y , in a real-life setting, the husband will be
in a very inauspicious position to truly strike a deal regarding how much he should
read in bed while his wife tries to sleep. It is because equation (18.19) implies that
Y could potentially take a very large value which to the wife is not realistic at all.
Now, by looking at Theorem 18.2, the difficulty in this scenario becomes obvious:
Becker’s rotten kid theorem will not work automatically unless a negotiation takes
place. At the same time, the analysis above indicates that the wife might give in
temporarily when she imagines a foreseeable a-value. However, if such an imagined
a-value were not to materialize in a timely fashion, the final decision would be
“absolutely no bedtime reading allowed.” QED
Similar to the last example, let us pay a revisit to the following example. Once
again, by making use of Theorem 18.2, the situation can be readily analyzed and
understood.

Example 18.3 (A Prodigal Son) Based on the work of Lindbeck and Weibull (1988),
Bergstrom (1989) constructed this example to show that Becker’s rotten kid theorem
is not generally true.

Specifically, consider a selfish kid k and a benevolent parent p, who live in two
nonoverlapping time periods 1 and 2. In period 1, kid k has a certain allotment of
wealth to live on with the option of spending it all now or spending some now while
saving the rest for future use. In period 2, the kid knows that he will receive a lump
sum of money as a gift from parent p, assuming that the parent cannot make any
precommitment to punish k for any of his profligate period 1 behavior. Suppose that
the utility functions Uk and Up of kid k and parent p are
 
Uk  Uk ck1 , ck2  lnck1 + lnck2

and
 
Up  Up cp1 , cp2 , ck1 , ck2  lncp1 + lncp2 + αUk  lncp1 + lncp2 + αlnck1 + αlnck2 ,

where cki and cpi are, respectively, k’s and p’s consumptions in period i, and α > 0
a fixed number. If in period 2, the parent allocates his wealth to k to maximize his
utility Up subject to the budget constraint:

cp2 + ck2  wp2 + wk2 , (18.20)

where wj2 stands for j’s own wealth in period 2, j  p, k. Then the first-order conditions
of this optimization problem imply that to continue to maximize his utility in period
290 18 Some Issues of the Family
 
2, p will divide the family’s total wealth wp2 + wk2 in such a way that the fraction
1/(1 + α) of the total will go to p for his own consumption, and the fraction α/(1 + α)
will go to k. That implies that if k puts some money away in savings in period 1,
he will have a greater sum of wealth in the second period. That is, to obtain more
money from the parent in period 2, the kid should have spent all or more than his
period 1 allotment during period 1. In terms of Pareto efficiency, it will be wise for
the kid to squander as much as he can during period 1 so that his altruistic parent will
gift more to him during period 2. This end implies that Becker’s rotten kid theorem
is not true—for example, this particular selfish kid is not automatically motivated to
maximize the family wealth and consumption.
For why Becker’s rotten kid theorem fails in this case, Theorem 18.2 explains
clearly. First, in Becker’s original Theorem 18.1, “the family income” means the
income of the benevolent head only. Secondly, if the kid k saves some money in
period 1, he would naturally not want this savings being used to figure how much
he will be given in period 2. Since in equation (18.20) the kid’s wealth wk2 is used
against the kid’s will, Theorem 18.2 implies that at the very least, kid k would not
be motivated to maximize the parent’s income. And this example actually shows a
scenario where the kid is motivated to behave much worse. QED

18.3 Why Is Any Established Value System Imperfect?

By looking at Figure 18.3 from a different angle, we have the following result.

Theorem 18.3 [The Theorem of Never-Perfect Value Systems (Lin and Forrest
2008a)] In a family of at least two members, one member h is the head who is
benevolent and altruistic toward all other members. The head establishes a value
system for all members of the family to follow so that if a selfish member k mea-
sures up well to the value system, he will then be positively awarded by the head h.
Unfortunately, the more effort member k puts into measure up to the value system,
the more he will be punished by the award system.

Before the proof, let us first look at the background on which Theorem 18.3 is
composed. As depicted in Figure 18.1, we do not look at resource transfers as a
one-shot deal. Instead, the benevolent head transfers his resources to other family
members in a periodic fashion over time without an end in sight. The introduction
of a value system implies that the head tells other members at some time moment
that starting at a certain pay period, each member’s behavior will affect how much
he will receive from the head at the end of that period. And, the head will design his
response to each member’s behaviors in such a way as to maximize his own utility.
In other words, Theorem 18.3 explains a sequential behavioral change and reflection
over time, or a feedback mechanism, between the head and his family members.
Proof of Theorem 18.3. Assume that Yk is an index, satisfying 0 < Yk < 1, estab-
lished to check how well member k measures up to the value system predetermined
18.3 Why Is Any Established Value System Imperfect? 291

by the family head h. This index satisfies that the greater Yk is the better member
k measures up to the value system. Because k needs to put in additional effort to
increase the value of Yk , his utility function Uk  Uk (Xk , Yk ) satisfies:

∂Uk /∂Xk > 0 and ∂Uk /∂Yk < 0,

where Xk stands for k’s total consumption of numeraire good. The reason why
∂Uk /∂Yk < 0 is because all people are lazy to a certain degree.
Let the utility function of the head h be Uh  Uh (Xh , U1 , . . . , Un ), assuming that
other than h, the family contains n other members who are all selfish, Xh is h’s own
consumption, and Ui the utility function of family member i, i  1, 2, …, n.
Assume that the award received by member k from the head h is determined by
hk  hk (Yk ) such that ∂hk /∂Yk > 0. Now, the total consumption of member k is
given by Xk  Ik + hk  Ik + hk (Yk ), where Ik is k’s own income unrelated to Yk .
To reflect the fact that the index Yk represents k’s performance on the head’s value
system, which is assumed to make k better off, we did not explicitly list Yk as an
independent variable in the head’s utility function. Instead, Yk is shown in the total
consumption Xk  Ik + hk (Yk ) of k. That is, an increase in the value of Yk has a
complex effect on k, both positive and negative. Being positive is in the short term
because the head h will help to bring k’s consumption to a higher level, and it could
be negative because k has to put in more effort to raise the Yk -value. So, to make our
model more plausible, we can add the assumption that
   
 ∂Uk ∂Xk   ∂Uk 
   
 ∂X · ∂Y  >  ∂Y .
k k k

That is, the increased utility of k, brought forward by the higher-level consumption
Xk  Ik + hk (Yk ), is more than enough to offset the decreased utility of k when he
has to put in additional effort to incrementally raise the Yk -value.
To the head h, his distribution plan of his own income to all other members needs
to maximize his utility function subject to the following budgetary constraint:
n n   n
Xh + Xj  Xh + Ij + hj  Ih + Ij .
j1 j1 j1

If we ignore all members except the head h and member k, then the first-order
condition for the head’s optimization problem is given by
⎡ ∂U ⎤ ⎡ ⎤
h ∂Uh ⎡ ⎤
∂Xh ∂Xh 1
⎢ ⎥ ⎢ ⎥
⎢ ∂Uh ⎥ ⎢ ∂Uh ∂Uk ⎥ ⎢ ⎥
⎢ ∂Xk ⎥  ⎢ ∂Uk ∂Xk ⎥  λ⎣ 1 ⎦.
⎣ ⎦ ⎣ ⎦ ∂hk
∂Uh ∂Uh ∂Uk ∂Yk
∂Yk ∂Uk ∂Yk

And, member k chooses such a Yk -value Yk∗ to satisfy his first-order condition
292 18 Some Issues of the Family

∂Uk ∂Uk ∂Uk ∂hk (Yk )

 + · 0
∂Yk ∂Yk ∂Xk ∂Yk

so that he maximizes his utility.

Now, the third equation in the head’s first-order condition implies that when the
head’s expected Yk -value is greater than Yk∗ , his award hk (Yk ) for member k would
have a negative rate of increase. That is, we have shown that the more effort member
k puts in to measure up to the value system, the more he will be punished by the
award system. QED

Example 18.4 Here, we show how the result in Theorem 18.3 materially acts out. To
this end, assume that the family of concern consists of two members: the benevolent
and altruistic head h and a selfish member k, and that Yk is the index given in the
proof of Theorem 18.3. Let k’s utility function be:

Uk  Uk (Xk , Yk )  Xk (1 − Yk ),

where Xk stands for k’s total goods consumption and (1 − Yk ) his degree of laziness,
and the head h’s utility function be
 
Uh  Uh (Xh , Uk )  Xh + Uk  Xh + Xk (1 − Yk ).

Assume that the award from the head h to member k is determined by

hk  wYk ,

where w is a fixed constant >0. Then the total consumption of k is

Xk  Ik + hk  Ik + wYk ,

where Ik is member k’s own income unrelated to Yk .

To the family head h, he needs to maximize his utility function subject to the
following budgetary constraint:

Xh + Xk  Xh + (Ik + hk )  Ih + Ik . (18.21)

The first-order condition of this optimization problem is

⎡ ⎤
⎡ ∂U ⎤ 1
h
⎢ ⎥ ⎡ ⎤
⎢ ∂Xh ⎥ ⎢ ⎥ 1
⎢ ∂Uh ⎥ ⎢ 1−Yk ⎥
⎢ ∂Xk ⎥  ⎢ √ ⎥  λ⎣ 1 ⎦, (18.22)
⎣ ⎦ ⎢ 2 Xk (1−Yk ) ⎥
∂Uh ⎣ w(1−Y )−X
⎦ −w
∂Yk √ k k
2 Xk (1−Yk )

where λ  1 the Lagrange multiplier. So, from equation (18.22), it follows that
18.3 Why Is Any Established Value System Imperfect? 293

1 − Yk Xk − w(1 − Yk )
 2 and √  2w.
Xk Xk (1 − Yk )

So, w  1/8 and the selfish member k chooses Yk∗  1

2
− 4Ik to maximize his
utility, and when Yk  Yk∗ , we have
1
Xk  (1 − Yk ). (18.23)
4
Substituting equation (18.23) into equation (18.21) and then solving for hk lead
to hk  (1 − Yk )/4 − Ik . This end implies that member k’s award hk from the family
head h is a decreasing function of Yk , an index which measures how well k is doing
in terms of the established value system. QED
Theorem 18.3 indicates that if a divorced father wants to stay in good terms with
his kids who live with their mother under a different roof, his monetary gifts to the
kids should not be permanently tied to a value-based award system. If they were,
the consequence would be infelicitous. To illustrate this conclusion, the following
example shows how the “resourceful” father tries to stay out of his children’s personal
affairs in order to maintain a good relationship with them.

Example 18.5 [The “Resourceful” Father, (Lin 2009)]. Suppose that the family of
our concern has a father and two children, named 1 and 2. The utility functions of
the children are, respectively, given by the following in eyes of the father:

Ui  mi , i  1, 2, (18.24)

where mi stands for child i’s consumption in the marketplace using his father’s
monetary gifts only. The reason why Ui does not involve child i’s spending of his
own income is because the father purposely ignores the effect of those incomes so that
all the kids would be as happy as they can be when dealing with himself. That is, the
children’s utility functions in equation (18.24) are the father’s subjective functions.
Let the father’s utility function be

Uh  U1 U2 , (18.25)

with the following budget constraint

m1 + m2  Ih , (18.26)

where Ih is the father’s income, and his own consumption, when compared to his
children’s, is roughly zero.

By maximizing the father’s utility Uh subject to the constraint in equation (18.26),

we see that the father will distribute his income Ih to his 2 children evenly. That is,
Ui  0.5Ii , i  1, 2.
294 18 Some Issues of the Family

To make a comparison, assume that the father also worries about how the kids
spend their own resources in the marketplace and how much they enjoy their leisure.
In this case, the father’s subjective utility functions of the children are, respectively,

Ui  mi + Li (Yi ), i  1, 2, (18.27)

where mi is kid i’s consumption in the marketplace and Li (Yi ) his spending on leisure
with Yi being his effort invested in his work to earn his income Ii (Yi ).
Assume that the father’s utility function is still given in equation (18.25). So, the
total income of the family, consisting of the kids and the father, is Ih + I1 (Y1 ) + I2 (Y2 ).
From equation (18.27), it follows that to maximize Uh in equation (18.25) subject to
the constraint

U1 + U2  Ih + I1 (Y1 ) + I2 (Y2 ),

the father will distribute his own income in such a way that

Ui  0.5[Ih + I1 (Y1 ) + I2 (Y2 )]. (18.28)

Because Li (Yi ) is a decreasing function in Yi , the more effort to put into his work,
the less opportunity for him to have leisure, Ii (Yi ) – Li (Yi ) is an increasing function of
Yi . That is, the father now faces an incentive problem, similar to the one in Example
18.1. Specifically, say I2 (Y2 )  0, due to the fact kid 2 did not go to work at all,
and L1 (Y1 )  0, because kid 1 worked all his available time so that he did not have
any time to enjoy any leisure. Now, equation (18.28) implies that U1  U2  total
consumption of either kid 1 or 2  0.5[Ih + I1 (Y1 )]. That is, kid 2 enjoyed his leisure
to the fullest amount possible, while his total consumption is the same as that of kid
1. If we see how much each of the kids gets from the father, we find that

Kid 2 gets 0.5[Ih + I1 (Y1 )],

and

Kid 1 gets m1 − I1 (Y1 )  0.5[Ih − I1 (Y1 )].

That is, the indolent kid is handsomely awarded for his indolence, while the
assiduous kid is punished for his assiduousness as severely as giving half of his own
income to the lazy sibling and receiving the monetary gift from the father by less
than half of his own income.
What happens here is that if the father gifts his children by using an established
value system, none of the kids would be happy with him (Theorem 18.3). Conse-
quently, now the father only cares about how equally his kids would be happy by
gifting his money to them accordingly. However, as a consequence, either (1) he
creates an environment for the lazy kid to be lazier and the hardworking kid to work
18.3 Why Is Any Established Value System Imperfect? 295

harder or (2) both kids learned their lesson over time and eventually become welfare
recipients. QED
Up to this point, the gifting of either m1 or m2 in Figure 18.3 is only understood
as that the family head H distributes his resources voluntarily to other members.
However, the gifting can also be seen as the possibility that the selfish kid K takes
his initiative and proactively grabs as much of the head’s resources as possible, as in
the case of the Samaritan dilemma (Buchanan 1975; Lagerlof 2004). Corresponding
to such an understanding of Figure 18.3, we have the following result.

Theorem 18.4 (Lin 2009). If a family has such a head who cares about all other
members so much that he transfers his resources to them as long as they need to
satisfy their own desires of consumption, then each selfish member will devote as
little effort in their income-generating work as possible in order to maximize their
amounts of transfers from the head.

Proof. Let H denote the family head and the n selfish kids be named 1, 2, …, n.
Because the kids are selfish and only care about satisfying their own consumption
desires, their utility functions can be written as

Ui  Ui (mi , Yi ), ∂Ui /∂mi > 0 and ∂Ui /∂Yi < 0, i  1, 2, . . . , n,

where mi is kid i’s commodity consumption in the marketplace and Yi his effort
invested in his work to earn his income Ii (Yi ) satisfying ∂Ii /∂Yi > 0, i  1, 2, . . . , n.
Because the head H cares about all other members of his family, his utility can
be written as UH  UH (mH , U1 , U2 , . . . , Un ), where mH stands for the head’s own
consumption of numeraire good. The family’s total income is


n 
n
IH + Ii (Yi )  mH + mi .
i1 i1

To make each and every one of the selfish kids as happy as he could help, the head
H transfers his resources to maximize his utility. So, we have
⎡ ⎤ ⎡ ⎤  
∂UH ∂UH ∂Ui
∂mi ∂Ui ∂mi 1
⎣ ⎦⎣ ⎦λ , i  1, 2, . . . , n, (18.29)
∂UH ∂UH ∂Ui − ∂I∂Y
i (Yi )
∂Yi ∂Ui ∂Yi i

where λ is the Lagrange multiplier. By dividing the second row entry equation by
that of the first row entry equation in equation (18.29), we obtain
∂mi ∂Ii (Yi )
− < 0, i  1, 2, . . . , n,
∂Yi ∂Yi

which implies that as long as the selfish kids could obtain financial care from the
head H, the kids would devote as little effort in their work as possible. QED
296 18 Some Issues of the Family

As a matter of fact, the result in Theorem 18.4 is empirically supported by count-

less real-life evidence documented by Stanley and Danko (1996) in their over twenty
years of research on financially successful Americans.

18.4 Conclusions

This chapter demonstrates how the systemic yoyo model can be practically employed
to uncover insights in the study of household economics. As a matter of fact, when
circumstance involves a force or a pair of acting and reacting forces, the yoyo model
can be used to model the situation, because each such pair of forces means an eddy
motion as implied in Newton’s second law of mechanics (Lin 2007). Because of this
very useful realization, one can reasonably expect that he should be able to revisit
many or most of the basic concepts in economics in light of acting and reacting
forces, spinning yoyos, and their spin fields.
In summary, considering the significance of Becker’s rotten kid theorem, which
has found applications in household economics and designs of large-scale social
programs (Becker 1991), this chapter develops a long-waited sufficient and neces-
sary condition (Theorem 10.2) under which Becker’s rotten kid theorem actually
holds true in general. Based on the Theorem of Never-Perfect Value Systems (The-
orem 18.3), this chapter discovers the following paradox within the growth of a
family:
To raise healthy kids, the parents need to teach them how to take care of themselves. To
achieve this, they will need to award the kids for certain desirable habits. That will involve
establishing a value system. Theorem 10.3 implies that, consequently, the kids would not be
happy with how they are expected to live their lives.

To raise socially responsible kids, a similar consequence appears. To pass on

family value and beliefs down to the next generation, Theorem 18.3 indicates that
beyond monetary awards, a lot of convincing and negotiations have to be done. In
other words, the ultimate goal for the parents is to establish award systems and value
systems in such a way that Becker’s rotten kid theorem holds true. So, an extremely
important practical question arises: How can the parents achieve an operative balance
between reinforcing their value systems and keeping the kids motivated to help
maximize the family utility?
One implication of Theorem 18.3 is that to accomplish what the established value
system of the family likes to achieve, the parents have to employ available resources
from outside the family. External services and opportunities may very well lessen the
burden of the award system. With kids less angry toward the establishment within the
family, agreements between the parents and the kids can be relatively easily estab-
lished. And, examples, given throughout the chapter, show that generally conflicts
among individuals’ consumption preferences may not necessarily create any dis-
agreement between how the family resources are distributed and individual desires
of family members. However, one problem that is left unsettled here is: When a spe-
18.4 Conclusions 297

cific conflict in consumption preferences indeed leads to a problematic distribution

system, what can be done to make the family functional?
Overall, this chapter clearly shows that to raise responsible and independent kids
the parents should never overextend their altruism, even when the kids suffer from
setbacks. If they do, the kids would most likely become welfare recipients; instead
of helping the family to prosper, the kids would exploit and abuse the altruism of the
parents.
Further investigations along the lines given in this chapter are expected to expand
the range of applicability of Becker’s rotten kid theorem and to bring about brand
new insights on the relevant topics and areas of knowledge.

References

Becker, G.: A theory of social interactions. J. Polit. Econ. 82(6), 1063–1093 (1974)
Becker, G.: A Treatise of the Family. Harvard University Press, Cambridge, Massachusetts (1991)
Bergstrom, T.: A fresh look at the Rotten Kid Theorem and other household mysteries. J. Polit.
Econ. 97(5), 1138–1159 (1989)
Bernheim, B., Shleifer, A., Summers, L.: The strategic bequest motive. J. Labor Econ. 4, S151–S182
(1986)
Dijkstra, B.R.: Samaritan versus rotten kid: another look. J. Econ. Behav. Organ. 64(1), 91–110
(2007)
Fultz, D., Long, R.R., Owens, G.V., Bohan, W., Kaylor, R., Weil, J.: Studies of thermal convection
in a rotating cylinder with some implications for large-scale atmospheric motion. Meteorol.
Monographs (American Meteorological Society) 21(4) (1959)
Hide, R.: Some experiments on thermal convection in a rotating liquid. Quart. J. Roy. Meteorol.
Soc. 79, 161 (1953)
Hill, M.S., Duncan, G.: Parental family income and the socioeconomic attainment of children. Soc.
Sci. Res. 16, 36–73 (1987)
Lin, Y.: Systemic yoyo model and applications in Newton’s, Kepler’s laws, etc. Kybernetes: Int. J.
Cybern. Syst. Manag. Sci. 36(3–4), 484–516 (2007)
Lin, Y.: Systemic Yoyos: Some Impacts of the Second Dimension. Auerbach Publications, an imprint
of Taylor and Francis, New York (2009)
Lin, Y., Forrest, D.: Kybernetes: Economic yoyos and never-perfect value system. Int. J. Syst.
Cybern. 37(1), 149–165 (2008a)
Lin, Y., Forrest, D.: Economic yoyos and Becker’s rotten kid theorem. Kybernetes: Int. J. Syst.
Cybern. 37(2), 297–314 (2008b)
Lindbeck, A., Weibull, J.: Altruism and time consistency: The economics of fair accompli. J. Polit.
Econ. 96, 1165–1182 (1988)
Norberg-Schonfeldt, M.: Children’s school achievements and parental work: an analysis for Sweden.
Working Paper: Umea Economic Studies 645, Umea University, Department of Economics (2004)
Stanley, T., Danko, W.: The Millionaire Next Door: The Surprising Secrets of America’s Wealthy.
Simon and Schuster, New York (1996)
Chapter 19
Some Remarks on Organizational
Efficiencies

This chapter, mostly based on the work by Forrest and Orvis (2016), introduces
two important principles of efficiency—management efficiency and organizational
inefficiency. The former addresses the problem of how management efficiency can
be achieved, while the latter investigates the structure of employees’ efforts and
devotion toward realizing the mission of the organization. These results help to
enrich the managerial understanding on what can be improved and what cannot.
On the basis of the systemic yoyo model, all discussion and reasoning in this
chapter are established on the traditional methods of microeconomics. In particular,
the yoyo model is employed as the intuition on which conclusions are seen first;
and then traditional methods are utilized to present the exact details underneath the
established conclusions.
All results established in this chapter are expected to be applicable in all business
circumstances and scenarios. By understanding these results, frontline managers
could simply devote more of their time and effort on being flexible in terms of
management styles and focusing on the “big” picture of the corporation instead of
dwelling on how to improve employees’ efficiencies.
The rest of this chapter is organized as follows: Section 19.1 describes the problem
to be addressed in the rest of this chapter. Section 19.2 reviews the relevant literature in
order to show the contribution of this work. Section 19.3 establishes a general theory
on how people develop their individual philosophical and value systems. Section 19.4
shows why not all employees work toward materializing their company missions.
Section 19.5 presents a possible way to achieve managerial efficiency. Section 19.6
introduces the principle of organizational inefficiency. Then the chapter is concluded
in Section 19.7.

© Springer Nature Switzerland AG 2018 299

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_19
300 19 Some Remarks on Organizational Efficiencies

19.1 The Question that Is to Be Addressed

When a person is new to a company, he generally is able to see much inefficiency

within the management, the business operation, and other areas of the business. And
like many of us, he most likely has ideas about how things could potentially improve.
Of course, such experience is not only limited to new hires.
One reason why people can easily see abundant inefficiencies within their envi-
ronment is because each person, as a living being that is severely limited by its
sensing organs, looks at the world with his particular set of values and philosophical
assumptions about the world (Lin and Forrest 2011; Villalobos and Vargas 2015;
Terán et al. 2015). In other words, because each person has his unique philosophical
assumptions and value system, the physical world becomes extremely beautiful and
multi-colored when people individually try to describe what they see and what the
world is really about.
This chapter attempts to address what is just described by addressing the following
questions: Is there such thing as management efficiency? If there is, how can it be
achieved? Why in real-life organizational efficiency can only be a conceptual dream?
To accomplish this goal, this chapter will first establish a systemic theory about
how personal values and philosophical assumptions are developed and established
over time since when a person is born into this physical world and why they evolve
with time and changes of the environment. Then, on the basis of this new under-
standing, it becomes clear for why different people mostly likely have diverse value
systems. With such a general theory in place, this chapter then turns its focus to
the concept of organizational efficiency and tries to answer the question of whether
or not an organization could ever be efficient by looking at several examples that
are constructed rigorously by using methods of microeconomics. Through stren-
uous reasoning based on the intuition of the systemic yoyo model, it is discovered
that inconsistencies between employees’ personal values and between these personal
values and the organization’s mission always lead to organizational inefficiencies.
That is, organizational inefficiencies in real life always exist. At the same time, the
analyses underlying the established theory suggest that managerial efficiency can be
potentially achieved by being flexible regarding management styles.

19.2 Literature Review

Different authors have studied the concepts of organizational efficiency and man-
agement efficiency from various angles. For example, Pawłowski et al. (2012) treat
modern management as a series of decision makings and creations of conditions
for effective realization of the decisions. When diagnosing a given situation and
attempting to improve the organization’s efficiency, these authors suggest employ-
ing an approach at three different efficiency levels: at the organization level, at the
process level, and at the workstation level. They treat the degree of decision effec-
19.2 Literature Review 301

tiveness as dependent on the rationality of the involved problem-solving process,

and determined by the primary and supporting goals, strategies developed for their
realization, and the quality of internal efforts. Here, the quality of internal efforts
includes readiness for cooperation and flexibility toward changes. Ren and Xiong
(2010) investigate how to measure management efficiency from the angle of systems
that involve many mutual coupling and unknown or uncertain factors by using the
principle of information entropy. By utilizing matrix theory, these authors provide
relevant evidence and a basic method for probability assignment. Their case study,
using data that is stored in the database of enterprise management information sys-
tem and information offered by management experts, indicates that their method and
model can decouple the relevant mutual-coupling factors and can objectively and
accurately measure management efficiency.
Because management increases its company’s functionality and competitive-
ness and impacts the efficiency and efficacy of the company’s organization, Laura-
Georgeta (2011) considers the performance management by using an approach that
joins both organization efficiency and efficacy and the grounds for achieving orga-
nization’s competitiveness. Burton et al. (1991) research the relationship between
organizational size and performance. They use three specific dimensions of size,
as measured by the number of headquarter resources and policies, the number of
products or services, and the number of resources and services of subunits, to ana-
lyze organizational performance efficiency by using a computer simulation. Their
results suggest that some dimensions of size, particularly the number of headquarter
resources and policies, are related significantly to performance efficiency. Cummins
et al. (1999) introduce the technique of cross-frontier analysis for estimating the
relative efficiency of alternative organizational forms in an industry. These authors
analyze a sample of stock and mutual property-liability insurers using nonparametric
frontier efficiency methods. Their technique of analysis measures the relative effi-
ciency of each organizational form by computing the efficiency of each stock (mutual)
firm relative to a reference set that consists of all mutual (stock) firms. Their results
indicate that stocks and mutuals operate on separate production and cost frontiers
and represent distinct technologies, where the stock technology dominates the mutual
technology for producing stock outputs, while the mutual technology dominates the
stock technology for producing mutual outputs, and the stock cost frontier dominates
the mutual cost frontier.
Ismail et al. (2011) provide an empirical study on the relationship between effi-
ciency and organizational structure for takaful operators of the dual financial system
in Malaysia by using a sample of 19 firms chosen over the time period of 2004–2009.
These authors employ the method of data envelopment analysis, a nonparametric
approach, with the input orientation measurement used to estimate the technical effi-
ciency for both industries. They apply both constant return to scale and variable
return to scale to separate the scale efficiency from the technical efficiency, while
a Mann Whitney test is employed to examine any significant difference in the effi-
ciency between takaful and insurance industry. Their findings indicate that there is a
significant difference in technical efficiency between takaful industry and insurance
industry that takaful has lower technical efficiency than the conventional insurance
302 19 Some Remarks on Organizational Efficiencies

and that the organization form has an influence on efficiency, while the conventional
insurers have higher-scale efficiency than takaful industry. Alvesson (1989) surveys
some of the common conceptualizations of organizational culture as a building block
in organizational design, as the outcome of symbolic management, as a diagnostic
instrument, and as a paradigmatic concept. He examined assumptions and hypotheses
on the relationship between organizational culture and various performance-related
outcomes, indicating that many works exaggerate the central role played by corpo-
rate culture as an independent variable affecting organizational efficiency. This work
concludes that although it is understandable that scholars for technical and prag-
matic reasons emphasize culture as a key to organizational performance and then
narrow down the concept of culture to a variable or factor, this reductionistic view
on culture is unfruitful, and a broader view, conceptualizing culture as a metaphor
for organization, appears to be more promising.
By recognizing the fact that benchmarking for decision-making units (DMUs) is
more than a purely monitoring process and includes a component of future planning,
Stewart (2010) extends the standard data envelopment analysis model to include
longer-term top management goals. This new model uses a goal programming struc-
ture to find points on the efficient frontier that are realistically achievable by DMUs
while at the same time achieving a closer approach to long-term organizational goals,
as distinct from the local performance of individual DMUs. Stewart considers the
possibility of adjusting constraints on the DMUs by investment in extended inputs or
new technologies, in which case minimizing the associated investment costs becomes
an additional management objective. By identifying the manager of an organization
as a systems designer who plays the role of self-organization both within and outside
the organization of concern, Kasianiuk (2016) presents identification models useful
for understanding self-organization processes within and outside the organizations
facilitated by leaders.
In comparison, this chapter formulates the concepts of management and orga-
nizational efficiency at the height of theoretical abstraction and carries the existing
literature on these concepts steps forward with a much wider range of applicability.

19.3 Individuals’ Underlying Philosophical and Value

Systems

This section addresses the following questions: (1) Why do different people have dif-
ferent underlying values and philosophical assumptions about how the world works?
And (2) why is it extremely difficult to find two people with the same, identical value
system. By underlying values and philosophical assumptions, it means the value
system of a person that consists of his/her beliefs about how the world functions and
his/her moral codes with which he/she is recognized with his/her particular identity
and integrity.
19.3 Individuals’ Underlying Philosophical and Value Systems 303

As a matter of fact, from the systemic yoyo model of systems, it follows that each
human being lives in a vast ocean of spinning fields. These are the fields of other
people, various physical objects, abstract thoughts, and myriad of other things and
matters. After a person is born into this physical world, his/her yoyo field starts to
interact with many different yoyo fields. It is these interactions with people, with
physical objects, with abstract thoughts, and the myriad of other things and matter
that shape the person’s philosophical assumptions and values. The situation is similar
to how a civilization formulates its value system (Lin and Forrest 2011). Within the
interactions experienced by one person from those by another person, there are always
subtle differences. That explains why each person has his/her own unique set of
very specific philosophical assumptions and values. It is these specific philosophical
assumptions and values that dictate the behaviors and decision making of the person
for the rest of his/her life. Comparing to the magnificent scale of the entire ocean of
spin fields of the world, although the differences might be “subtle,” they are generally
major to the individuals involved. They cause important differences in the relevant
personal value systems. That is why children who grow up in the same household
possess quite different personalities, characteristics, and thinking processes.
In the following, let us look at some of the relevant details of the underlying sys-
temic reasoning behind this general theory of how personal philosophical assump-
tions and values are formed.
In 1898, over one hundred years ago, when he studied the eddy effects of density
changes in the media of the atmosphere and ocean, Bjerknes established his well-
known circulation theorem; see Hess (1959) for details. This theorem shows how in
terms of mathematics nonlinearity stands (mostly) for singularities and in terms of
physics it represents eddy motions. Such motions represent structural evolutions, as
a natural consequence of uneven evolutions of things, such as information, energy,
investment, talents, etc.
In particular, a circulation means a closed contour in a fluid, Figure 19.1. Here,
the term “fluid” can be understood generally either as fluids in the conventional
sense, as such water or gas, or as information, knowledge, money, etc., because the
movement of information, knowledge, money, etc., possesses the basic characteristics
of fluids of the conventional sense. Rigorously, each circulation  is defined as the
line integral about the contour of the component of the velocity vector locally tangent
−→
to the contour. In symbols, let V be the speed of a moving fluid, S an arbitrary closed
curve, δr the vector difference of two neighboring points of the curve S (Figure 19.1),
 −→
then a circulation  is defined as follows:   S V δ r .
Through some ingenious manipulations (Wu and Lin 2002), the following well-
known Bjerknes’ Circulation Theorem is obtained:
→ ¨  
−
dV 1 dσ
 ∇ × (−∇p) · δσ − 2 , (19.1)
dt ρ dt
σ
304 19 Some Remarks on Organizational Efficiencies

Figure 19.1 A closed circulation

Figure 19.2 A diagram for equal p- and ρ-plane distributions

where σ is the projection area on the equator plane of the area enclosed by the closed
curve S, p the atmospheric pressure, ρ the density of the atmosphere, and  the
earth’s rotational angular speed.
In terms of the meaning of this theorem, the left-hand side of equation (19.1)
stands for the acceleration of the moving fluid. According to Newton’s second law
of motion, it is equivalent to the force acting on the fluid. On the right-hand side,
the first term is a solenoid term, as such known in meteorology. It is originated from
the interaction of the p- and ρ-planes due to uneven density ρ. So, a twisting force
is consequently created. As a result, the movement of the fluid must be a rotation
with the rotating direction determined by the equal p- and ρ-plane distributions
(Figure 19.2). The second term in equation (19.1) comes from the rotation of the
earth.
From another angle that is different from the systemic yoyo model, this theo-
rem reveals the commonly existing and practically significant eddy effects of fluid
motions. It explains why uneven eddy motions are the most common form of move-
ments observed in the universe. Because twisting forces are created by uneven den-
sities, fields of spinning currents (of water, information, knowledge, money, etc.) are
naturally produced. Such fields exist as currents of different types, where clockwise
and counterclockwise eddy pools always coexist, leading to destructions of the initial
smooth, if any, fields of currents.
What’s very important is that the concept of uneven eddy evolutions reveals that
forces exist within the structures of evolving objects and do not exist independently
outside of the objects. That shows indirectly why the systemic yoyo model is impor-
tant in terms of its power to reveal characteristics of spinning motions and their inter-
19.3 Individuals’ Underlying Philosophical and Value Systems 305

actions, because throughout the Newtonian theory of motion acting forces always
exist outside the objects involved.
Now, we are ready to look at the questions we like to address in this section.
At the start of a new life, when it is still living in the tender and loving care of
others, due to the existing conditions of disability and the limitations of available
resources within the environment, the newborn forms his/her elementary beliefs,
basic values, and fundamental philosophical assumptions, on which he/she reasons in
order to acquire what is needed, explains whatever inexplicable, develops approaches
to overcome hardships, and establishes methods to manage personal affairs.
With time, his/her increasing physical and mental capabilities make him/her able
to handle more advanced tools, more innovative thoughts, and more intelligent meth-
ods to deal with personal and interpersonal affairs. As the spinning strength of the
yoyo field of the person increases, the person natural desires to acquire better condi-
tions, more control, and more recognition that pave the way for the person to invent
or master new tools to deal with issues appearing in the person’s interaction with the
environment, discover new methods to reason, and introduce more efficient ways to
deal with various affairs. In other words, a circulation of information, knowledge,
capability, etc., starts to form within and around the person.
With rapidly growing awareness of new frontiers in the environment and within the
personal being of the person, various practically useful skills also become parts of the
circulation. As soon as such a circulation appears, the result of Bjerknes’ Circulation
Theorem starts to apply and guarantees the appearance of abstract eddy motions
within the mind of the person. These abstract eddy motions consist of the appearance
of new people, expansion of knowledge, improvement of old skills, acquirement
of new skills, and accumulation of wealth. In other words, with age the person
gradually forms his/her underlying assumptions and values of philosophy on how
the world functions, what kinds of behaviors are socially acceptable, how he/she
should appropriately interact with others and the environment in order to achieve
desired results.
As the desire of controlling more resources and influencing more people in the
environment grows, the yoyo field of the person gradually becomes stronger than
before and matures over time. That makes the person’s yoyo structure more powerful
and more able to fight off different beliefs and value systems that might very well
destroy his/her own beliefs and system. It is through these nonstopping struggles
with different beliefs and value systems, the person enriches his/her philosophical
assumptions and values and redefines his/her identity throughout life. In other words,
the person’s basic philosophical assumptions and value system reflect the fundamen-
tal structure of his/her yoyo field. Although his/her yoyo field changes and evolves
with the environment and time, its basic characteristics stay the same throughout the
person’s life.
This general theory of how a person’s philosophical assumptions and value system
grow explains why different people have different underlying philosophical assump-
tions and value systems . It is because
306 19 Some Remarks on Organizational Efficiencies

• First, no two people grow up within a perfectly identical environment.

• Second, with age, people’s philosophical assumptions and values evolve according
to their, respectively, changing environments.
For more related and detailed discussions based on the systemic yoyo model regard-
ing the four human endowments—self-awareness, imagination, conscience, free will,
the concepts of character, thought, desire, enthusiasm, happiness, fear, and forced
struggle, and the structure of human mind, see Lin and Forrest (2011).
A related but different concept well studied in the literature of economics is that
of leisure (Becker 1991). In comparison, the concept of person values is much more
general than that of leisure. As a matter of fact, the concept of leisure, as how it
is used in the literature of economics, stands for only a minor part of the concept
of personal value systems. For example, two people with different value systems
can all be working hard or finding pleasure in their works. However, their different
value systems generally lead to drastically different outcomes. To this end, in real
life, we can readily find two such university professors that one believes that higher
quality education is achieved by passing on to students as much known knowledge as
possible, similar to what is done in most of the current high schools, while the other
recognizes the fact that a high-quality education is accomplished only by cultivating
critical thinking and desire to accomplish in students. That is actually one of the
major differences between craftsmen and innovators, the former attempt to make
better and constantly increasing known products, while the latter focus more on
introducing new products with improved functionality.

19.4 Forever Existence of Inefficiency

In this chapter, by organization, it means such an economic entity that consists of

people who are connected by some common purpose while somehow the entity has
to maintain its economic viability through its members’ efforts and works. There-
fore, any group of randomly selected, unorganized individuals is not treated as an
organization. In terms of the concept of general systems (Lin 1999), an organization
considered in this chapter has to be a nontrivial system that is made up of, at least
partially, people, while the system has to produce its outputs in order to generate
revenue to sustain its operation. For example, a university is an organization that
exists for the purpose of educating people while sustained by financial viability; a
business firm is an organization that exists for the purpose of generating positive cash
flows through producing and/or selling products and services, while all the people
who are leaving a theater right after having finished watching a movie are not an
organization, because these people are no longer related to each other in terms of
any common purpose.
Additionally, assume that each organization has a fixed mission at any chosen
moment of time. As guided the mission, all employees put in their efforts and work to
19.4 Forever Existence of Inefficiency 307

produce the predetermined products and/or services in order to sustain the economic
viability of the organization from falling into an entity of nonorganization.
For any chosen organization, its organizational efficiency is defined as how well its
employees help the organization to reach its defined mission. An organization is said
to be efficient, if all employees work toward the common goal and help approach or
materialize the stated mission. Otherwise, the organization is said to be not efficient
(or inefficient).
The question this section addresses is: Can an organization ever be efficient?
To answer this question, let us first see two examples.

Example 19.1. What follows in the paragraphs below in this case analysis is based on
the reading nightlight example initially constructed by Becker (1974), later modified
by Bergstrom (1989), and then revisited by Lin (2009).

Assume that the organization of concern has two such employees i and j among
others that a particular effort of i, which will potentially help increase the income
of i, really bothers j, while for some reason, j cannot simply leave the organization.
In the eyes of i, this particular effort is potentially profitable to the organization,
potentially strengthens the financial viability of the organization, and will help the
organization to approach its mission.
In order to calm j down and hopefully make j become supportive of his/her effort,
i will compensate j for his/her frustration by transferring monetary contribution to j
from his increased income, a result of his particular effort. Now, if the organization
is efficient for this scenario, then a natural question arises: How much and how long
will both i and j be better off without selfish j starting to interfere with what i is
doing, which will make the organization inefficient?
In particular, assume that the organization has the following production function
     
Pc  P Ui Xi , Xj , Y , Uj Xj , Y , . . .  Ui Uja , 0<a< 1, (19.2)

satisfying
∂Ui ∂Ui ∂Uj ∂Uj
> 0, k  i, j, > 0, > 0, and < 0,
∂Xk ∂Y ∂Xj ∂Y

where the condition 0 < a < 1 reflects the assumption that i’s effort is potentially
beneficial to the organization in several ways, while it really bothers j, because a
< 1 is the root reason, the dots represent an abbreviation of the utility functions of
all employees of the organization other than i and j, and  the product of all other
employees’ utility functions, Ui is the utility function of employee i, and Uj the utility
function of j, defined as follows:

Ui  Xi Xj (Y + 1) (19.3)

and
308 19 Some Remarks on Organizational Efficiencies

Uj  Xj e−Y , (19.4)

where X k is the total consumption of k, k = i, j, and Y an index that measures the effort
employee i puts into his work. Here, we assume that while employee j is so selfish
that he/she does not care about the well-being of any other coworkers, employee i is
so altruistic that he includes j’s utility in his utility function, while knowing that his
effort really bothers j.
In equation (19.3), the factor (Y +1) indicates that other than receiving a portion of
utility from the joint consumptions of both i and j, i also enjoys additional utility that
is proportional to how much effort he/she puts into his/her work. In equation (19.4),
the factor e−Y models the fact that j is really bothered by i’s effort Y exponentially.
Substituting equations (19.3) and (19.4) into equation (19.2) produces

Pc  Ui Uja   Xi Xj1+a (Y + 1)e−aY .

Now, the organization wants to maximize Pc subject to the following budgetary

constraint:


n 
n
Xk  I1 + I2 + · · · + In  Ik , (19.5)
k1 k1

where Ik is the personal income of employee k, k  1, 2, …, n, assuming that the

organization has n employees total. The first-order conditions for this optimization
problem is
⎡ ∂P ⎤ ⎡ ⎤
c
Xj1+a (Y + 1)e−aY  ⎡ ⎤
∂Xi 1
⎢ ⎥ ⎢ ⎥
⎢ ∂Pc ⎥ ⎢ (1 + a)X a X (Y + 1)e−aY  ⎥ ⎢ ⎥
⎢ ∂Xj ⎥⎢ j i ⎥  λ⎣ 1 ⎦, (19.6)
⎣ ⎦ ⎣ ⎦
∂Pc Xi Xj1+a e−aY [1 − a(Y + 1)] − dY
dIi

∂Y

where λ > 0 is the Lagrange multiplier. The (3.1)-entries in equation (19.6) show that
only when
1
Y > − 1, (19.7)
a
employee i’s income Ii has an up-trend with an increase of Y . Dividing the (1.1)-
entries in equation (19.6) by the corresponding (2.1)-entries provides

Xj  (1 + a)Xi .

By solving this equation for Xi and inserting into equation (19.5), we obtain j’s
consumption as follows:
19.4 Forever Existence of Inefficiency 309
⎛ ⎞
1 + a⎜ 
n 
n
⎟
Xj  ⎝ Ik − Xk ⎠. (19.8)
2+a
k1 k1
ki,j

Because the value of a in equation (19.7) is really not known to either i or j, and
because j’s consumption in equation (19.8) does not have a clear connection with Y ,
in a real-life setting employee i will be in a very inauspicious position to truly strike
any deal regarding how much he/she should devote himself/herself to the particular
effort while j continues to be bothered. It is because equation (19.7) implies that Y
could potentially take a very large value, if a is a value close to 0, which, to employee
j, is not realistic at all in terms of getting compensated for his/her frustration, since
no one knows when i will get any additional pay for his/her particular effort.
Regarding organizational efficiency, the difficulty in this scenario becomes obvi-
ous: To achieve the desired organizational efficiency, a negotiation between i and j
(not between the organization and employee j, because the organization has done its
part by including j’s utility within its production function) needs to take place. At
the same time, the analysis above indicates that j might give in temporarily when
he/she imagines a foreseeable a-value, which means when j foresees when he/she
might get compensated for his/her frustration. However, if such an imagined a-value
was not materialized in a timely fashion, then a frustrated j will sooner or later start
to interfere with i’s effort, making the organization inefficient. QED
Example 19.2. This case analysis is constructed based on the works of Lindbeck and
Weibull (1988) and Bergstrom (1989) for the purpose of showing that organizational
efficiency cannot be generally achieved or maintained.
Consider an organization consisting of n different departments. Each department
receives a lump sum of money as its budget at the start of each time period to cover all
the operational expenses of the department throughout the period. Let us single out
one department, named k, and two nonoverlapping time periods 1 and 2. In period
1, department k has a budget of a certain amount of money to operate on. It has
the option to spend it all now or to spend some now and save the rest for future
use. In period 2, department k knows that it will receive a lump sum of money as
its new budget from the parent organization, assuming that the organization cannot
make any precommitment to punish department k for any of its profligate period 1
behavior. Suppose that the production functions Pk and Po of the department k and
the organization are defined by

Pk (ck1 , ck2 )  ln ck1 + ln ck2

and

Po (co1 , co2 , ck1 , ck2 )  ln co1 + ln co2 + αUk  ln co1 + ln co2 + α ln ck1 + α ln ck2 ,

where cki and coi are, respectively, department k’s and the organization’s direct
expenses in period i, and α > 0 a fixed number, i  1, 2.
310 19 Some Remarks on Organizational Efficiencies

If in period 2, the organization allocates its total available budget to department k

to maximize its production Po subject to the following budget constraint:

co2 + ck2  wo2 + wk2 , (19.9)

where w2j stands for the available money in j’s budget in period 2, j  o, k. Then the
first-order conditions of this optimization problem are given as follows:
⎡ ⎤ ⎡ ⎤
∂Po 1  
⎢ ∂co2 ⎥ ⎣ co2 ⎦ 1
⎣ ∂Po ⎦  α  λ , (19.10)
c2
1
∂ck
2
k

where λ is the Language multiplier. Dividing the first entry equation in equa-
tion (19.10) by the second gives

ck2  αco2 .

So, equation (19.9) implies that

1  2  α  2 
co2  wo + wk2 and ck2  wo + wk2
1+α 1+α
That is, to maximize its production in period 2, the organization will divide the
organization’s total budget (w2p + w2k ) in such a way that the fraction 1/(1 + α) of the
total will go to the organization’s budget for its direct expenses, and the fraction
α/(1 + α) will go to department k.
This end implies that if department k controls its expenses with frugality in period
1 and possibly rolls over the surplus into period 2, then it will have a greater sum of
money available for spending in the second period. That is, to obtain more money
from the parent organization in period 2, department k should have spent all or more
than its period 1 allotment during period 1. That is, in terms of Pareto efficiency, it will
be wise for department k to squander as much as it can during period 1 so that its parent
organization will provide more to it during period 2. If the organization’s mission can
only be further materialized with budgetary support, then this end implies that with
the budgetary arrangement of the organization no department will be automatically
motivated to maximize the overall pool of monetary asset of the organization. In
other words, the organization is not efficient.
Analyzing the afore-described scenario in further details provides the reason why
organizational inefficiency appears here. Firstly, the budget of the organization is
not the sum of all the available money supplies of the departments. Secondly, if
department k saves some money in period 1, the department would naturally not
want this money being used to figure how much it will get in period 2. Because in
equation (19.9) department k’s budget w2k is used, an inconsistency appears between
the desires of the department k and the organization, that is, the root cause behind why
19.4 Forever Existence of Inefficiency 311

department k is not motivated to further achieving the mission of the organization.

QED
Riding on the case studied in Examples 19.1 and 19.2, the following general result
can be established:

Theorem 19.1. Inefficiency naturally exists within any organizational system that
has at least one full-time employee whose personal value is not in total agreement
with the organization’s mission.
Proof. Assume by contradiction that there is such an organizational system that
satisfies the conditions listed in the theorem while it is fully efficient, where the
mission of the organization is not in total agreement with the personal value of full-
time employee k. Let Y be a variable measuring one aspect of employee k’s personal
value such that the utility of k increases with Y while the work efficiency of k in terms
of helping realize the mission of the organization decreases with Y . Symbolically,
we have
∂Uk ∂Uk
Uk  Uk (Xk , Y ), satisfying > 0 and > 0,
∂Xk ∂Y

where Uk is the utility function of k, Xk the total consumption of k, and the production
function of the organization is
∂P ∂P
P  P(Xc , Uk , . . .), satisfying > 0, > 0, (19.11)
∂Xc ∂Uk

where Xc is the total expenditure of the organization, including the monetary expenses
on all employees except k, and the dots represent the abbreviation of all the utility
functions of the other employees of the organization. The fact that employees’ utility
functions enter into the organization’s production function implies that the organi-
zation keeps its employees’ well-being as part of its objectives. Now, the monetary
bonus that measures the work efficiency of k is expressed by
dhk
hk  hk (Y ), satisfying < 0.
dY
Note that in real life, such a variable Y might only exist implicitly and indirectly
and cannot be measured readily. However, its negative effect on the work quality and
efficiency generally can be clearly seen. So, for our purpose, assume without loss of
generality that Y can be measured and used in determining the monetary bonus.
The organization distributes its resources to its employees in such a way that
maximizes its production function P in equation (19.11) subject to the following
budgetary constraint:

Xc + Xk  Xc + (Ik + hk ), (19.12)
312 19 Some Remarks on Organizational Efficiencies

where Ik is the base personal income of k earned from his work at the organization.
Maximizing the production function in equation (19.11) subject to the constraint in
equation (19.12) leads to the following contradiction: ∂X k /∂Y > 0 and ∂Xk /∂Y 
dhk /dY < 0. This contradiction implies that the assumption that the organizational
system that satisfies the conditions listed in the theorem is fully efficient is incorrect.
QED

19.5 One Possible Way to Achieve Managerial Efficiency

Because each organization exists for a particular purpose, to materialize the purpose
the organization most likely has to hire employees with particular talents. So, the
following natural question arises: When employees have conflicting philosophical
assumptions and values, can the organization still operate smoothly while keeping
the best interests of these employees in mind at the same time? To see how widely
employees with conflicting philosophical assumptions and values exist in real life,
let us consider two such employees, for example, that one of them enjoys life in any
way possible while treating work as a nuisance, while the other is totally submerged
in his/her work, believing the work is his/her life and purpose of life.
Discussions in this chapter will show that the answer to this question is: Yes, it is
possible. To reach this end result, consider the following example.

Example 19.3. Assume that an organization has two such employees i and j that
they hold one conflicting personal value, denoted by Y. Assume also that the utility
functions of i and j are

Ui  Xi − Y and Uj  Xj + Y ,

where Xk is the consumption of k, k  i, j. Then, the organization’s production

function could be defined by
   
Pc  Xc Ui + Uj  Xc Xi + Xj , (19.13)

where Xc represents the total expenditure of the organization. In this case, equa-
tion (19.13) indicates that although there is conflict in the personal values of i and j,
the organization can still be completely neutral, because variable Y totally disappears
from the organization’s production function in equation (19.13). QED

A second natural question that arises at this junction is that when the organization’s
mission is in conflict with the personal value of an employee i, can the operation of
the organization still be smooth while keeping the well-being of employee i in mind?
As what the next example shows, the answer to this question is: Yes, it is possible.
19.5 One Possible Way to Achieve Managerial Efficiency 313

Example 19.4. Assume that Y stands for a particular aspect of employee i’s personal
value that is in disagreement with the mission of his/her company. Let the utility
function of i be given by

Ui  Xi Y

and the organization’s utility function from running its business be

Uc  Xc /Y

where Xi stands for the total consumption of employee i and Xc the total expenditure
of the organization. If the product function of organization is given by

Pc  Pc (Uc , Ui )  Ui · Uc  Xi · Xc ,

which implies that although the organization’s mission is in conflict with the personal
value of employee i, in the production function, the organization still cares about
employee i as much as if they did not have any conflict. QED

Jointly, Examples 19.3 and 19.4 imply that no matter whether there exists a con-
flict regarding individual personal values of the employees or between the personal
values of some of the employees and the mission of the company, the organization
could still operate with all employees’ well-being in mind. In other words, if the
efficiency of management is defined as keeping all employees’ well-being in mind
while materializing the mission of the organization, then the following result holds
true.
Principle of Management Efficiency: Management flexibility in terms of manage-
rial style is the key for maintaining management efficiency.
As what Examples 19.3 and 19.4 suggest, how to flexibly define the organization’s
production function is the key for eliminating any potential and existing conflicts
between employees’ personal values and between employees’ personal values and the
mission of the organization. In these two case analyses, different ways of formulating
the product function are seen as varying approaches of management. That is, if the
management of a company has a fixated method of measurement of success, then it
will have difficulties to deal with the multi-faceted personalities, and consequently,
some of the crucial talents, badly needed for the success of the company, will be
forced out of the organization.

19.6 Principle of Organizational Inefficiency

From Theorem 19.1, it readily follows that if a company has two employees with
conflicting personal philosophical assumptions and value systems, then the business
314 19 Some Remarks on Organizational Efficiencies

entity will have to suffer from organizational inefficiency. It is because one of the two
conflicting personal philosophical and value systems has to be not in total agreement
with the organization’s mission. Hence, a natural question arises: In order to increase
the efficiency of the company of concern, why can we not hire only those people
whose personal values are in complete agreement with the organizational mission?
First of all, from Section 19.3 on how personal philosophical assumptions and
value systems are formed, it follows that finding employees with identical personal
philosophical assumptions and value systems is practically impossible. Secondly,
personal philosophical assumptions and value systems evolve with time and the
changing environment. So, personal philosophical assumptions and value systems,
which are initially similar, will definitely diverge over time. Thirdly, if employees
who have the desirable identical personal value can be found, then what is observed
in the dishpan experiment suggest that differences among the personal philosophical
assumptions and value systems will inevitably appear within the smooth operation
of the organization. In particular, Raymond Hide (1953) of Cambridge University,
England, and Dave Fultz and his colleagues of University of Chicago (1959) inde-
pendently conducted the so-called dishpan experiment. To make the story short, Fultz
and his colleagues partially filled a cylindrical vessel with water, placed it on a rotat-
ing turntable, and subjected to heating near the periphery and cooling near the center.
The bottom of the container is intended to simulate one hemisphere of the earth’s
surface, the water the air above this hemisphere, the rotation of the turntable simu-
lates the earth’s rotation, and the heating and cooling simulate the excess external
heating of the atmosphere in low latitudes and the excess cooling in high latitudes.
In order to observe the pattern of flows at the upper surface of the water, which
was intended to simulate atmospheric motion at high elevations, Fultz sprinkled some
aluminum powder. A special camera that effectively rotated with the turntable took
time exposures so that a moving aluminum particle would appear as a streak and
sometimes each exposure ended with a flash, which could add an arrowhead to the
forwarded end of each streak. The turntable generally rotated counterclockwise, as
does the earth when viewed from the North Pole.
Although everything in the experiment is arranged with perfect symmetry about
the axis of rotation, such as no impurities added in the water, the bottom of the
container is flat, both expected flow patterns as shown in Figure 19.3a, b are observed,
and the choice between the two flow patterns depends on the speed of the turntable’s
rotation and the intensity of the heating. Briefly, with fixed heating, a transition from
the circular symmetry in Figure 19.3a to that in Figure 19.3b will take place as the
rotation speed increase past a critical value, and when the rotation speed passes yet
another higher critical value, a transition from the chaotic pattern in Figure 19.3b
back to that in Figure 19.3a appears. On the other hand, with the sufficiently rapid, but
fixed speed of rotation, similar transitions will occur when the temperature difference
between the heating along the periphery and the cooling at the center of the dish
reaches a critical strength, while another transition back to the symmetry would
occur when the difference reaches another higher critical level.
19.6 Principle of Organizational Inefficiency 315

(a) Symmetric flow at the upper surface (b) Asymmetric flow at the upper surface

Figure 19.3 Patterns observed in Fultz’s dishpan experiment

By associating this laboratory observation with our current context, we can natu-
rally establish the following model:
• The entire dishpan stands for the mission of the organization of concern.
• The spin of the dishpan represents the operation of the organization.
• Individual employees’ personal philosophical assumptions and value systems are
drops of the water inside the dishpan.
Then, this dishpan experiment indicates that although the organization of our con-
cern could supposedly find and hire employees with identical personal philosophical
assumptions and value systems, this initial uniformity in individual personal philo-
sophical assumptions and values will soon be destroyed by the operation, no matter
whether which is smooth or not, of the organization. Speaking in the ordinary lan-
guage, what this conclusion means is that the uniformity in personal values is mate-
rialistically destroyed by the interactions and conflicts of interests of the employees
of the organization. Therefore, Theorem 19.1 directly implies that some inefficiency
in the organizational functions starts to appear.
Based on the discussion in the previous paragraphs, Theorem 19.1 actually implies
the following general conclusion.

Principle of Organizational Inefficiency: Inefficiency always exists in any organi-

zation.

In our discussion here, by organizational efficiency, it includes all aspects of

running the business of the organization beyond from what is specified earlier. In
particular, it includes, but not limited to, efficiencies of communication (both inter-
nal and external), management efficiency, public relations, employee satisfactions,
etc. Specifically, communication efficiency consists of the efficiencies of the media,
the receivers, and the senders of information. When the media is somehow defi-
cient, defective, or even dysfunctional, due to diverse perceptions and background
knowledge involved, the receivers and senders of information can easily misunder-
stand each other in many different ways. Management efficiency is very delicate.
More often than not, any managerial push for reaching a predetermined goal could
316 19 Some Remarks on Organizational Efficiencies

(a) Both harmonic N and M are divergent (b) When N is divergent and M is convergent

(c) Both harmonic N and M are convergent (d) When N is convergent and M is divergent

Figure 19.4 Different consequences appear when two systems interact

easily make some employees upset, leading to purposefully delay work progress or
unconscious slowdown.
As a matter of fact, no matter if the organization of our concern has full-time
employees or not, this proposed principle of organizational inefficiency still holds
true. To see why, it suffices to address the situation where the organization has only
part-time employee(s) and none of them has a conflicting personal value with the
organization’s mission. In this case, there are the following scenarios to consider:
1. The organization has one part-time employee who is the founder of the company.
2. The organization hires at least one employee who is not the founder.
For scenario 1, it is reasonable to imagine that the founder employee is the one
who formulates the mission statement of his/her organization according to what
he/she desires to achieve in reference to his/her personal philosophical assump-
tions and value system. Now, there are three systems involved here: the founder
him/herself, his/her organization (a formal system), and the environment, with the
former two being input–output systems. When the founder interacts with the envi-
ronment, inevitable consequences appear. As shown in Figure 19.4, let N stand for
the yoyo field of the founder and M the yoyo field of the environment. By being an
input–output system, N has to interact with M so that some unexpected subeddies
appear, the small pools shown in Figure 19.4a–d. That means that the previously
stated mission of the organization, as time goes on, becomes inadequate while the
desire of the founder who now likes to reach another mission, which might be slightly
different from the one stated in the formulated mission earlier. So, from the definition
of organizational efficiency, it follows that this organization experiences inefficiency.
For scenario 2, assume that different from the founder the organization hires only
one employee. In this case, there are four systems that interact with each other:
the founder himself, the hired employee, the organization (a formal system), and
the environment. The first three systems are input–output systems. Similar to what
is analyzed above, as input–output systems, the founder and the employee have to
19.6 Principle of Organizational Inefficiency 317

interact with the environment, creating unexpected consequences, as indicated by

the subeddy pools in the relevant interacting yoyo fields. So, similar organizational
inefficiencies appear as in the analysis of scenario 1.
Let us next look at a real-life-like case to illustrate how widely useful the principles
established in this chapter are. To this purpose, let us consider a university with the
mission of producing as many graduates who are assets to the society as possible.
First, wrapping around the mission statement, organizational inefficiencies will
naturally appear. In particular, professors, who believe in passing on book knowledge
to students is of the ultimate importance than anything else, will focus on doing so,
while those professors, who believe in cultivating the spirit and desire to succeed in
life in students, will, beyond passing on book knowledge, place additional emphases
on motivating students to work hard and smart in order to achieve personal and career
successes. When students take classes from professors of these different types, these
different professional orientations and emphases will surely create chaos in at least
some of the students. That surely leads to organizational inefficiency.
Second, suppose that the chairperson of one particular department is a faithful
believer of that education only means passing on book knowledge to students. Due
to various reasons at the university level, such as recruiting, third-party university
ranking, etc., the faculty of the department also consists of some believers of that
it is more important to inspire students than simply passing on book knowledge to
students. Now, in order to manage the department efficiently, the chairperson has to be
flexible in his style of management; otherwise, the department will be dysfunctional
in no time.

19.7 Concluding Remarks

By employing the systemic yoyo model as the intuition, this chapter first develops
a systemic view of how personal values are established and evolving with time and
how different people most likely have dissimilar personal value systems. On the
basis of this point of view of personal values, this work looks at the concept of
organizational efficiency and the question of whether or not an organization could
ever be efficient. Then, it is found that inconsistencies between employees’ personal
values and between these personal values and the organization’s mission always lead
to organizational inefficiencies. Based on this result and the underlying analyses, the
principle of management efficiency and the principle of organizational inefficiency
are established. However, what remains unsettled is that in practical situations, how
can one actually become efficient in terms of his/her management styles and how
organizational inefficiencies be improved.
318 19 Some Remarks on Organizational Efficiencies

Professor Bruce Orvis, School of Business, Slippery Rock University, Slippery

Rock PA 16057-1399, USA, e-mail: bruce.orvis@sru.edu, also contributed to the
research work of this chapter.

References

Alvesson, M.: Concepts of organizational culture and presumed links to efficiency. Omega 17(4),
323–333 (1989)
Becker, G.: A theory of social interactions. J. Polit. Econ. 82(6), 1063–1093 (1974)
Becker, G.: A Treatise of the Family. Harvard University Press, Cambridge (1991)
Bergstrom, T.: A fresh look at the Rotten Kid Theorem and other household mysteries. J. Polit.
Econ. 97(5), 1138–1159 (1989)
Burton, R.M., Minton, J.W., Obel, B.: Organizational size and efficiency: an information processing
view. Scand. J. Manag. 7(2), 79–93 (1991)
Cummins, J.D., Weiss, M.A., Zi, H.M.: Organizational form and efficiency: the coexistence of stock
and mutual property-liability insurers. Manag. Sci. 45(9), 1254–1269 (1999)
Forrest, J.Y.L., Orvis, B.: Principles of management efficiency and organizational inefficiency.
Kybern. Int. J. Cyber. Syst. Manag. Sci. 45(8), 1308–1322 (2016)
Fultz, D., Long, R.R., Owens, G.V., Bohan, W., Kaylor, R., Weil, J.: Studies of Thermal Convection
in a Rotating Cylinder with Some Implications for Large-Scale Atmospheric Motion. Meteorol.
Monographs (American Meteorological Society) vol. 21, no. 4 (1959)
Hess, S.L.: Introduction to Theoretical Meteorology. Holt, Rinehart and Winston, New York (1959)
Hide, R.: Some experiments on thermal convection in a rotating liquid. Quart. J. Roy. Meteorol.
Soc. 79, 161 (1953)
Ismail, N., Alhabshi, D.S.O., Bacha, O.: Organizational form and efficiency: the coexistence of
family takaful and life insurance in Malaysia. J Glob Bus Econ (Glob Res Agency) 3(1), 122–137
(2011)
Kasianiuk, K.: White box, black box and self-organization: a system-to-environment approach to
leadership. Kybernetes 45(1), 126–140 (2016)
Laura-Georgeta, T.: Organizational management, efficiency, efficacy and competitiveness. In: Ovid-
ius University Annals, Economic Sciences Series, vol. XI, no. 2, pp. 1261–1264 (2011)
Lin, Y.: General Systems Theory: A Mathematical Approach. Plenum and Kluwer Academic Pub-
lishers, New York (1999)
Lin, Y.: Systemic Yoyos: Some Impacts of the Second Dimension. Auerbach Publication, an imprint
of Taylor and Francis, New York (2009)
Lin, Y., Forrest, B.: Systemic Structure behind Human Organizations: From Civilizations to Indi-
viduals. Springer, New York (2011)
Lindbeck, A., Weibull, J.: Altruism and time consistency: the economics of fair accompli. J. Polit.
Econ. 96, 1165–1182 (1988)
Mathematical Sciences: A Unifying and Dynamic Resource. Notices of the American Mathematical
Society, vol. 33, pp. 463–479 (1985)
Pawłowski, M., Pi˛atkowski, Z., Żebrowski, W.: Management efficiency. Found Manag 1(1), 95–110
(2012)
Ren, J., Xiong, Y.J.: Measurement of enterprise management efficiency based upon information
entropy and evidence theory. Int J Appl Manag Sci 2(1), 93–106 (2010)
Stewart, T.J.: Goal directed benchmarking for organizational efficiency. Omega 38(6), 534–539
(2010)
References 319

Terán, O., Sibertin-Blanc, C., Gaudou, B.: The influence of moral sensitivity on organizational
cooperation. Kybernetes 44(6/7), 1067–1081 (2015)
Villalobos, J., Vargas, M.: Towards a simple dynamical model of citizens’ perception. Kybernetes
44(6/7), 1134–1145 (2015)
Wu, Y., Lin, Y.: Beyond Nonstructural Quantitative Analysis: Blown-Ups, Spinning Currents and
Modern Science. World Scientific, River Edge (2002)
Chapter 20
Dealing with Indecisive Customers
Through Pricing

One characteristic of the market economy is the competition that widely exists in the
marketplace. To continuously find a new profitable market niche, an entrepreneurial
behavior that makes competition increasingly intensive is the key for success in
running business enterprises. So, a natural question arises for anyone who dreams to
acquire success in the business world: What characterizes a market that begets new
opportunities or competition?
To this end, this chapter, based on Forrest and Anderson (2017), attempts to
address the question by investigating a condition under which market competition
will intensify and how the existing loyal consumer base of an established firm will
start to deteriorate. In particular, we will show the following main results: (1) When
the consumer surplus of the market is sufficiently large, market competition will
intensify with additional profits to be made; (2) when there is a large number of
competitors, the number of loyal customers will approach zero; (3) How different
pricing strategies could be employed to potentially double expected profits.
The literature on market competition is quite intensive. For example, Belu and
Caragin (2008) analyze the possibilities that a company has when considering to
enter a foreign market, and the decision that is very important and involves mar-
ket assessment and analysis. Kopalle and Lehmann (2006) develop a two-period
model that determines consumers’ satisfaction in the first period formed on the gap
that exists between the actual quality and the expectations; that satisfaction in turn
impacts the sales of the second period. Siebert (2015) investigates the optimal entry
strategies for firms to use when entering into new or empty markets by considering
how many products of different qualities the firms should introduce. Guo et al. (2012)
addresses the role of downstream market competition under symbiotic production
and demonstrates that incorporating different types of competition in product mar-
kets will partially eliminate the inefficiency as caused by double marginalization.
Chang et al. (2015) consider the effect of market competition on the relationship
that exists between corporate governance and capital structure dynamics. Debruyne
and Reibstein (2005) investigate when incumbents should enter into new market
niches created by technological innovation. Kumar and Sudharshan (1988) investi-
© Springer Nature Switzerland AG 2018 321
J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_20
322 20 Dealing with Indecisive Customers Through Pricing

gate the development of optimal defensive strategies based on an understanding of

the possible reactions of all the defenders to an optimal attack. Similar results are also
obtained by Hauser and Shugan (1983), where different consumer response models
and a different equilibrium assumption are used.
In comparison, conclusions in this chapter add additional results of very different
kinds to the existing literature on market competition and carry that knowledge many
steps forward. In particular, we show the following main results among others:
• How a firm can potentially double its expected profits by being strategic.
• When all incumbent firms could potentially double their profits.
• Under exactly what conditions the market competition will intensify,
• When the existing loyal consumer base of an established firm will start to deteri-
orate.
Additionally, another main contribution of this chapter is that it shows vividly
how the relevant reasoning and intuitive thinking in marketing research can be based
on systems science in general and systemic yoyo model in particular.
The rest of this chapter is organized as follows: Section 20.1 considers the case
when only one firm becomes strategic with its loyal customers and price switchers.
Section 20.2 looks at the case when all firms become strategic. And Section 20.3 stud-
ies which price strategy leads to higher expected profits. And Section 20.4 concludes
this chapter.

20.1 One Strategic Firm

In terms of customer behaviors, indecisiveness is a commonly seen phenomenon

(McWilliams 2004; Mittal et al. 2008; Shin et al. 2012), where customers do not seem
to know exactly what they want in the eyes of the sales people. This end in fact can
be fundamentally explained by using the concept of personal value systems (Forrest
and Orvis 2016), where the indecisiveness is rooted in the fact that differences in
personal value systems lead to varied perceptions about how goods could be used and
services enjoyed because generally no good or service perfectly matches an existing
personal value system and the consequent perception.
At the same time, the systemic yoyo model provides another explanation. Specif-
ically, when the yoyo field of the customer is obligated to absorb whatever necessary
from the environment in order to fill an existing void, since her field is located within
a vast, dynamic ocean of different yoyo pools, she is not exactly sure what she needs
in the first place and then she is not sure where she could acquire whatever that might
fit the need about which she is uncertain.
From the previous two paragraphs, it follows that when a customer is indecisive in
her purchase decision making, there generally is a whole host of reasons for her inde-
cisiveness. So, corresponding to different situations, appropriate sales’ approaches
need to be taken by the sales associates in order to complete successful sales. In the
20.1 One Strategic Firm 323

Figure 20.1 Spinning field

in equilibrium

following, let us focus on the reason of pricing. That is, we assume that the indeci-
siveness of a potential purchase reflects the fact that the particular customer is a price
switcher who makes purchase decisions based on the best price available to her (for
a related study, see Homburg et al. (2012)). To this end, we assume that the oligopoly
market of our concern has m firms, named 1, 2, …, m, which provide horizontally
differentiated products at constant marginal costs, which we set to zero without loss
of generality. Assume that each Firm k has a market share α of the loyal customer
s such that these customers only purchase the product of Firm k provided that the
price is no more than their reservation value, which is set to 1. Let β  1 − mα
be the portion of the switchers who base their purchase decisions on which price is
lower. The risk-neutral managements of these m firms are well aware of the pricing
strategies of the other firms and have established their best responses by playing the
Nash equilibrium through pure self-analyses.

Lemma 20.1. In the symmetric Nash equilibrium, each of the existing firms’
expected profits stays constant and is equal to α, although the firms try to attract
as many switchers as possible.
From the angle of the systemic yoyo model, this result is quite intuitively clear. In
particular, let us imagine the market as the entire spinning yoyo field in Figure 20.1,
while each of the m firms is one of the local eddy leaves so that the m local eddies in
the spinning dish have reached their balance in their movements. Hence, no matter
how hard each of the local eddy spins in order to pull more of the fluid particles into its
field, some of the fluid particles always tend to wonder around the large field without
actually becoming part of any of the local eddies. In other words, as soon as such a
pattern of fluid movement as the one shown in Figure 20.1 is formed and becomes
stabilized, it will be difficult or impossible for the local eddies to grow any larger
(note: The pattern shown in Figure 20.1 was initially discovered by Hide (1953) of
University of Cambridge, England, and then by Dave Fultz and his colleagues of
University of Chicago (1959). The evolution of the pattern has been well observed
and studied by scholars in fluid dynamics and meteorology.)
Proof. Firstly, this game does not have any pure strategy equilibrium, where each
firm’s available strategy is the selling price of the firm’s product. In fact, for any pure
strategy portfolio (x1 , x2 , . . . , xm ), if there is a unique index i ∈ {1, 2, . . . , m} such
that
324 20 Dealing with Indecisive Customers Through Pricing

xi < xj , for j ∈ {1, 2, . . . , m} and j  i, (20.1)

then Firm i has successfully attracted all the price switchers and can therefore slightly
raise its price xi to bring in additional profits as long as the new price still satisfies the
condition in equation (20.1), while all other firms maintain their respective strategy
unchanged. So, the pure strategy portfolio (x1 , x2 , . . . , xm ) is not a Nash equilibrium.
If the set
 
m
I  i ∈ {1, 2, . . . , m}: xi  min{xj }
j1

has cardinality |I | great than 1, then the Firms k, k ∈ I , have absorbed all the price
switchers. Due to the fact that everything in this scenario is set up symmetrically, we
can see that each of these firms would have taken in β/|I | portion of the switcher
segment of the market. So, Firm k’s profits, k ∈ I , are αxk + (β/|I |)xk . So, one of
these firms, say Firm k, can lower its price slightly to xk , satisfying

xk α + β/|I |
> ,
xk α+β

to bring in additional profits by attracting all the switchers. That is, the pure strategy
portfolio (x1 , x2 , . . . , xm ) is not a Nash equilibrium. Similar argument can be used
to show that even when not all firms k ∈ I share the same portion of the price
switchers, then the firm with the fewest price switchers could slightly lower its price
to increasing its profits by attracting all the price switchers. That is, once again, the
pure strategy portfolio (x1 , x2 , . . . , xm ) is not a Nash equilibrium.
Lastly, if for any i ∈ {1, 2, . . . , m}, xi  1, then Firm j, for any chosen j ∈
{1, 2, . . . , m}, can slightly lower its price from the reservation value 1 to anywhere
in the interval (α/(α + β), 1) to produce more profits by taking in the entire segment
of the switchers. So, (1, 1, …, 1) is not a Nash equilibrium, either. That is, the game
we are looking at does not have any pure strategy Nash equilibrium.
Secondly, this game does have a symmetric mixed strategy Nash equilibrium
(Zhou et al. 2015). To this end, let Fi (P) stand for the price distribution of Firm i,
i ∈ {1, 2, . . . , m}. Then, the profits i Firm i generate when the firm sells its product
at price P are


m
 
αP + 1 − Fj (P) βP,
ji

and the objective function of Firm i is

⎧ ⎫
1 ⎨ 
m ⎬
maxFi (P) E(i )  αP + [1 − Fk (P)]βP dFi (P)
⎩ ⎭
0 ji
20.1 One Strategic Firm 325

Firm i can earn α by charging the reservation value 1, because all the firm’s loyal
customers will purchase its product at that price. However, to maximize the profits,
each firm likes to adjust the price P in order to take in as many of the switchers as
possible. At the same time, each firm does not have any incentive to price its product
below the price of α/(α + β), because any price below α/(α + β) will yield profits
less than α despite of attracting all the switchers.
The equilibrium indifference condition for Firm i is


m
α×P+ [1 − Fk (P)]β × P  α × 1, (20.2)
ji

α
for α+β ≤ P ≤ 1, i, j  1, 2, …, m, and i  j. So, the symmetric equilibrium price
distribution is
  m−1
1
(1 − P)α
F(P)  Fi (P)  Fj (P)  1 − , (20.3)
P(1 − mα)

and each firm’s expected profits are

⎧ ⎫
+∞⎨ 
m ⎬
E()  αP + [1 − Fk (P)]βP dF(P)
⎩ ⎭
−∞ ji

1
 αdF(P)
α
α+β

 αF(P)|1 α  α. QED
α+β

What Lemma 20.1 indicates is that if an identical pricing strategy is employed by

all firms, the competition will intensify while the consequent profits do not increase.
So, a natural question is: While holding on to the intensifying competition, can a
firm adopt a different pricing strategy that is additional to the one already used in
the competition at the same time so that it can increase its profits? According to
the following result, the answer to this question is YES, it can be done when the
consumer surplus β is big enough.

Theorem 20.1. Assume that the consumer surplus β of the market place satisfies
β  1 − mα ≥ α, and that other than Firm i no other existing firms plan to market
their products differently. Then the particular Firm i could potentially double its
expected profits by uniformly randomizing its price P over the interval [0, 1] when
dealing with price switchers.

Proof. What is given here implies that Firm i now employs two separate and unrelated
pricing strategies: one, which is that shown in the proof of Lemma 20.1, for its loyal
customers, and the other for price switchers. So, for the sake of communication
326 20 Dealing with Indecisive Customers Through Pricing

convenience, let us theoretically treat Firm i as two separate firms, named Firm i
and Firm ia, where the former stands for the original Firm i that employs the firm’s
originally adopted pricing strategies, while the latter deals only with price switchers
by uniformly randomizing its price P over the interval [0, 1].

From β  1 − mα ≥ α, it follows that α/β ≤ 1. So, for any price P, satisfying

α/β ≤ P ≤ 1, the following function F(P),
  m−1
1
α
F(P)  1 − , (20.4)
βP

represents a well-defined strategy for each of the m existing firms, which satisfies
the following equilibrium indifference condition of Firm k, k  1, 2, …, m,


m
 
α×P+β ×P (1 − P) 1 − Fj (P)  α × 1, (20.5)
jk

which implies that for the m existing firms, their lowest allowed price is α/β.
Next, we show that the expected profits of Firm i are still α, the same as that before
adopting the new pricing strategy for price switchers. In fact, from equation (20.5)
we have
⎧ ⎫
1 ⎨  m ⎬   m−1 1
α
Ei ()  α × P + β × P(1 − P) [1 − F(P)] dF(P) + α ×
⎩ ⎭ β
α/β ji

1   m−1
1
α
 αdF(P) + α  α. (20.6)
β
α/β

For the rest of this proof, it suffices based on equation (20.6) to show that there
is such an opportunity that Firm ia could expect to make at least as much profits as
1
α. To this end, because limP→1− F(P)  1 − (α/β) m−1  F(1)  1, the cumulative
price distribution function F(P) has a jump discontinuity at the reservation value P
1
 1, where the amount of jump is (α/β) m−1 . So, the expected profits of Firm ia are
the following:

α/β +∞
Eia ()  βPdP+ βP[1 − F(P)]m dP
0 α/β

α/β 1   m−1
m
α
 βPdP+ βP[1 − F(P)] dP + β
m
β
0 α/β
20.1 One Strategic Firm 327
⎧
⎪   m−1
m
⎪
m
⎪
⎨
−m α 2
+ m−1 α m−1
+β α
, if m ≥ 3
2(m−2) β m−2 m−11 β
β
   m−1
⎪
⎪
m
⎪ α2
⎩ 2β − α2
ln βα +β α
, if m  2
β β

∂
And because ∂α [Eia () − α] > 0 and when α  1/(m + 1)  β, Eia () − α > 0, it
∗ ∗
follows that there is α ∈ (0, 1/(m + 1)) such that when α ≥ α , Eia () > α. QED

What Theorem 20.1 indicates is that when confronting with indecisive customers,
first find out what is beneath the purchase indecisiveness. If it is the price, then the
firm could use such sales’ strategy that it materially randomizes the selling price
uniformly over the interval [0, 1], where 0 stands for the cost and 1 the reservation
price charged to loyal customers. By doing so, the firm could potentially achieve two
advantages in the market place:
1. Increase its consumer base.
2. Grow its baseline of profits.
At the same time, Theorem 20.1 and its proof imply that if the consumer surplus
β is not taken care of by the existing firms, potentially new competitions will enter
the market. In particular, Firm ia, introduced in the proof of Theorem 20.1, can be
a completely different firm not related to Firm i. When such a situation occurs, it
means a new firm enters into the market to compete with the existing firms while
expecting to potentially make more profits than any of the existing firms.

20.2 All Firms Are Strategic

Although the situation described in Theorem 20.1 would not appear in real life
according to the structure of the spinning field in Figure 20.1, where all the eddy
leaves move and behave in concert, this result still naturally leads to the belief that
many of the existing firms, if not all, would like to adopt a similar strategy as Firm i
in that theorem. To this end, we have the following result.

Theorem 20.2. A sufficient and necessary condition for each existing Firm i, i 
1, 2, …, m, to increase its expected profits by uniformly randomizing its price P
over the interval [0, 1] when dealing with price switchers, while still employing its
mixed pricing strategy at the Nash equilibrium with the loyal customers, is that the
consumer surplus β and the market share α of Firm i satisfy the condition

mm α < (m − 1)m−1 β. (20.7)

Proof. (⇒) Assume that Firm i, i = 1, 2, …, m, can increase its expected profits by
uniformly randomizing its price P over the interval [0, 1] when dealing with price
switchers. Just as what is done in the proof of Theorem 20.1, let us conceptually slip
328 20 Dealing with Indecisive Customers Through Pricing

Firm i into two firms, named Firm i and Firm ia, where the former is run as before
without adopting the pricing strategy of uniformly randomizing its price P over the
interval [0, 1] when dealing with price switchers, while the latter deals with price
switchers only by uniformly randomizing its price P over the interval [0, 1].
The profits of Firm i are given by


m
 
αP + βP(1 − P) 1 − Fj (P)
m

ji

and the objective function of Firm i is

⎧ ⎫
1 ⎨  ⎬
m
 
maxFi (P) E(i )  αP + βP(1 − P)m 1 − Fj (P) dFi (P)
⎩ ⎭
0 ji

The equilibrium indifference condition for Firm i is


m
 
α × P + β × P(1 − P)m 1 − Fj (P)  α × 1. (20.8)
ji

So, the symmetric equilibrium price strategy of each Firm i, i  1, 2, …, m, is

  1
1 α m−1
F(P)  Fi (P)  1 − . (20.9)
1 − P βP

For this strategy to be valid, we must have F(P)  0, for P ≤ P, F(P)  1, for
P ≥ P, and F(P) ≥ 0, for P ≤ P ≤ P, where P and P are some fixed price levels
such that 0 ≤ P < P ≤ 1.
Next, let us see when such price levels P and P can exist. To this end, define
  m−1
1
1 α
h(P)  (1 − P)P m−1 −
β
1
so that F(P)  h(P)/(1 − P)P m−1 . Then it can be readily checked that F(P) is well
defined if and only if h(P) ≥ 0, for P ≤ P ≤ P, where P and P are some fixed price
levels such that 0 ≤ P < P ≤ 1, and that h(0) < 0 and h(1) < 0. Since
1
 P m−1
h (P)  (P −1 − m),
m−1

it can be shown that h(P) reaches its maximum at P  1/m. That is, in order for the
previously mentioned price levels P and P to exist, h(P) must satisfy
20.2 All Firms Are Strategic 329

    m−1
1
1 m−1 α
h  − > 0. (20.10)
m 1
m1+ m−1 β

That is, equation (20.7) holds true.

(⇐) Assume that equation (20.7) holds true. Then the reasoning in the previous
proof can be deducted backward to show that there are price levels P and P, satisfying
0 ≤ P < P ≤ 1, such that F(P), as defined in equation (20.9), is a valid symmetric
pricing strategy for the existing firms i, i  1, 2, …, m, and that F(P) satisfies the
Nash equilibrium indifference condition in equation (20.8). In other words, each
Firm i, i  1, 2, …, m, can be conceptually run as two firms with Firm i employing
the mixed strategy F(P), while the shadow Firm ia deals with price switchers only
by uniformly randomizing its price over the interval [0, 1].
Now, for Firm i, its expected profits E(i ) are
⎧ ⎫
1 ⎨  ⎬
m
 
E(i )  αP + βP(1 − P)m 1 − Fj (P) dFi (P)
⎩ ⎭
0 ji
⎧ ⎫
P ⎨  ⎬
m
 
 αP + βP(1 − P)m 1 − Fj (P) dFi (P)
⎩ ⎭
P ji

P
 αdFi (P)
P

 α Fi (P)|PPP  α, (20.11)

and the profits E(ia ) of Firm ia are

P P
E(ia )  (1 − P) m−1
βPdP + (1 − P)m−1 [1 − F(P)]m βPdP
0 P

1
+ (1 − P)m−1 βPdP, (20.12)
P

where the first and third terms, respectively, represent the expected profits of Firm
ia, when its price is lower than or higher than that of Firms j, j  1, 2, …, m, and the
second term is the expected profits of Firm ia when it is in direct competition with
all other firms, be they real or shadow.
Because the first term in equation (20.12) is given by
330 20 Dealing with Indecisive Customers Through Pricing

P  P
emln(1−P) (mp + 1)
(1 − P)m−1 βPdP  β >0
m2 + m P0
0

while the second term in equation (20.11) is nonnegative and the third term is 0,
because no price switchers exist in that price interval, we conclude that E(ia ) > 0.
So, from equations (20.11) and (20.12), we can conclude that each existing firm can
increase its expected profits by uniformly randomizing its price P over the interval
[0, 1] when dealing with price switchers, while still employing its symmetric mixed
pricing strategy at the Nash equilibrium with the loyal customers. QED
Corollary 20.1. When the competition of a market grows with an increasing number
of firms entering the market, the base of loyal customers for each firm will gradually
diminish.
Proof. From Theorem 20.2, it follows that the consumer surplus β and the market
share α of Firm i satisfy the condition
 
α 1 m 1 1
< 1− → × 0  0, for m → ∞.
β m m−1 e

That is, the ratio /β → 0, as m → ∞. QED

Before we move on to the next result, let us first take a closer look at equa-
tion (20.7). If m  2, (α/β)m2 < 1/4; if m  3, (α/β)m3 < 27 4
; if m  4,
(α/β)m4 < 27/256; in general, we can prove that (m − 1) /m is a decreas-
m−1 m

ing function in m. So, if we look at the market with the concept of evolution in mind,
where the market could be either expanding or shrinking, then the dynamics of mar-
ket competition is described by Theorems 20.1 and 20.2 very well. For example, if
the market is expanding quickly, then that means that the size β of consumer surplus
segment increases quickly when compared to the magnitude of α, the base of loyal
customers of an existing firm. So, the existing firms would have to expand their con-
sumer bases by employing competitive pricing strategies; otherwise, Theorem 20.1
implies that new firms would enter the market to grab the segment of consumer sur-
plus. And, as more firms enter the market, Corollary 20.1 implies that the consumer
bases of the existing firms would dwindle, while the traditional pricing strategies
would have to be replaced by new strategies.
The systemic intuition for this scenario is given in Figure 20.2. Different from the
model in Figure 20.1, the expanding spinning field in Figure 20.2 does not have a
solid periphery. Instead, various elements from the neighborhood, which is relatively
motionless when compared to the center of the field, constantly join into the field,
making the field grow bigger. Another strikingly different characteristic is that the
eddy leaves in Figure 20.2 spin in directions opposite to those in Figure 20.1 due
to the effect of the relatively motionless neighborhood. Now, if the existing eddy
leaves do not grow bigger, more eddy leaves have to appear with the expansion of
the overall field in order to have the overall field occupied due to uneven distribution
of forces.
20.2 All Firms Are Strategic 331

Figure 20.2 Yoyo field

model for an expanding
market

On the other hand, if the market shrinks, it initially means that the segment of
consumer surplus disappears gradually first, followed by the decline of the loyal
consumer bases of the existing firms. That means that as soon as the segment of
consumer surplus shows clear signs of contraction, the existing firms need to inves-
tigate their exit strategies from the market and explore new market options. To this
end, the research on personal values (Forrest and Orvis 2016) comes into play nicely:
Those employees with personal value systems not in total agreement with their firm’s
mission would potentially discover new market directions for the firm, because their
perceptions of the world, which are different from that of the firm, might help the
firm to uncover one or several new markets with enough consumer surplus for the
firm to enter with the potential of making more profits than any of the existing firms
within the newly discovered market. This discussion actually shows that organiza-
tional inefficiencies (Forrest and Orvis 2016), which exist at different time moments
throughout the development of an organization, might not be all bad in the growth
of the organization in terms of evolutions and developments.

20.3 Which Price Strategy Leads to Higher Expected

Profits?

Example 20.1. Let us look at the special case of m  2 when equation (20.7)
becomes 4α < β, where each Firm i is divided theoretically into the real Firm i
and the shadow Firm ia as described in the proof of Theorem 20.2. In particular,
Firm i deals only with its loyal customers by employing the mixed strategy F(P) in
equation (20.9), which is simplified as follows:
α 1
F(P)  1 − · , 0 < P ≤ P ≤ P < 1.
β (1 − P)P

while the shadow Firm ia deals with price switchers only by uniformly randomizing
its price over the interval [0, 1].
332 20 Dealing with Indecisive Customers Through Pricing

√
By solving F(P)  0, we get P  (1− 1 − 4α/β)/2. And because F(P) reaches
its maximum at P  1/2, we have P  1/2. Since F(P) has a mass point of size
4α/β at P  1/2, the expected price of Firm i is
 
P 1 − 1 − 4α α 1− 1− β
4α
1 4α β 2α
Ei (P)  P × dF(P) + ×  − ln  +
2 β 2 β 1 + 1 − 4α β
P β
   
α 1− 1−
4α
1 4α 4α β
 − − 1− − ln  ,
2 β β β 1 + 1 − 4α
β

where the second term on the left-hand side is positive and could be very close to 0
if 4αβ.
Define the following auxiliary function
√
H (x)  x − 1 − x, 0 < x < 1,

then H (x) is an increasing function satisfying limx→1− H (x)  1 and limx→0+ H (x) 
−1. So, there is a unique x ∈ (0, 1) such that H (x)  0, which leads to x 
√ √
−1+ 5
(−1 + 5)/2. In other words, when 2 ≤ 4α β
< 1, Ei (P) > Eia (P)  1/2.
The expected profits of Firm ia are
P P    
1 β 4α 2
E(ia )  (1 − P)βPdP + (1 − P)[1 − F(P)]2 βPdP + 1 −
2 2 β
0 P

P P  2
α2 1 β 4α
 (1 − P)βPdP + dP +
β (1 − P)P 4 β
0 P
⎛  ⎞2    
 
β 1− 1− α2 1 − 1 − β
4α 4α
β 4α β 4α 2
 ⎝ ⎠ 2+ 1− − ln  +
6 2 β β 1 + 1 − 4α 4 β
β
⎧ ⎛  ⎞2   ⎫
⎪   ⎪
⎨ 2 1− 1− 4α
β 4α α 1− 1 − 4α
β 4α ⎬
α ⎝ ⎠ 2+ 1− − ln  +
⎪
⎩ 3(4α/β) 2 β β 1 + 1 − 4α β ⎪⎭
β

√
where −1+2 5 ≤ 4α β
< 1 such that E(ia )|4α/β(−1+√5)/2  0.94409α and
lim4α/β→1 + E(ia ) → 4α/3. So, the continuity of E(ia ) in α/β implies that there
√
is −1+2 5 ≤ P ∗ < 1 such that E(ia ) > α when P ∗ ≤ 4α
β
< 1. QED
What is analyzed
√ in Example 20.1 implies that there is an interval [P ∗ , 1) ⊂
∗
[P, 1)  [(−1 + 5)/2, 1) such that when 4α/β ∈ [P , 1), the expected price Ei (P)
of Firm i is greater than the expected price Eia (P)  1/2 of the shadow Firm ia,
while the expected profits E(i ) of Firm i are less than that of the shadow Firm
ia E(ia ). In other words, what is confirmed in Example 20.1 is that when the two
20.3 Which Price Strategy Leads to Higher Expected Profits? 333

existing firms occupy consumer bases of a good size, they might become content
with what they could make from their loyal customers; the sufficiently large segment
of consumer surplus would conveniently encourage new competitions to foray into
the market with the potential of making more profits than any of the existing firms.
As for the general case of m ≥ 3, let us see how to establish Theorem 20.3 below.

Lemma 20.2. The upper limit P, 0 ≤ P < P ≤ 1, for the pricing strategy F(P) 
  m−1
1
α
1 − 1−P
1
βP
, 0 ≤ P ≤ P ≤ P ≤ 1, as given in equation (20.9), is equal to 1/m.
m   1
And, F(P) has a mass point of size mm−1 βα
m−1
at P  1/m.
m−1

Proof. It is ready to show that

  m−1
1
α 1 − mP
F  (P)  , (20.13)
β 1
(m − 1)(1 − P)2 P m−1 +1
  
satisfying F  m1  0, F  m1 > 0, for P ≤ P < 1/m, and F  m1 < 0, for m1 < P < 1.

Therefore, F m1 reaches a local maximum so that we know P  1/m. And because
  m   1
1 m m−1 α m−1
F 1− ,
m m−1 β

m   m−1
1
α
the pricing strategy F(P) has a mass point of size m m−1
m−1 β
at P  1/m. QED

Lemma 20.3. The lower bound P, as given in Lemma 20.2, satisfies the following
properties:
dP 1
> 0, lim + P  0, and lim P .
d (α/β) α/β→0 α/β→((m−1)m−1 /mm )
− m

Proof. The lower bound P satisfies the equation F(P)  0. So, we have
  m−1
1
1 1 α
P m−1 −P 1+ m−1
 . (20.14)
β

Differentiating this equation with respect to α/β produces

  m−1
1
−1
dP 1  α
· P m−1 P −1 − m 
d (α/β) β
  m−1
1
−1
α
> 0, and P < 1/m (so that P −1 − m > 0), we get
1
Because P m−1 > 0, β
dP
> 0.
d (α/β)
By rewriting equation (20.14), we have
334 20 Dealing with Indecisive Customers Through Pricing

  m−1
1
1 α
P m−1 (1 − P)  , (20.15)
β

So, we obtain
  m−1
1
1 α
lim P  lim + P m−1  lim +  0.
α/β→0+ α/β→0 α/β→0 β

And, equation (20.15) implies that when α/β → ((m − 1)m−1 /mm )− , we have
1 1
(mP) m−1 (1 − P)  1 −
m

so that P  1/m. That is, we have shown limα/β→((m−1)m−1 /mm )− P  1

m
. QED

So, Lemma 20.2 implies that the expected price of Firm i is

1/m m   1
 1 m m−1 α m−1
Ei (P)  P × F (P)dP + ×
m m−1 β
P

1/m m   1
1/m 1 m m−1 α m−1
 PF(P)|P − F(P)dP + ×
m m−1 β
P

1/m
1
1− +P+ [1 − F(P)]dP
m
P

which, according to Lemma 20.3, is greater than the expected price Eia (P)  1/2 of
the shadow Firm ia.
For the expected profits of Firm ia, we have
P 1/m m   1
m m−1 α m−1
E(ia )  (1 − P)m−1 βPdP+ (1 − P)m−1 [1 − F(P)]m βPdP + α ·
m−1 β
1 P

Because ∂E( ia ) > 0, lim α/β→((m−1)m−1 /mm )− P  1 (Lemma 20.3), and E(ia ) →
∂ βα m
! 1/m
1 (1 − P) βPdP + α when α/β → ((m − 1)m−1 /mm )− , it follows that there is
m−1

a γ satisfying 0 < γ < (m − 1)m−1 /mm such that the expected profits E(ia ) of the
shadow Firm ia are greater than the expected profits α of Firm i, when γ < α/β <
(m − 1)m−1 /mm . That is, by combining Example 20.1 and the discussion above we
have shown the following general results.

Theorem 20.3. If each existing Firm i, i  1, 2, …, m, increases its expected profits

by uniformly randomizing its price P over the interval [0, 1] when dealing with price
20.3 Which Price Strategy Leads to Higher Expected Profits? 335

switchers, while still employing its mixed pricing strategy at the Nash equilibrium
with the loyal customers, then there is γ satisfying 0 < γ < (m − 1)m−1 /mm such
that the expected price Ei (P) of Firm i is greater than the expected price Eia (P)  1/2
of the shadow Firm ia, while the expected profits E(i ) of Firm i are less than that
of the shadow Firm ia E(ia ). QED

20.4 Some Final Words

This chapter looks at an issue that is closely related to the very survival of the firm
about how to compete with other firms by finding the market characteristics that
beget new opportunities and competition.
To this end, this chapter establishes several interesting results: a condition under
which market competition will intensify, how the existing loyal consumer base of an
established firm will start to deteriorate, among others. At the same time, this section
shows the importance of systemic thinking and the yoyo model in visualizing the
dynamics of the market competition. In particular, the yoyo model, different from that
of Cartesian coordinate system, provides another platform for scholars and decision-
making managers to think and to imagine before they can rigorously prove what they
first see intuitively.
Other than the fact that the results established in this section can be readily
employed in practice, this work contributes to the existing literature in two noticeable
ways: signs of the market that is ripe for new competition and an innovative tool for
visualization. Based on the development history of science, either natural or social,
being able to visualize is the key for new discoveries and reliable quick decision
making. In this regard, it is very reasonable to expect that this work will play the role
of a flying pebble that will attract many beautiful gemstones in the areas of marketing
research, consumer behaviors, business decisions, etc.
Dr. Melanie Anderson (School of Business, Slippery Rock University, Slippery
Rock, PA 16057, U.S.A.; e-mail: Melanie.anderson@sru.edu) also participated in
the research of this chapter.

References

Belu, M., Caragin, A.R.: Strategies of entering new markets. Rom. Econ. J. 11(27), 1–16 (2008)
Chang, Y.K., Chen, Y.L., Chou, R.K., Huang, T.H.: Corporate governance, product market compe-
tition and dynamic capital structure. Int. Rev. Econ. Financ. 38, 44–55 (2015)
Debruyne, M., Reibstein, D.J.: Competitor see, competitor do: incumbent entry in new market
niches. Mark. Sci. 24(1), 55–66 (2005)
Forrest, J.Y.L., Anderson, M.: Customers’ indecisiveness: a pricing strategy approach. In: Pennsyl-
vania Economic Association 2017 Conference Proceedings, pp. 80–95 (2017)
Forrest, J.Y.L., Orvis, B.: Principles of management efficiency and organizational inefficiency.
Kybern. Int. J. Cybern. Syst. Manag. Sci. 45(8), 1308–1322 (2016)
336 20 Dealing with Indecisive Customers Through Pricing

Fultz, D., Long, R.R., Owens, G.V., Bohan, W., Kaylor, R., Weil, J.: Studies of Thermal Con-
vection in a Rotating Cylinder with Some Implications for Large-Scale Atmospheric Motion.
Meteorological Monographs. vol. 21, no. 4. American Meteorological Society (1959)
Guo, W.C., Lai, F.C., Liu, C.J., Mai, C.C.: Symbiotic production and downstream market compe-
tition. Atl. Econ. J. 40(3), 329–340 (2012)
Hauser, J.R., Shugan, S.M.: Defensive marketing strategies. Mark. Sci. 2(4), 319–360 (1983)
Hide, R.: Some experiments on thermal convection in a rotating liquid. Q. J. R. Meteorol. Soc. 79,
161 (1953)
Homburg, C., Jensen, O., Hahn, A.: How to organize pricing? Vertical delegation and horizontal
dispersion of pricing authority. J. Mark. 76(5), 49–69 (2012)
Kopalle, P.K., Lehmann, D.R.: Setting quality expectations when entering a market: what should
the promise be? Mark. Sci. 25(1), 8–24 (2006)
Kumar, K.R., Sudharshan, D.: Defensive marketing strategies: an equilibrium analysis based on
decoupled response function models. Manag. Sci. 34(7), 805–815 (1988)
McWilliams, G.: Minding the store: analyzing consumers, best buy decides not all are welcome.
Wall Str. J. A1 (2004)
Mittal, V., Sarkees, M., Murshed, F.: The right way to manage unprofitable consumers. Harv. Bus.
Rev. 2–10 (2008)
Shin, J., Sudhir, K., Yoon, D.-H.: When to “fire” consumers: consumer cost-based pricing. Manag.
Sci. 58(5), 932–947 (2012)
Siebert, R.: Entering new markets in the presence of competition: price discrimination versus can-
nibalization. J. Econ. Manag. Strat. 24(2), 369–389 (2015)
Zhou, B., Mela, C.F., Amaldoss, W.: Do firms endowed with greater strategic capability earn higher
profits? J. Mark. Res. 52, 325–336 (2015)
Chapter 21
An Idea on How to Heighten
the Competitive Spirits of Sales
Associates

This chapter, based on Forrest and McCarthy (2017), looks at the problem of sales
associates’ compensation (or sales force compensation). Its importance is manifested
in the fact that competition is one spirited characteristic of the market economy and
continuously finding a new profitable market niche, an entrepreneurial behavior that
makes competition increasingly intensive, is the key for success in running business
enterprises. When looking at the internal operation of a firm, the following question
naturally arises: How should a firm motivate and compensate its sales force to achieve
additional success? This problem has been an active area of research with a great
potential for direct applications in real life. And what is presented in this chapter
attempts to address this problem from a particular angle.
The literature on market competition is quite intensive. For example, Belu and
Caragin (2008) analyze the possibilities that a company has when considering to
enter a foreign market and the decision that is very important and involves market
assessment and analysis. Kopalle and Lehmann (2006) develop a two-period model
to determine consumers’ satisfaction formed on the gap that exists between the actual
quality and the expectations; that satisfaction in turn impacts the sales of the second
period. Siebert (2015) investigates the optimal entry strategies for firms to use when
entering into new or empty markets by considering how many products of different
qualities the firms should introduce. Guo et al. (2012) address the role of downstream
market competition under symbiotic production and demonstrate that incorporating
different types of competition in product markets will partially eliminate the ineffi-
ciency as caused by double marginalization. Chang et al. (2015) consider the effect
of market competition on the relationship that exists between corporate governance
and capital structure dynamics. Debruyne and Reibstein (2005) investigate when
incumbents should enter into new market niches created by technological innova-
tion. Kumar and Sudharshan (1988) investigate the development of optimal defensive
strategies based on an understanding of the possible reactions of all the defenders to
an optimal attack. Similar results are also obtained by Hauser and Shugan (1983),
where different consumer response models and a different equilibrium assumption
are used.
© Springer Nature Switzerland AG 2018 337
J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_21
338 21 An Idea on How to Heighten the Competitive Spirits …

As for the problem of sales associates’ compensation (or sales force compensa-
tion), it represents an area of research with a great potential for direct applications in
real life. And a good number of scholars have contributed to this important research.
For example, Basu et al. (1985) develop a theory of sales force compensation plans,
where the sales of a product depend not only on the sales associates’ effort but also on
the uncertainty in the selling environment. Joseph and Thevaranjan (1998) analyze a
framework that simultaneously examines the role of both monitoring and incentives
in the design of sales force control systems. Lal and Srinivasan (1993) modify the
agency theory approach for understanding sales force compensation plans by incor-
porating the intra-temporal nature of the sales person’s effort-rate decision. Inderst
and Ottaviani (2009) analyze the implications of the inherent conflict between two
tasks performed by direct marketing agents: prospecting for customers and advising
on the product’s “suitability” for the specific needs of customers. Sheth and Sharma
(2008) examine the evolution of sales organizations as a result of the shift from
product-to-service-focused commerce that took place in the past three decades.
Misra and Nair (2011) present an empirical framework to analyze real-world
sales force compensation schemes and report on a multi-million dollar, multi-year
project involving a large contact lens manufacturer in the U.S.A. Frenzen et al. (2010)
develop and empirically test a framework of important drivers of price delegation to
salespeople based on agency-theoretical research and investigate the impact of price
delegation on firm performance. Chan et al. (2014a, b) examine how compensation
systems impact peer effects and competition in collocated sales teams by using
department store sales data and show that compensation systems influence worker
incentives to help and compete with peers within the same firm, which in turn changes
the capability of the firm to compete with rivals.
Lo et al. (2011) use data on individual sales associate compensation contracts to
show that firms design their pay plans to both discriminatingly select (i.e., attract
and retain) salespeople and provide them with the right level of incentives. Daljord
et al. (2016) focus on the constraints faced by firms in fine-tuning contracts to the full
heterogeneity distribution of their employees and explore the implication of these
constraints for the provision of incentives within the firm. Chung et al. (2013) estimate
a dynamic structural model of sales force response to a bonus-based compensation
plan. Kräkel and Schöttner (2016) explain when the compensation plans that are
most common in practice, such as fixed salaries, quota-based bonuses, commissions,
or a combination thereof, are optimal.
In comparison, conclusions in this chapter add additional results of very different
kinds to the existing literature on sales force compensation and carry that knowledge
many steps forward. In particular, the following major conclusions are established:
• How sales associates’ compensation should be aligned with the profit maximiza-
tion of the firm;
• How by revealing private cost information of individual sales associates within
the firm increases the expected incomes of the associates; and
• Specifically how the firm should assign incentive weights to each sales associates
in order to maximize its profits.
21 An Idea on How to Heighten the Competitive Spirits … 339

Additionally, another main contribution of this chapter is that it shows vividly

how the relevant reasoning and intuitive thinking in sales associates’ compensation
can be based on systems science in general and systemic yoyo model in particular.
The rest of this chapter is organized as follows: Section 21.1 establishes an
abstract maximization problem for the sales associates’ compensation of con-
cern. Section 21.2 studies the amounts of sales of individual associates at Cournot
equilibrium. Section 21.3 looks at the associates’ payoffs at Cournot equilibrium.
Section 21.4 investigates the relationship between the associates’ private cost infor-
mation and expected incomes. Section 21.5 specifically formulates how associates’
incentive weights should be determined at the firm level. And Section 21.6 concludes
the presentation of this chapter.

21.1 Sales Associates’ Compensation

Forrest et al. (2017) show that the dynamics of market competition affects firms that
exist within the market or intent to enter the market both externally and internally
in terms of pricing strategies. In other words, although the existing firms might not
grow their expected profits by participating in market competition through adjusting
their prices, they still have to compete in order to prevent new competition from
entering the market by sufficiently reducing the magnitude of the consumer surplus.
So, a natural question arises for the managers of individual firms: How could each
incumbent firm encourage its sales associates to sell more?
To this end, Forrest and Orvis (2016) indicate that sales associates’ compensation
has to be aligned with the associates’ personal value systems in order to promote
organizational efficiency. So, when dealing with the problem of sales associates’
compensation, let us assume that the goal of the associates is to provide quality service
to the consumers while maximizing their personal incomes, and that the mission of
the company is the same: maximize its profits by providing quality products to as
many satisfied consumers as possible. In order to make the goal of maximizing
personal incomes and the goal of optimizing the company profits consistent, let us
consider aligning sales associates’ compensation with the company’s profits [Note:
Our model discussed in this section is inspired by Bagnoli and Watts (2015)].
To address the problem of sales associates’ compensation, let us assume that the
firm of concern has two sales associates, named 1 and 2. The portion of the variable
compensation beyond the fixed base salary of associate i is proportional to

Pi  πi + λi Si , i  1, 2, (21.1)

where πi stands for the portion of the firm’s profits realized by associate i, Si associate
i’s completed sales, and λi > 0 the parameter the firm chooses to maximize its
expected profits.
The fixed base salary for the associates is employed to encourage associate i to
work together with the other associate as a team so that when necessary, one will be
340 21 An Idea on How to Heighten the Competitive Spirits …

willing to provide the needed support to the other in order for the latter to complete
a successful sale. The first term πi in equation (21.1) is employed to encourage
associate i to generate as much profits as possible, while the second term is used
to stimulate more sales, because after all in practice the firm can only observe the
number of completed sales and the amount of profits generated without any effective
way to measure efforts of the associates (Joseph and Thevaranjan 1998). Based on
Forrest et al. (2017), increasing sales are important in terms of bringing in additional
profits with the potential of more than doubling the profits. And together, the sum
of these two terms encourages the associates to complete as many and as profitable
sales as possible. Here, the parameter λi is contractually determined by the firm at
the start of the employment cycle of associate i and is used to get rid of the problem
of free riders.
In this model, we introduced the information/decision structure to investigate the
impact of associates’ efforts in the product market on the profit sharing decision of
the firm. One feature of this model is that the firm offers its associates contracts
that depend on both of the firm’s profits and revenues, delegates sales choices, and
rewards the associates based on revenues and profits as a means of precommitting
to sell more and generate more profits than would be optimal absent delegation and
precommitment. The firm adopts such contracts to specifically change the marginal
benefit/marginal cost calculation of its associates in order to motivate them to sell
more and to generate more profits. The second feature is the assumption that the
incentive contracts chosen by the firm are observable by all parties. This observability
of the contracts simplifies the analysis without changing the qualitative nature of the
results, for details, see Kockesen and Ok (2004), Kockesen (2007).
In the second stage of employment, given the incentive weights λi , i  1, 2, chosen
by the firm, each associate engages in Cournot competition in her sales efforts by
choosing the sales level qi in the product market, i  1, 2, that maximizes her
expected payoffs conditional on all information available to her. We assume that the
risk-neutral associates make their sales decisions to maximize their own expected
payoffs.
Assume that the demand curve of the firm is

p  α − qi − qj , i, j  1, 2, i  j (21.2)

where α is the intercept of the firm’s demand curve. Because both of the associates
work for the same firm, we assume that they sell identical or completely substitutable
products. Both associates know this demand curve, and each of them only knows
their own marginal costs ci of sales, assuming that there are no fixed costs of sales.
Here, we introduce the private information ci to take into account the reservation
utility of associate i.
To complete the description of the model, we assume that the firm and the asso-
ciates have common priors such that all variables are drawn from known distributions
with finite and positive means and variances. We also assume that the marginal costs
ci and cj are independent so that the distribution of ci given cj is independent of cj ,
for i, j  1, 2, i  j. To ensure interior solutions in the second stage of the game, the
21.1 Sales Associates’ Compensation 341

firm’s demand function satisfies the condition that the value of the intercept exceeds
the largest value of each associate’s marginal costs. We seek a perfect Bayes–Nash
equilibrium of the model and begin by solving the second stage of the game, in which
the associates compete in the product market.
So, associate i solves the following maximization problem

max qi E[ πi + λi Si |φi ], (21.3)

where φi represents the information associate i has about ci and cj at the time of her
sales’ choice, πi  the profits generated by associate i  (price − marginal cost of
associate i) ×qi , and Si  the revenue produced by associate i  price × qi .

21.2 Associates’ Sales at Cournot Equilibrium

The maximization problem in equation (21.3) can be rewritten as follows:

    
max qi E (1 + λi ) α − qi − qj qi − ci qi φi . (21.4)

The first-order condition of this maximization problem is

  
1  ci 
qi  E α − qj − φ i . (21.5)
2 2(1 + λi ) 

If both associates know both ci and cj , then the associates play a game of complete
information. To find the Nash equilibrium at this second stage, we solve the following
system of first-order conditions:
⎧  
⎪
⎨ qi  21 α − qj − 2(1+λ ci
i)

⎪ cj
⎩ qj  2 (α − qi ) − 2(1+λj )
1

and obtain

  1 2ci cj
qi ci , cj  α− + , (21.6)
3 1 + λi 1 + λj
 
where qi ci , cj stands for the amount of sale of associate i if she knows both ci and
cj .
If associate i knows ci while associate j knows both ci and cj , then the associates
play such a game that associate i only knows her own private information ci while
associate j has the complete information. To find the Nash equilibrium at this second
stage, we solve the following system of first-order conditions by substituting the
second equation into the first one:
342 21 An Idea on How to Heighten the Competitive Spirits …
⎧    
⎪
⎨ qi  21 α − E qj ci − ci
2(1+λi )

⎪ cj
⎩ qj  2 (α − qi ) − 2(1+λj )
1

and obtain
  
  1 2ci E cj
qi ci , not cj  α− + . (21.7)
3 1 + λi 1 + λj
 
where and qi ci , not cj stands for the amount of sale of associate i if she knows ci
but not cj .
If associate i knows both ci and cj while associate j knows cj , then the associates
play such a game that associate i has the complete information while associate j only
knows her own private information cj . To find the Nash equilibrium at this second
stage, we solve the following system of first-order conditions:
⎧  
⎪
⎨ qi  21 α − qj − 2(1+λ
ci
i)
  
⎪ cj
⎩ qj  2 α − E qi |cj − 2(1+λj )
1

 
where qi  qi not ci , cj represents the amount of sale of associate i if associate
 j
does not know ci while associate i knows both ci and cj , and qj  qj cj , not ci the
amount of sale of associate j if associate j does not know ci , and the formula of the
latter
 is given  by equation (21.7) by using symmetry. Substituting this expression for
qj cj , not ci into the first equation of this system, we obtain
  
  1 E[ci ] E cj ci
qi not ci , cj  α− + − (21.8)
3 2(1 + λi ) 1 + λj 2(1 + λi )

If both associates only know their own private information ci and cj , respectively,
then to find the Nash equilibrium at this second stage, we solve the following system
of first-order conditions:
⎧    
⎪
⎨ qi  21 α − E qj ci − 2(1+λci
i)
  
⎪ cj
⎩ qj  2 α − E qi |cj − 2(1+λj )
1

where qk  qk (not ck , not c ), k,   1, 2, k  , stands for the amount of sale

  of
associate k if she only knows her own private information c . Because E q 
j ci 
       k
E qj not cj , not ci ci  E qj not cj , ci , the symmetry between associates i and j
and equation (21.8) lead to the following:
21.2 Associates’ Sales at Cournot Equilibrium 343

Figure 21.1 Competition

between sales associates
within a firm

     
E qj ci  E qj not cj , ci
     
1 E cj E[ci ] cj
E 2α − + −  
3 1 + λj 1 + λi 2 1 + λj
   
1 5E cj E[ci ]
 2α −  +
3 2 1 + λj 1 + λi

Therefore, we have
  
1 E[ci ] + 3ci 2E cj
qk (not ck , not ck )  2α − + . (21.9)
6 1 + λi 1 + λj

Proposition 21.1. At Cournot equilibrium, the amount of associate i’s sales is an

increasing function of λi and decreasing function of λj .

Proof: The result follows readily from equations (21.6) to (21.9), where each term

with denominator (1 + λi ) is negative, while each term with denominator 1 + λj is
positive. QED

What this proposition says is that when the firm’s demand curve is given in
equation (21.2), the efforts and achievements of associates i and j are limiting and
constraining each other. It is just as what is depicted in Figure 21.1, where the
spinning field represents the operation of the firm of our concern, and when one
of the associates, say Ai , grows larger, then the out of proportional growth of Ai
will definitely affect the development of Aj . In other words, assigning incentives to
revenues generated by individual associates really helps boost sales. In particular,
assume that associate i is provided with an incentive weight λi > 0 while associate
j is not (that is, λj  0). Then, the revenue incentive would cause associate i to
sell more than that of associate j, even when associate j attempts to maximize her
profits. And the demand curve in equation (21.2) implies that associate j would have
to accommodate the effects of the incentives by reducing her amount of sales. As a
result, associate i’s sales would be disproportionally larger than those of associate j.
344 21 An Idea on How to Heighten the Competitive Spirits …

21.3 Associates’ Payoffs at Cournot Equilibrium

Based on what has been established in the previous sections, we have

Theorem 21.1. At Cournot equilibrium, the equilibrium payoffs to associate i are

proportional to

Pi  (1 + λi )qi2 , (21.10)

and the equilibrium profits of the firm are

  πi + πj  qi2 + qj2 , (21.11)

where πi  qi2 , for i  1, 2.

Proof: From equation (21.1), it follows that

  
ci 
Pi  E[ πi + λi Si |φi ]  (1 + λi )E qi p − φi . (21.12)
1 + λi 

For the case when associates i and j know both ci and cj , from equation (21.6),
we have

  ci
Pi ci , cj  πi + λi Si  (1 + λi )qi p −
1 + λi

where p − 1+λ ci
i
 qi . So, we have Pi  (1 + λi )qi2 .
For the case when associate i only knows ci while associate j knows both cj and
cj , we have from equation (21.12) that
  
    ci 
Pi ci , not cj  (1 + λi )qi ci , not cj E p − ci
1 + λi 

       E[ci ]
 (1 + λi )qi ci , not cj α − qi ci , not cj − E qj not cj , ci ci −
1 + λi
  2
 (1 + λi ) qi ci , not cj .

For the case when associate i knows both ci and cj while associate j only knows
cj , we have from equations (21.12) to (21.8)
  
        ci 
Pi not ci , cj  (1 + λi )E qi not ci , cj α − qi not ci , cj − qj cj , not ci −  φi
1 + λi 

      ci
 (1 + λi )qi not ci , cj α − qi not ci , cj − qj cj , not ci −
1 + λi
  2
 (1 + λi ) qi not ci , cj .

For the case when associates i and j only knows their own individual ci and cj ,
respectively, we have from equations (21.12) to (21.9) the following:
21.3 Associates’ Payoffs at Cournot Equilibrium 345
      
Pi not ci , not cj  (1 + λi )qi not ci , not cj · α − qi not ci , not cj

    ci

− E qj not ci , not cj ci −
1 + λi
  2
 (1 + λi ) qi not ci , not cj .

So, what is produced above proves equation (21.10).

For the equilibrium profits of the firm   πi + πj , where because
    2
Pi di , dj  (1 + λi ) qi di , dj
  2   2
 qi di , dj + λi qi di , dj
 πi + λi Si ,
 
for any di ∈ {ci , not ci } and dj ∈ cj , not cj . If we treat the last two expressions as
polynomials in variable λi , then comparing the corresponding coefficients leads to
  2
πi  qi di , dj . Therefore, equation (21.11) is established. QED

21.4 Private Cost Information and Expected Incomes

In terms of the relationship between private cost information of the associates and
their individual incomes, we have

Theorem 21.2. When the associates i and j are Cournot competitors, making the
private cost information ci known to associate j does expectedly increase the incomes
of i (= 1, 2, i  j).
Intuitively speaking, when the firm makes the associates’ private information not
that private among its members, it actually encourages competition between its sales
associates. For example, when the firm spends more on associate Ai in Figure 21.1,
it means the area of Ai half is greater than that of Aj half. That naturally puts pressure
on associate Aj . Now, for Aj , she has two choices: being squeezed out of her job
or measuring up to the challenge. If Aj chooses the former choice, then she will be
simply replaced by another sales associate, which is an optimal personnel decision
for the firm. That is, no matter who particular Aj is, the systemic field structure in
Figure 21.1 always exists in terms of Ai and Aj individually. On the other hand, Aj
does not want her existence to cost too much to the firm, either; otherwise, she runs
the risk of being let go or facing a tougher challenge of a newly hired more capable
colleague. For related studies, see Chan et al. (2014a, b).

Proof: For the situation in hands, it suffices to show that

     
E Pi ci , not cj > E Pi not ci , not cj (21.13)
346 21 An Idea on How to Heighten the Competitive Spirits …

and
     
E Pi ci , cj > E Pi not ci , cj , (21.14)

for i  1, 2. That is, what equations (21.13) and (21.14) imply is that no matter
whether or not associate j’s cost information cj is known to associate i, making her
own cost information ci known to j expectedly leads to greater amount of profits for
associate i. According to Theorem 21.1., instead of equations (21.13) and (21.14), it
suffices to show
  2    2 
E qi ci , not cj > E qi not ci , not cj (21.15)

and
  2    2 
E qi ci , cj > E qi not ci , cj , (21.16)

for i  1, 2. To this end, we have

  2    2 
E qi ci , not cj − E qi not ci , not cj
  2   2 
 E qi ci , not cj − qi not ci , not cj
   
1 E[ci ] 4E ci2
 X − 4X
2
+
9 1 + λi (1 + λi )2
   
1 E[ci ] 7E[ci ]2 E[ci ] 9E ci2
− 4X − 4X
2
+ − 12X +
36 1 + λi (1 + λi )2 1 + λi (1 + λi )2
1   2  1
 E ci − E[ci ]2  var[ci ] > 0,
2(1 + λi ) 2
2(1 + λi )2
   
where X  α + E cj / 1 + λj . That is, equation (21.15) holds true. For equation
(21.16), we have
  2    2    2   2 
E qi ci , cj − E qi not ci , cj  E qi ci , cj − qi not ci , cj
   
1 var cj 7var[ci ]
   + > 0.
9 1 + λj 2 4(1 + λi )2

That is, equation (21.16) holds true. QED

What is implied by Theorem 21.2. is that no matter whether individual sales

associates like or not, the firm needs to make sure somehow its associates know
about other associate’s cost information in order to stimulate more profits and more
revenue.
21.5 Associates’ Incentive Weights 347

21.5 Associates’ Incentive Weights

Next, let us look at the firm’s decision on the incentive weights λi , i  1, 2, at the
first stage of the game. At this stage, the firm chooses λi , i  1, 2, to maximize its
expected profits,
  2    2 
E[]  E[πi ] + E[πi ]  E qi di , dj + E qj dj , di

by taking into account of how the sales associates subsequently find it optimal to
compete in the firm’s product market. In particular, the firm solves the following
maximization problem:
  2   2 
maxλi (di ,dj ),λi (di ,dj ) E qi di , dj + qj dj , di ,

for di  ci and dj  cj , because as the employer the firm knows exactly how much
each sales’ associate would cost itself. The first-order conditions are
   
  ∂qi ci , cj   ∂qj cj , ci
2qi ci , cj + 2qj cj , ci 0 (21.17)
∂λi ∂λi

and
   
  ∂qi ci , cj   ∂qj cj , ci
2qi ci , cj + 2qj cj , ci 0 (21.18)
∂λj ∂λj

Now, equations (21.17) and (21.18) reduce to the following system of equations:
⎧
⎪
⎨ α − 1+λ
5ci 4c
+ 1+λj j  0
i

⎪
⎩α − 5cj
+ 4ci
 0.
1+λj 1+λi

By solving this system we establish the following result:

Theorem 21.3. At the Bayes–Nash equilibrium, the firm would define its incentive
weights by λi  ci /α − 1, i  1, 2. QED
What Theorem 21.3 indicates is that when an associate costs more to the firm,
the firm expects her to generate a greater level of revenue by providing a greater
incentive weight to her. Systemically, this result is clear: If the Ai half in Figure 21.1
costs more to maintain within the overall spinning field, then the Ai half in the three-
dimensional structure in Figure 1.1 in Chapter 1, which is reproduced here for the
convenience of the reader, has to give out more to balance its structural stability.
348 21 An Idea on How to Heighten the Competitive Spirits …

Figure 21.2 Eddy motion

model of the general system
(Figure 1.1 from Chapter 1)

Although the results in this section are established for two sales associates, if we
are willing to get caught with tedious symbolic computations, we will be able to
develop the same results for m many sales associates with their individual incentive
weights (Figure 21.2).

21.6 Remarks

This chapter looks at one important issue facing a retail firm, which is closely related
to the very survival of the firm on how to bring the best out of all sales associates.
By employing systems logic and systemic intuition, this chapter shows some the-
oretically very significant results, such as why firms should somehow make private
cost information of their sales associates known to their respective members and
what incentive weights a firm should give to its individual sales associates in order
to optimize its profits. At the same time, it needs to be pointed that what is done
in this chapter cannot be directly generalized to other situations of workers’ com-
pensation. For example, when individuals’ performance is highly dependent on the
collaboration and support of others, such as the situation with a basketball team,
then what is established above needs to be reconsidered and enriched with relevant
details (Antonietti 2006). The major difference involved here is the following: The
sales associates considered in this chapter are treated mostly as isolated individuals,
whose achievements are not heavily dependent on other sales associates’ support
(that is, the conflict of interest in terms of individual performance is minimal), while
for an athlete on the basketball team, his performance is highly correlated to a host
of factors, including, but not limited to,
• How much the coach allows him to play, or communication between the coach
and the athlete (Rosca 2010);
• Rivalries between teammates on and off the court (Tiago et al. 2016; Arai et al.
2014);
• Teammates collaborations (Jane 2015);
• Commercial endorsements (Fizel et al. 2008);
• How well the athlete could stay healthy.
21.6 Remarks 349

In particular, the basic setup in equations (21.1) and (21.2) needs to be altered
when studying a different setup of employees’ performance. In other words, although
the presentation in this chapter is limited in terms of how widely it can be applied in
real life, it is expected to open up a wide area of study of workers’ compensation for
different employment scenarios.
Mr. Larry McCarthy, School of Business, Slippery Rock University, Slippery
Rock, PA 16057, USA; e-mail: larry.mccarthy@sru.edu, also took part in the research
of this chapter.

References

Antonietti, R.: Human capital, sport performance, and salary determination of professional athletes.
SSRN Electron. J. (2006). https://doi.org/10.2139/ssrn.888535
Arai, A., Ko, Y.J., Ross, S.: Branding athletes: exploration and conceptualization of athlete brand
image. Sport Manag. Rev. 17(2), 97–106 (2014)
Bagnoli, M., Watts, S.G.: Delegating disclosure and production choices. Account. Rev. 90(3),
835–857 (2015)
Basu, A., Lal, R., Srinivasan, V., Staelin, R.: Salesforce compensation plans: an agency theoretic
perspective. Mark. Sci. 4, 267–291 (1985)
Belu, M., Caragin, A.R.: Strategies of entering new markets. Rom. Econ. J. 11(27), 1–16 (2008)
Chan, T.Y., Li, J., Piece, L.: Compensation and peer effects in competing sales teams. Manage. Sci.
60(8), 1965–1984 (2014a)
Chan, T.Y., Li, J., Pierce, L.: Learning from peers: knowledge transfer and sales force productivity
growth. Mark. Sci. 33(4), 463–484 (2014b)
Chang, Y.K., Chen, Y.L., Chou, R.K., Huang, T.H.: Corporate governance, product market compe-
tition and dynamic capital structure. Int. Rev. Econ. Finance 38, 44–55 (2015)
Chung, D.J., Steenburgh, T., Sudhir, K.: Do bonuses enhance sales productivity? A dynamic struc-
tural analysis of bonus-based compensation plans. Mark. Sci. 33(2), 165–187 (2013)
Daljord, Ø., Misra, S., Nair, H.S.: Homogeneous contracts for heterogeneous agents: aligning sales
force composition and compensation. J. Mark. Res. 53(2), 161–182 (2016)
Debruyne, M., Reibstein, D.J.: Competitor see, competitor do: incumbent entry in new market
niches. Mark. Sci. 24(1), 55–66 (2005)
Fizel, J., McNeil, C., Smaby, T.: Athlete endorsement contracts: the impact of conventional stars.
Int. Adv. Econ. Res. 14(2), 247–256 (2008)
Forrest, J.Y.L., Amatucci, F., Markman, G.D.: Dynamics of the firm population within an established
market. In: Proceedings of the 48th annual meeting of the Decision Sciences Institute, Washington,
DC, 18–20 Nov 2017
Forrest, J.Y.L., McCarthy, L.: Sales associates’ compensations: a game theoretic approach. In:
Pennsylvania Economic Association 2017 conference proceedings, pp 52–64
Forrest, J.Y.L., Orvis, B.: Principles of management efficiency and organizational inefficiency. Int.
J. Cybern. Syst. Manag. Sci. 45(8), 1308–1322 (2016)
Frenzen, H., Hansen, A.K., Krafft, M., Mantrala, M.K., Schmidt, S.: Delegation of pricing authority
to the sales force: an agency-theoretic perspective of its determinants and impact on performance.
Int. J. Res. Mark. 27(1), 58–68 (2010)
Guo, W.C., Lai, F.C., Liu, C.J., Mai, C.C.: Symbiotic production and downstream market compe-
tition. Atlantic Econ. J. 40(3), 329–340 (2012)
Hauser, J.R., Shugan, S.M.: Defensive marketing strategies. Mark. Sci. 2(4), 319–360 (1983)
Inderst, R., Ottaviani, M.: Misselling through agents. Am. Econ. Rev. 99(3), 883–908 (2009)
Jane, W.J.: Peer effects and individual performance. J. Sports Econ. 16(5), 531–539 (2015)
350 21 An Idea on How to Heighten the Competitive Spirits …

Joseph, K., Thevaranjan, A.: Monitoring and incentives in sales organizations: an agency-theoretic
perspective. Mark. Sci. 17(2), 107–123 (1998)
Kockesen, L.: Unobservable contracts as precommitments. Econ. Theor. 31(3), 539–552 (2007)
Kockesen, L., Ok, E.: Strategic delegation by unobservable incentive contracts. Rev. Econ. Stud.
71(2), 397–424 (2004)
Kopalle, P.K., Lehmann, D.R.: Setting quality expectations when entering a market: what should
the promise be? Mark. Sci. 25(1), 8–24 (2006)
Kräkel, M., Schöttner, A.: Optimal sales force compensation. J. Econ. Behav. Organ. 126, 179–195
(2016)
Kumar, K.R., Sudharshan, D.: Defensive marketing strategies: an equilibrium analysis based on
decoupled response function models. Manage. Sci. 34(7), 805–815 (1988)
Lal, R., Srinivasan, V.: Compensation plans for single- and multi-product salesforces: an application
of the Holmstrom-Milgrom model. Manage. Sci. 39(7), 777–793 (1993)
Lo, D., Ghosh, M., Lafontaine, F.: The incentive and selection roles of sales force compensation
contracts. J. Mark. Res. 48(4), 781–798 (2011)
Misra, S., Nair, H.S.: A structural model of sales-force compensation dynamics: estimation and
field implementation. Quant. Mark. Econ. 9(3), 211–257 (2011)
Rosca, V.: The coach-athlete communication process: towards a better human resources manage-
ment in sport. Manag. Res. Pract. 2(3), 275–283 (2010)
Sheth, J.N., Sharma, A.: The impact of the product to service shift in industrial markets and the
evolution of the sales organization. Ind. Mark. Manage. 37(3), 260–269 (2008)
Siebert, R.: Entering new markets in the presence of competition: price discrimination versus can-
nibalization. J. Econ. Manag. Strat. 24(2), 369–389 (2015)
Tiago, T., Tiago, F., Faria, S.D., Couto, J.P.: Who is the better player? Off-field battle on Facebook
and Twitter. Bus. Horiz. 59(2), 175–183 (2016)
Chapter 22
Reducing the Adverse Effects
of Speculative Capital

As the final chapter of this book, let us look back at the previous chapters and see
how they are organically associated with each other. First, the three parts present a
sequential order of learning and dynamic development of a new theory. In particular,
Part 1 shows how a cohesive and systematic theory of general systems has been
developed by using set theory, a language that is very versatile and universal. The
results and conclusions of this theory of general systems can be powerfully employed
in any specific area of learning and application, producing meaningful and practi-
cally useful outcomes. Filling a gap existing in the reasoning process of systemic
logic, Part 2 introduces the yoyo model, a badly needed systemic intuition, based on
a holistic analysis of different definitions of the concept of (general) systems. By ref-
erencing the magnificent successes of Euclidean geometry and calculus throughout
the scientific history, Part 2 clearly shows the fact that the cohesive body of results
and conclusions in Part 1 now becomes such knowledge that is useful for resolving
problems in real life and to meet challenges in various areas of human endeavor. That
end is demonstrated by the chapters in Part 3, where various successful applications
in business decision making are presented. Other than presenting another success of
application, this final chapter of the book also demonstrates how the large-scale sys-
tems, studied in Part 1, can be practically used jointly with the systemic yoyo model
of Part 2. In particular, this chapter looks at how to manipulate the flow of specu-
lative hot money and how to introduce monetary/fiscal policies in order to reduce
the severity of the effects of currency attacks. In other words, this final chapter of
the book not only shows how the theory developed in this volume can be employed
for practical purposes, as what has been done in other chapters of Part 3, but also
demonstrates why what is presented in this book is indeed a holistic, organic body
of knowledge.
Based on the theory of how a currency war could be potentially raged against a
nation or a region (Forrest et al. 2013a), the rest of this chapter, which is mainly
based on Forrest et al. (2013b, 2018), develops self-defense strategies that could
conceivably protect the nation or region under siege by using the concept and results
of feedback systems (Lin 1994).
© Springer Nature Switzerland AG 2018 351
J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6_22
352 22 Reducing the Adverse Effects of Speculative Capital

The rest of this chapter is organized as follows: Section 1 introduces a conceptual

way to divide the economy into three different sectors and economic policies accord-
ingly. Section 2 employs the method of large-scale, complex systems, as studied in
Part 1, to separate the economic sectors and economic policies into three relatively
closed systems. Section 3 presents the main theoretical result on when the economic
system modeled by the particular control-theoretical model is stable. Section 4 details
a strategy for national defense against possible currency attacks. Then the chapter is
concluded in Section 5.

22.1 Categorizing Purchasing Power

The following model describes the purchasing power of money within a national
economy:
dP
 k(D − S) (22.1)
dt
where D and S, respectively, stand for the demand for and supply of money, and
P the purchasing power of money, k > 0 is a constant, and t represents time. What
this model says is that when all other variables stay constant, if the money supply S
increases so that S > D, then the purchasing power of the money P decreases.
Now, conceptually divide the overall national economy into three sectors E 1 , E 2 ,
and E 3 as follows: E 1 stands for the goods, services, and relevant production of
these goods and services that are needed for maintaining the basic living standard,
E 2 the goods, services, and relevant productions that are used to acquire desired
living conditions, and E 3 the goods, services, and relevant productions that are used
for the enjoyment of luxurious living conditions. According to Allen and Goldsmith
(1972), the following are four basic requirements for a society to achieve and maintain
stability:
1. Minimum disruption of ecological processes;
2. Maximum conservation of materials and energy, or an economy of stock rather
than flow;
3. A population in which recruitment equals loss; and
4. A social system in which individuals can enjoy rather than feel restricted by the
first three conditions.
So, in terms of economics, for the society to maintain its stability, sector E 1 has to
be kept relatively stable first.
Accordingly divide the overall demand D of money into three corresponding
categories as follows: D1  the demand of money for meeting the minimum
requirement to maintain the basic living standard; D2  the demand of money for
acquiring desired living conditions; and D3  the demand of money for enjoying
luxurious living conditions, so that D  D1 + D2 + D3 . To stabilize the economy, the
22.1 Categorizing Purchasing Power 353

purchasing power of money in category D1 needs to stay relatively constant, while in

D2 increases somehow slightly, and in D3 it should be allowed to increase in order to
attract and trap the additional money supply away from category D1 . So, let P1  the
purchasing power of money in category D1 ; P2  the purchasing power of money
in category D2 ; and P3  the purchasing power of money in category D3 . Similarly,
define S 1  the money supply that goes into category D1 ; S 2  the money supply
that goes into category D2 ; and S 3  the money supply that goes into category D3 .
Now, equation (22.1) is rewritten as follows:
⎧
⎪
⎪ dp1 n
⎪
⎪  k11 (D1 − S1 ) + k12 (D2 − S2 ) + k13 (D3 − S3 ) + q1j xj
⎪
⎪
dt
⎪
⎪
j1
⎪
⎨
dp2 n
 k21 (D1 − S1 ) + k22 (D2 − S2 ) + k23 (D3 − S3 ) + q2j xj (22.2)
⎪
⎪
dt
j1
⎪
⎪
⎪
⎪
⎪ dp1  k31 (D1 − S1 ) + k32 (D2 − S2 ) + k33 (D3 − S3 ) +  q3j xj
⎪ n
⎪
⎩ dt
j1

where kij , and qij are constants, and xj stands for monetary or fiscal policies, i  1,
2, 3, and j  1, 2, …, n. By using matrix notations, equation (22.2) is equivalent to:

Ṗ  Kz + Qx (22.3)
 T
where P  [P1 P2 P3 ]T , Ṗ  dP dt
1 dP2 dP3
dt dt
, K  [kij ]3×3 the coefficient matrix of
the variables (Di − Si ), i  1, 2, 3, Q  [qij ]3×n the coefficient matrix of the variables
xj , j  1, 2, …, n.
Intuitively, the mode in equation (22.3) is really a feedback system, Figure 22.1,
where S is the initial state of the economy. After introducing the monetary and or fiscal
policies x 1 , x 2 , …, x n , the participants of the economy introduce either consciously
or unconsciously a feedback component system S f so that the overall system with
the added feedback produces the desired output P1 , P2 , and P3 . In other words, when
the market participants are allowed to design their own methods (without violating
the established laws) to achieve their individually defined financial successes, the
economy constitutes a “rotational” (or yoyo) field. Here, the word of rotation means
that when a monetary/fiscal policy is introduced, the market participants will find
ways to take advantage of the policy so that the policy and the individually designed
methods jointly produce the individually desired outputs.
354 22 Reducing the Adverse Effects of Speculative Capital

Figure 22.1 How

monetary/fiscal policies and
categorized purchasing
powers are connected

22.2 Economic Sectors Defined Through Feedback

Mechanism

Considering the fact that when economic policies are concerned with, the policy
makers generally only think about how to regulate the economy for the moment.
That is, we have the following linearized model
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
P1 D1 (t) − S1 (t) c1
⎢ ⎥ ⎢ ⎥
⎣ P2 ⎦  A3×3 ⎣ D2 (t) − S2 (t) ⎦ + ⎣ c2 ⎦ (22.4)
P3 D3 (t) − S3 (t) c3

where ci  0 stands for mathematical expectation of noise, i  1, 2, 3. Substituting

equation (22.4) into equation (22.3) produces

R3×3 ż  Kz + Qx (22.5)

Without loss of generality, assume that R3×3 is invertible, which means that in
general the categorized purchasing power of money is completely determined by the
categorized differences of demand and supply of money. Then, equation (22.5) can
be rewritten as follows:

ż  Az + Bx (22.6)

where A  R−1 K and B  R−1 Q. To make equation (22.6) technically manageable,

assume without loss of generality that B is a 3 × 3 matrix, meaning that the mone-
tary/fiscal policies x 1 , x 2 , …, x n are accordingly categorized into three groups: X 1
 the set of all those monetary/fiscal policies that deal with the population meeting
the minimum need to maintain the basic living standard; X 2  the set of all those
monetary/fiscal policies that deal with the population’s need for acquiring desired
living conditions; and X 3  the set of all those monetary/fiscal policies that deal with
the population’s need for enjoying luxurious living conditions. And we will con-
 T
tinue to use the same symbol x to represent the vector X1 X2 X3 of categorized
monetary/fiscal policies.
22.2 Economic Sectors Defined Through Feedback Mechanism 355

Similar to the concept of consumer price index (CPI), introduce an economic

 T
index vector y  y1 y2 y3 such that yi measures the state of the economic sector
Ei , i  1, 2, 3. Then equation (22.6) implies that the national economy of concern
can be modeled as follows:
⎧
⎪
⎨ ż  Az + Bx
y  Cz + Dx (22.7)
⎪
⎩ z(0)  0

where z is the 3 × 1 matrix [D1 − S1 D2 − S2 D3 − S3 ]T , referred to as the state

of the economic system, A, B, C, and D are, respectively, constant 3 × 3 matrices,
satisfying that D is nonsingular (meaning that each introduction of monetary/fiscal
policies does have direct, either positive or negative, effect on the performance of
the economy), and the input space X and output space Y are following:
 
X  Y  r: [0, +∞) → R3 : r is a piecewise continuous function (22.8)

where R stands for the set of all real numbers and R3 the three-dimensional Euclidean
space.
What is just described is that the state of the national economy can be approx-
imately represented by using the state variable z that helps the economy to absorb
the positive and negative effects of the monetary/fiscal policies X 1 , X 2 , and X 3 . The
internal mechanism z of the economy and the monetary/fiscal policies x jointly have
a direct effect on the overall performance y of the economy.
The condition, imposed on the input space X and the output space Y , means
that monetary/fiscal policies, which form the input space X, are introduced based
on the effects of the previously implemented policies, while the overall perfor-
mance, indexes of which constitutes the output space Y , of the economy evolves
from previous states mostly continuous. Historically speaking, such assumptions are
not always true, while failures occur rarely in terms of frequencies. For instance, in
modern China, the so-called Cultural Revolution occurred abruptly (MacFarquhar
and Schoenhals 2008); and several times in the history of Russia, the Tsars had
introduced rushed social reforms, creating a torn country (Huntington 1996).
The geometry of equation (22.7) is depicted in Figure 22.2, where the mone-
tary/fiscal policy input X is split into two portions F 1 F 2 F 3 F 4 and F 5 F 6 F 7 F 8 . The
former directly affects the performance Y of the economy, and the latter is fed into
the initial state of the economy S, leading to the introduction of the feedback compo-
nent system S f . This component system stands for market reactions to the introduced
monetary/fiscal policies and the formation of a feedback loop within the economy.
In real life, this feedback loop constitutes the main body of the economy, while the
overall performance vector Y is merely an artificially designed measure.
Note that within each of the economic sectors E 1 , E 2 , and E 3 , market reactions,
which constitute parts of the feedback component system S f , to monetary/fiscal
356 22 Reducing the Adverse Effects of Speculative Capital

Figure 22.2 Geometry of

the systemic model of a
national economy

policies in general are unique and economic sector specific. That is, market reactions
in one economic sector are different from those of another sector. The loop F 1 F 2 F 3 F 4
is the second stage of the economy after the introduction of new monetary/fiscal
policies. By the joint effect of the market reactions (S f ) and the nonreactionary
aspects (F 1 F 2 F 3 F 4 ), the final numerical readings Y of the economy are produced.
As an example, let us see how monetary/fiscal policy input X could have two
aspects: one is reactionary and the other nonreactionary. Assume that X is such an
economic policy that allows the inflation to inch higher. So, the price of crude oil
is expected to rise. Now, the reactionary aspect of the policy X makes the gradual
and calculated increase in the price of crude oil more or less random. On the other
hand, the expected outcome of the policy is the theoretical, nonreactionary aspect of
the policy. When the actual outcome deviates from the expectation, the difference is
caused by the market reactions to the policy.
According to what is established in Part 1 or (Lin 1994), the system in equa-
tion (22.7) can be decoupled into three independent systems of the same kind with
one-dimensional input and output. To this end, let

S  {(x, y) ∈ X × Y : ∃z ∈ Z such that x, y, z satisfy equation (22.7)}

 the system of all the ordered pairs (x, y) satisfying equation (22.7)

where the state space Z  {r: [0, +∞) → R3 : r is a piecewise continuous function},
and for each i  1, 2, 3, define a system S i as follows:
⎧
⎨ ż  Az + Bi xi
⎪
yi  Ci z + Di xi (22.9)
⎪
⎩
z(0)  0

where Bi is the ith column of B, Ci the ith raw of C, Di a nonzero constant, and the
input space X i and the output space Y i are given as follows:

Xi  Yi  {r: [0, +∞) → R: r is a piecewise continuous function}.

22.2 Economic Sectors Defined Through Feedback Mechanism 357

Then the system S can be decoupled through feedback into the factor systems S i , i
 1, 2, 3, as follows: In particular, if
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
A0 0 B1 0 0 C1 0 0 D1 0 0
⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥
α  ⎣ 0 A 0 ⎦, β  ⎣ 0 B2 0 ⎦, γ  ⎣ 0 C2 0 ⎦, δ  ⎣ 0 D2 0 ⎦.
0 0A 0 0 B3 0 0 C3 0 0 D3

Then, the Cartesian product system S d  S 1 × S 2 × S 3 is represented by the set of

all ordered pairs (x, y) satisfying
⎧
⎨ ż  αz + βx
⎪
y  γ z + δx (22.10)
⎪
⎩
z(0)  0

Because δ is nonsingular, the inverse system Sd−1 is obtained as follows:

Sd−1  (y, x): (y, x) satisfies equation (22.11),

where
⎧
⎪ −1 −1
⎨ ż  (α − βδ γ )z + βδ y
⎪

⎪ x  −δ −1 γ z + δ −1 y (22.11)
⎪
⎩ z(0)  0

The particular feedback component system S f : Y → X used in this decoupling is

given as follows:
⎧      
⎪
⎪
⎪ ż
⎪ A − BD−1 C 0 z BD−1
⎪
⎪  + y
⎪
⎪ ż  0 α − βδ −1 γ z βδ −1
⎪
⎪
⎪
⎨   z 
⎪ x  −D C δ γ−1 −1 + (D−1 − δ −1 )y (22.12)
⎪
⎪ z
⎪
⎪
⎪
⎪  
⎪
⎪ z
⎪
⎩ z  (0)  0
⎪

In terms of economic policy making, decoupling the three-dimensional system S

into the component systems S i , i  1, 2, 3, above implies that when monetary/fiscal
policies are established individually and, respectively, for each of the three economic
sectors E 1 , E 2 , and E 3 , although these economic sector specific policies will most
definitely have joint effects on the economy, there is at least one way to design a
feedback component system S f so that the overall feedback system F(S, S f ), which
358 22 Reducing the Adverse Effects of Speculative Capital

represents the whole system as depicted in Figure 22.1, can be controlled through
adjusting individually each of the economic sectors E 1 , E 2 , and E 3 .

22.3 The Main Result

Lemma 22.1. For any positive definite symmetric matrix W and any constant y > 0,
the following inequality holds true

0
−y ẋT (t + ξ )W ẋ(t + ξ )d ξ ≤ (xT (t) − xT (t − y))W (x(t) − x(t − y)) (22.13)
−y

Proof: Because W is a positive definite symmetric matrix, there is D > 0 such that
W  DT D. Let G  (g1 , . . . , gn ) ∈ Rn ,
G
 1, be a constant vector G of length
one. Then in light of Cauchy inequality, we have:
⎛ ⎞⎛ ⎞
0 0 0 0
⎜ ⎟⎜ ⎟
Dẋ(t + ξ )T Dẋ(t + ξ )d ξ G T Gd ξ ≥ ⎝ G T Dẋ(t + ξ )d ξ ⎠⎝ Dẋ(t + ξ )T Gd ξ ⎠
−y −y −y −y
(22.14)
Therefore, we have
⎛ 0 ⎞⎛ 0 ⎞
0  
1
ẋT (t + ξ )W ẋ(t + ξ )d ξ ≥ ⎝ G T Dẋ(t + ξ )d ξ ⎠⎝ Dẋ(t + ξ )T Gd ξ ⎠
y
−y −y −y
1  
 xT (t) − xT (t − y) DT G · G T D (x(t) − x(t − y))
y
1
 (xT (t) − xT (t − y))W (x(t) − x(t − y)). QED
y

Next, let us consider the following situation with a general polynomial lag
⎧
⎪
⎪ n
⎪ ẋ  Ax(t) + Ai x(t − hi ) + Bw(t) + B1 u(t) (22.15a)
⎪
⎪
⎪
⎪
⎨ i1
n
(22.15)
⎪
⎪ z  Cx(t) + Ci x(t − hi ) + Dw(t) + D1 u(t) (22.15b)
⎪
⎪
⎪
⎪
i1
 
⎪
⎩ x(t)  ϕ(t), t ∈ −h̄, 0

and choose the following V function:

22.3 The Main Result 359

m t m t
V  x Px +
T T
x Qi xdt + (hi − t + ξ )ẋT (hRi )ẋdt (22.16)
i1 t−h i1 t−h
i i

Then the following result holds true:

Theorem 22.1. If the decision matrix M  M 1 + M 2 + M 3 of system in equa-

tion (22.15) satisfies M 1 + M 2 + M 3 < 0, then system in equation (22.15) is stable.

The proof of this result is very technical and is omitted. If interested, please see
Forrest et al. (2018).

22.4 A Strategy for National Defense

Based on the theory developed earlier, this section looks at how a strategy for national
defense against currency warfare can be designed for the case that initial “friendly”
foreign investments turn out to be an aggressive act of war by suddenly withdrawing
all or significant amount from the hosting economy.
Assume that at time t, the total money supply of a specific nation is $X, 10%
of which is issued based on foreign direct investments (FDI). This 10% additional
money supply (due to the foreign investments) helps to increase the moment of
circulation of money, as signaled by rising spending and elevated levels of economic
activities. If these FDIs suddenly leave the nation, more than 10% of the economic
activities of the nation will be affected adversely due to the abrupt cutoff of financial
backing. It is because supplementing the FDIs, there are domestic matching funds
attached too, making the total investments in the national development more than
10% of the economy.
Now, if before FDIs suddenly depart, leaving behind disastrous aftermath, the
monetary authority of the hosting economy is able to keep the exchange rate stable
(currently not all nations in the world are able to do this successfully), while gradually
and strategically increases its money supply, say, to the level of $10X, then in the
entire money supply the proportion of FDIs shrinks to about 1% from the original
10%. And if the money supply had been increased to the level of $100X, then the
proportion of FDIs would have shrunk from the original 10% to about 0.01%, which
is nearly zero. So, if at this moment of time FDIs are suddenly withdrawn as an
aggressive act of war, only around 1% or 0.01% of the consumption and economic
activities of the hosting nation will be adversely affected and the overall health of
the hosting economy will be relatively stable.
At the same time, with drastically increasing money supply (due partially to FDIs
and partially to strategic planning), the hosting economy has to face a mountainous
inflation pressure. To deal with this inevitable inflation pressure, the national gov-
ernment has to work on a wise distribution of the additional money supply so that the
increasing money supply will not cause social upheaval, when FDIs suddenly depart.
360 22 Reducing the Adverse Effects of Speculative Capital

However in reality, any distribution of the extra money supply can easily lead to major
societal instabilities because the rich become richer while the poor become poorer.
The economic and political situation here is similar to that of the normal inflation,
which has been employed to make the structure of the economy more uneven than
before so that the yoyo structure of the economy spins with more strength. So, if the
theory developed earlier in this chapter is adopted to counter the damaging effects of
sudden departure of FDIs, the nation needs to develop a practical plan to distribute
the extra money supply in order to: (1) maintain the economic stability and normal
growth and (2) increase the nation’s economic prosperity no matter when FDIs leave
either slowly or suddenly.
Specifically, the systemic model in the previous section suggests that to help
protect the nation from suffering from potential economic turmoil, the national mon-
etary authority could purposely divide the economy into three sectors E 1 , E 2 , and
E 3 as described earlier so that the sector performance of E 1 evolves as normally as
possible, the sector performance of E 2 could outpace that of E 1 by a large amount,
and critically, most of the additional money supply needs to be trapped in E 3 . The
systemic model of the national economy discussed in the previous sections indicates
that by managing the market reactions correctly, that is the design of the feedback
component system S f , these three economic sectors can be well separated from each
other.
When sector E 1 evolves normally, the citizens would not worry about their basic
living and survival. That naturally leads to the desired societal stability and peace.
Even though prices in sector E 2 increase drastically when compared to those in E 1 ,
that in general would not affect the mood or the happiness of the public, because
desired living conditions change from time to time and vary from one family to
another. In other words, isolated, temporary personally desires, which are not yet
satisfied and which are generally inconsistent with each other and even contradictory
to each other, could not amalgamate into a political force to cause turmoil. The key, of
course, is how to keep most of the additional money supply in sector E 3 . To this end,
let us pay attention to the assumption that a large amount of FDIs entered the nation.
It means that new products and services are expected to and will become available
soon. Because these products and services are brand new to the nation, high prices
can of course be charged for these commercial goods and services. The situation
is similar to the scenarios with developed nations, where the focus has been placed
on innovations, design and production of unprecedented products and services. It is
these never-seen-before products and services that generate the most profits.
For practical purposes, let us see how to classify a product or a service into one of
the three sectors E 1 , E 2 , and E 3 . First of all, goods and services that go into E 1 are
clear. They are such commodities and services that directly relate to the survival of
any human being such as foods, utilities, and shelters of the basic quality. If a family
lives in a rented apartment and wants to own a house, then houses to that particular
family will be in sector E 2 . If the people in a community go to work, shopping,
and entertainment mostly on foot, then individually owned transportation tools, such
as bicycles, motorcycles, cars, private jets, etc., will belong to the economic sector
E 3 . That is, with the elapsing of time, what is in sector E 3 gradually lowers itself
22.4 A Strategy for National Defense 361

into sector E 2 , and what used to be in E 2 also gradually moves into E 1 . That is, the
classification of one particular product or service is a function of time and space. From
the setup of shopping centers, one can see vividly that afore-described classification
of commercial goods and services has been done by merchants throughout the world
without any trouble.
If the hosting nation develops quickly into a super economic power such that FDIs
cannot jointly amount to an influencing percentage in the nation’s domestic consump-
tion and/or economic activities, then the nation will be safe no matter whether any
of the FDIs depart suddenly or not.

22.5 A Few Final Words

There are definitely many different ways a nation under siege by a currency attack
can protect itself. What is presented in this chapter is only one such method designed
by using the concepts of Cartesian product systems, feedback systems, and control
systems.
Other than the method of defense studied in this chapter, there is only another
known method that is established on sound scientific logic and reasoning for the
same purpose. That means that more scholarly efforts are needed for developing a
mature theory on how to defend an economy from financial or general economic
attacks in order to protect those innocent people who work hard in their daily jobs
and try to simply make a living.
Thank you very much for reading my book. I surely hope that you have enjoyed
learning general systems theory presented in this book and its applications in areas of
business decision making. I look forward to hearing from you if you are so generous
that you are willing to share your thoughts, comments, and feedbacks with me. You
can find my e-mail addresses in the preface of this book.

References

Allen, R., Goldsmith, E.: Towards the stable society: Strategy for change. The Ecologist Archive,
The Ecologist, located at http://www.theecologist.info/page32.html. Accessed on 22 May 2012
(1972)
Forrest, J., Hopkins, Z., Liu, S.F.: Currency wars and a possible self-defense (I): how currency wars
take place. Adv. Syst. Sci. Appl. 13, 198–217 (2013a)
Forrest, J., Hopkins, Z., Liu, S.F.: Currency wars and a possible self-defense (II): a plan of self-
protection. Adv. Syst. Sci. Appl. 13(4), 298–315 (2013b)
Forrest, J., Ying, Y.R., Gong, Z.W.: Currency Wars: Offense and Defense Through Systemic Think-
ing. CRC Press, an imprint of Taylor and Francis, New York (2018)
Huntington, S.P.: The Clash of Civilizations and the Remaking of World Order. Simon & Schuster,
New York (1996)
Lin, Y.: Feedback transformation and its application. J. Syst. Eng. 1, 32–38 (1994)
MacFarquhar, R., Schoenhals, M.: Mao’s Last Revolution. Harvard University Press, Cambridge
(2008)
Index

A Basic living standard, 352

Abelian group, 77 Behaviors, 303
Absolute truths, 26 customer, 323
Acting force, 169 Benevolent head, 278
Acupuncture, 189 Big river culture, 7
Adaptability, 259 Bjerknes’ Circulation Theorem, 222, 224, 270
Adaptive advantage, 259 Black box, 10
Additive inverse, 78 Blown-up, 193
Agency theory, 338 non-transitional, 195
Algebra, 31, 187 periodic, 246
abstract, 188 periodic transitional, 199
Altruism, 14, 286 solution, 168, 194
Altruist, 278 theory, 6, 260
Analysis transitional, 194
cross-frontier, 301 Blow-up, 167
data envelopment, 301 Bohr radius, 230
stability, 245 Book of Changes, 7, 191
whole-evolutionary, 245 Box product, 51
Analytical tool, 8, 15 Business decision making, 13
Analytic analysis, 18 Business organization, 12
Anaxagoras’ “seeds”, 28
Ancient medicine, 188 C
Angular momentum, 230 Calculation spill, 168
orbital, 231 Calculus, 7, 188
Aristotelian teleology, 3 Callback services, 258
Aristotle, 3 Cancellation, 149
Asset distribution, 278 Cantor’s set theory, 27
Atom, 7 Capital structure, 258, 321
Atomic world, 229 Cartesian coordinate system, 12
Attractor, 16, 106 Cartesian product, 30, 357
strange, 107 Cascade (connecting) operation, 118
Castle-like environment, 7
B Category, 93
Background radiation, 209 Causality, 5
Base salary, 339 Causal relation, 4

© Springer Nature Switzerland AG 2018 363

J. Y.-L. Forrest, General Systems Theory, IFSR International Series in Systems Science
and Systems Engineering 32, https://doi.org/10.1007/978-3-030-04558-6
364 Index

Cause-effect chains, 29 innovation, 259

Celestial, 232 Customer linking, 259
Chain of object systems, 35 Customers’ indecisiveness, 14
Chaos, 16, 103 Cybernetics, 10
Chinese traditional medicine, 2
Circulation, 222 D
Civilization, 6, 303 Data mining, 14
Collective action, 13 Decision
Collocated sales teams, 338 effective realization of, 300
Commercial endorsements, 348 making, 166, 303
Compensation process, 281
bonus-based, 339 profit sharing, 340
variable, 339 purchase, 322
Competition, 225 style, 259
Cournot, 340 Decision theory, 10
downstream market, 258, 321 Decomposability, 162
dynamics of market, 330 Decomposition, 16
external, 254 Decoupling, 16
free, 243 by feedback, 137
internal, 254 feedback, 137
market, 321, 339 Defensive strategies, 258
Competitive advantage, 14, 165 Demand, 243
readiness, 259 curve, 340
sustainable, 268 function, 341
transient, 268 Demand/supply interactions, 14
Complexity, 6 Descartes’ second principle, 3
large-scale, 166 Desired living conditions, 352
Component, 64 Development of science, 6
Computational uncertainty, 189 Development of technology, 5
Computer simulation, 301 Dialectic structure, 3
Concept of Differential equation, 113, 139
demand and supply, 244 Direct bargaining, 286
energy, 12 Disastrous event, 194
general systems, 25, 226 Disastrous weather forecasting, 12
numbers, 4 Discontinuity, 167, 194, 216
quality, 238 Discontinuous
Tao, 17 singularity, 203
Conflicts, 313 Discrete, 30
Conjectured conclusion, 15 Dishpan experiment, 175, 250
Consumer price index (CPI), 355 Disruptive technologies, 167
Consumption, 278 Disturbance, 210
preference, 281 Divorced father, 293
Contour, 222 Double marginalization, 258
Control function, 82 Duality of rotation, 209
Controllabilities of systems, 16
Coordinated monopoly, 17 E
Corporate governance, 258, 321 Earthquakes, 236
Cournot competitors, 345 Earth’s atmosphere, 234
Covering, 152 Eastern logic, 16
input-output system, 152 Ecological processes, 352
trivial, 153 Economic cycles, 17
Crude oil, 356 Economic disasters, 236
Cultural Revolution, 355 Economic events, 18
Culture Economic phenomena, 18
Index 365

Economic sector, 247 structural, 169, 208

Economy uneven, 169
knowledge-based, 268 uneven eddy, 208
market, 321 whole, 168, 217, 249
national, 352 Evolutionary transition, 209
open-market, 226 Exchange rate, 359
state of, 353 Experience
Eddy accumulated, 259
converging, 171
currents, 212, 216 F
diverging, 171 Factor set, 131
effects, 189, 222 Feedback, 351, 353
field, 172 decoupled by, 137
irregularities, 190, 213 invariance, 121
motion, 169 transformation, 16
source, 6 Feedback component, 353
Educational achievements, 286 Field, 111
Efficiency bond, 181
management, 299 eddy, 171
organizational, 300, 307, 315 meridian, 171
performance, 301 potential pit, 180
scale, 302 potential rampart, 180
technical, 301 Figurative expression, 176
Electromagnetic interaction, 229 Figurative structures, 7
Electron, 227 Final section, 84
Entry Financial setback, 278
strategies, 321 First push, 209, 216
timing, 258 Fluid motions, 7
Environment, 174 Force, 170, 215
Epimorphism, 148 Coriolis, 223
Epistemology of natural sciences, 190 electromagnetic, 229
Equal-quantitative effect, 189, 212 gradient, 215
Equal quantitative movements, 189, 212 materialism of, 209
Equation pressure gradient, 223
discriminant of, 245 static electrical, 227
Euler, 223 stirring gradient, 225
evolution, 202 twisting, 215, 225
kinematic, 208 uneven arms of, 172
nonlinear evolution, 200 Formal
quasi-linear hyperbolic, 211 analysis, 211
Riccati, 196 logic reasoning, 18
Equilibrium Formalism, 6
Bayes-Nash, 341, 347 Foundations of mathematics, 12
Cournot, 343 Free sum, 45
demand-supply, 244 Fronts, 282
Nash, 262, 264, 341 Fundamental elements, 28
Error-value spiral, 198 Fundamental particles, 28
Euclidean geometry, 188 Fuzzy systems, 32
Euclidean space, 166, 355
Evolution G
model, 6 Galilean conception, 4
nonlinear, 213 Galileo’s method, 3
of competitions, 243 Game theory, 10, 249
of the universe, 235 Generalized Lenz’s Law, 178
366 Index

General system, 16, 30 noise, 237

Geological time measurement, 236 Informational infrastructure, 8
Geometry, 187 Initial condition, 190
Globalization, 269 Initial section, 84
Gradient, 170, 215 Innovation, 226
Grand unification, 229 business model, 259
Gravitation, 227 technological, 321
collapse, 235 Input space, 45, 103
deflection, 235 Integrity, 253
Grey systems theory, 11 Interaction between systems, 17
Growth Interaction of demand and supply, 17
explosive, 198 Inter-firm hostility, 249
G-system, 33 Internal structural solos, 268
Gustav Fechner, 3, 30 Internal structure
unevenness of, 170
H International tariffs, 258
Hegel, 3 Internet of knowledge, 5
Hierarchy, 69 Internet telephones, 258
a-type, 69 Inverse limit, 93
treelike, 75 1-level, 100
Homomorphism, 42 Investment
n –level, 42 foreign, 359
Honesty, 238 foreign direct, 359
Household economics, 277 Invisible particle, 7
Hubble Irregularities
age, 229 structural, 216
constant, 228 Isomorphism, 149
distance, 229
Human behavior, 18 J
Human thought, 26 Jet stream, 282

I K
Identity, 253 Keynesian income analysis, 2
Incentive weight, 338
Indecisive customers, 17 L
Indecisiveness, 322 Laboratory experiment, 18
Indeterminacy, 168, 203 Land-ocean breeze, 224
Inefficiency, 299, 311, 321 Landscape of knowledge, 11
Inflation, 356 Lao Tzu, 213
pressure, 359 Large number hypothesis, 227
Information Large-scale systems, 16
content of, 227 Lattice, 32
entropy, 301 Law of conservation, 8, 226
highway, 5 of informational infrastructure, 235
irregular, 216 Laws of motion, 7
know-how, 273 Laziness, 286
private cost, 338 Lazy rotten kids, 279
small probability, 216 Left-hand rule, 181
structural, 229 Leibniz, 3
theory, 10 Leisure, 306
Informational Leucippus–Democritus atom, 28
infrastructure, 234 Linear equation, 113
Index 367

Linear space, 111, 114 scale, 235

Living things, 2 world, 228
Luxurious living conditions, 352 Mind, 12, 305
Minimum cost, 5
M Minkowski space, 233
Management Mission, 300
approaches, 313 Mixed strategy, 262
efficiency, 301, 313 Model
flexibility, 313 business, 258
goals, 302 dynamic structural, 338
modern, 300 for an infinite universe, 235
performance, 301 linear elasticity, 202
strategic, 258 nonlinear elasticity, 202
Management efficiency, 14 quadratic evolution, 205
Manufacturing production, 6 Moments of forces, 170
Mapping, 41 Monetary authority, 360
from hierarchies into hierarchies, 74 Money
identity, 61 demand for, 352
limit, 97 purchasing power of, 352
linkage, 69 speculative hot, 351
product, 147 supply of, 352
restriction, 41 transfer, 283
s-continuous, 44 Monomorphism, 150
similarity, 75 Moral codes, 253
Marginalization, 321 Morphism, 93
Market Morphological change, 209
analysis, 321 Multilevel structure, 26
assessment, 321 Mutual forbearance, 249
competition, 14, 337 Mutual restrictions, 246
consumer, 269 Mystery of nonlinearity, 224
coordinately monopolized, 260
dynamics of, 260 N
foreign, 321 N-ary operation, 31
local, 272 National defense, 359
niches, 258 National economic security, 166
oligopoly, 249 Negative news, 268
regional, 272 Never-perfect value system, 17, 290
sensing, 258 Newtonian physics of motion, 12
Market-driven orientation, 258 Newton’s second law of motion, 5
Marketing, 264 Nicholas of Cusa, 3
agents, 338 Nightlight controversial, 287
Marx, 3 n-level similarity, 42
Mass production, 247 Nonlinear evolution, 6
Mathematical indeterminacy, 168 essential characteristics of, 197
Mathematical law, 4 fundamental characteristics of, 203
Maximum economic effect, 5 Nonlinearity, 208, 246
Merit goods, 286 Nonlinear mutual reactions, 208
Method Nonlinear space, 169
nonparametric frontier efficiency, 301 Number, 166
of linearity, 194
Methodology O
management, 273 Object, 29
of systems science, 10 isolated, 60
Microscopic Object set, 30
368 Index

Ocean of eddies, 170 free riders, 340

One-element Ionians, 27 household incentive, 281
Open market, 243 many-body, 29
Operator principal-agent, 246
shift, 124 three-body, 4, 176
Organization Prodigal Son, 289
culture of, 258 Product, 48
strategy of, 258 Production
Organizational level of, 247
adaptation, 259 theory, 2
behaviour, 18 Profit
culture, 302 maximization, 338
design, 302 opportunities, 247
efficiency, 307, 315 Projection, 131
form, 301 Properties of systemhood, 10
inefficiency, 17, 315 Property
performance, 302 of additivity, 11
size, 301 transferable, 132
structure, 4, 301 Proton, 227
Overall development, 249
Q
P Quantitative infinity, 203
Panattractor, 110 Quantum mechanics, 7, 227
Panchaos, 110 Quark structure, 174
Panrelativity theory, 213 Quasi-equal acting force, 8
Pansystems, 28 Quasi-equal mutual constraints, 189, 246
Parasite, 278 Quiet and unspoken evaluation, 238
Pareto optimum, 278
Particle assumption, 208, 213 R
Partition Reading nightlight, 307
input-output system, 152 Reciprocal impressions, 238
trivial, 153 Reciprocating movement, 209
Pattern Reductionism, 11
harmonic spin, 247 Relation, 27
inharmonic, 247 equivalence, 42
Peer effects, 338 length of, 30
Period, 78 mass-energy, 234
Personal value, 253 modeling, 147
Philosophical assumption, 253, 300 n-ary, 30
Polarization, 179 set, 29, 30
Policies Relational systems, 30
economic, 354 Relativity effects, 235
fiscal, 353 Relativity theory, 7
monetary, 351 Reliability of information, 18
Prediction, 190 Representation
financial market, 190 complex system, 137, 145
science, 194 global system, 146
Price delegation, 338 injective complex system, 156
Pricing strategies, 14 isomorphic complex system, 156
Principle surjective complex system, 156
management efficiency, 313 Resourceful” father, 293
organizational efficiency, 313 Retail, 348
Private information, 17 Reversal change, 7
Problem Riemann ball, 168, 204
Index 369

Rigorous tool of reasoning, 14 analysis, 203

Rivalry, 249 tool, 15
Rotten kid theorem, 17, 277 Strong interaction, 229
Russell’s paradox, 29 Structure, 27, 31
figurative, 212
S information/decision, 340
Sales associates, 17 internal, 169
compensation, 337 of layers, 26
contracts, 338, 340 spatial, 226
effort, 338 S-truth, 29
Sales force, 337 Subsystem
compensation plans, 338 dense, 63
compensation schemes, 338 Supply, 243
control systems, 338 intensities of, 244
response, 338 Supply chain, 178
Samaritan’s dilemma, 278, 295 Support, 57
Satisfaction Symbolism, 7
consumers’, 321 System, 2
Saturation constant, 244 Cartesian product, 49
Science celestial, 231
first dimension, 167 centralizable, 42
second dimension, 167 centralized, 36
Scientific revolution, 4 collectively controllable, 83
Second stir, 208, 216 complex, 166
Self-defense, 351 component, 132
Selfish kids, 277 connected, 57
Silent communication, 237 maximal, 64
Single-relation system, 16 control, 361
Singularity, 169 controllable, 82
discontinuous, 197 decoupled through feedback, 357
problem, 6 designer, 302
Small probability, 216 disastrous weather, 190
events, 171, 216 discrete, 30
information, 216 equal, 33
Solar system factor, 132
mystery of, 231 feedback, 114
stability of, 231 feedback component, 114
Spatial dynamics, 169 functional, 114
Speculative capital, 166 general, 9
Speed of communication, 5 general dynamic, 207
Speed of light, 228 general feedback, 119
Spin general linear, 114
concept of, 173 input–output, 45, 103
field, 171 isomorphic, 152
Spinning 1-level Cartesian product, 51
currents, 210 1-level product, 51
inwardly, 216 linear, 111
outwardly, 216 linear time, 124
pool, 170 causal, 126
Star neighborhood, 59 precausal, 126
State of motion strongly stationary, 125
first Law on, 182 time invariably realizable, 127
fourth Law on, 187 linked, 70
second Law on, 183 MT-time, 123
third Law on, 185 noninteracted, 132
Statistical nontrivial, 36
370 Index

nth-level object, 34 Unequal quantitative movements, 213

open, 166 Uneven internal structure, 6
organizational, 254 Unevenness, 208
original, 114 internal structural, 213
over a group, 77 structural, 212
overall, 132 Uni-directionality, 210
partial, 34 Unification
periodic, 79 of mass and energy, 234
quotient, 42 of space, 233
similar, 42 structural, 235
subsystem, 34 Unified atmospheric dynamics, 236
symmetric, 104 Universal
time, 77 scale, 230
value, 254, 302 truth, 29
v-controllable, 82 Universal mapping property, 93
weather, 193 Universe, 228
Systemic intuition, 15 age of, 229
System-problem solving, 167 radius of, 229
System’s efficiency, 10 Utility
Systems methodology, 4 reservation, 340
Systems movement, 12 Utopian project, 4
Systems science, 2
Systems thinking, 13 V
Value
T personal, 300
Talent, 313 system, 14
Teleological cosmos, 4 Variable
Theory of general systems, 25 quantitative, 166
Thread, 70
Time W
set, 77 War
uneven, 169 aggressive act of, 359
Time–space distribution, 169 currency, 351
Topological space, 31 Wave motion, 210
Torrential rains, 236 Wealth of shareholders, 258
Total quality management, 258 Well-posedness, 199
Transformation, 78 Western thinking, 16
feedback, 114 White box, 10
heat-kinetic energy, 213 Whole evolution, 194
implicit, 169, 218 World commerce, 269
left, 78 World economy, 170
of rotations, 216
right, 78 Y
spatial dynamic, 204 Yellow Emperor, 2
Transitional change, 7, 167, 194 Yoyo, 8, 166
Traveling vortices, 282 dipoles, 179
3-quark, 174
U Yoyo field, 16
Uncertainties, 193
computational, 214 Z
in the selling environment, 338 Zermelo–Russel paradox, 34
measurement, 213
Unequal-quantitative effect, 7