Experience Seamless Full Ebook Downloads for Every Genre at textbookfull.

com

Applied Evolutionary Algorithms in Java Ghanea

Hercock Robert

https://textbookfull.com/product/applied-evolutionary-
algorithms-in-java-ghanea-hercock-robert/

OR CLICK BUTTON

DOWNLOAD NOW

Explore and download more ebook at https://textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Polymorphism in Java: Methods and polymorphic algorithms

applied to computer games 1st Edition Privitera

https://textbookfull.com/product/polymorphism-in-java-methods-and-
polymorphic-algorithms-applied-to-computer-games-1st-edition-
privitera/
textboxfull.com

Stable Mutations for Evolutionary Algorithms Andrzej

Obuchowicz

https://textbookfull.com/product/stable-mutations-for-evolutionary-
algorithms-andrzej-obuchowicz/

textboxfull.com

Genetic Algorithms in Java Basics Jacobson Lee Kanber

Burak

https://textbookfull.com/product/genetic-algorithms-in-java-basics-
jacobson-lee-kanber-burak/

textboxfull.com

Beginning Java with WebSphere Expert s Voice in Java

Janson Robert W

https://textbookfull.com/product/beginning-java-with-websphere-expert-
s-voice-in-java-janson-robert-w/

textboxfull.com
Metaheuristic and Evolutionary Computation: Algorithms and
Applications Hasmat Malik

https://textbookfull.com/product/metaheuristic-and-evolutionary-
computation-algorithms-and-applications-hasmat-malik/

textboxfull.com

Optimization Using Evolutionary Algorithms and

Metaheuristics: Applications in Engineering 1st Edition
Kaushik Kumar
https://textbookfull.com/product/optimization-using-evolutionary-
algorithms-and-metaheuristics-applications-in-engineering-1st-edition-
kaushik-kumar/
textboxfull.com

Java Closures and Lambda Fischer Robert

https://textbookfull.com/product/java-closures-and-lambda-fischer-
robert/

textboxfull.com

Applied Fluid Mechanics Robert L. Mott

https://textbookfull.com/product/applied-fluid-mechanics-robert-l-
mott/

textboxfull.com

Java pocket guide Fourth Edition Robert Liguori

https://textbookfull.com/product/java-pocket-guide-fourth-edition-
robert-liguori/

textboxfull.com
Applied Evolutionary Algorithms
in Java
Springer Science+Business Media, LLC
Robert Ghanea-Hercock

Applied Evolutionary
Algorithms in Java

With 57 Illustrations

" Springer
Robert Ghanea-Hercock
BTexact Technologies
lntelligent Systems Laboratory
Main Lab Block, First Floor, pp12
Adastral Park
Martlesham Heath, lpswich IP5 3RE
UK
[email protected]

Library of Congress Cataloging-in-Publication Data

Ghanea-Hercock, Robert.
Applied evolutionary algorithms in Java / Robert Ghanea-Hercock.
p. cm.
Includes bibliographical references and index.
ISBN 978-1-4684-9526-3
1. Evolutionary programming (Computer science) 2. Genetic algorithms. 3. Java
(Computer program language) 1. Title.
QA76.618 .G453 2003
005.I--dc21 2002042740

Printed on acid-free paper.

Additional material to this book can be downloaded from http://extra.springer.com.

ISBN 978-1-4684-9526-3 ISBN 978-0-387-21615-7 (eBook)

DOI 10.1007/978-0-387-21615-7

© 2003 Springer Science+Business Media New York

Originally published by Springer-Verlag New York, Inc. in 2003
Softcover reprint of the hardcover 1st edition 2003
Ali rights reserved. This work may not be translated or copied in whole or in part without the
written permission of the publisher ( Springer Science+Business Media, LLC ),
except for brief excerpts in connection with reviews or scholarly analysis. Use
in connection with any form of information storage and retrieval, electronic adaptation, computer
software, or by similar or dissimilar methodology now known or hereafter developed is forbidden.
The use in this publication oftrade names, trademarks, service marks, and similar terms, even if
they are not identified as such, is not to be taken as an expres sion of opinion as to whether or not
theyare subject to proprietary rights.
Authorization to photocopy items for intern al or personal use, or the internal or personal use of
specific clients, is granted by Springer-Verlag New York, Inc., provided that the appropriate fee is
paid directly to Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, USA (Tele-
phone (508) 750-8400), stating the ISBN number of the volume, the volume ti tie, and the first and
last page numbers of each chapter copied. The copyright owner's consent does not include copying
for general distribution, promotion, new works, or resale. In these cases, specific written permisison
must first be obtained from the publisher.

987 6 5 4 3 2 I SPIN 10891089

www.springer-ny.com

Springer-Verlag New York Berlin Heide1berg

A member of BertelsmannSpringer Science+Business Media GmbH
Preface

The Quantum Weather Butterfly (Papilio tempestae) is an undistinguished

yellow colour. Its outstanding feature is its ability to create weather. This
presumably began as a survival trait, since even an extremely hungry bird
would find itself inconvenienced by a nasty localized tornado.

Terry Pratchett, Interesting Times

The principle of adaptive evolution is strongly intuitive, as we all perceive its

effec~ in the natural world and the Darwinian perspective has permeated
common culture. Computing researchers working in the fields of Artificial
Intelligence and Machine Learning realised over 30 years ago that such a
principle could be simulated in software, and hence exploited as a tool for
automating problem solving. The breakthrough in understanding how to apply
evolution is normally credited to the group led by John Holland and their
seminal work on the Genetic Algorithm, in 1975 at the University of Michigan.
However, it has taken several decades for evolution-based methods to
become an accepted element of computing techniques. Part of the reason for this
is the relative complexity of the theories underlying evolution-based algorithms,
which are still poorly understood.
This text is aimed at providing the reader with a practical guide to applying
evolutionary algorithms to engineering and scientific problems. The concepts
are illustrated through clear examples, ranging from simple to more complex
problem domains.
A major aspect of this text is the use of a Java toolkit for exploring Genetic
Algorithms, with examples in digital image processing and mobile robot control.
A hands-on approach is encouraged throughout the text, in order to develop an
intuitive understanding of evolution-driven algorithms. The toolkit provides an
easy-to-use visual interface, with integrated graphing and analysis tools, which
aid in the task of visualising the underlying processes. Full code for the toolkit is
available from the author's web site, where regular updates can be found
(http://www.cybernetics.org.uk), and on the included CD-ROM.
Chapter 1 provides a basic introduction to the history and background of
evolution-based algorithms. Chapter 2 covers some basic but essential
background material and principles from biological evolution. Chapters 3 and 4
cover the theory and operation of Genetic Algorithms and Genetic
vi Preface

Programming, respectively, in greater detail, as these represent the main

techniques in common use. Chapter 5 provides a detailed pair of examples from
the engineering domain in image processing and mobile robot control. Chapters
6 and 7 finish with a review of future directions in the field and some
conclusions on the benefits of evolution-based processing. Appendix A provides
a listing of some useful web-based references and software. Appendix B
describes the specific genetic algorithm software used in the text and the Eos
evolution toolkit from the Intelligent Systems laboratory at BTexact
Technologies. Appendix C provides some essential background knowledge in
Fuzzy Logic, which is required for the mobile robot control example. Appendix
D is a detailed user guide for the excellent mobile robot simulator, which has
been generously contributed by Gary Lucas.
This book is intended for anyone interested in evolution-based software at
the graduate and postgraduate levels. No mathematics beyond basic algebra and
Cartesian graph methods is used, as the aim is to encourage use of the Java
toolkit in order to develop an appreciation of the power of these techniques.
Several further texts are referenced that provide extensive coverage of the theory
and mathematical basis of genetic operators. The author welcomes suggestions
and feedback on the text and software.

Description and Use of CD-ROM

The accompanying CD-ROM in the back cover of the book contains the Java
code for the example applications from Chapter 5. It also contains a selection of
associated open-source Java code libraries for Evolutionary Applications and
robot simulators from other sources. In addition the papers directory contains a
selection of useful papers on Evolutionary Computing. Installation instructions
are contained in the CD file ReadMe.txt.

Acknow ledgements

I would like to acknowledge the following people who for various reasons were
involved in contributing to the creation of this book. I must particularly thank
Gary Lucas, the author of the Rossum mobile robot simulator, which forms the
basis for the second application in Chapter 5, as it was ideally suited to the
purpose of this text to promote Java and evolutionary algorithms in graduate
studies. Thanks also go to members of the Future Technologies group at
BTexact Technologies - in particular to Richard Tateson, Rob Shipman and
Paul Marrow for contributions, Erwin Bonsma for the Eos toolkit and appendix
material, and Mark Shackleton for allowing me time to work on this project.

Special thanks also go to my parents, Joyce and Derek Hercock. Finally,

thanks to my wife for tolerating my obsessive interest in artificial evolution.
Contents

Preface ........................................................................................................... v

1 Introduction to Evolutionary Computing ............................................... 1

1.1 Evolutionary Computation ................................................................... 1
1.2 History of Evolutionary Computing ..................................................... 2
1.3 Obstacles to Evolutionary Computation ............................................... 3
1.4 Machine Learning ................................................................................. 3
1.5 Problem Domains ................................................................................. 4
1.5.1 Search Spaces ................................................................................ 4
1.5.2 Optimisation Versus Robustness.................................................... 6
1.5.3 Expert Systems and AI ................................................................... 6
1.5.4 Fuzzy Logic .................................................................................... 7
1.5.5 Bayesian Networks ........................................................................ 9
1.5.6 Artificial Neural Networks ........................................................... 10
1.5.7 Feedforward Networks ................................................................ 11
1.6 Applications ........................................................................................ 12
1.6.1 Problems ...................................................................................... 13
1.7 Evolution-Based Search ..................................................................... 14
1.7.1 Languages for Evolutionary Computing ...................................... 14
1.7.2 C and C++ .................................................................................. 14
1.7.3 Pascal and Fortran ...................................................................... 15
1.7.4 Visual Basic ................................................................................. 15
1.7.5 Java ............................................................................................. 15
1.7.6 Object-Oriented Design ............................................................... 16
1.7.8 Java and 00 Design .................................................................... 16
1.8 Summary ............................................................................................ 17
Further Reading ................................................................................... 18
2 Principles of Natural Evolution .............................................................. 19
2.1 Natural Selection ................................................................................ 19
2.1.1 Genes and Chromosomes ............................................................ 19
2.1.2 Biological Genes .......................................................................... 20
viii Contents

2.2 DNA Structure .................................................................................... 20

2.2.1 Transcription -from DNA to RNA .............................................. 21
2.2.2 Translation -from RNA to Protein ............................................. 21
2.2.3 Genotype ...................................................................................... 21
2.2.4 No Lamarckianism! ..................................................................... 22
2.2.5 Evolution and Variation .............................................................. 22
2.2.6 Redundancy ................................................................................. 22
2.2.7 Self-Maintenance ......................................................................... 22
2.2.8 Evolvability .................................................................................. 23
2.2.9 Mutation ...................................................................................... 23
2.2.10 Sexual Recombination ............................................................... 24
2.2.11 Nonselectionist Issues ................................................................ 24
2.2.12 Epigenesis .................................................................................. 24
2.2.13 Dynamics and Morphogenesis ................................................... 25
2.3 Summary ............................................................................................ 25
Further Reading ................................................................................... 26
3 Genetic Algorithms .................................................................................. 27
3.1 Genetic Algorithms ............................................................................ 27
3.2 GA Basics ........................................................................................... 27
3.2.1 Fitness and Evaluation Functions ............................................... 27
3.3 GA Theory .......................................................................................... 30
3.3.1 Deception ..................................................................................... 31
3.3.2 Messy Genetic Algorithm ............................................................. 31
3.4 GA Operators ..................................................................................... 32
3.4.1 Mutation ...................................................................................... 32
3.4.2 Crossover..................................................................................... 33
3.4.3 Multipoint Crossover ................................................................... 34
3.4.4 Selection ...................................................................................... 34
3.4.5 Fitness-Proportionate Selection .................................................. 36
3.4.6 Disadvantages of Fitness-Proportionate Selection ..................... 37
3.4.7 Rank Selection ............................................................................. 37
3.4.8 Tournament Selection .................................................................. 38
3.4.9 Scaling Methods .......................................................................... 39
3.5 Pros and Cons of Genetic Algorithms ................................................ 39
3.6 Selecting GA methods ........................................................................ 40
3.6.1 Encoding Choice ......................................................................... 40
3.6.2 Operator Choice .......................................................................... 42
3.6.3 Elitism .......................................................................................... 42
3.7 Example GA Application .................................................................. .42
3.8 Summary ............................................................................................ 45
Further Reading ................................................................................... 46
 Contents ix

4 Genetic Programming ............................................................................. 47

4.1 Genetic Programming ......................................................................... 47
4.2 Introduction to Genetic Programming ................................................ 47
4.2.1 Variable-length and Tree-Based Representations ....................... 49
4.2.2 GP Terminal Set .......................................................................... 49
4.2.3 GP Function Set .......................................................................... 49
4.2.4 Function Closure ......................................................................... 50
4.2.5 Tree Structure Processing ........................................................... 50
4.2.6 Linear Structure Encoding .......................................................... 50
4.2.7 Graph Structure Encoding ........................................................... 51
4.2.8 GP Initialisation .......................................................................... 51
4.3 GP Operators ...................................................................................... 52
4.3.1 GP Crossover .............................................................................. 52
4.3.2 Mutation ...................................................................................... 52
4.3.3 Selection Operators in GP ........................................................... 52
4.3.4 Controlling Genome Growth ....................................................... 53
4.4 Genetic Programming Implementation ............................................... 54
4.4.1 Advances in GP - Automatically Defined Functions ................. 54
4.5 Summary ............................................................................................ 55
Further Reading ................................................................................... 56
5 Engineering Examples Using Genetic Algorithms .•.••••••••••••••.•••.•••••••••• 57
5.1 Introduction ........................................................................................ 57
5.2 Digital Image Processing .................................................................... 57
5.3 Basics of Image Processing ................................................................ 58
5.3.1 Convolution ................................................................................. 58
5.3.2 Lookup Tables ............................................................................. 59
5.4 Java and Image Processing ................................................................. 60
5.4.1 Example Application - VEGA ................................................... 61
5.4.2 Operator Search Space ................................................................ 61
5.4.3 Implementation ............................................................................ 61
5.5 Spectrographic Chromosome Representation ..................................... 67
5.6 Results ................................................................................................ 68
5.6.1 GA Format ................................................................................... 68
5.7 Summary - Evolved Image Processing ............................................... 70
5.8 Mobile Robot Control.. ....................................................................... 71
5.8.1 Artificial Intelligence and Mobile Robots.................................... 72
5.8.2 Planning ...................................................................................... 72
5.8.3 Static Worlds ............................................................................... 73
5.8.4 Reactive and Bottom-up Control ................................................. 74
5.8.5 Advantages of Reactive Control .................................................. 76
5.8.6 Alternative Strategies .................................................................. 76
5.9 Behaviour Management. ..................................................................... 77
x Contents

5.9.1 Behaviour Synthesis Architecture ................................................ 78

5.9.2 Standard Control Methods - PlD .............................................. 81
5.10 Evolutionary Methods ...................................................................... 82
5.10.1 GAs and Robots ......................................................................... 82
5.10.21nferencing Problem .................................................................. 82
5.10.3 Behaviour priorities: Natural Agents. ....................................... 83
5.11 Fuzzy logic Control .......................................................................... 84
5.11.1 Fuzzy Control of Subsumption Architectures ............................ 87
5.12 Evolved Fuzzy Systems .................................................................... 87
5.13 Robot Simulator ................................................................................ 89
5.13.1 The Robot Control Architecture ................................................ 90
5.13.2 Related Work ............................................................................. 93
5.13.3 Results ....................................................................................... 94
5.14 Analysis ............................................................................................ 98
5.15 Summary - Evolving Hybrid Systems ............................................ 99
Further Reading ................................................................................... 99
6 Future Directions iu Evolutiouary Computing •.••.•.•••.••.•.••.•••.•.•••.••••.• 101
6.1 Developments in Evolutionary Algorithms ...................................... 101
6.2 Evolvable Hardware ......................................................................... 101
6.3 Speciation and Distributed EA Methods .......................................... 103
6.3.1 Demetic Groups and Parallel Processing ................................. 103
6.3.2 Parallel Genetic Programming with Mobile Agents.................. 104
6.3.3 Mobile Agents ............................................................................ 104
6.3.5 EA Visualisation Methods .......................................................... 107
6.4 Advanced EA techniques .................................................................. 108
6.4.1 Multiobjective Optimisation ...................................................... 109
6.4.2 Methods: Weighted Sum Approach............................................ 109
6.4.3 Minimax Method ........................................................................ 109
6.4.4 Parameter Control.. ................................................................... 110
6.4.5 Diploid Chromosomes ............................................................... 110
6.4.6 Self-Adaptation .......................................................................... 110
6.5 Artificial Life and Coevolutionary Algorithms ................................ 111
6.6 Summary .......................................................................................... 113
Further Reading ................................................................................. 114
7 The Future of Evolutionary Computing ......•.•.•.•.•.•.•.•.............•.•••.•••••• 115
7.1 Evolution in Action .......................................................................... 115
7.2 Commercial value of Evolutionary Algorithms ................................ 115
7.3 Future Directions in Evolutionary Computing ................................. 116
7.3.1 Alife and Coevolution ................................................................ 116
7.3.2 Biological Inspiration ................................................................ 117
7.3.3 Developmental Biology.............................................................. 117
7.3.4 Adaptive Encoding and Hierarchy ............................................ 118
 Contents xi

7.3.5 Representation and Selection .................................................... 118

7.3.6 Mating Choice ........................................................................... 118
7.3.7 Parallelism ................................................................................ 118
7.4 Conclusion ........................................................................................ 119
Bibliography .............................................................................................. 121

Appendix A ..................................................................•....................•....... 133

A.1 Java-based EA Software .................................................................. 133
A.2 CIC++ based EA Software ............................................................... 134
A.3 General Evolution and Robotics References ................................... 134
A.4 Java Reference Guides ..................................................................... 136
A.5 Useful References ............................................................................ 136
Appendix B ........................................•........•..........................................•... 137
A Genetic Algorithm Example and the GPSYS GP Library .................. 137
B.1 Basic Genetic Algorithm .................................................................. 137
B.2 Simple Java Genetic Algorithm ....................................................... 137
Exercises ............................................................................................. 147
B.2.1 Vectors and Arraylists ............................................................... 147
B.3 Application Design .......................................................................... 148
B.4 Eos: An Evolutionary and Ecosystem Research Platform ............... 149
Authors: Erwin Bonsma, Mark Shackleton and Rob Shipman ........... 149
B.4.1 Introduction ............................................................................... 150
B.4.2 Design Overview ....................................................................... 150
B.4.3 Key Classes ............................................................................... 152
B.4.4 Configuration ............................................................................ 153
B.4.5 Illustrative Example Systems ..................................................... 154
B.4.6 Aerial Placement for Mobile Networks ..................................... 154
B.4.7 An Ecosystem Simulation Based on Echo ................................. 155
B.4.S Coevolutionary Function Optimisation ..................................... 156
B.4.9 Telecommunications Research using Eos ................................. 157
B.4.lO NetGrow: A Telecommunications Application ........................ 158
B.4.11 Eos Summary ........................................................................... 158
B.5 Traveling Salesman Problem ........................................................... 159
B.5.1 EOS Traveling Salesman Problem ............................................ 159
B.6 Genetic Programming ...................................................................... 163
B.6.1 Observations from Running GPsys - Lawnmower Problem: .... 164
Eos References.................................................................................... 167
Appendix C .........................................................••...•....•................•....•..... 169
C.1 Fuzzy Logic ..................................................................................... 169
C.2 Fuzzy Set Theory ............................................................................. 170
xii Contents

C.2.I Fuzzy Operators ........................................................................ 171

C2.2 Linguistic Variables .................................................................. 171
C.2.3 Fuzzy IF .................................................................................... 172
C2.4 Fuzzy Associative Memories ..................................................... 173
C2.5 Fuzzy Control Systems .............................................................. 174
C.2.6 Defuzzification .......................................................................... 175
C.2.7 Fuzzy Applications .................................................................... 178
C.3 Limitations of Fuzzy Control ........................................................... 178
C3.I Advantages of Fuzzy Systems .................................................... 179
C.4 Summary .......................................................................................... 180
Further Reading ................................................................................. 180
Appendix D ............................................................................................... 181
Introduction ............................................................................................ 181
System Overview .................................................................................... 181
Use and License ...................................................................................... 181
Programming Language and Run-Time Environment.. .......................... 182
Top-Level Directory Files and Hierarchy ............................................... 182
Units of Measure .................................................................................... 183
The Client-Server Architecture .............................................................. 183
Network/Local Connections Versus Dynamically Loaded Clients ..... 184
Why a Client-Server Architecture? ..................................................... 185
Client-Server Communications ........................................................... 185
Network and Local Connection Issues ............................................... 186
Communication via Events and Requests ........................................... 187
Keeping the RP I Protocol Language-Independent ............................ 188
Configuration Elements and Properties Files ......................................... 188
The "port" and "hostName" Properties ............................................ 189
Overriding Properties ........................................................................ 189
Loading RsProperties Files as a Resource ......................................... 190
The Server .............................................................................................. 191
Server Properties Files ....................................................................... 191
Accepting Clients ................................................................................ 191
The Scheduler ..................................................................................... 192
The Floor Plan ........................................................................................ 194
Syntax and Semantics ....................................................................... ,. 194
Building a Virtual Robot ........................................................................ 196
Introduction ........................................................................................ 196
Thinking About Client Design ............................................................ 196
The Demonstration Clients ..................................................................... 196
Life Cycle of the Demonstration Clients ................................................ 197
How ClnMain Extends RsClient and Implements RsRunnable .............. 199
The RsRunnable Interface .................................................................. 200
 Contents xiii

Building RsRunnable and RsClient into ClnMain.. ............................ 201

DemoMain Implements RsRunnable, But Does Not Extend RsClient.202
The Execution of ClnMain .................................................................. 202
Uploading the Body Plan ................................................................... 204
Registering Event Handlers ................................................................ 204
Running the Event Loop ..................................................................... 205
How the Demo Clients Work .............................................................. 205
Physical Layout of ClientZero ................................................................ 207
The RsBody and RsBodyPart Classes .................................................... 207
RsBodyShape ...................................................................................... 209
RsWheeISystem ................................................................................... 210
The Sensor Classes ............................................................................. 212
RsBodyTargetSensor .......................................................................... 213
RsBodyContactSensor ........................................................................ 213
Events and Requests ............................................................................... 214
Index ...................................................................................................... 216
 1
Introduction to Evolutionary Computing

Four billion years ago, the Earth was a molecular Garden of Eden. There
were as yet no predators. Some molecules reproduced themselves
inefficiently, competed for building blocks and left crude copies of
themselves. With reproduction, mutation and the selective elimination of the
least efficient varieties, evolution was well under way, even at the molecular
level. As time went on, they got better at reproducing. Molecules with
specialized functions eventually joined together, making a kind of molecular
collective - the first cell.
Carl Sagan, Cosmos, 1980

1.1 Evolutionary Computation

Evolution is a ubiquitous natural force that has shaped all life on Earth for
approximately 3.2 billion years. For several thousand years humanity has also
utlised artificial selection to shape domesticated plant and animal species. In the
past few decades, however, science has learned that the general principles at
work in natural evolution can also be applied to completely artificial
environments. In particular, within Computer Science the field of automated
machine learning has adopted algorithms based on the mechanisms exploited by
natural evolution.
One driving motivation for many researchers in Evolutionary Algorithms
(EA) and other machine learning fields is to create a set of automatic processes
that can translate high-level task descriptions into well-coded solutions. The
problem is that almost all commercially available software relies on handcrafted
and custom-written lines of code. This is always a time-consuming, expensive,
and error-prone process. For example, a typical modem operating system
contains several million lines of code and thousands of bugs, most of which are
tolerated and slowly weeded out by consumer effort (i.e., you). This may work
at present, but the cost and success rate of large and complex pieces of software
are a major concern to software developers. The result is a widening gap
between the capabilities of computing hardware and the quality and efficiency
of the software it runs. One solution being pursued is the open source movement

R. Ghanea-Hercock, Applied Evolutionary Algorithms in Java

© Springer Science+Business Media New York 2003
2 1 Introduction to Evolutionary Computing

in which code is opened to a large community of developers who contribute in

an ongoing process to its development (the principal example being
development of the Linux operating system). Good solutions are then
incorporated into the final product and poor solutions culled, which is not
entirely dissimilar to the approch of evolution-based algorithms.
There are many flavours of machine learning algorithms, from Case Based
Reasoning to Artificial Neural Networks, each of which has particular strengths
and weaknesses. However, there is a particular attraction of EA to many
computer scientists. This is difficult to convey until you have experimented with
such algorithms and been fascinated by their ability to produce a functioning
solution to a problem out of initially random bit sequences. It is the aim of this
text to enable readers to quickly acquire confidence in building and developing
their own EA programs and hence to gain some understanding of the power of
EA as a machine learning method.

1.2 History of Evolutionary Computing

It is particularly useful to consider the history of evolution within computing as

it covers much of the timeframe of computing itself. Some of the earliest work
can be traced back to Friedberg (1958), who introduced the idea of an
evolutionary algorithm approach for automatic programming. Later significant
developments included the creation of evolutionary programming by Fogel
(Fogel et aI., 1966). John Holland (1967) founded the initial work on genetic
algorithms at the University of Michigan. Parallel work was also initiated by
Bienert, P. Rechenberg, and Schwefel in evolutionary strategies (1966).
However, the first evolutionary algorithm in the form most commonly
utilised today was the genetic algorithm (GA) created by John Holland and his
students at the University of Michigan. The methods defined by the GA are the
basis for this text as they represent the most commonly used EA methods and
have proved to be highly flexible in real-world applications. Endless variations
of the standard GA operators have also been developed, such as variable-length
chromosomes, multipoint crossover, and many other flavours. Chapter 3
provides an introduction to the core GA concepts and some of the common
variations in present use.
The next major development in EA systems was Genetic Programming,
introduced by John Koza (1992). Koza formulated a general-purpose system for
program induction based on a tree-based representation scheme, which has been
applied to a very wide range of machine learning problems. Genetic
programming is allocated to Chapter 4 precisely because of its broad
applicability and to present an alternative approach to evolving computer
algorithms. In addition, it allows a more complex genome data representation,
which can be mapped to the object-based approach presented in the applications
in Chapter 5.
 1.1 Evolutionary Computation 3

The remainder of this chapter will first review the challenges encountered in
EA and the software and hardware issues that impact the exploitation of
artificial evolution as a problem-solving method. Second, a brief overview of the
history and methods of machine learning is provided, as this introduces some
core concepts required later in the text.

1.3 Obstacles to Evolutionary Computation

A major barrier to the early adoption of EA in the computing domain came from
opposition within the computer science community itself. This was often based
on the mistaken belief that such algorithms, with probabilistic processes as a
core mechanism, would not be amenable to producing functional code.
Particularly as normal computer code is generally brittle, and very susceptible to
minor variations in structure. In Chapters 2 and 3 we develop the necessary
background to EA methods to understand why probabilistic selection· methods
can generate functional code. As Goldberg states:
In our haste to discount strictly random search techniques, we must be careful
to separate them from randomised techniques. The genetic algorithm is an
example of a search procedure that uses random choice as a tool to guide a
highly exploitative search through a coding of a parameter space. (Goldberg
1989,p.5)
The second barrier to EA development was the problem that contemporary
computing technology in software, and particularly hardware, in the early 1970s
was barely capable of generating useful results in acceptable time scales (i.e.,
less than a few weeks). This problem added to the belief that such methods,
while theoretically interesting, would never be capable of useful applications.
Fortunately the exponential development in computing systems has generated
both software and hardware that offer the necessary processing power and also
software sophistication in terms of efficient compilers and object-based design.
A modem high-specification desktop can generate useful results from an EA
within a few hours, in a typical application such as those described in Chapter 5.
In Chapter 6 some speculative ideas are discussed on the possibilities future
hardware will enable, such as using EA within real-time adaptive applications.
In particular, the possibility of using reconfigurable hardware based on Field
Programmable Gate Array devices, opens entirely new applications to EA
methods. This area is explored in Chapter 6.

1.4 Machine Learning

Machine learning explores the mechanisms by which knowledge or skills can be

acquired through computing processes. It includes models for learning based on
statistics, logic, mathematics, neural structures, information theory, and heuristic
4 1 Introduction to Evolutionary Computing

search algorithms. It also involves the development and analysis of algorithms

that identify patterns in observed data in order to make predictions about unseen
data. Mitchell gives a useful definition:
".. machine learning is the study of computer algorithms that improve
automatically through experience." (Mitchell, 1996}.

In contrast, the early history of computer intelligence was dominated by the

field of 'Expert Systems' in which the emphasis was in precisely defining human
knowledge into machine processable tasks. Some success was achieved with the
expert systems approach, but it became increasingly obvious that such systems
were highly task-specific with minimal ability to generalise the knowledge
supplied. Hence increased interest developed during the 1980s in giving
machines the ability to learn new knowledge, leading to a revival of machine
learning processes.
Longer-term goals that the pioneers of the field considered achievable, such
as general learning and cognition, have been shelved until far more work has
been completed in understanding the basics of specific machine skills. Expert
and knowledge-based systems will therefore continue to develop in parallel with
adaptive and soft computing processes, such as EA and Fuzzy Logic systems.

1.5 Problem Domains

There is an increasing range of complex problems across commerce and science,

which require automated and adaptive computational techniques. For example,
relational databases with millions of records are now common in business,
astronomy, engineering, and the sciences. The problem of extracting useful
information from such data sets is an important (and lucrative) commercial
problem. Evolution-based algorithms have been applied to a wide range of
technical problems, including handwriting recognition, scheduling in
manufacturing, speech and image processing. Other examples of EA
applications include

• Gas pipeline optimisation (Goldberg, 1983)

• Robot kinematics (Davidor, 1991)
• Image Processing (Harvey & Marshall, 1996)
• Evolved art (Todd & Latham, 1992)
• Network design (Choi, 1996)

1.5.1 Search Spaces

A central aspect of machine learning is the requirement of an algorithm to
perform an automated search of a complex and often multidimensional problem
space, (as shown in figure 1.1). The problem space is effectively a mapping of
the problem parameters into an abstract n-dimensional state space. In state space
 1.5 Problem Domains 5

search, a programmer first assigns a computer an initial state. A set of operators

is automatically used to move from one state to the next. In addition, some
halting condition is required, which specifies when a sufficiently close solution
has been achieved. Common search algorithms applied to AI search strategies
include breadth-first and depth-first search.

0.5

FIGURE 1.1 A typical mUltipeak search space, in which a machine learning

algorithm can become trapped on a suboptimal solution.

A useful starting point for developing search algorithms is the text

Fundamentals of Data Structures in C++ by Horowitz et aI, (Horowitz et aI ,
1995). Simple hill-climbing or related calculus-based methods may have great
difficulty in solving such multioptima spaces, and are of even less use in
discontinuous and irregular spaces. Unfortunately, many if not all real-world
problems are of this form. All too often researchers test new algorithms and
strategies on relatively simple test cases and then fail to achieve any qualitative
improvement when running the algorithm on a real data set. It is therefore
6 1 Introduction to Evolutionary Computing

instructive to consider the idea that in all machine search methods the no free
lunch theorem applies (Wolpert and Macready, 1995) that is:
all algorithms that search for an extremum of a cost function perform exactly
the same, when averaged over all possible cost functions. In particular, if
algorithm A outperforms algorithm B on some cost functions, then loosely
speaking there must exist exactly as many other functions where B
outperforms A. (Wolpert and Macready, 1995)

1.5.2 Optimisation versus Robustness

It is of some importance when considering the value of any machine learning
algorithm to determine the correct measure of effectiveness (or MOE as defined
in military strategy studies). Real-world search spaces as discussed are
invariably complex and ill defined, hence spending large resources on locating a
precise optimum may be futile. A better approach may be to employ a spread of
algorithms, which generate a robust and usable set of good solutions. The
lessons learned from the underachievement of expert systems were exactly that
they frequently generated optimised but fragile and nonadaptive solutions. EA
can also experience the same problems if poorly designed.
Another way of reading this is to apply quick and cheap searches if they
approach a working solution. Fuzzy logic and EA are examples of broad and
robust search algorithms, and their robustness attributes will be considered in
some detail later in the text.

1.5.3 Expert Systems and AI

This section provides a very cursory review of rule-based AI methods, as a
comparison with soft computing and EA. Expert systems utilise rule-based
programming rules to represent heuristics, or "rules of thumb," which specify a
set of actions to be performed for a given situation. A rule is composed of an if
portion and a then portion. The if portion of a rule is a series of patterns that
specify the facts (or data) that cause the rule to be applicable (the precondition).
The process of matching facts to patterns is called pattern matching. The expert
system tool provides a mechanism, called the inference engine, which
automatically matches facts against patterns and determines which rules are
applicable. The if portion of a rule can actually be thought of as the whenever
portion of a rule since pattern matching always occurs whenever changes are
made to facts. The then portion of a rule is the set of actions to be executed
when the rule is applicable (the postcondition). Hence the actions of applicable
rules are executed when the inference engine is instructed to begin execution.
The inference engine selects a rule and then the actions of the selected rule are
executed (which may also affect the list of applicable rules by adding or
removing facts).
 1.5 Problem Domains 7

A useful publicly available rule-based expert system is CLIPS (the C

Language Integrated Production System), which was originally developed at
NASA. CLIPS is available from www.ghg.netlclips/CLIPS.html. (A useful
Java-based rule system is the IBM AbleBeans code library, available from
http://alphaworks.ibm.com.) For further material a good introductory text to
modem AI is that by Russell and Norvig (1995).
Current work in expert systems has now shifted to Case Based Reasoning
and agent-based systems, each of which extends and adapts the use of heuristic
inferencing. An interesting (and profitable) application domain example
involves online e-commerce systems (e.g., Maes et aI., 1999, and Ardissono et
aI., 1999), in which agents and CBR are combined to support intelligent user
services. However, the majority of working business and e-commerce systems
still rely heavily on simple rule-based inferencing.

1.5.4 Fuzzy Logic

One highly successful computational intelligence method that has emerged is
Fuzzy Logic. Fuzzy logic (Zadeh, 1965) has enjoyed wide popularity in
engineering as an advanced control and AI technique; particularly within the
Japanese and Asian markets. However, it is only in the past two decades that it
has been recognised in Europe and America as a useful AI system. Driankov et
aI., (Driankov, Hellendoorn, and Reinfrank, 1996) provides an excellent
introduction to applied fuzzy control methods. It provides a particularly clear
insight into the theory and application of fuzzy techniques.
Since Zadeh's original work on Fuzzy Logic (Zadeh, 1965), intense debate
has surrounded the subject, with vigorous defenders and opponents on both
sides. In terms of the underlying theory the argument has revolved around
whether fuzzy logic is just another form of probabilistic reasoning or whether
probability is a subset of fuzziness (Kosko, 1992).
A fuzzy rule system effectively embeds expert knowledge into a set of
flexible overlapping rules in order to create an adaptive data processing system.
Its prime application area is in advanced control systems, such as robotics,
industrial plant control, transport speed controllers, and washing machines! Its
strength lies in its ability to control a complex system, even if no precise
mathematical model of the underlying processes is available. The key issue
revolves around designing the required input and output rule sets.
Since a fuzzy system normally contains no learning ability in itself, it is
often combined with evolutionary algorithms or neural networks, which act to
learn the necessary rule sets from some training data. Appendix C provides an
introductory tutorial to fuzzy systems, as they form an important part of the
mobile robot control application presented in Chapter 5. Figure 1.2 provides an
illustration of a fuzzy clustering algorithm applied to an e-commerce
application.
8 1 Introduction to Evolutionary Computing

As with all of the machine learning methods fuzzy logic is not a panacea, but
it is highly effective in the control of complex engineering systems and can
provide a useful interface between other adaptive algorithms and physical
systems. An excellent introduction to fuzzy systems, which includes material on
neural networks and how the two systems can be combined, is Kosko's text
(Kosko, 1992). The advantages of fuzzy control systems can be divided into
theoretical and practical issues.

, SelectedC
,
I

d Product

30

x-axis, Cost
.,
••
FIGURE 1.2 Example fuzzy clustering technique applied to data points for an e-
commerce application. Using 4 input attributes to generate 4 clusters, from a range of 64
laptop computers. In this example a fuzzy rule system can help identify a product based
on a customers combined set of specified preferences. (Copyright BT ISR Laboratories,
1999.)

First, the arguments for fuzzy control include

• They permit the incorporation of linguistic knowledge into a control

architecture.
• It is a model-free system, hence no underlying mathematical model of the
control process or plant is required .
• It provides an efficient nonlinear control mechanism, which is justified by
the universal-approximation theorem.
• Ease of implementation.
 1.5 Problem Domains 9

• Simple to design, compared to an equivalent conventional controller for the

same system (particularly true for nonlinear systems).
• Cheap to develop, with well-established software packages available, and
fast to prototype new code if required.
• Possible to implement as a VLSI chip, which are now widely available (in
contrast to the inherent complexity of hardware neural network systems).

1.5.5 Bayesian networks

A Bayesian network is basically a model representation based on reasoning with
uncertainty, using a statistical approach. A problem domain is divided into a
number of entities or events, which are represented as separate variables. One
variable could represent the event that a server in a communications network has
failed. The variables representing different events are then connected via
directed edges to describe relations between events. An edge between two
variables X and Y then simply represents a possible dependence relation
between the events represented by X and Y, as illustrated in Figure l.3.
An edge could, for instance, describe a dependence relation between a
disease and a symptom (a common application domain for Bayesian net
software). Hence, edges can be used to represent cause-effect relations. These
dependence relations between entities of the problem domain may be organised
as a graphical structure. This graphical structure describes the possible
dependence relations between the entities of the problem domain. The advantage
of such Bayesian networks is their compactness and efficiency.
The uncertainty of the specific problem domain is then represented through
conditional probabilities. Conditional probability distributions specify our belief
about the strengths of the cause-effect relations; for example, a server crash may
not be due to network traffic but could also be due to a power outage, each of
which is assigned some probability value.
Thus, a Bayesian network consists of a qualitative part, which describes the
dependence relations of the problem domain, and a quantitative part, which
describes our belief about the strengths of the relations. The framework of
Bayesian networks offers a compact, intuitive, and efficient graphical
representation of dependence relations between the entities of a problem
domain.
The graphical structure reflects properties of the problem domain in an
intuitive way, which makes it easy for nonexperts to understand and build this
kind of knowledge representation. It is possible to utilise both background
knowledge such as expert knowledge and knowledge stored in databases when
constructing Bayesian networks.

(A freeware tool for experimenting with Bayesian networks is available from

the Microsoft research web site at http://research.microsoft.com/msbn/.)
10 1 Introduction to Evolutionary Computing

FIGURE 1.3 Example graphical representation of a Bayesian network showing a simple

example belief network for diagnosing why a car won 't start, based on spark plugs,
headlights, main fuse, etc. (Copyright Norsys Software Corp., 2001.)

1.5.6 Artificial Neural Networks

A dominant field in machine learning is that of artificial neural networks
(ANN). Original work in this field can be traced back to Mc Culloch and Pits
(1943) and their development of artificial models of neuron behaviour. In this
model the neuron receives one or more inputs and produces one or more
identical outputs. Each output is then a simple nonlinear function of the sum of
the inputs to the neuron .
Given the obvious fact that the most intelligent computer on the planet
weighs about 2.5 pounds and is composed of a jellylike mass of biological
neurons, there has been significant interest in the computer science community
in learning how to replicate the capabilities of the biological neuron. Strong
interest in the AI community continued until 1969, when a text by Marvin
Minsky and Seymour Papert emphasised the limitations of the Perceptron model
 1.5 Problem Domains 11

of ANN, which was dominant at the time; this virtually halted further research in
the subject.
In 1982, however, a seminal paper by Hopfield (Hopfield, 1982) reignited
interest in the field with a new model of a bidirectional ANN. Since then a large
research community has grown around the investigation of ANN models and
their application. One useful definition is
a neural network is a system composed of many simple processing elements
operating in parallel whose function is determined by network structure,
connection strengths, and the processing performed at computing elements or
nodes. (DARPA Neural Network Study, 1988, AFCEA Int. Press.)

1.5.7 Feedforward Networks

The commonest ANN is the three-layer feedforward network, which has an
input layer, one hidden layer, and an output layer. The network operates as a
transfer function acting on a vector applied to the input layer, as illustrated in
Figure 1.4.

Hidden Layer
Ho
Xo Yo

Inputs Outputs

XII

Input Output
weights Wij weights Wj .;

FIGURE 1.4 Example of a three-layer feedforward network (Welstead, 1994).

The network's behaviour is determined by the number of layers and nodes

within each layer and the value of each weight linking the nodes. Through a
training process involving the presentation of large sets of input-output pairs the
network is able to learn the required transfer function. The popularity of these
systems is based on the mathematical result that such networks are universal
approximators, hence they can model any reasonable function (i.e., one having a
small number of discontinuities and defined on a closed, bounded subset of RN)
12 1 Introduction to Evolutionary Computing

based on a deterministic process (see Kosko, 1992). The most frequently used
transfer function for a neural network is a sigmoidal function, as shown in
Figure 1.5.
1
f(x) = 1+ e(-a.x)
(Ll)

The use of weighted threshold connections is a basic approximation of the

way in which biological synapses operate. Nonlinear transfer functions increase
the computational range of a parallel network and help in noise suppression.
However, they increase the complexity of analysing the trained network.
Learning in natural organisms utilises adjustments in the strength of the synaptic
connections, and similarly ANNs are also able to encode learned responses in
the set of weight vectors.

OL-__ -*~~~ ____L __ _ ~ _ _ _ _~_ __ _L __ _~_ _ _ _~

o 2 4

FIGURE 1.5 Example sigmoidal threshold function, used in ANN nodes, based on the
logistic sigmoid function in (1.1) a = gain factor, i.e., response time of the network.

1.6 Applications

The prime benefit of ANN is the ability to act as pattern recognisers and noisy
data classifiers. They can generalise decisions from poor-quality input data and
hence are ideal for a range of classification problems, such as
• Handwriting recognition
• Speech processing
• Image processing
• Digital and analogue signal processing
 1.6 Applications 13

• Robot control

ANN may be categorised into two general types; feedforward and recurrent
(utilising feedback). They can also be divided into whether they use training
data to learn the required response (supervised) or use a self-organising method
(unsupervised) to learn. An example of the later is Kohenen networks (Kohenen,
1984). Unsupervised networks operate by clustering data into related groups
based on measuring the features acting as inputs to the network.

1.6.1 Problems
While there is an intuitive and appealing relationship between ANN and
biological neural systems, the artificial variety suffers from several
disadvantages. First, the universal approximation property may be theoretically
useful, but it fails to inform us how many hidden layers and nodes the network
may actually require for a given function. Second, the number of pairs of
training data pairs required to train even a simple network is often several
thousand. This generally requires a significant amount of computation. In
addition, the algorithms used to compute the weights are nontrivial and require
extensive processing. Finally, the most significant problem relates to
interpretation of the final trained network and its set of activation weights. It is
extremely difficult to extract the logical inference, which relates a given input
vector to an output response, namely, to understand how the network is
performing its specific task. This has led to the formation of the new subfield of
Rule-Extraction from ANN. In effect, the trained network is a black-box
mechanism. In contrast, fuzzy rule-based systems are intrinsically transparent to
inspection and analysis of which rules relate to specific behaviours of the
system.
It is also very important to note that contrary to popular opinion, biological
neurons are not simple function transducers (Koch, 1997); rather they should be
modeled as individual agents in their own right; or at least as having adaptive
properties. Cortical neurons contain a nucleus and cellular body and are
specialised but fully functional cells. Constructing simple three-layer
feedforward networks based on sigmoid functions and expecting them to model
natural neural computational structures is simply unrealistic. Significantly more
complex neural systems have, however, been designed. These may be based on
a cellular model, which can act in a co-operative manner to process complex
signals. Recent work on using cellular neural systems for controlling
autonomous software agents with a more realistic biological model has been
demonstrated by Cyberlife Technology Inc., in their "Creatures" product
(Cyberlife, Grand, 1996).
However, ANN systems do possess intrinsic learning and adaptive
capabilities, and these have been exploited in a range of advanced self-
organising controllers. An interesting example is mobile robot control systems.
14 1 Introduction to Evolutionary Computing

Robotic systems are a useful test platform for a wide range of machine learning
methods and frequent use of ANN has been made in this domain, as they offer a
way to investigate sensory integration and processing. They may also be used to
directly implement the sensor-actuator link or to control the interaction of lower-
level reactive behaviours.

1.7 Evolution-Based Search

The central concept of EA is that repeated mixing, mutation, and a selective

filtering action on a population of individual bit string solutions (i.e.,
chromosomes) is an effective method for locating a solution point within a
complex search space. Chapters 2 and 3 cover the field of EA in greater detail.

1.7.1 Languages for Evolutionary Computing

In order to implement an evolutionary algorithm, we need two components, a
sufficiently powerful computer and a suitable programming language.
Fortunately both of these components have matured since Holland's early work,
and modem computers and high-level languages make the task of building
evolutionary algorithms far easier. However, the software aspect of
implementing EA is frequently a source of frustration for new researchers and
students in the field. It is a nontrivial exercise to code even a minimal EA. This
is particularly the case with languages that are not object-oriented, as an EA is
best represented as a collection of interacting data objects, each possessing
multiple parameters. In addition, the classic programming problem of memory
management is of importance, as individual chromosomes are being added and
deleted from the pool of evolving code at some rate, leading to the need for
careful deallocation of objects from memory. Since the advent of new object-
based languages such as Java, C++, and more advanced compilers, these have
greatly simplified the programming effort required to develop EA systems.
The relative merits of some suitable languages wiIl now be considered. All
of the examples and exercises in this book are based on the Java programming
language from Sun MicroSystems, in its current version JDK1.3. There are
several strong reasons why this language was selected for this text, and these
wiIl be discussed with comparisons to alternative languages.

1.7.2 C and C++

The principal computing language used to date for implementing EA has been C
or its derivative, C++. It offers several advantages well suited to EA work;
• High speed: from being a compiled medium-level language.
• Compact size.
• Wide range of efficient compilers and support tools available.
• C++ is an early object-based language that builds on the strengths of C.
 1.7 Evolution based search 15

As a result there are now a significant number of code libraries in both C and
C++ available to perform a vast range of EA operations. In Appendix A, a few
of the principal libraries are listed, particularly those available for open source
distribution.
However, there are also some disadvantages to these languages. First, C is a
relatively low-level language with few operators for complex data manipulation.
It can also be difficult for new students to learn and it can be difficult to
implement large-scale programs. C++ was an attempt to add object-based
functionality to C, which has been partially successful. However, both languages
offer little in the way of automatic memory management, which places a burden
on the developer to manually code explicit deallocation of used memory
resources. C++ has also been criticised for its complexity and lack of
standardisation. It can be a major task to port code developed on one specific
hardware platform to another.

1.7.3 Pascal and Fortran

Pascal and more frequently Fortran have also been used to develop EA systems.
However, there are far fewer code libraries available than C or C++. Both
languages suffer to some extent from the disadvantages of C and usually lack
the object capabilities of C++ (Delphi, a strong object-based language derived
from Pascal, is an exception).

1.7.4 Visual Basic

A modern object oriented language that could be used for EA work is Visual
Basic; however, it is relatively slow, compared to fully compiled languages, and
quite platform-specific. Some source code for EA in VB is available on the web.

1.7.5 Java
Java has several major advantages for developing EA applications:
• Automatic memory management.
• Pure object-oriented design.
• High-level data constructs: for example, the Vector and ArrayList objects
offer dynamically resizable arrays.
• Platform independent code is easily ported among Unix, Linux, Windows,
and Mac systems.
• Several complete EA libraries are available for EA systems (see Appendix
A).
Since life is rarely perfect, however, there is a price to pay! Java is an
interpreted language, meaning it is converted from its bytecode format into a
platform-specific executable at run-time. This significantly slows down Java
applications, relative to equivalent C or C++ code. However, recent work by
Sun and third-party companies has resulted in "Just in Time" compilers for Java
16 1 Introduction to Evolutionary Computing

that preload and compile the Java bytecode. The end result is performance,
which is within 30-50% of an equivalent C application. (See the Bibliography
for references in this area.)

1.7.6 Object-Oriented Design

One of the fundamental advances in software development was the introduction
of object-oriented (00) design. The following quotes attempt to define the basic
idea behind 00 design:

The first principle of object oriented programming might be called

intelligence encapsulation: view objects from outside to provide a natural
metaphor of intrinsic behavior. (Rentsch et aI., 1982)

The basic support a programmer needs to write object-oriented programs

consists of a class mechanism with inheritance and a mechanism that allows
calls of member functions to depend on the actual type of an object (in cases
where the actual type is unknown at compile time). (Stroustrup, 1991)

The object-oriented approach combines three properties: encapsulation,

inheritance, and organization. (Nguyen, 1986)

In basic terms rather than use a procedure or structured programming format,

00 design works by encapsulating related methods and variables within a single
code entity, i.e., an object. Hostetter (Hostetter, 2002) gives a useful description
of the evolution of 00 languages. One of the first such languages was
SmallTalk, but the first widespread use of 00 came in 1983 when the first
version of C++ was released. More features have been continually added, until
an ISO standard version of C++ was agreed, in 1998.

1.7.7 Java and 00 Design

Java was originally designed for programming embedded systems (e.g., smart
toasters). Because of this, the ideas of platform independence and run-time
safety are key aspects of its design. However, it is also completely object-
oriented in every respect. The Java online tutorial provides an excellent
definition of the 00 approach:
Real-world objects share two characteristics: They all have state and
behaviour. For example, dogs have state (name, colour, breed, hungry) and
behaviour (barking, fetching, and wagging tail). Bicycles have state (current
gear, current pedal cadence, two wheels, number of gears) and behaviour
(braking, accelerating, slowing down, changing gears). Software objects are
modelled after real-world objects in that they too have state and behaviour. A
software object maintains its state in one or more variables. A variable is an
Other documents randomly have
different content
which took the initiative in calling the international conference
which arranged for a great reduction in naval armaments.
General References
Charles A. Beard, American Government and Politics, pp. 315-341;
Ibid., Readings in American Government and Politics, pp. 291-307;
Everett Kimball, National Government of the United States, pp. 540-
573;
A. B. Hart, Actual Government, pp. 430-445;
John W. Foster, The Practice of Diplomacy, especially pp. 34-54;
P. S. Reinsch, Readings in American Federal Government, pp. 651-
682;
E. S. Corwin, The President’s Control of Foreign Relations, passim;
Gaillard Hunt, The Department of State.
Group Problems
1. The Monroe Doctrine. Is it obsolete? The international situation
during the years 1815-1823. The Holy Alliance, its organization and aims.
Spain in America. The revolt of the Spanish Colonies. Preliminaries of the
declaration. Canning’s suggestion. Scope of the doctrine as announced.
Subsequent applications and extensions. The French in Mexico. The
Venezuela controversy. Present scope of the doctrine. Attitude of Europe
toward it. Attitude of the Spanish-American states. Its value for the future.
Conclusion. References: Hiram Bingham, The Monroe Doctrine: An
Obsolete Shibboleth, pp. 3-55; A. B. Hart, The Monroe Doctrine, pp. 55-
83, and passim; A. C. Coolidge, The United States as a World Power,
pp. 95-120; C. H. Sherrill, Modernizing the Monroe Doctrine, pp. 64-76;
C. L. Jones, Caribbean Interests of the United States, pp. 323-351; J. H.
Latané, The United States and Spanish America, pp. 292-334; D. C.
Gilman, James Monroe (American Statesmen Series, Standard Library
Edition), pp. 156-174; Theodore Roosevelt, American Ideals, pp. 220-
237; Cyclopedia of American Government, Vol. II, pp. 456-468; Dexter
Perkins, “Europe, Spanish America, and the Monroe Doctrine” in
American Historical Review (January, 1922).
2. The diplomatic service and how it can be improved. References:
J. W. Foster, The Practice of Diplomacy, pp. 34-54; John A. Fairlie,
National Administration, pp. 77-91; E. Van Dyne, Our Foreign Service,
pp. 45-113; Cyclopedia of American Government, Vol. I, pp. 593-595; P.
S. Reinsch, Readings on American Federal Government, pp. 651-658;
675-682.
3. The chief rules of international law; how can their enforcement
be ensured? References: G. B. Davis, Elements of International Law,
pp. 19-30; T. J. Lawrence, Principles of International Law, pp. 119-138;
G. G. Wilson and G. F. Tucker, International Law (7th ed.), pp. 44-60;
A. S. Hershey, The Essentials of International Public Law, pp. 143-169;
A. H. Snow, The American Philosophy of Government, pp. 113-154; 267-
283. See also the General References to Chapter XXX.
Short Studies
1. The rights and duties of neutrals. G. B. Davis, Elements of
International Law, pp. 376-395 (Rights of Neutrals); pp. 396-445 (Duties
of Neutrals).
2. The privileges of diplomats. J. W. Foster, The Practice of
Diplomacy, pp. 159-174.
3. How treaties are made. G. B. Davis, Elements of International Law,
pp. 223-249.
4. The power of the Senate in relation to treaties. Ralston Hayden,
The Senate and Treaties, especially pp. 169-195; J. W. Foster, The
Practice of Diplomacy, pp. 262-283.
5. The Venezuelan controversy. Grover Cleveland, Presidential
Problems, pp. 173-281.
6. Arbitration as a method of settling International disputes. R. L.
Jones, International Arbitration as a Substitute for War between Nations,
pp. 218-269; J. W. Foster, Arbitration and The Hague Court, pp. 39-57;
J. B. Moore, American Diplomacy, pp. 200-222.
7. The Hague Conferences. G. B. Davis, Elements of International
Law, pp. 258-263; 519-524; 525.
8. The proposed codification of international law. A. H. Snow, The
American Philosophy of Government, pp. 395-418.
Questions
1. What is international law? Is it properly a system of law? Explain the
sense in which you use the term law in the following expressions: law of
gravitation; law of the land; law of supply and demand; law of fashion.
2. Look up and explain the following terms: belligerent, contraband,
unneutral service, filibustering, blockade, three-mile limit, diplomatic
immunity.
3. Make a list of (a) the rights of neutrals; (b) the duties of neutrals, and
show how each right involves a duty.
4. Draw up, in the form of a diary, a day’s happenings in the American
embassy at Tokyo, putting down at least six things done by the
ambassador during the day.
5. Explain what is meant by secret diplomacy. To what extent has the
United States avoided it and why?
 6. Give an account (from your studies in American History) of some
important treaty to which the United States was a party. Tell how it was
negotiated, signed, and ratified.
7. Is the principle set forth by Washington and Jefferson concerning the
true policy of the United States in foreign affairs applicable at the present
time?
8. Are the following statements true of the United States today:
(a) “In the wars of European powers in matters relating to themselves
we have never taken any part.”
(b) “With the existing colonies or dependencies of any European power
we have not interfered”?
9. What is meant by the saying that “the covenant of the League of
Nations does not destroy the Monroe Doctrine but extends it to the whole
world”? Is that statement correct?
10. What seems to you to be the most important among American
contributions to international law?
Topics for Debate
1. All members of the diplomatic service, including ambassadors,
should be chosen under civil service rules.
2. A majority vote in the Senate should be made sufficient for the
ratification of treaties.
3. It would be a violation of the Monroe Doctrine if Great Britain were to
sell the island of Jamaica to Germany.
 CHAPTER XXX
THE UNITED STATES AS A WORLD POWER

The purpose of this chapter is to answer the question: What are the
relations of the United States to the rest of the world?

The Old Policy of Isolation.—For more than one hundred

years it was the settled policy of the United States to keep
aloof from all entanglements in the affairs of the rest of the
world. This tradition of aloofness was The doctrines of
given a definite form by Washington, Washington and
who solemnly warned his countrymen Jefferson.
against getting mixed up in the “ordinary” conflicts of
European states, and it was subsequently endorsed by
Jefferson.[296] Yet even in Jefferson’s own administration it
became apparent that if the United States intended to carry
on trade with all parts of the world, the government must
intervene for the protection of its own citizens whenever this
should become necessary. So, in 1803, the American fleet
was sent to the Mediterranean, where it bombarded a nest of
pirates who had been interfering with American commerce.
Then came the War of 1812, which grew out of foreign
interference with American trade. On several subsequent
occasions during the nineteenth century the policy of
protecting and promoting foreign trade drew the United
States into negotiations with various countries of Europe and
Asia. In a sense, therefore, the United States has never
pursued a policy of complete isolation; on the other hand no
permanent alliances have been made with any country, and
the principle of independence in all matters of foreign policy
 has been consistently maintained. So far as diplomatic
matters did not directly concern North, Central, or South
America, the statesmen of the world could safely leave the
United States out of their reckonings during the greater part
of the nineteenth century. In diplomacy the United States
belonged, so to speak, to a different world.

THE SPIRIT OF LIGHT. By Edwin A. Abbey

Copyright by Edwin A. Abbey. From a Copley Print, copyright

by Curtis & Cameron, Boston. Reproduced by permission.

THE SPIRIT OF LIGHT

By Edwin A. Abbey
From a mural painting in the Pennsylvania
State Capitol at Harrisburg.
This is a very striking picture, one of the
artist’s best. In the background are the huge
 derricks which lift the oil from the bowels of the
earth. In front of them golden-haired figures,
robed in gauze with torch in hand, are swirling
upward in joyous energy like a swarm of fireflies.
In making this picture the artist took infinite
pains. Each figure was first drawn from a living
model. Each was then photographed and by the
use of a lantern the figures were projected upon
the canvas where they were manœuvred into
place for the artist’s guidance. The whole picture
is successful in conveying the impression of
spontaneity combined with lightness and grace.

Why Isolation was Possible.—This substantial isolation

was made possible for more than a hundred years by three
features. The first is the favored 1. The fortunate
geographical position of the country. The geographical
United States, as a strong nation, has position of the
stood alone in the Western Hemisphere. United States.
Her only neighbors were European colonies and the
struggling states of Latin-America. So long, therefore, as the
powerful nations of Europe could be held at arm’s length
there was no reason why the United States should give much
thought to problems of defence, alliances, and diplomacy.
Nature gave the United States an advantage in this respect
which is not possessed by any other strong nation with the
exception of Japan. Countries like England, France, and
Germany could not have pursued a policy of isolation even if
their people had desired it, for they are too close to each
other.
In the second place the United States 2. The abundance
was encouraged to hold aloof from the of land.
older countries of the world by the fact that there was plenty
of room for expansion at home. For a hundred years there
was no need to go abroad seeking new territories. It took the
United States a whole century to develop and populate the
solid block of country which extends from the Atlantic to the
Pacific. When other countries desired places of overflow for
their population and new fields of investment for their capital,
they engaged in a race for colonial possessions. The United
States had no such need or ambition; there was quite
enough opportunity at home.
Finally, the traditional policy of 3. The absence of
isolation was made possible by the good European
fortune which prevented European intervention.
interference at critical times, notably during the American
Civil War, when there was serious danger that Great Britain
and France might combine to aid the South. If that had
actually happened, it is not unlikely that Russia would have
come to the aid of the North, and the Civil War would then
have developed into a world conflict. In that case American
isolation would have ended more than a half century ago. But
good fortune, aided by competent diplomacy, enabled the
United States to settle its own troubles without foreign
interference and to continue the traditional policy of incurring
no obligations to any other country. In a word there was the
will to keep aloof and, what is quite as important, the
opportunity to do it.[297] From the War of Independence down
to the year 1917 the United States entered into no military
alliance or association with any other country; when the
American armies fought, they fought alone.
America’s Entry into the World War.—The World War
created a situation which the United States had never faced
before. All Western Europe burst ablaze; The old policy of
one country after another was drawn in; isolation comes to
and hostilities soon spread beyond the an end.
borders of the Old Continent. From the outset the United
States endeavored to maintain a strict neutrality; but
American commerce was subjected to interference by the
belligerents on both sides. Particularly offensive to the United
States, however, was the German practice of sinking without
warning passenger vessels upon which American citizens
were traveling. The torpedoing of the British liner Lusitania,
and the consequent loss of many American lives, stirred
public opinion throughout the United States. This and other
offences against the law of nations moved President Wilson
to demand from the German government a pledge that the
practice of sinking vessels without warning should cease,
and this pledge was conditionally given. Early in 1917,
however, the German government decided to inaugurate, as
a desperate stroke, a campaign of “unrestricted submarine
warfare”, and the government of the United States was
informed that even neutral vessels, unless they observed
certain strict precautions, would be torpedoed without
warning.
This action settled the matter of America’s continued
neutrality. Diplomatic relations with The declaration of
Germany were broken off and in April, war in 1917.
1917, Congress passed a declaration of war. The events of
the next eighteen months are still fresh in everyone’s mind.
America entered the struggle with a determination to turn the
scale, and on November 11, 1918, the German military
authorities were brought to terms. By signing an armistice
they acknowledged defeat and agreed to terms dictated by
the Allied and Associated Powers.
The Fourteen Points.—Some months before the signing
of this armistice President Wilson, in an address to
Congress, set forth the principal aims of the United States in
the war. These aims were grouped under fourteen heads and
soon came to be known as the Fourteen Points. Every one of
them had to do with matters which, prior to the war, would
have been deemed of no immediate concern to the United
States. Taken as a whole, however, they outlined the
principles upon which, in President Wilson’s opinion, a
durable peace could be erected and the future security of the
world maintained. The German government, in asking for an
armistice, declared its acceptance of these principles.
The Treaty of Versailles.—After the armistice had been
signed on behalf of the various belligerents a conference was
convened at Versailles to draw up a definite treaty of peace.
This conference included delegates from the countries which
had shared in the winning of the war. Germany and her
allies, the vanquished, were not represented. For several
months the conference wrestled with the problems involved
in the making of a treaty—the rearrangement of boundaries,
the recognition of new states, the disposal of German
colonies, the payment of reparations, and, most difficult of all,
the forming of a league of nations to prevent future wars.
When the work was finished the German representatives
were called in and were required to sign the treaty
substantially without any changes. The treaty was then
communicated to the various countries to be ratified and in
due course it was ratified by all the important countries
except the United States.
The New World Order.—The war and Why isolation is no
the changes which accompanied it longer possible:
served to alter the whole world environment. America was
brought into more intimate contact with Europe than ever
before. Even before the war, however, it had become
apparent that the traditional policy of isolation could not be
permanently maintained. To all intents and purposes the
world has become much smaller in these latter days. In point
of miles America is just as far away from Europe as ever, but
a thousand miles count for less nowadays than did a
hundred in our great-grandfathers’ time. During the summer
of 1918 the United States transported to Europe in less than
four months a million men. Fifty years ago that would have
been deemed to be an utterly impossible achievement. The
fast steamship of today can cross the 1. The annihilation
ocean in a hundred hours; in of distance.
Washington’s time the fleetest sailing-ships could not skim
the Atlantic in less than three weeks on the average. The
time is soon coming, in all probability, when men can be in
London one day and in New York the next. This is not a mere
dream; it is well within the range of possibilities. So we can
no longer talk of geographical isolation. The progress of
mankind has virtually annihilated distance.
 Again, the United States is no longer, 2. The acquisition of
as in the old days, devoid of tangible overseas
interests in distant parts of the earth. possessions.
Beginning in 1898, Porto Rico, the Philippines, and Guam
were acquired from Spain, and Hawaii was annexed. Later
the Panama Canal was built and a zone of territory on both
sides of it acquired. More recently, the Virgin Islands were
purchased from Denmark. All this has involved a departure
from the traditional policy of acquiring local interests only. It
has given America, in the case of the Philippines, an outpost
several thousand miles away. Whatever, therefore, concerns
the Malay Archipelago or, indeed, any part of the Far East,
concerns the interests of the United States. Isolation is no
longer possible because the United States has surrendered,
in this case at any rate, the geographical advantage of
isolation.
Finally, during the past few years, the 3. The acquisition of
relation of the United States to the rest of interests through
the world has been changed by reason the war.
of the interests acquired through the war. The fact that the
Treaty of Versailles did not receive the approval of the
Senate does not in any way impair the rights and interests
which the United States acquired as one of the victors in the
war. Those interests, obtained at great sacrifice and
acknowledged by Germany in the separate treaty which the
United States made with that country in 1921, are spread
over virtually the entire world. They are of incalculable value,
present and future. No policy of isolation is now possible
unless the country is ready to abandon these privileges
altogether, and, for reasons which will presently be stated,
the surrender of these various American interests is out of
the question. In the new world order the United States cannot
hold off from the rest of the world. The policy of a nation is
determined by what it regards as its own vital interests.
The New American Interests.— Wide scope of
Some important interests in various parts these interests.
of the world were acquired by the United States before the
war; others have been obtained or intensified as a result of it.
The scope and nature of these interests may best be
explained, perhaps, by grouping them under four main
heads, according to their general geographical location,
namely, Europe, Central and South America, the Far East,
and the Near East. It is not possible to arrange them in the
order of their relative importance, for only the future can
determine what this order of importance may turn out to be.
Certain it is, however, that in all four world-areas the interests
of the United States are of vast consequence not only to the
American people but to the cause of world peace and
prosperity. Above and beyond all, moreover, is the vital
interest of America in the maintenance of international amity.
Apart from the loss of life, America’s participation in the
World War cost the country, directly and indirectly, more than
thirty thousand million dollars. That is indeed a heavy price to
pay for helping to settle a quarrel which the United States
had no part in promoting. It surely requires no argument to
prove that America has a vital interest in avoiding another
such calamity.
America and Europe.—The war America’s interests
resulted in placing Great Britain, France, now extend to
Italy, and the other victorious countries of everything that may
Europe under heavy obligations to the threaten peace.
United States. To a certain extent these obligations are
sentimental; in return for America’s help towards winning the
war the other victorious countries are under a natural
obligation to give the United States an adequate share in
determining the permanent conditions of peace. This they
have been willing to do; but it involves responsibilities which
the United States has shown no great willingness to accept.
The old tradition of non-interference in strictly European
affairs is still strong and this has led the American
government to distinguish, wherever possible, between
questions of local and of world-wide concern. The distinction,
however, is practically impossible to make. The boundaries
of some small European state may seem to be a matter of no
concern at Washington; but if a disagreement over this
question should bring once more a general European clash
of arms, the importance of the issue would speedily be
recognized. So long as the general preservation of world-
peace is among the primary interests of the United States, as
it seems bound to be, no menace to peace, anywhere, at any
time, can be lightly regarded by the people of America.
The Loans to Associated Nations.—But the war did not
result in the creation of sentimental obligations only.
Obligations of great importance and a tangible nature on the
part of Europe to America grew out of it. During the conflict
the United States loaned large sums of money to Great
Britain, France, Italy, Russia, and the other Associated
Powers.[298] These loans were made generously, in the midst
of a grave emergency; but nothing definite was arranged as
to when or how they should be repaid. In How the loans were
view of the disorganized conditions in made.
Europe created by the war no request for the payment of
interest was made by the United States during the conflict or
for some years after its close. In 1922, however, Congress
authorized the President to appoint a commission of five
persons to arrange with the European countries for the
funding of the debts by the issue of bonds. These bonds will
be given to the United States.
Now it must be reasonably clear to anyone who gives the
matter a moment’s thought that until these bonds are paid off
by the various European countries (which will be thirty or
forty years hence) the United States will be vitally interested
in what Washington called the “vicissitudes” of the Old World.
America, in effect, holds a mortgage on America’s mortgage
Europe, and it is the practice of on Europe.
mortgage-holders to keep a sharp eye on their invested
funds. Great Britain, France, and the other debtor countries
expect to redeem these loans, in considerable part, out of
reparation payments made to them by Germany. If Germany
does not pay them, it will be much harder for them to pay
America.[299] In this roundabout way, therefore, the United
States has acquired a tangible interest in the pledges made
by the German government.
America’s Interest in the Industrial Reconstruction of
Europe.—The sum total of America’s interest in the peace
and prosperity of Europe is not represented, however, by
these ten billion dollars of loans. The commercial relations of
the two continents have become so intimate that whatever is
an injury to the one is a detriment to the other. Europe is
America’s best customer. Our exports there are greater than
to all the rest of the world put together. Importance of
The farmer, the cotton grower, the Europe as a
manufacturer—all depend in part upon market.
the European market. There is not sufficient demand at
home for all the foodstuffs, materials, and manufactured
goods which the United States can now produce. The
European market, however, has been broken down as the
result of the long conflict and it is greatly to the interest of the
United States that it should be built up again. This can only
be done by keeping the world at peace until the damage
done by the war has been repaired. For that commercial
reason, if for no other, the United States cannot well afford to
remain entirely isolated from the rest of the world.
The United States and Latin-America.—The relations of
the United States with most of the Latin-American states
have been at all times friendly. We have never been at war
with any of them except Mexico. When these various
countries revolted against Spanish control about a hundred
years ago, the people of the United States, remembering
their own experience, were in sympathy with them. The
announcement of the Monroe Doctrine was regarded by
Latin-America as an act of friendliness. And for more than a
century since that time the United States has served as a
protector to the sister republics of the southern continent.
When the War with Spain began in 1898 Congress
announced that the United States had no intention to annex
Cuba and this pledge, at the close of the war, was kept.
Cuba was given her independence. Naturally this evidence of
good faith made a strong and favorable impression upon the
Central and South American states.
With Mexico, however, relations have Relations with
not been cordial for several years. Ever Mexico.
since the invasion of their country by an American army in
1846 the Mexican people have been suspicious of American
aggression; but the relations between the two governments
remained cordial enough so long as President Diaz
continued in power south of the Rio Grande, which was from
shortly after the close of the American Civil War until well into
the twentieth century. Diaz ruled Mexico in the fashion of a
dictator; but he kept the country peaceful as well as on good
terms with the outside world. Since the expulsion of Diaz the
Mexicans have had several changes in the presidency and
for ten years the government has been denied recognition by
the United States. The successors of Diaz have professed
their desire to place the government of the country on a truly
democratic basis and to some extent they have succeeded in
doing so; but they have not managed to maintain order and
justice with a firm hand. Twice during the past decade it has
been deemed necessary to send American troops into the
country. The government of Mexico is republican in form, but
elections have not, as a rule, been fairly conducted. The
leaders who have control of the government try to manipulate
the elections so as to maintain their own hold upon the
country, and they usually succeed.
Between Mexico and the United The situation today.
States there are today no questions of
great importance in dispute. The United States is ready to
recognize the existing government of Mexico but only upon
condition that certain pledges are made in writing. These
include assurances that payments will be made by Mexico as
compensation for the lives and property of American citizens
destroyed during the troubles of the past ten years; that there
shall be no confiscating of property without legal reason in
the future; and that payments of interest on Mexico’s foreign
obligations shall be resumed. These do not appear to be
unreasonable conditions.
There are large American investments in Mexico,
particularly in the oil and mining districts. Some of those who
hold these investments would like to see the United States
intervene by force of arms, but it is quite unlikely that there
will be anything of the kind unless all other means of
securing the rights of Americans in Mexico prove unavailing.
The United States has a certain moral responsibility for the
good behavior of Mexico, even though the Mexican
government may not recognize the existence of such an
obligation. If the Monroe Doctrine gives Mexico and the
the United States the right to keep Monroe Doctrine.
European countries from interfering in Mexican affairs, even
when their citizens have been wronged, it may also be said
to carry the duty of seeing that Mexico does not abuse this
protectorship.
In the region of the Isthmus the Panama.
interests of the United States are
especially important because of the Canal. The Panama
Canal is not only of commercial but of military value to the
United States, and no serious disturbance of the peace in
this section of Central America can well be tolerated.
Many years ago the United States The Pan-American
government made the suggestion that Congress.
from time to time a Pan-American Congress made up of
delegates from all the republics of the New World should be
held to discuss matters of common interest. The suggestion
was accepted and several Congresses have been held
during the past three decades. There has also been
established at Washington a Bureau of American Republics
whose function it is to carry out the resolutions of each
Congress and to spread information concerning the common
interests of all the countries.
The United States and the Far East.—In ordinary usage
the term “Far East” includes the Japanese and Chinese
empires, Siberia and the other Russian territories to the north
of China, and the Malay Archipelago to the south. Until a
quarter of a century ago the interests of the United States,
whether political or commercial, were relatively small in this
part of the world. But the acquisition of The Philippines.
the Philippines and the growth of
American trade with the Orient have combined to alter the
situation. Another factor which has impelled the United
States to pay greater attention to the Orient today is the
progress of Japan. The rapid growth of this empire in military
and naval strength means that the United States has a rival
for the mastery of the Pacific. During the nineteenth century
the eyes of America were turned entirely towards Europe; in
the twentieth they will have to be turned towards Asia as
well.
Apart from affairs in the Philippines the China and Japan.
problems of the Far East, so far as the
United States is concerned, center around two present-day
international phenomena, the weakness of China and the
strength of Japan. China is a vast country with at least three
or four times the population of the United States. Although
nominally a republic its government is weak, inefficient,
unable to exercise firm control over all parts of the country,
and without effective means of national defence. Quite
naturally, therefore, China offers a temptation to any strong
country desiring exclusive trade advantages for itself. Her
nearest neighbor, Japan, would speedily be able to secure
entire control of the Chinese Republic and make China a
vassal state were it not for the deterring influence of the other
great powers of the world.
In 1899, after the close of the Spanish War, the
government of the United States addressed a note to all the
great powers urging that they agree to seek no further
special trade advantages in China, that the integrity of
Chinese territory be preserved, and that the principle of
“equal and impartial trade” should be adopted. To this
suggestion all the powers agreed. This The “open door.”
policy thus accepted has become known
as the policy of the “open door”, and until the outbreak of the
World War it was substantially followed, except that the
various powers retained the commercial advantages that
they had already acquired.
During this war, however, Japan attacked and captured
Kiao-Chao, a port which had been leased by China to
Germany for a long term of years, and this territory the
Japanese continued to hold after the war was over. Not until
the Washington conference of 1922 did Japan agree to give
it up. In 1918, moreover, the government Recent
of Japan made a list of twenty-one developments.
demands upon China for special privileges, and although
some of these demands were later modified or withdrawn
entirely, several important privileges were wrung from the
Chinese. In connection with these negotiations the United
States government gave assurance in the so-called Ishii-
Lansing agreement that the United States would recognize
the “special interest” of Japan in Chinese affairs. It is
avowedly the policy of Japan to acquire, if she can, the same
predominance in Asia that the United States has exercised in
North and South America.
The Conference on Pacific Problems. Regarding it as
highly desirable that all controversies affecting the Far East
and the Pacific should be amicably settled, thus forestalling
the growth of large naval armaments on both sides of the
Western ocean, President Harding in the summer of 1921
proposed that the Washington conference should discuss
these questions and should endeavor to secure a
satisfactory solution of them. The The Washington
conference did so, and embodied the conference.
results of its negotiations in certain agreements, particularly
in what is commonly known as the “Four Power” treaty. By
the terms of this treaty the United States, Great Britain,
France, and Japan mutually agree to respect the integrity of
each other’s possessions in the islands of the Pacific.[300] The
nations represented at the conference also agreed to refrain
from the erection of fortifications in certain places now
unfortified.
Out of the negotiations at Washington, moreover, came the
agreement on the part of Japan to restore Kiao-Chao and the
adjacent province of Shantung to China. First and last,
therefore, the Washington conference succeeded in
promoting an amicable agreement on most of the questions
at issue. It did not, however, take up the question of
Japanese immigration to the Pacific Coast of America, nor
did it discuss the grievances of the Japanese immigrants
already there. These matters are left for further negotiation
through the regular diplomatic channels.
America and the Near East.—The expression “Near
East” is commonly regarded as including the areas which lie
at the eastern end of the Mediterranean and thereabouts; it
comprises Turkey, Armenia, Syria, Palestine, Arabia,
Mesopotamia, and other territories in the same general
region. As a result of the war the Turkish Empire has been
disintegrated; most of its territories have been virtually placed
under the control of Great Britain, France, Italy, and Greece
through the instrumentality of mandates (see p. 636). The
United States was offered the mandate for Armenia, but
declined to accept it.
Now some of these territories are rich The oil question.
in natural resources. Mesopotamia, for
example, is known to possess extensive oil fields. The
question arises, therefore, whether the European countries
which hold the mandates are to have the lion’s share of this
natural wealth. And it is a question of considerable
importance when one bears in mind the fact that the oil fields
of the United States will probably be exhausted before many
decades have passed (see p. 330). The direct interest of the
United States is less immediate, perhaps, than in the other
areas (Europe, Central and South America, and the Far
East), but it is sufficiently vital to deserve mention.
The Wide Scope of America’s Interests.—From this brief
and general survey some idea of the scope of American
interests can be gained. But the preceding paragraphs have
not listed them all. The people of the United States have a
sentimental interest in many foreign problems where no
economic considerations are at stake. Ireland is an example.
America’s interest in a just and peaceful settlement of the
Irish question is not inspired by economic motives. It arises in
large part from the sentimental desire to see a people, with
whom there are close ties of kinship, attain contentment and
prosperity. So with Poland and the new Slavic countries of
Continental Europe. America would regret to see them lose a
status of independence which was gained at so great a
sacrifice.
There is no part of the world, in fact, to which the interest
of the United States, direct or indirect, sentimental, political,
or economic, does not now extend. The enormous strength
and prestige of America, as disclosed during the war, have
made a profound impression in every part of the globe and
have given the United States a potent influence upon the
destinies of mankind. The United States has become a world
power of the first order. Whether the American people like it
or not, that inexorable fact remains.
General References
A. T. Mahan, The Interest of America in International Conditions, pp.
127-185;
W. E. Weyl, American World Policies, passim;
A. C. Coolidge, The United States as a World Power, pp. 95-120;
C. E. Jones, Caribbean Interests of the United States, pp. 148-192;
J. H. Latané, America as a World Power, pp. 255-268; Ibid., From
Isolation to Leadership, pp. 3-39;
H. H. Powers, America Among the Nations, pp. 197-239;
W. A. Dunning, The British Empire and the United States, pp. 357-371;
A. B. Hart, Foundations of American Foreign Policy, pp. 1-52;
Woodrow Wilson, State Papers and Addresses, pp. 464-479;
P. S. Reinsch, World Politics, pp. 327-362.
Group Problems
1. How the United States became a world power. Early relations with
Europe. The Monroe Doctrine. The opening of Japan. The war with Spain
and the new acquisitions. John Hay and the “open door.” The World War
and its aftermath. Scope of American interests today. References: J. H.
Latané, America as a World Power, pp. 3-28; 63-81; Ibid., The United
States and Latin America, pp. 61-291; A. C. Coolidge, The United
States as a World Power, pp. 121-147; J. W. Foster, American
Diplomacy in the Orient, pp. 399-438; W. M. Fullerton, Problems of
Power, pp. 11-43; P. S. Reinsch, World Politics, pp. 309-336.
2. How foreign trade affects the national welfare. Foreign trade and
national prosperity. Foreign trade and world power. “Dollar Diplomacy.”
Trade and imperialism. References: C. M. Pepper, American Foreign
Trade, pp. 3-32; 62-89; 110-139; J. D. Whelpley, The Trade of the
World, pp. 391-425; A. J. Wolfe, Theory and Practice of International
Commerce, pp. 495-522; C. L. Jones, Caribbean Interests of the United
States, pp. 1-16; L. C. and T. F. Ford, The Foreign Trade of the United
States, pp. 1-27.
3. How the building of the Panama Canal extended American
interests abroad. References: Lincoln Hutchinson, The Panama
Canal and International Trade Competition, pp. 46-97; F. A. Ogg,
National Progress, pp. 246-265; A. B. Hart, The Monroe Doctrine: An
Interpretation, pp. 340-348; W. M. Fullerton, Problems of Power, pp.
300-315.
4. The open door in China. What it means. Obstacles in its way.
References: K. K. Kawakami, Japan in World Politics, pp. 117-166; J. H.
Latané, America as a World Power, pp. 100-119; W. W. Willoughby,
Foreign Rights and Interests in China, pp. 245-266; A. B. Hart, The
Monroe Doctrine: An Interpretation, pp. 282-298; W. R. Thayer, The Life
and Letters of John Hay, Vol. II, pp. 231-249; A. C. Coolidge, The
United States as a World Power, pp. 327-374; K. K. Kawakami, Japan
and World Peace, pp. 160-196; John Dewey, China, Japan, and the
United States (New Republic Pamphlets, No. 1).
Short Studies
1. Pan-Americanism. What it means. R. G. Usher, Pan-Americanism,
pp. 203-231; J. V. Noel, The History of the Second Pan-American
Congress, pp. 9-30; Pan-American Union, Bulletins, March, 1911.
2. The United States as an international policeman. Theodore
Roosevelt, Autobiography, pp. 543-553; D. C. Munro, The Five
Republics of Central America, pp. 227-264; A. B. Hart, The Monroe
Doctrine: An Interpretation, pp. 223-242.
3. Our Mid-Pacific possessions. E. J. Carpenter, America in
Hawaii, pp. 192-251; J. M. Callahan, American Relations in the Pacific
and the Far East, pp. 114-145.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com