Artificial Intelligence 1st edition by Peggy

Thomas ISBN 1590184378  ‎ 978-1590184370

download

https://ebookball.com/product/artificial-intelligence-1st-
edition-by-peggy-thomas-
isbn-1590184378-aeurz-978-1590184370-19590/

Download more ebook instantly today - Get yours now at ebookball.com

 Get Your Digital Files Instantly: PDF, ePub, MOBI and More
Quick Digital Downloads: PDF, ePub, MOBI and Other Formats

Introduction To 80 86 Assembly Language And Computer Architecture 1st

Edition by Detmer ISBN 0763717738 9780763717735

https://ebookball.com/product/introduction-to-80-86-assembly-
language-and-computer-architecture-1st-edition-by-detmer-
isbn-0763717738-9780763717735-12404/

Introduction to 80 86 Assembly Language and Computer Architecture 1st

Edition by Richard C Detmer ISBN 0763746622 9780763746629

https://ebookball.com/product/introduction-to-80-86-assembly-
language-and-computer-architecture-1st-edition-by-richard-c-
detmer-isbn-0763746622-9780763746629-9016/

Machine Learning for Cybersecurity Cookbook Over 80 recipes on how to

implement machine learning algorithms for building security systems
using Python 1st edition by Emmanuel Tsukerman 9781838556341
1838556346
https://ebookball.com/product/machine-learning-for-cybersecurity-
cookbook-over-80-recipes-on-how-to-implement-machine-learning-
algorithms-for-building-security-systems-using-python-1st-
edition-by-emmanuel-tsukerman-9781838556341-1/

Artificial Intelligence and Education Frontiers in Artificial

Intelligence and Applications 1st Edition by Bierman, Breuker,
Sandberg ISBN 9051990146 9789051990140

https://ebookball.com/product/artificial-intelligence-and-
education-frontiers-in-artificial-intelligence-and-
applications-1st-edition-by-bierman-breuker-sandberg-
isbn-9051990146-9789051990140-19708/
AI 2011 Advances in Artificial Intelligence Lecture Notes in
Artificial Intelligence 1st edition by Dianhui Wang, Mark Reynolds
7106 9783642258329 3642258328

https://ebookball.com/product/ai-2011-advances-in-artificial-
intelligence-lecture-notes-in-artificial-intelligence-1st-
edition-by-dianhui-wang-mark-
reynolds-7106-9783642258329-3642258328-11140/

KI 2006 Advances in Artificial Intelligence 1st Edition by Anique ISBN

https://ebookball.com/product/ki-2006-advances-in-artificial-
intelligence-1st-edition-by-anique-isbn-13700/

Artificial Intelligence Illuminated 1st edition by Ben Coppin ISBN

0763732303 9780763732301

https://ebookball.com/product/artificial-intelligence-
illuminated-1st-edition-by-ben-coppin-
isbn-0763732303-9780763732301-11676/

KI 2007 Advances in Artificial Intelligence 1st Edition by Anique ISBN

https://ebookball.com/product/ki-2007-advances-in-artificial-
intelligence-1st-edition-by-anique-isbn-13698/

Artificial Intelligence 2nd Edition by Margaret Boden ISBN

9780080527598 0080527590

https://ebookball.com/product/artificial-intelligence-2nd-
edition-by-margaret-boden-isbn-9780080527598-0080527590-19698/
 Artificial
Intelligence
Other books in the Lucent Library of Science and Technology series:

Bacteria and Viruses

Black Holes
Comets and Asteroids
Energy Alternatives
Exploring Mars
Genetics
Global Warming
Lasers
Space Stations
Telescopes
Virtual Reality
 Artificial
Intelligence
by Peggy Thomas
On cover: The robot Kismet responds with a look of wonder to re-
searcher Cynthia Breazeal, who programmed it to communciate its
moods through complex facial expressions.

© 2005 Thomson Gale, a part of The Thomson Corporation.

Thomson and Star Logo are trademarks and Gale and Lucent Books are registered trademarks used
herein under license.

For more information, contact

Lucent Books
27500 Drake Rd.
Farmington Hills, MI 48331-3535
Or you can visit our Internet site at http://www.gale.com

ALL RIGHTS RESERVED.

No part of this work covered by the copyright hereon may be reproduced or used in any form or by any
means—graphic, electronic, or mechanical, including photocopying, recording, taping, Web
distribution or information storage retrieval systems—without the written permission of the publisher.

Every effort has been made to trace the owners of copyrighted material.

LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA

Thomas, Peggy.
Artificial intelligence / by Peggy Thomas.
p. cm. — (Lucent library of science and technology)
Includes bibliographical references and index.
ISBN 1-59018-437-8 (lib. bdg. : alk. paper)
1. Artificial intelligence—Juvenile literature. I. Title. II. Series.
Q335.4.T48 2005
006.3—dc22 2004016477

Printed in the United States of America

 Table of Contents
Foreword 6
Introduction 9
Fact Follows Fiction
Chapter 1 13
The First Thinking Machines
Chapter 2 28
Mind Versus Metal
Chapter 3 41
Everyday AI
Chapter 4 54
AI and Robotics
Chapter 5 68
In Pursuit of the Mechanical Man
Chapter 6 84
The Future of AI
Notes 98
Glossary 100
For Further Reading 102
Works Consulted 103
Index 106
Picture Credits 111
About the Author 112
 Foreword

“The world has changed far more in the past 100 years
than in any other century in history. The reason is not
political or economic, but technological—technologies
that flowed directly from advances in basic science.”
— Stephen Hawking, “A Brief History
of Relativity,” Time, 2000

T he twentieth-century scientific and technological

revolution that British physicist Stephen Hawking
describes in the above quote has transformed virtually
every aspect of human life at an unprecedented pace.
Inventions unimaginable a century ago have not only
become commonplace but are now considered necessi-
ties of daily life. As science historian James Burke writes,
“We live surrounded by objects and systems that we take
for granted, but which profoundly affect the way we be-
have, think, work, play, and in general conduct our
lives.”
For example, in just one hundred years, transporta-
tion systems have dramatically changed. In 1900 the
first gasoline-powered motorcar had just been intro-
duced, and only 144 miles of U.S. roads were hard-
surfaced. Horse-drawn trolleys still filled the streets of
American cities. The airplane had yet to be invented.
Today 217 million vehicles speed along 4 million miles
of U.S. roads. Humans have flown to the moon and
commercial aircraft are capable of transporting passen-
gers across the Atlantic Ocean in less than three hours.
The transformation of communications has been just
as dramatic. In 1900 most Americans lived and worked
on farms without electricity or mail delivery. Few peo-
ple had ever heard a radio or spoken on a telephone. A
hundred years later, 98 percent of American homes have
6
Foreword 7

telephones and televisions and more than 50 percent

have personal computers. Some families even have more
than one television and computer, and cell phones are
now commonplace, even among the young. Data
beamed from communication satellites routinely pre-
dict global weather conditions, and fiber-optic cable,
e-mail, and the Internet have made worldwide telecom-
munication instantaneous.
Perhaps the most striking measure of scientific and
technological change can be seen in medicine and pub-
lic health. At the beginning of the twentieth century, the
average American life span was forty-seven years. By the
end of the century the average life span was approach-
ing eighty years, thanks to advances in medicine in-
cluding the development of vaccines and antibiotics, the
discovery of powerful diagnostic tools such as X-rays, the
life-saving technology of cardiac and neonatal care, and
improvements in nutrition and the control of infectious
disease.
Rapid change is likely to continue throughout the
twenty-first century as science reveals more about phys-
ical and biological processes such as global warming, vi-
ral replication, and electrical conductivity, and as people
apply that new knowledge to personal decisions and gov-
ernment policy. Already, for example, an international
treaty calls for immediate reductions in industrial and
automobile emissions in response to studies that show
a potentially dangerous rise in global temperatures is
caused by human activity. Taking an active role in de-
termining the direction of future changes depends on
education; people must understand the possible uses of
scientific research and the effects of the technology that
surrounds them.
The Lucent Books Library of Science and Technology
profiles key innovations and discoveries that have trans-
formed the modern world. Each title strives to make a
complex scientific discovery, technology, or phenome-
non understandable and relevant to the reader. Because
scientific discovery is rarely straightforward, each title
8 Artificial Intelligence

explains the dead ends, fortunate accidents, and basic sci-

entific methods by which the research into the subject
proceeded. And every book examines the practical appli-
cations of an invention, branch of science, or scientific
principle in industry, public health, and personal life, as
well as potential future uses and effects based on ongoing
research. Fully documented quotations, annotated bib-
liographies that include both print and electronic sources,
glossaries, indexes, and technical illustrations are among
the supplemental features designed to point researchers
to further exploration of the subject.
 Introduction

Fact Follows
Fiction

A s far back as the ancient Greek civilization, peo-

ple have imagined machines and mechanical men
that could work and think like any human. One Greek
myth, for example, tells of the Greek god Hephaestus,
who built mechanical men to forge powerful weapons
and spectacular jewelry. When the king of Crete re-
quested that he make a giant man to guard his island,
Hephaestus constructed the metallic warrior Talos.
Talos patrolled the shores of Crete until Jason and the
Argonauts defeated it.
In the late 1800s, as the genre of science fiction de-
veloped, books about intelligent machines serving hu-
man masters sparked the public’s imagination, but not
until the 1920s did such machines acquire the name
robot. In 1920 Czech playwright Karel Capek published
the play R.U.R. (Rossum’s Universal Robots) in which
the character Professor Rossum manufactures artifi-
cial men to do all the menial chores, called robota in
the Czech language. Rossum’s plan backfires and
armies of robots are purchased by warring nations.
Eventually the robots themselves revolt and attempt
to take over all of humankind.
In the 1950s and 1960s a series of movies and books
portrayed the blessings but more often the horrors of
intelligent machines, which took all forms. Some were
Talos-like robots made of shiny metal, like Gort in The
9
10 Artificial Intelligence

Day the Earth Stood Still (1951); others were intelligent

supercomputers that threatened to take over the world
in movies like Colossus: The Forbin Project (1969). In
this film the supercomputer that ran the U.S. national
defense systems overrode all human control. But per-
haps the most famous malevolent supercomputer was
HAL 9000, the sinister manager of the spaceship
Discovery in 2001: A Space Odyssey (1968). HAL could
learn and act independent of human input and in so
doing it killed all but one crew member.

Filmmaker George
Lucas created the
lovable robots R2-D2
and C-3PO for his
popular science fiction
saga Star Wars.
Fact Follows Fiction 11

Not all robots or supercomputers have been por-

trayed as evil. In the 1970s moviegoers fell in love with
the flighty C-3PO and quirky trashcan-shaped R2-D2
in the Star Wars series. Without the help of these au-
tonomous beings, the heroes would not have prevailed
over the evil empire. And in 2001 audiences connected
with the boy robot in Steven Spielberg’s film AI when
he expresses emotions and feelings of love toward his
adoptive human parents.
These fictional stories reflect humans’ dreams and
fears about intelligent machines. And there is an odd
correlation between science fiction and science fact.
Writers have envisioned worlds that are technologi-
cally several steps ahead of reality. They imagined ro-
bots on Mars long before the first moon landing and
portable minicomputers long before the invention of
laptops.
Scientists who work in the field of artificial intelli-
gence (AI), the study of intelligent machines, also fore-
see a future filled with independent, thinking robots
and computer companions. Unlike fiction writers,
however, they face the daunting challenge of making
their dreams a reality. Each year AI researchers come
closer to realizing their dreams with new develop-
ments in computer programming and robotic engi-
neering.

Real-life AI
As in fictional stories, the pursuit of intelligent ma-
chines has taken two forms. Researchers have created
robotic bodies—like that of C-3PO—that look and
move like a human. They have even created artificial
material similar to human skin that will mask the com-
puter chips and wiring inside a robot’s head. Other re-
searchers are working on robots that learn from ex-
perience—like those in Isaac Asimov’s classic story I,
Robot—and can express emotions like the robotic boy
in director Steven Spielberg’s feature film AI.
12 Artificial Intelligence

The second form of intelligent machines is the all-

knowing and all-powerful “beings without bodies” like
HAL. Artificial intelligence experts have already cre-
ated supercomputers that navigate and control the
space shuttles; other computer systems are powerful
enough to effectively run major companies. But AI is
not just reserved for grand space exploration or high
finance. In fact, much of a person’s daily life is affected
by some form of artificial intelligence. AI computer
programs keep track of a person’s banking, translate
foreign languages, locate a car’s position, and put once
hard-to-find information at a person’s fingertips with
the Internet. Each year smart machines become more
proficient at chores we used to do for ourselves, and
people purchase the newest electronic AI gadgets in
hopes of making their lives easier. But what happens
to a society that gives more and more control to ma-
chines? Will fact follow fiction? Will people enjoy a
life of friendly companionship with robots like R2-D2
and live blissfully by relying on all-knowing machines
to help them get through their days?
Anyone who has been beaten by a computer at a
game of chess knows the unsettling feeling of dealing
with an intelligent system. Will that unease grow? And
as more jobs are given to computers, how will that
change humanity? Will people lose skills like remem-
bering phone numbers or calculating large sums? Will
people become slaves to the very machines they have
created and lose their humanity in a world of mech-
anization? Will the fictional cautionary tales be
heeded, or will the future hold a wondrous collabo-
ration between man and machine? Only time will tell
as scientists continue to forge ahead and attempt to
make real the amazing dreams of fiction writers.
 Chapter 1

The First Thinking

Machines

G aak was missing. No one knew what it was think-

ing or if it was thinking at all. The only thing that
robotics expert Noel Sharkey knew was that the small
robotic unit he had just repaired had disappeared.
Gaak had been injured in battle during a “survival of
the fittest” demonstration at the Living Robots exhi-
bition in Rotherham, England, in 2002. In this con-
test predator robots sought out prey robots to drain
their energy, and prey robots had to learn to avoid
capture or be inactivated.
Gaak, a predator robot, may have had enough of the
competition. Only fifteen minutes after Sharkey left the
robot’s side, the autonomous machine forced its way
out of its corral, sidled past hundreds of spectators, ma-
neuvered down an exit ramp, and left the building.
Gaak’s bid for freedom was stopped short when it was
almost run over by a car as it fled toward the exit gate.
Although this may sound like another sci-fi night-
mare, it is actually an example of artificial intelligence
at work today. AI is the study and creation of machines
that can perform tasks that would require intelligence
if a human were to do the same job. This emerging
and constantly changing field combines computer
programming, robotics engineering, mathematics,
neurology, and even psychology. As a blend of many
sciences, AI has a scattered history. It has almost as
13
14 Artificial Intelligence

many processes as there are researchers active in the

field. Like the limbs of a tree, each new idea spawns
another, and the science of artificial intelligence has
many branches. But its roots were planted by scien-
tists and mathematicians who could imagine all the
possibilities and who created amazing machines that
startled the world.

The Analytical Engine

The first glimmer of a “thinking machine” came in the
1830s when British mathematician Charles Babbage en-
visioned what he called the analytical engine. Babbage
was a highly regarded professor of mathematics at
In the 1830s, British Cambridge University when he resigned his position
mathematician to devote all of his energies to his revolutionary idea.
Charles Babbage In Babbage’s time, the complex mathematical tables
envisioned the
used by ship’s captains to navigate the seas, as well as
world’s first
intelligent machines. many other intricate computations, had to be calcu-
lated by teams of mathematicians who were called
computers. No matter how painstaking these
human computers were, their tables were
often full of errors. Babbage wanted to
create a machine that could automati-
cally calculate a mathematical chart
or table in much less time and with
more accuracy. His mechanical com-
puter, designed with cogs and gears
and powered by steam, was capa-
ble of performing multiple tasks by
simple reprogramming—or chang-
ing the instructions given to the
computer.
The idea of one machine per-
forming many tasks was inspired by
the giant industrial Jacquard looms
built by French engineer Joseph-Marie
Jacquard in 1805. These looms performed
mechanical actions in response to cards that
had holes punched in them. Each card provided
The First Thinking Machines 15

Babbage’s Difference Engine

Although no one may ever see a real example of Charles Babbage’s
ingenious Analytical Engine, a Difference Engine was re-created in
the 1990s from Babbage’s original drawings by a team of engineers
at London’s Science Museum. Made of cast iron, bronze, and steel,
the workable machine stands ten feet wide and six feet high and
weighs three tons. The Difference Engine performs mathematical
computations that are accurate up to thirty-one digits. But unlike the
Analytical Engine, Babbage’s Difference Engine is powered by hand.
The computer operator has to turn a crank hundreds of times to per-
form one calculation.

This recent
reproduction
of Babbage’s
Difference
Engine performs
mathematical
calculations
accurate to
thirty-one digits.

the pattern that the loom would follow; different cards

would instruct the loom to weave different patterns.
Babbage realized that the instructions fed into a ma-
chine could just as easily represent the sequence of in-
structions needed to perform a mathematical calcula-
tion as it could a weaving pattern.
His first attempt was called the Difference Engine.
It could translate instructions punched on input cards
into arrangements of mechanical parts, store variables
16 Artificial Intelligence

in specially positioned wheels, perform the logical op-

erations with gears, and deliver the results on punched
output cards. Only one small version of the Difference
Engine was created before Babbage turned his atten-
tion to a more ambitious machine that could perform
more abstract “thinking.”
Babbage’s second design was a larger machine called
the Analytical Engine. It was designed to perform
many different kinds of computations such as those
needed to create navigational tables and read symbols
other than numbers. His partner in this venture was
Lady Ada Lovelace, daughter of the poet Lord Byron.
Unlike her father, who had a talent for words, the
countess had a head for numbers. She is often cred-
ited with inventing computer programming, the
process of writing instructions that tell a computer or
machine what to do. Of Babbage’s machine she said,
“The Analytical Engine weaves algebraical patterns
just as the Jacquard Loom weaves flowers and leaves.”1
Unfortunately Lovelace was never able to program
one of Babbage’s machines. As a result of financial
problems and the difficulty of manufacturing preci-
sion parts for his machine, Babbage had to abandon
his project. Almost a century and a half would pass
before a similar machine was assembled. Seldom has
such a long time separated an idea and its technolog-
ical application.

The Turing Machine

The next and perhaps most influential machine to
mark the development of AI was, once again, never
even built. It was a theoretical machine, an idea that
existed only on paper. Devised by the brilliant British
theoretical mathematician Alan Turing in 1936, this
simple machine consisted of a program, a data stor-
age device (or memory), and a step-by-step method of
computation. The mechanism would pass a long thin
tape of paper, like that in a ticker-tape machine,
through a processing head that would read the infor-
The First Thinking Machines 17

mation. This apparatus would be able to move the pa-

per along, read a series of symbols, and produce cal-
culations based on the input on the tape. The so-called
universal computer, or Turing machine, became the
ideal model for scores of other researchers who even-
tually developed the modern digital computer. Less
than ten years later, the three most powerful nations
in the world had working computers that played an
integral part in World War II.
Dr. J.W. Mauchly
The Giants of the Computer Age makes an adjustment
to ENIAC, the
Teams of British mathematicians, logicians, and en-
massive computer he
gineers created a computer called Colossus, which was designed to assist the
used to decode secret enemy messages intercepted U.S. military during
from Germany. The advantage of Colossus was that World War II.
18 Artificial Intelligence

the British military did not have to wait days for a

team of human decoders to unravel secret plans.
Colossus could decode messages within hours. It
helped save lives and was a key factor in the Allied
forces’ defeat of the Germans.
In the meantime, German mathematicians and
computer programmers had created a computer to
rival Colossus. Called Z3, the German computer was
used to design military aircraft. The United States also
entered the computer age with the development of a
computer called the Electronic Numerical Integrator
and Computer, or ENIAC, which churned out accu-
rate ballistic charts that showed the trajectory of bombs
for the U.S. Navy.
ENIAC, Z3, and Colossus were monsters compared
to modern-day computers. ENIAC weighed in at thirty
tons, took up three rooms, and consisted of seventeen
thousand vacuum tubes that controlled the flow of elec-
tricity. However, these tubes tended to pop, flare up,
and die out like fireworks on the Fourth of July. Six full-
time technicians had to race around to replace bulbs
and literally debug the system: Hundreds of moths, at-
tracted to the warm glow of the vacuum tubes, would
get inside the machine and gum up the circuits.

The Binary Code

ENIAC and all other computers that have followed
have spoken the same basic language: the binary code.
This is a system of symbols used to program a com-
puter. Proposed by U.S. mathematician Claude Shannon
and expanded on by Hungarian-born mathematician
John von Neumann in the 1940s, the binary code is
a language with only two symbols, 0 and 1. Shannon
and von Neumann showed that the simplest instruc-
tion was yes/no, or the flicking on and off of a switch,
and that any logical task could be broken down to this
switching network of two symbols. The system is bi-
nary, based on two digits, and combinations of these
two digits, or bits, represent all other numbers.
The First Thinking Machines 19

Binary Code Computers use a binary number system to quickly process

information. Electrical pulses travel through millions of
transistors, which are miniature electric switches that
have two positions: on and off. When a switch is on,
it represents binary digit 1. When a switch is off, it
represents binary digit 0.

The English alphabet in binary code:

A 01000001 G 01000111 M 01001101 S 01010011 Y 01011001
B 01000010 H 01001000 N 01001110 T 01010100 Z 01011010
C 01000011 I 01001001 O 01001111 U 01010101
D 01000100 J 01001010 P 01010000 V 01010110
E 01000101 K 01001011 Q 01010001 W 01010111
F 01000110 L 01001100 R 01010010 X 01011000

The words ARTIFICIAL INTELLIGENCE look like this in binary code:

01000001 01010010 01010100 01001001 01000110 01001001 01000011 01001001
01000001 01001100 00100000 01001001 01001110 01010100 01000101 01001100
01001100 01001001 01000111 01000101 01001110 01000011 01000101

Inside the computer, electrical circuits operate as

switches. When a switch is on, it represents the binary
digit 1. When a switch is off, it represents the binary
digit 0. This type of calculation is fast and can be ma-
nipulated to count, add, subtract, multiply, divide,
compare, list, or rearrange according to the program.
This simple digital system can be used to program a
computer to do even the most complex tasks. The re-
sults of electrical circuits manipulating strings of bi-
nary digits are then translated into letters or numbers
that people can understand. For instance, the letter A
is represented by the binary number 01000001. Today,
by reading and changing binary digits, a machine can
display a Shakespearean sonnet, play a Mozart melody,
run a blockbuster movie, and even represent the en-
tire sequence of human DNA.
20 Artificial Intelligence

Smaller and Faster

The basic digital language of early computers was fast,
but the hardware that performed the tasks was not.
Vacuum tubes were unreliable and broke down fre-
quently. Computing was made faster with transistors,
a crucial invention developed in the 1950s. A transis-
tor is a miniature electronic switch that has two op-
erating positions: on and off. It is the basic building
block of a computer that enables it to process infor-
mation. Transistors are made from silicon, a type of
material that is called a semiconductor because cer-
tain impurities introduced to the silicon affect how
an electrical current flows through it.
These early transistors were lighter, more durable,
and longer lasting. They required less energy and did
not attract moths, as ENIAC’s vacuum tubes did. They
were also smaller, about the size of a man’s thumb. And
smaller translated to faster. The shorter the distance the
electronic signal traveled, the faster the calculation.
The next improvement came with the invention of
the integrated circuit, an arrangement of tiny transis-
tors on a sliver, or chip, of silicon that dramatically
reduced the distance traveled. The amount of infor-

A Science Gets a Name

In pursuit of a machine that could pass the Turing Test, a group of
researchers held a conference in 1956. Called the Dartmouth
Conference, it was an open invitation to all researchers studying or
trying to apply intelligence to a machine. This conference became
famous not necessarily because of those who attended (although
some of the most well-known AI experts were there) but because of
the ideas that were presented. At the conference the researchers put
forth a mission statement about their work. Quoted in Daniel Crevier’s
1993 book AI: The Tumultuous History of the Search for Artificial Intelli-
gence, the mission statement said, “Every aspect of learning or any
other feature of intelligence can in principle be so precisely described
that a machine can be made to simulate it.” And the researchers
came up with a name for their pursuits that became accepted world-
wide and solidified a science—artificial intelligence.
The First Thinking Machines 21

mation held in a given amount of space also increased

drastically. A vacuum tube, for example, could fit one
bit of information in a space the size of a thumb. One
of the first transistors could hold one bit in a space
the size of a fingernail. Today a modern silicon mi-
crochip the size of a grain of rice can contain millions
of bits of information. These rapid improvements in
computing technology allowed for rapid advance-
ments in the pursuit of artificial intelligence.

The Turing Test

In 1950 Alan Turing solidified his place as the grand-
father of AI with his paper “Computing Machinery
and Intelligence.” The now-famous report claimed
that computing technology would one day improve
to the point where machines would be considered in-
telligent. He knew this claim would be difficult to
prove, so he also put forth the idea of a standardized
test he called the Imitation Game. Now known as the
Turing Test, it is set up in the following manner: An
interrogator or judge sits in front of two computer ter-
minals. One terminal is connected to a person in an-
other room; the other terminal to a computer in a
third room. The interrogator types questions on both
terminals to try to figure out which terminal is con-
trolled by a human and which is controlled by a ma-
chine. If the interrogator cannot decide which con-
testant is human, or chooses incorrectly, then the
computer would be judged intelligent.
Turing’s paper and theoretical test was another mile-
stone in the development of AI. He predicted that by
the year 2000 a computer “would be able to play the
imitation game so well that an average interrogator
will not have more than a 70 percent chance of mak-
ing the right identification (machine or human) after
five minutes of questioning.”2
The year 2000 has come and gone and no machine
has yet passed the test. But the lack of a winner has
not deterred AI programmers in their quest. Today an
22 Artificial Intelligence

annual contest called the Loebner Prize offers $100,000

to the creator of the first machine to pass the Turing
Test.
Some people argue that the Turing Test must be
flawed. With all that artificial intelligence can do, it
seems illogical to them that a machine cannot pass
the test. But the importance of the Turing Test is that
it gives researchers a clear goal in their quest for a
thinking machine. Turing himself suggested that the
game of chess would be a good avenue to explore in
the search for intelligent machines, and that pursuit
led to the development of another machine that made
early AI history.

Deep Blue
The complex game of chess involves intellectual strat-
egy and an almost endless array of moves. According
to one calculation there are 10120 (or 10 followed by
120 zeros) possible moves. In comparison, the entire
universe is believed to contain only 1075 atoms. Of
course, no player could consider all possible moves,
but the best players are capable of thinking ahead, an-
ticipating their opponent’s play, and instantaneously
selecting small subsets of best moves from which to
choose in response. Prompted by Turing’s suggestion
that chess was a good indicator of intelligence, many
AI groups around the world began to develop a chess-
playing computer. The first tournament match that
pitted a computer against a human occurred in 1967.
But it was not until the 1980s that computers became
good enough to defeat an experienced player. A com-
puter called Deep Thought, created by students at
Carnegie Mellon University in Pittsburgh, beat grand
master Brent Larsen at a single game in 1988.
A team from IBM took over Deep Thought, recon-
figured its programming, dressed it in blue, and re-
named the new chess-playing computer Deep Blue.
Deep Blue’s strength was sheer power: It was so fast it
could evaluate 200 million positions per second and
look fifteen to thirty moves ahead. Its human oppo- Chess champion Gary
nent, the world chess champion Gary Kasparov of Kasparov executes a
move during a 1997
Russia, could consider only three moves per second.
match against the
Deep Blue, outweighing its human opponent by al- highly sophisticated
most a ton, won its first game in 1996. Kasparov computer Deep Blue
fought back and managed to win the match four as the computer’s
games to two. A rematch held on May 10, 1997, again designer watches.
pitted a souped-up Deep Blue against Kasparov. Deep
Blue won every game in the match. In an interview
afterward, Kasparov admitted that he “sensed a new
kind of intelligence”3 fighting against him.

Shakey the Robot

While some researchers were making headlines play-
ing computer chess, others were exploring the com-
bination of robotics and AI. In 1969 Nils Nilsson at
Stanford University created an early AI robot called
Shakey. This five-foot-tall boxy robot earned its nick-
name from the video camera and TV transmitter mast
24 Artificial Intelligence

extending from its top, which shook back and forth

as the robot moved. Shakey’s world was limited to a
set of carefully constructed rooms. It was programmed
Computer scientist with an internal map of the dimension and position
Charles Rosen poses of every object in the rooms so that it would not bump
next to Shakey, an
into anything. Shakey moved around at the snail-like
early AI robot that
was able to move speed of one foot every five minutes, stopping peri-
about without odically to “see” its environment through the camera
bumping into objects. lens, which transmitted images to the computer. It
The First Thinking Machines 25

would then “think” of its next move. If someone came

in and moved objects in the room while Shakey was
“thinking,” the robot would not notice and inevitably
bump into the moved objects.
Shakey’s program was so immense that it could not
carry the necessary computer power around with it.
It had to be attached by a thick cable to its electronic
brain, located in another room. Because of its tethered
existence and programmed environment, Shakey
could not function outside of its own little world. But
its importance lay in what it taught AI researchers
about intelligence and the real world.
Machines like Deep Blue and Shakey represent the
traditional era of AI. Traditional AI specialists thought
—and some still do think—that human intelligence
is based on symbols that can be manipulated and
processed. Intelligence was equated with knowledge.
The more facts people or computers knew, the more
intelligent they were. That is why early AI projects fo-
cused on things that most people, including college
professors, found challenging, like playing chess, prov-
ing mathematical theorems, and solving complicated
algebraic problems.
Early AI researchers believed that the human brain
worked like a computer, taking in information and
converting it to symbols, which were processed by the
brain and then converted back to a recognizable form
as a thought. Every aspect of Shakey’s world was pro-
grammed using the digital binary code so that the ma-
chine could perceive changes and operate within its
mapped world. Deep Blue’s chess program was created
the same way. This traditional AI, although very rigid,
had amazing success. It was the basis of programs
called expert systems, the first form of AI put to prac-
tical, commercial use.

Expert Systems
An expert system is an AI program that imitates the
knowledge and decision-making abilities of a person
26 Artificial Intelligence

with expertise in a certain field. These programs pro-

vide a second opinion and are designed to help peo-
ple make sound decisions. An expert system has two
parts—a knowledge base and an inference engine.
The knowledge base is created by knowledge engi-
neers, who interview dozens of human experts in the
field. For example, if an expert system is intended to
help doctors diagnose patients’ illnesses, then knowl-
edge engineers would interview many doctors about
their process of diagnosing illnesses. What symptoms
do they look for, and what assumptions do they make?
These questions are not easy to answer, because a per-
son’s knowledge depends on something more than
facts. Often experts use words like intuition or hunch
to express what they know. Underlying that hunch,
however, are dozens of tiny, subconscious facts or rules
of thumb that an expert or doctor has already learned.
When those rules of thumb are programmed into a
computer, the result is an expert system. “Real [stock
market] analysts think what they do is some sort of
art, but it can really be reduced to rules,”4 says Edgar
Peters of PanAgora Asset Management. This is true for
most decision-making procedures.
The second part of an expert system, the inference
engine, is a logic program that interprets the instruc-
tions and evaluates the facts to make a decision. It op-
erates on an if-then type of logic program—for ex-
ample, “if the sun is shining, then it must be daytime.”
Every inference engine contains thousands of these
if-then instructions.
The first expert systems were used in medicine, busi-
ness, and finance to help doctors and business man-
agers make better decisions. Expert systems are still
used today in the stock market, banking, and military
defense.

Intelligence or Imitation?
Traditional AI represents only one type of intelligence.
It is very effective at applying knowledge to a single
The First Thinking Machines 27

problem, but over time, it became apparent that tra-

ditional expert systems were too limited and special-
ized. A chess program like Deep Blue could play one
of the most difficult strategy games in the world, but
it could not play checkers or subtract two-digit num-
bers, and it performed its operations without any un-
derstanding of what it was doing.
In 1980 John Searle, a U.S. philosopher, described
artificial intelligence with an analogy he called the
Chinese room. Suppose an English-speaking person is
sitting in a closed room with only a giant rule book
of the Chinese language to keep him company. The
book enables the person to look up Chinese sentences
and offers sentences to be used in reply. The person
receives written messages through a hole in the wall.
Using the rule book, he looks up the sentences, re-
sponds to them, and slips the response back through
the hole in the wall. From outside the room it appears
that the person inside is fluent in the Chinese lan-
guage, when in reality, the person is only carrying out
a simple operation of matching symbols on a page
without any understanding of the messages coming
in or going out of the hole.
Searle’s analogy made researchers wonder. Was AI
just an imitation of human intelligence? Was intelli-
gence simply the processing of information and spit-
ting out of answers, or was there more to it than that?
 Chapter 2

Mind Versus Metal

T he activities that kindergarten-age children per-

form effortlessly, like knowing the difference be-
tween a cup and a chair, or walking from one room
to the next without bumping into the wall, were not
thought of as intelligent behavior or worthy of study
by traditional AI researchers. But when traditional sys-
tems did not perform as they had expected, experts
in AI began to wonder what intelligence really meant.
They also began to think about different ways to show
intelligence in a machine. Although the definition of
intelligence is still debated today, scientists understand
that intelligence is more than the sum of facts a per-
son knows; it also derives from what a person experi-
ences and how a person perceives the world around
him or her. Neil Gershenfeld, author of When Things
Start to Think, believes that “we need all of our senses
to make sense of the world, and so do computers.”5
As ideas about human intelligence changed, so did ap-
proaches to creating artificial intelligence.
In the 1980s AI experts working in robotics began to
realize that the simple activities humans take for granted
are much more difficult to replicate in a machine than
anyone thought. As expert AI researcher Stewart Wilson
of the Roland Institute in Cambridge explains:

AI projects were masterpieces of programming

that dealt with various fragments of human in-
telligence. . . . But they were too specialized. . . .
They couldn’t take raw input from the world
28
Mind Versus Metal 29

around them; they had to sit there waiting for a

human to hand them symbols, and they then ma-
nipulated the symbols without knowing what
they meant. None of these programs could learn
from or adapt to the world around them. Even
the simplest animals can do these things, but they
had been completely ignored by AI.6

Breaking away from traditional AI programming,

some researchers veered off to study the lower-level
intelligence displayed by animals. One such person is
Rodney Brooks, the director of the Computer Science
and Artificial Intelligence Laboratory at the Massachu-
setts Institute of Technology (MIT).

Insect Intelligence
Brooks started at the bottom of the evolutionary lad- The six-legged
der, with insects, which were already capable of do- insectlike robot
known as Genghis
ing what the most sophisticated AI machines could
is programmed to
not do. Insects can move at speeds of a meter or more home in on objects
per second, avoid obstacles in their path, evade preda- by tracking the
tors, and seek out mates and food without having to heat they give off.
30 Artificial Intelligence

have a mental map as Shakey the robot did. Instead

of preprogramming behaviors, Brooks programmed in
less information, just enough to enable his AI robots
to adapt to objects in their path. He felt that naviga-
tion and perception were key to mastering higher-level
intelligence. This trend became known as the bottom-
up approach, in contrast to the top-down approach
of programming in all necessary information.
Brooks’s focus was creating a machine that could
perceive the world around it and react to it. As a re-
sult, he created Genghis, a six-legged insectlike robot.
According to Brooks, “When powered off, it sat on the
floor with its legs sprawled out flat. When it was
switched on, it would stand up and wait to see some
moving infrared source. As soon as its beady array of
six sensors caught sight of something, it was off.”7 Its
six sensors picked up on the heat of a living creature,
such as a person or a dog, and triggered the stalking
mode. It would scramble to its feet and follow its prey,
moving around furniture and climbing over obstacles
to keep the prey in sight.
Brooks’s machine could “see” and adapt to its en-
vironment, but it could not perform higher-level in-
telligent behaviors at the same time. A man, for ex-
ample, can make a mental grocery list while he is
walking down the street to the store; a woman can
carry on a conversation with a passenger while dri-
ving safely down the road and looking for an address.
Researchers began to ask, how is the human brain able
to perform so many tasks at the same time?

How the Brain Works

The human brain has close to 100 billion nerve cells,
called neurons. Each neuron is connected to thou-
sands of others, creating a neural network that shut-
tles information in the form of stimuli, in and out of
the brain constantly.
Each neuron is made up of four main parts: the
synapses, soma, axon, and dendrites. The soma is the
Mind Versus Metal 31

Biological Neural Network

The human brain has approximately 100 billion nerve cells, called neurons, each connected to
thousands of others. Senses and thoughts trigger electrical impulses that quickly travel through
the neural network. When a neuron receives information, it can send the message on or stop
it from traveling forward.

Dendrites
(accept input)

Synapse
(tiny gap between two neurons
where information is transfered)

Neurotransmitters
(chemicals released
Axon
by one cell to trigger or
(turns processed
stop electrical impulses
input into output)
in the next cell)

Nucleus

Cell Body, or Soma Axon Terminals

(processes input) Nerve Impulse (send information to next cell)

body of the cell where the information is processed.

Each neuron has long, thin nerve fibers called den-
drites that bring information in and even longer fibers
called axons that send information away. The neuron
receives information in the form of electrical signals
from neighboring neurons across one of thousands of
synapses, small gaps that separate two neurons and
act as input channels.
Once a neuron has received this charge it triggers
either a “go” signal that allows the message to be
passed to the next neuron or a “stop” signal that pre-
vents the message from being forwarded. When a per-
son thinks of something, sees an image, or smells a
scent, that mental process or sensory stimulus excites
a neuron, which fires an electrical pulse that shoots
32 Artificial Intelligence

out through the axons and fires across the synapse. If

enough input is received at the same time, the neu-
ron is activated to send out a signal to be picked up
by the next neuron’s dendrites. Most of the brain con-
sists of the “wiring” between the neurons, which
makes up one thousand trillion connections. If these
fibers were real wire, they would measure out to an es-
timated 63,140 miles inside the average skull.
Each stimulus leads to a chain reaction of electrical
impulses, and the brain is constantly firing and
rewiring itself. When neurons repeatedly fire in a par-
ticular pattern, that pattern becomes a semipermanent
feature of the brain. Learning comes when patterns are
strengthened, but if connections are not stimulated,
they are weakened. For example, the more a student
repeats the number to open a combination lock, the
more the connections that take in that information
are bolstered to create a stronger memory that will be
easily retrieved the next time. At the end of the school
year, when a student puts the lock away, that number
will not be used for a couple of months. Those three
numbers will be much harder to recall when fall comes
and that student needs to open the lock again.

Artificial Neural Networks

The branch of AI that modeled its work after the
neural network of the human brain is called connec-
tionism. It is based on the belief that the way the brain
works is all about making the right connections, and
those connections can just as easily be made using sil-
icon and wire as living neurons and dendrites.
Called artificial neural networks (ANNs), these pro-
grams work in the same way as the brain’s neural net-
work. An artificial neuron has a number of connec-
tions or inputs. To mimic a real neuron, each input is
weighted with a fraction between 0 and 1. The weight
indicates how important the incoming signal for that
input is going to be. An input weighted 0.4 is more
important than an input weighted 0.1. All of the in-
Mind Versus Metal 33

coming signals’ weights are added together and the

total sum equals the net value of the neuron.
Each artificial neuron is also given a number that
represents the threshold or point over which the ar-
tificial neuron will fire and send on the signal to an-
other neuron. If the net value is greater than the
threshold, the neuron will fire. If the value is less than
the threshold, it will not fire. The output from the fir-
ing is then passed on to other neurons that are
weighted as well. For example, the computer’s goal is
to answer the question, Will the teacher give a quiz
on Friday? To help answer the question, the pro-
grammer provides these weighted inputs:

The teacher loves giving quizzes = 0.2.

The teacher has not given a quiz in two weeks = 0.1.
The teacher gave the last three quizzes on Fridays = 0.3.

The sum of the input weights equals 0.6. The threshold

assigned to that neuron is 0.5. In this case, the net value
of the neuron exceeds the threshold number so the ar-
tificial neuron is fired. This process occurs again and
again in rapid succession until the process is completed.
If the ANN is wrong, and the teacher does not give a
quiz on Friday, then the weights are lowered. Each time
a correct connection is made, the weight is increased.
The next time the question is asked, the ANN will be
more likely to answer correctly. The proper connections
are weighted so that there is more chance that the ma-
chine will choose that connection the next time. After
hundreds of repeated training processes, the correct
neural network connections are strengthened and re-
membered, just like a memory in the human brain. This
is how the ANN is trained rather than programmed with
specific information. A well-trained ANN is said to be
able to learn. In this way the computer is learning much
like a child learns, through trial and error. Unlike a child,
however, a computer can make millions of trial-and-
error attempts at lightning speed.
 34 Artificial Intelligence

Artificial Neural Network (ANN)

Artificial neurons simulate the basic functions of biological Simple ANN
neurons: input, processing, output, and passing information
to other neurons. Each input is given a weight to signify how
important it is compared to other input.

Output
Layer
Processing
Input Elements
Layer

Input Information
(with varying
weights of Sum
importance) of Transfer Output
Input

Processing Element

Whereas traditional AI expert systems are special-

ized and inflexible, ANN systems are trainable and
more flexible, and they can deal with a wide range of
data and information. They can also learn from their
mistakes. This kind of AI is best for analyzing and rec-
ognizing patterns.

Pattern Recognition
Pattern recognition may seem obvious or trivial, but
it is an essential, basic component of the way people
learn. Looking around a room, a child learns the pat-
terns of the room’s layout and recognizes objects in
the room. People know that a pencil is a pencil be-
cause of the pattern the pencil presents to them. They
Mind Versus Metal 35

learn and become familiar with that pattern. Learning

a language is actually the activity of learning nu-
merous patterns in letters, syllables, and sentences.
The first ANN prototype, Perceptron, created in the
1950s, was trained to perform the difficult task of
identifying and recognizing the letters of the alpha-
bet. Today more sophisticated ANNs are also capable
of finding patterns in auditory data. ANN software or
programs are used to analyze handwriting, compare
fingerprints, process written and oral language, and
translate languages.
Incorporated into expert systems, ANNs can pro-
vide a more flexible program that can learn new pat-
terns as time goes on. They are especially useful when
all the facts are not known. These programs have
proven more reliable than humans when analyzing
applications for credit cards or home mortgage loans.
An ANN system can find patterns in data that are un-
detectable to the human eye. Such tracking is called
data mining, and oftentimes the outcome is surpris-
ing. For example, the sophisticated computers that
keep track of Wal-Mart’s sales discovered an odd rela-
tionship between the sale of diapers and the sale of
beer. It appeared that on Friday nights the sales of both
products increased.
Some police departments use an ANN search engine
called Coplink to search multiple case files from dif-
ferent locations and criminal databases to find pat-
terns to seemingly unrelated crimes. Coplink helped
catch the two snipers convicted of a string of shoot-
ings in the Washington, D.C., area in 2002.
The Chicago police department uses another ANN
program called Brainmaker to predict which police of-
ficers would be more likely to become corrupt and not
perform their job according to the oath they took upon
becoming an officer. Another software product called
True Face recognizes human faces by comparing video
images to thousands of stored images in its memory
in spite of wigs, glasses, makeup, or bad lighting.
36 Artificial Intelligence

Breeding Programs
Another model of machine learning is based on the
biological system of genetics, in which systems change
over time. Introduced by John Henry Holland at MIT
in the 1950s, this kind of system uses genetic algo-
rithms. An algorithm is simply a step-by-step process
of solving a specific problem. An algorithm comes in
many forms, but usually it is a set of rules by which a
process is run. In a genetic system an algorithm is
made up of an array of bits or characters, much like a
chromosome is made up of bits of DNA. Each bit is
encoded with certain variables of the problem or with
functions used in solving the problem. Just as a gene
would carry specific information about the makeup
of an organism (blue or green eyes), a bit might select
for functions of addition or subtraction.
In traditional AI, algorithms would be programmed
into a system to perform the same task over and over
again without variation. But in a genetic system the
computer is given a large pool of chromosomes or bit

A Commonsense Computer
According to expert Marvin Minsky, as quoted in the Overview of
Cycorp’s Research and Development, the one problem with artificial in-
telligence is that “people have silly reasons why computers don’t re-
ally think. The answer is we haven’t programmed them right; they
just don’t have common sense.” Common sense allows a person to
make assumptions, jump to conclusions, and make sound decisions
based on common knowledge.
Doug Lenat took up the challenge and created Cyc, a program
named after the word encyclopedia. For more than twenty years, Lenat
and his team at Cycorp has fed the Cyc system with every bit of knowl-
edge that a typical adult would or should know; for example, that
George Washington was the first president, that plants perform pho-
tosynthesis, that it snows in Minnesota, that e-mail is an Internet com-
munication, and that beavers build dams. The idea behind this long-
term project is that Cyc will be able to make reasonable assumptions
with the knowledge it has been given. For example, if a person said
that he or she had read Melville, Cyc might assume that the person
had read the author Herman Melville’s most famous book, Moby Dick.
Mind Versus Metal 37

strings encoded with various bits of information. The

bit strings are tested to see how well they perform at
the task at hand. The algorithms that perform the best
are then bred using the genetic concepts of mutation
and crossover to create a new generation of algorithms.
Mutations occur by randomly flipping the location of
bits on the chromosome. Mutations can bring about
drastic random change that may or may not improve
the next generation. Crossover occurs when two par-
ent programs trade and insert fragments from one to
the other. This ensures that groups of bits or genes
that work well together stay together.
With each change, the procedure creates new com-
binations. Some may work better than the parent gen-
eration and others may not. Those algorithms that do
not perform well are dropped out of the system like a
weak fledgling from a nest. Those that perform better
become part of the breeding process. Genetic programs
can be run many times, creating thousands of gener-
ations in a day. Each time bits of information are
changed and the set of rules are improved until an ad-
equate solution to the problem is found.
The kinds of problems that genetic algorithms are
most useful at solving involve a large number of pos-
sible solutions. They are fondly called traveling sales-
man problems, after the classic example of a sales-
person finding the shortest route to take to visit a set
number of towns. With five towns to visit, there are
120 possible routes to take, but with twenty-five towns
to visit, there are 155 x 1023 (or 155 followed by 23 ze-
ros) possible routes and finding the shortest route be-
comes an almost insurmountable task.
Such a task is a common one in real life. It is a prob-
lem for airlines that need to schedule the arrivals and
departures of hundreds of planes and for manufac-
turers who need to figure out the most efficient order
in which to assemble their product. Phone companies
also need to figure out the best way to organize their
network so that every call gets through quickly.
38 Artificial Intelligence

Genetic algorithms are also used to program the way

a robot moves. There are an almost infinite number
of potential moves, and genetic algorithms find the
most efficient method. After all of the genetic algo-
rithms have evolved, the genetic program produces
an overall strategy using the best moves for certain sit-
uations. The best performing programs are then taken
out of the robots and reassembled to produce a new
generation of mobile robots.
In the future, these genetic systems may even evolve
into software that can write itself. Two parent pro-
grams will combine to create many offspring programs
that are either faster, more efficient, or more accurate
than either of the two parents.

Logic and Fuzzy Logic

Regardless of the technique that is used to perform
certain AI functions, most software is created using
complex mathematical logic. A common symbolic
logic system used in computer programming was de-
veloped in the mid-1800s by mathematician George
Boole, and many search engines today use Boolean
operators—AND, NOT, and OR—to logically locate ap-
propriate information. Boolean logic is a way to de-
scribe sets of objects or information. For example, in
Boolean logic:
Strawberries are red is true.
Strawberries are red AND oranges are blue is false.
Strawberries are red OR oranges are blue is true.
Strawberries are red AND oranges are NOT blue is true.
Boole’s logic system fit well into computer science be-
cause all information could be reduced to either true
or false and could be represented in the binary num-
ber system of 0s and 1s—with true being represented
by a zero, and false by the number one, or vice versa.
Computers tend to view all information as black or
white, true or false, on or off. But not everything in
Mind Versus Metal 39

life follows the strict true-false, if-then logic. In the Computer scientist
1960s mathematician Dr. Lotfi Zadeh of the University Alfonso Perez
modifies Fuzzy Cat,
of California at Berkeley developed the concept of
a robot that uses
fuzzy logic. Fuzzy logic systems work when hard and fuzzy logic, a form
fast rules do not apply. They are a means of general- of reasoning that
izing or softening any specific theory from a crisp and works when hard
precise form to a continuous and fuzzy form. and fast rules do
Programs that forecast the weather, for example, not apply.
deal in fuzzy logic. A fuzzy logic system understands
40 Artificial Intelligence

air temperature as being partly hot rather than being

either hot or cold, and the temperature is expressed
in the form of a percentage. Fuzzy logic programs are
used to monitor the temperature of the water inside
washing machines, to control car engines, elevators,
and video cameras, and to recognize the subtle dif-
ferences in written and spoken languages.

Alien Intelligence
Although traditional expectations of artificial intelli-
gence are to duplicate human intelligence, author and
computer specialist James Martin believes that that
expectation is unrealistic. He believes that in the fu-
ture AI should more properly be called alien intelli-
gence, because the way computers “think” is vastly
different from the way a human thinks.
AI is faster and has a larger capacity for storage and
memory than any human. The largest nerves in the
brain can transmit impulses at around 90 meters per
second, whereas a fiber optics connection can trans-
mit impulses at 300 million meters per second, more
than 3 million times faster. A human neuron fires in
one-thousandth of a second, but a computer transis-
tor can fire in less than one-billionth of a second. The
brain’s memory capacity is some 30 billion neurons,
while the data warehouse computer at Wal-Mart has
more than 168 trillion bits of storage with the capac-
ity to grow each year. Such a computer can process
vast amounts of data that would bury a human proces-
sor and can quickly find patterns invisible to the hu-
man eye. The logical processes that some systems go
through are so complex that even the best program-
mers cannot understand them. These computers, in a
sense, speak a language that is understood only by an-
other computer. Martin suggests that AI researchers of
the past, who predicted a robot in every kitchen,
promised more than could be delivered. But what they
did create is much more than anyone could have
dreamed possible.
 Chapter 3

Everyday AI
T he real success of AI is that most people are sim-
ply unaware of how significantly it affects and en-
ables the routines of daily life. A man gets up in the
morning to the smell of coffee already brewing. This
is thanks to a microchip inside the coffee machine
that allows him to program his coffeemaker to turn
itself on while he is still sleeping. Another microchip
keeps his toast from burning and remembers which
setting from light to dark he likes best.
At one time these were all novel forms of AI.
According to Douglas Hofstadter, a German researcher,
“Once some mental function is programmed, people
soon cease to consider it as an essential ingredient of
‘real thinking.’ AI is whatever hasn’t been done yet.”8
Now that these features are commonplace, they are
thought of as just another necessary product that
makes life easier.
Already there are ovens on the market that keep
food chilled until it starts to cook at a preset time.
There are refrigerators with Web pads embedded in
the doors to display recipes or to surf the Internet. The
company iRobot now manufactures a robotic vacuum
cleaner that buzzes around a room without a human
operator.
But AI in the home today offers just a glimpse of
what future household appliances will look like. “Most
people don’t realize fundamental changes in kitchen
gear are coming. But in a decade or so, they won’t
know how they lived without their e-kitchens,”9 says
41
The Roomba
Intelligent Floorvac
simplifies daily living
by cleaning household
floor surfaces at the
touch of a button.

Bob Lamson, vice president of a company that makes

high-tech gear. In the future, a cook may be able to
place the ingredients on the kitchen counter and a
built-in computer sensor will suggest recipes on the
basis of those ingredients. Before leaving for work, a
parent might place a frozen casserole in the oven and
program it to defrost the casserole, then keep the dish
chilled until it is time to start cooking. If traffic is tied
up, the harried cook might simply be able to call the
oven from the cell phone and reprogram the oven to
delay dinner an hour. After dinner, family members
might entertain themselves with one of the most so-
Everyday AI 43

phisticated examples of AI in the home by playing a

computer game.

Computer Games
Traditionally, AI computer programs have worked in
the background, making sure that the digital environ-
ment of forests, smoking volcanoes, and rambling paths
run smoothly. But AI is also used to make computer
games more challenging for their human participants.
Most games are rule-based, which means that the
computerized characters such as enemy guards, friendly
wingmen, or monsters follow a basic set of rules ac-
cording to what is happening around them. For ex-
ample, if an enemy guard encounters a monster, it will
attack the creature. If the creature flees, the enemy
guard will chase after it, and if the enemy guard is fired
on, it will dodge the bullet, spear, or laser ray. But these
behaviors can become predictable for an experienced
human player. By using AI, the computerized charac-
ters become less predictable and more difficult to over-
come. Many designers incorporate fuzzy logic into the
characters’ decision-making programs so that the set
of rules they operate under are less severe. All of the
criteria do not have to be present for a character to act.
There is flexibility in its behavior responses. Others use
random weighting of certain attributes in order to cre-
ate a more unanticipated chain of events.
More sophisticated games use AI software to ana-
lyze their human player’s style of play. These programs
actually learn to refine the game characters’ social in-
teractions so they can adapt to the human player’s be-
havior. A game called Black and White God, designed
by Richard Evans at Lionhead Studios, uses what he
calls empathic learning to achieve this. In this game,
if the human player throws fireballs at a tribal village,
the computerized companion creature will learn that
the player does not like the tribe. It will then act ac-
cordingly and spontaneously try to attack and stomp
on that village’s tribesmen.
44 Artificial Intelligence

Other games are goal based, which means that the

computerized characters’ behavior is determined by a
larger preprogrammed goal. In Civilization III, by
Firaxis Games, a player and various computer oppo-
nents fight military, economic, and cultural battles.
The computerized opposition’s goal is to stop the
spread of the player’s civilization by any means at their
disposal. The characters’ actions change depending
on the obstacles put in their path.
Some of the most popular games are open-ended.
Their characters do not have a finite path to follow or
an ultimate goal. The top-selling game The Sims fea-
tures a world of autonomous simulated people in an
open-ended game where the human player oversees
all the action. The characters, or Sims, interact based
on a set of simple needs like food or entertainment.
But the intelligence to interact with objects in the en-
vironment is not built into the Sims themselves. In a
novel twist in programming, the objects in the envi-
ronment advertise their ability to satisfy certain needs
to any Sim character that is wandering by. The designer,
Will Wright of Electronic Arts, calls this programming
smart terrain. For example, when a Sim is looking
around for fun, it may spot a beach ball. The Sim reg-
isters the beach ball’s ability to satisfy its entertainment
need. Wright explains, “All the instructions about how
to bounce the ball are there as part of the ball’s code.”10
Rather than taking more control over a game’s de-
sign, game designers are giving more control to the
human player. Software called middleware can be
downloaded and used to direct certain aspects of the
game. A player can make the game harder or easier to
play. For example, the precise noise level that would
normally activate an enemy guard’s attack pattern can
be tuned up or down so that a softly creaking door
would not alert the guard, but a weapon firing nearby
would give the player away.
All of these unique AI programs, originally designed
for mindless popular entertainment, are finding their
Other documents randomly have
different content
Tretower, near Crickhowell, where a Normr ■• shell-keep had been
utilised by a later owner as the outer wall of his fortress, a ve |
narrow tower being erected in the centre of the shell-keep, so as to
make a litt | "inner watd " of the ground between the new building
and the old shell. \
 dSo] TWELFTH-CENTURY IMPROVEMENTS 525
mpplemented by outer defences, either at their first construction 3r
at a later date. It is rare to find examples of them without my
additional walls outside — though Bowes Castle in North Yorkshire
seems to be such an exception. The original English 3r continental
mound-fortress was of small extent, but round t grew up the
dwellings of the owner's retainers, and presently >ome light
defences of ditch and hedge were drawn round them, 50 that the
burh or motte became only the citadel. The name Durh, as we know,
soon came to be applied to the settlement -ound the palisaded
mound as well as to the structure itself When the defences of the
suburb were made stronger, and walls supplanted ditch and hedge,
we have arrived at a very common eleventh- and twelfth-century
type of fortress — the keep surrounded by a curtain-wall containing
a considerable space of ground. The enclosed area may be large,
and a whole town Tiay be built within it. On the other hand, it may
be quite small, only affording room for the few buildings and
storelouses needed by the garrison of the keep. As a general rule
:he keep lies not in the middle of the space, but at one end of it, or
set in the wall. This was often due to the fact that the mound was
the end of the spur of a hill or rising ground, cut off from it by the
excavation of its ditch. The extension of the fortress was along the
top of the spur, not below that front of the mound which looked
towards the plain. So we often find a castle with its original keep on
the end of the spur, its first extension just beyond the original ditch,
and then a second extension, or " outer ward," still farther remote
from the early citadel. When a castle was not on a spur, but upon an
isolated mound in the plain, it must of course have been more or
less a matter of chance on which side the outgrowth began. But as a
general rule the keep stands at one end of the enclosed space, not
in its midst. The same is true of towns and their citadels — the
normal type has the castle at one end of the place, like London,
Winchester, or Oxford. It is rare to find it set right in the midst of the
inhabited space, though Ferrara and Evreux may serve as examples.
Obviously there was danger in the close juxtaposition of houses to
the citadel : they gave too much cover to an enemy, and if set on
fire might stifle the defenders of the stronghold which they
surrounded. Such was the stage at which fortification had arrived in
Western and Central Europe, when a new influence was brought
 5z6 THE ART OF WAR IN THE MIDDLE AGES [1150 to bear
upon it. The Crusades put the men of the twelfth century in touch
with the Levant, where they had the opportunity of studying the
splendid fortresses which the Eastern emperors had built, and of
which so many were now in the hands of the Turks and Saracens. To
have to undertake the sieges of great fenced cities like Nicaea,
Antioch, or Jerusalem was almost an education in itself to the
engineers of the West. Their feeble engines and their primitive
methods of attack were utterly unable to cope with such
strongholds, and as a rule famine or treachery alone enabled them
to win the place?; which they beleaguered. The essential features of
Byzantine military architecture were the erection of double and triple
defences round the core of the fortress, and the careful provision of
towers set at intervals in the " curtain " of the walls. Both were new
ideas to the Crusaders, whose notion of a fortress was nothing more
than a keep surrounded by a plain outer curtain not strengthened
with towers. Constantinople, the most perfect of all the Eastern
fortresses, struck the Franks as absolutely impregnable : it had a
triple enceinte, with a deep ditch in front of the outermost face. The
first wall was commanded by the second, and the second by the
third, each overtopping the line below it, and all three furnished with
military machines capable of playing on the siege-works of the
beleaguering army. Moreover, the two firsi walls were loopholed at a
stage below the battlements, so thai the garrison could fire not
merely from the parapets, but from i well-protected second line of
openings. The siege-artillery 0 the enemy would therefore have
before it at any point five separat( lines of engines, each rising
above the other, and all command ing the ground beyond the ditch
where the investing arm} must necessarily begin to erect its works.
As a matter of fact no hostile force ever dared to attempt a regular
attack on thi tremendous front till the days of the invention of
gunpowdei The Avars, Persians, and Saracens in the seventh and
eight) centuries only blockaded the place and tried to starve it oul i
The Crusaders of 1204 studied the tremendous triple enceinte found
that it was impregnable, and then turned all their energie against
the sea face of the city, where there was only a singl wall to oppose
them. Previous besiegers had never possesse that complete
command of the water approaches which mad such an attack
possible. In the days of Heraclius, Constantin
 204] THE SIEGE OF CONSTANTINOPLE 527 'ogonatus, and
Leo the Isaurian, the Byzantine fleet had always •een strong enough
to render regular assaults on the sea wall 00 hazardous. Even when
not in complete command of the traits (as, for example, during the
Saracen siege of 673), the mperial navy had invariably been present
in strong force ;ithin the Golden Horn, and any attempt to assail the
water ront would have caused it to sally out and fall upon the
)esiegers while their ships, crowded with land troops, were rying to
haul in under the wall. Hence such attempts were lever made : the "
navy in being " of the besieged rendered them 00 hazardous. But in
1204 the wretched emperors of the lOUse of Angelus had so
neglected the fleet that the Venetians rere able to draw under the
sea wall and assail it without any ear of interruption. Thus it was
that Constantinople, for the irst time in history, fell before an attack
by open force : before, t had never been captured save by treachery
from within.^ Constantinople was of course quite exceptional in
showing a riple line of defence extending over several miles of front
: as a ule, it was only citadels and not cities which displayed such a
ormidable series of walls. Even the wealthy Byzantine Goverrt-r nent
could not afford to surround places of large size with nore than a
single enceinte. For castles and fortresses, however, vhere the space
was moderate, the concentric lines were possible, ind often were
erected : the citadel of Antioch, for example, lad a double wall on
the north and west sides, though not on he more precipitous
southern and eastern fronts.^ The vast town vhich lay below it, on
the other hand, had but a single wall, but his was made very strong
by-its splendid diadem of towers. The fortifications of Antioch may
serve as an example of he Byzantine methods of guarding a city of
first-rate importmce. The place had been retaken from the Saracens
by Nicephorus Phocas in 968: in 976 both walls and city were
;erribly injured by an earthquake, and the whole enceinte had 0 be
repaired. It then remained in the hands of the Eastern emperors till
1086, when the Seljoulc Sultan Suleiman cap:ured it by treachery.
Thus we see that the Turks had only 3een in possession of the place
for a trifle more than ten years »vhen the Crusaders came against it.
The barbarian conquerors lad of course added nothing to the
Byzantine walls, and the ^ e.g. As when Alexius Comnenus took it in
1081. ' See the Plan in Key's Architecture Militaire des Croises en
Syrie, Paris, iSifiV^^
 528 THE ART OF WAR IN THE MIDDLE AGES [109J
fortifications erected by Justinian, and remodelled in the tent!
century by the engineers of John Zimisces and Basil ii., were those
with which the Franks had to deal in 1098. Wher Antioch fell, and
became the capital of Bohemund's principality the old walls needed
no repair — the siege operations had don( no harm to them. The
Byzantine enceinte protected the Latii princes for nearly two hundred
years : its remains are stil sufficient to enable us to reconstruct the
whole system o defence. It consisted of a line of curtain, in which
towers wer placed at frequent but irregular intervals : in the more
expose( parts of the wall the towers were no more than fifty yards
apart in the more inaccessible parts they were some eighty or
hundred yards from each other. Where the walls lie along th river
Orontes to the. north-east, and along precipices on th southern,
south-eastern, and south-western fronts (see Ma facing p. 250),
they are not furnished with a ditch, but on th north-western' and
northern fronts the channel of the Oronte had been diverted along
their foot, so as to form a large moat, c rather a broad marshy
depression. The curtain was solid, an not pierced with loopholes ; its
main protection came from th projecting towers set in it at such
close intervals. Thei formidable structures were about twenty yards
square; half( their bulk stood out beyond the curtain wall, and
commanded side view of the ditch, or of the ground at the foot of
the wal where no ditch existed. They were about sixty feet hig and
had three storeys ; each storey was loopholed both to tt front and to
the sides, so as to furnish a flanking fire along t\ ditch as well as a
direct fire towards the open country. Beir set in the curtain for half
their bulk, the towers blocked tl road round the walls at frequent
intervals. No one could wa' for a quarter of a mile along the enceinte
without passing throu§ six or seven towers, and, as each tower had
strong doors whe its second storey opened on to the ramparts, each
section •curtain could be isolated by the closing of these doors. So
by chance the besieger mastered a part of the curtain, the tv towers
oh each side prevented him from making his way right or left along
the walls, and, as there was no way of gettii down from the
ramparts to the interior of the town (all stai being within the
towers), the assailant would have gain( nothing but some sixty or
eighty yards of narrow ramp? w^lk. The Crusaders in 1098 were
admitted into one of t
 1098] EARLY CASTLES IN SYRIA 529 towers (that of the "
Two Sisters " ) by th3 treachery of the renegade Firouz,^ and by
means of the gate on the ground floor of the tower got into the
town. If they had merely scaled the curtain they would have gained
nothing ; but, emerging from the tower, they were able to break
open first a blocked posterni^ate and then the great bridge-gate
(see Map of Antioch facing p. 250) ; through these two entries the
main body of the Franks poured in, and the place was won. Once
established in Syria, the Franks not only repaired the castles and city
walls which the Moslems had left behind them, but erected an
infinite number of new strongholds, varying in size from small
isolated watch-towers to the most formidable fortresses of the first
class, capable of holding garrisons of two or three thousand men. To
trace the exact stages by which they perfected their military
architecture is not easy, as most of the castles were being
perpetually strengthened, and present now the appearance which
they showed in the thirteenth century, when they finally fell back
into Moslem hands and were dismantled or left to decay. The most
perfect ruins, such as those of Markab and Krak-des-Chevaliers, do
not therefore give us so much information as to the twelfth century
as could be wished. To ascertain the earlier developments of
Frankish architecture in the Holy Land, places must be studied which
were surrendered to Saladin after the battle of Tiberias and never
again were in possession of the Crusaders, such as Saona and
Blanche-Garde (captured in 1 1 87) and Kerak-in-Moab (surrendered
in 1188). An examination of such castles shows that in the twelfth
century the two great principles of Byzantine military architecture—
the defence of the curtain by towers and the construction of
concentric lines of fortification — were thoroughly well understood
and practised by the Frankish builders. The early strongholds differ
from the later mainly by their want of finish, and greater simplicity of
detail. In the thirteenth century castles^ were built not only with
more elaborate and ingenious defences, ^ The first sixty combatants
mounted by a rope ladder on to the curtain adjoining the tower
which Firouz commanded. He led them from thence into the tower..
Next some descended to break open the postern, while others
pushed right and left along the curtain. They were so swift and silent
that they were able to penetrate into the towers, whose doors were
not closed, and to massacre their sleeping garrisons before the
alarm was given. Masters of five hundred or six hundred yards of the
enceinte, they could not be withstood. 34
 530 THE ART OF WAR IN THE MIDDLE AGES [1150 but
also with a certain regard to decoration and ornament. They show
carvings, shields of arms, and occasional inscriptions, of which the
buildings of the preceding age are destitute. But the general
principles of construction are the same throughout the two centuries
during which the Franks held their footing in Syria. It was probably
quite early in the time of the existence of the kingdom of Jerusalem
that the crusading architects adopted and improved on the
Byzantine models. The shell-keeps or square donjons with a plain
towerless curtain- wall, which they had left behind them in the West,
were so obviously inferior tc the military architecture of the Levant
that there was no temptation to reproduce them without an
improvement. Thus a great change in the fundamental conception of
the castle took place early in the twelfth century : instead of being
considered as 2 keep provided with an outer wall, it becomes an
enceinte with or without a keep as final place of refuge. Formerly
the greal donjon was the more and the outer wall the less important
parin the scheme of defence. But now the main resistance was to bf
opposed by the enceinte with towers set in it at intervals, anc the
donjon was a last resort, to which the garrison only retiree in
desperate extremity. It might even be merely the greatesi of the
several towers of the enceinte. When King Amaury aboui 1 165
erected the small but strong fortress of Darum on th( borders of
Egypt, he merely built a square enceinte with fouj large towers at its
angles, of which one was larger than the others. Though this served
as a donjon, it only differed in size from I^B other three. Sj Another
deviation from the old practice of the West wa.' that the strongest
tower was sometimes built not in the mos secure and well-defended
part of the castle, as a place of fina refuge, but at the fore-front of
the most exposed side of the fort ress, so as to bear the brunt of the
attack. In this case the keep if keep we may call it, would be the first
part of the place whicl would be assaulted by the besieger, and the
first, perhaps, to fal into his hands. As an example of this kind of
castle we ma\ quote Athlit (Chateau Pelerin), a castle built on a
promontory ^ William of Tyre, xx. 19, describes it as '* castrum
modicae quantitatis, vix tanr spatium inter se continens quantum est
jactum lapidis, formae quadrae, turres habci. quatiuor in angulis,
quarum una grossior et munitior erat aliis." See Key's A nAi tec/iire
Militaire, etc., p. 125, for its present state.
The text on this page is estimated to be only 19.50%
accurate

PLATE XXL mwW\wm^ CAERPHILLY CASTLE i2ro

 ii5o] THE CASTLE OF KERAK-IN-MOAB 531 where the main
defensive structure consisted of two massive towers connected by a
short curtain and placed across the neck of the promontory. Behind
them, seaward, the rest of the castle was only protected by an
ordinary enceinte with a few small towers. All the strength of the
place lay in the two splendid towers at the isthmus. But Athlit was
built late (12 1 8), and must not be quoted as an example of
twelfth-century architecture.i As a fair example of the strongholds
which the Franks erected after they had been seated for a
generation in the Holy Land, we may describe Kerak-in-Moab, the
eastern bulwark of the kingdom of Jerusalem, built about 1 140 by
Payn of Nablous, the high-butler of King Fulk. It was only forty years
in Christian hands, and seems never to have been much altered
from its original shape. It stands on one of the two narrow crests
which connect the hill of Kerak with the mountains of Moab. To east
and west the slope of the crest is too steep to be accessible : to
north and south, where the danger is greater, two enormous ditches
have been hewn in the rock, so as to isolate the castle from the rest
of the ridge of which it forms part ; they were only to be crossed by
narrow bridges removable in time of war. The fortress consisted of a
donjon in the south-east angle of the oblong enceinte, and of an
upper and a lower ward, separated from each other by a strong wall.
The northern front af the castle was the most exposed : it consisted
of a curtain flanked by two large towers, which gave a lateral fire
into the ditch : the curtain contained at least two stages pierced with
oopholes. The only opening in it was by a gate close under the
vvestern flanking tower: it was closed by a portcullis, and opened lot
directly into the court of the castle, but into a long passage oetween
the curtain and a wall built at its back. Two more port:ulli5es were
placed at intervals in this passage, and it was only ifter passing them
that the court was reached. (See Plan facing 3age 530.) Kerak-in-
Moab proved utterly impregnable to all the attacks )f Saladin.
Though repeatedly assailed, it was never harmed, nor iid the
assailants even enter its lower ward. It held out for nany months
after the battle of Tiberias, and only surrendered vhen provisions
had failed and all hopes of relief were ibsolutely at an end (1188). ^
All this comes from M. Key's admirable and oft-quoted work.
 532 THE ART OF WAR IN THE MIDDLE AGES [1150 It is
safe to say that such a fine example of a fortress with a double line
of defence could not have been built anywhere save in the East so
early as 1 140. Nothing approaching it for completeness of design
was reared in England, France, or Germany till fifty years later, when
Richard Coeur de Lion planned his famous Chateau Gaillard on the
bluff above Les Andelys. Richard, we cannot doubt, was utilising his
Eastern experience when he erected this splendid and complicated
structure, whose arrangements pleased him so well that he boasted
that " it might be held even if its walls were made of butter."
Nevertheless, the influence of Eastern military architecture began to
be felt in the West not long after the first Crusade though the
Western builders worked on a smaller scale, and were for many
years timid copyists of the crusading architects. The old type of the
keep standing in a base-court girt by a plair curtain begins about 11
30 or 1140 to develop into a more com plicated structure. The
enceinte wall becomes more important towers are presently set in it,
and the outer line of defence be comes less wholly subordinate to
the keep. At the same tim( the keep itself ceases to depend entirely
on its passiv< strength, and requires a gate-house, and a larger
provision o loopholes. In a few important castles, instead of building
a mere shell keep or rectangular keep, the architects of the wealthier
baron began about 1 140-50 to erect a more complicated central pile
a the main feature of a new castle. At Alnwick, for example, th
powerful Eustace de Vesey set on the ancient mound which h found
there existing, not a shell-keep (such as his father woul have built),
but a circular cluster of towers, enclosing an ope court. His outer
enceinte was also probably furnished with few small towers, though
these have been so reconstructed b thirteenth- and fourteenth-
century holders of the place that it difficult to be certain on the
point.^ The Tower of Londo round which Rufus had drawn a plain
curtain-wall,^ began to I strengthened with towers under Henry ll.^
The Wakefiel tower, oldest of those of its inner ward, seems to
belong to th; time ; the others have been so pulled about by later
kings, th; it is impossible to attribute any of them with certainty to ;
early a date. ^ See Clark's Mili/ary Architecture, etc., 1. 176-185. ^
See p. 522. 3 Clark, ii. 224. -J
 1 196] EASTERN INFLUENCES IN THE WEST 533 It must
not be supposed that the " adulterine " castles erected in Stephen's
reign showed any such improvements. Built hastily by men of
precarious fortunes, they were often mere walled enclosures, or at
best rough shell-keeps. Hence it comes that they were so easily
destroyed by Henry if., and that the majority of their sites exhibit
very slight traces of masonry. Perhaps some may have been mere
palisaded mounds of the ancient type. If they had been fitted with
massive rectangular keeps of the tirst Norman model, or with the
more complicated defences introduced from the East, they would
undoubtedly have left far more solid ruins behind them. By the end
of the twelfth century the military architects of the West had learned
their lesson, and were utilising everywhere the notions which had
originally been borrowed from the Byzantines. Outer wards and fore-
works begin to appear beyond the original curtain-walls ; towers
grow numerous and strong, and flanking fire is always provided to
cover exposed fronts. It may be worth while to give a sketch of the
strongest fortress of the day, in order to show the enormous
advance which had been made since the first Crusade. Chateau
Gaillard, as we have already had occasion to mention, was
considered the masterpiece of the time. The reputation of its builder,
Coeur de Lion, as a great military engineer might stand firm on this
single structure. He was no mere copyist of the models which he had
seen in the East, but introduced many original details of his own
invention into the stronghold. It is therefore not exactly a typical
castle of the last years of the thirteenth century, but rather an
abnormally superior specimen of its best work. Chateau Gaillard was
placed in a splendid strategical position, covering Rouen from all
attacks along the line of the Seine. By the aid of its outworks and
the fortified bridge below, it completely blocked the main avenue of
invasion from France. But it is with the castle itself, not with its
dependencies, that we have to deal. Like so many mediaeval
strongholds, it lies on the end of a long spur of steep ground,
connected only by a narrow neck with the hills behind. The slopes
below it are so steep and lofty that it can only be attacked with
advantage along the cramped front of the isthmus which joins it to
the main block of the upland. Its fortifications are intended to
oppose four successive lines of defence to an enemy advancinor
agrainst the single accessible side. Thus it cannot be
 534 THE ART OF WAR IN THE MIDDLE AGES [1196 called a
" concentric " castle, though each of its wards dominates and
commands that below it. The first of its defences is a lower ward or
outwork at the narrowest point of the isthmus. This outwork forms
an isosceles triangle, with its point facing toward the enemy. The
acute angle at its apex is occupied by a great circular tower, which is
flanked and supported by two other towers placed a little distance
down the curtain. The brunt of the attack must therefore fall on
these three towers and the short front of curtain between them. If
the apex oi the triangle was beaten in, the outer ward was lost, and
the defenders could retire to the middle ward. This was separated
from the outwork by a ditch thirty feet deep, crossed only by s single
narrow causeway. Across the ditch lies the middle ward which
exposes to the enemy, when he has gained the outer ward two
massive towers joined by a curtain. Here lay the chape and many
other buildings, whose cellars only now remain Placed within the
northern half of the middle ward was the inner ward, to which King
Richard had devoted special attention Instead of composing it of
towers connected by curtains, he constructed the whole wall in
segments of circles, so that on i ground plan its outer defences
present a scalloped shape. Hi: idea was to give the enceinte all the
advantages of tower; without their heaviness, for the centre part of
each scallop s( advances as to command the space between it and
the nex segment. The general effect is as if he had cut towers int(
slices, and then placed the slices side by side along the steej edge of
the hillside. The donjon forms part of the western wall of the inner
ward it is not completely round, but has a broad spur projecting int<
the open court of the inner ward. It splays out towards thbottom —
a device adopted both to give greater thickness to it base and to
throw outward missiles dropped from its parape" Moreover,itis
furnished withmachicolations,intended to comman< the foot of the
wall ; i.e. a series of corbels carry round it narrow gallery with holes
pierced in its floor, from which th defenders could shoot downwards,
pour liquid combustibles o: the enemy, or drop stones on him. This
is a very early exampl of stone machicolation : the majority of
builders at the time wer only employing wooden galleries
{brattices)^ projecting so as t overlook the ground below the wall. It
seems that ston machicolation was invented in the Holy Land, where
larg
 ;f.204] SIEGE OF CHATEAU GAILLARD 535 timber was so
scarce that the architects of the Crusaders were forced to replace it
by solid masonry. It is interesting to note the methods by which
Chateau Gaillard was taken by Philip Augustus in 1204. King John
neglected it, and allowed it to stand or fall on its own resources
without making any vigorous attempt to raise the siege. The French,
therefore, were able to beleaguer it at leisure, and employed six
months in reducing it by formal siege-operations ^ (September
1203-March 1204). The gallant governor, Roger de Lacy, Constable
of Chester, made an obstinate defence, but, getting no help from
outside, was bound to succumb in time. King Philip appeared in front
of the place in August 1203, and captured the isolated defences in
the neighbourhood lying outside the castle. He spent the autumn in
erecting works of circumvallation and contravallation round it, and in
levelling a platform opposite the apex of the outwork, from which he
intended to begin his attack. The French army lay within its lines all
the winter, fearing that, if it did not remain before the place in force.
King John would appear with a relieving army and raise the
blockade. In February King Philip began the attack by erecting
military machines on the isthmus, and battering the great tower at
the apex of the outwork and the short curtains on each side of it. He
filled the ditch with rubbish, and then set miners to burrow their way
beneath the foundations of the masonry. They finally succeeded in
undermining part of the defences, which fell in, leaving a breach : ^
through this the outer ward was stormed. The garrison, m.uch
reduced by famine, were unable to hold their ground, and retired to
the middle ward. This line of defence did not protect them very long:
it fell, if Guillaume le Breton is to be believed, by a kind of escalade.
In the southwestern angle of the ward lay the chapel, whose outer
wall formed part of the western front of the enceinte. Where the
chapel looked out on the cliff, which lies immediately below it, there
were some small windows not very far above the foot of the wall. A
little party of French crept along the cliff, and ^ Elaborately
described in the Philippeis of Guillaume le Breton, book vii. ^ From
G. le Breton, vii. 705-10, we should conclude that they got in by
throwing down the great angle tower ; but Mr. Clark suggests that as
that building shows no signs of having been breached and repaired,
it must have been the curtain next it which fell in (Clark, i. 384).
 536 THE ART OF WAR IN THE MIDDLE AGES [1204
succeeded in clambering into one of these windows, the first to
mount pulling up his comrades. They found themselves in a crypt
below the chapel : when they had entered they raised their war-cry,
and at the same time the main body made a demonstration along
the causeway against the gate of the middle ward. The garrison,
seeing enemies within the walls, and not realising their small
numbers, did not exterminate the few men who had got in below
the chapel, but hastily evacuated the middle ward and took refuge in
the inner ward, the strongest of all the enceintes of the castle. The
small party in the chapel then came out and admitted their friends.
PhiHp now set to work to erect opposite the gate a perriere of
unwonted size, which, as Guillaume le Breton says, was called a
Cahulus) While thus distracting the attention of the garrison, he
advanced miners under cover of a large " cat," to sap the foot of the
walls This was successfully done, and then the perriere was set tc
work on the shaken masonry. Its discharges brought dowr a
considerable mass of stone, and Philip bade his knights attem.pt to
storm the breach. They would not in all pro bability have succeeded
had not the defenders been reducec to great extremities by hunger.
There only remained twent) knights and a hundred and twenty men
to guard the breach they failed to hold it, and then (if Matthew Paris
may b; trusted), instead of retiring into the donjon, tried to cut thei
way out by the postern-gate and to escape into the open. Ii this
they failed, and were all taken prisoners. (March 6, 1204.) The real
work in this siege, it will be seen, was done by th' miners : it was
they who broke two of the lines of defence while the third was taken
only by the unlikely chance of ai escalade. The siege-engines only
contributed an inconsiderabl part to the main result: the " Cabulus"
might have battered fc ever at the scalloped walls of the inner ward
if the way had nc been prepared for it by the pick of the engineers.
Rounded keeps like that of Chateau Gaillard were ju.^ commencing
to supersede the old square Norman shape whe Richard built his
great castle. The probable reason for the adoption was that such a
shape is better adapted to resist th battering-ram, and even the
Welcome to Our Bookstore - The Ultimate Destination for Book Lovers
Are you passionate about books and eager to explore new worlds of
knowledge? At our website, we offer a vast collection of books that
cater to every interest and age group. From classic literature to
specialized publications, self-help books, and children’s stories, we
have it all! Each book is a gateway to new adventures, helping you
expand your knowledge and nourish your soul
Experience Convenient and Enjoyable Book Shopping Our website is more
than just an online bookstore—it’s a bridge connecting readers to the
timeless values of culture and wisdom. With a sleek and user-friendly
interface and a smart search system, you can find your favorite books
quickly and easily. Enjoy special promotions, fast home delivery, and
a seamless shopping experience that saves you time and enhances your
love for reading.
Let us accompany you on the journey of exploring knowledge and
personal growth!

ebookball.com