Metabook - 266download Ebook Discovering Abstract Algebra 1St Edition John K Osoinach JR Online PDF All Chapter
Metabook - 266download Ebook Discovering Abstract Algebra 1St Edition John K Osoinach JR Online PDF All Chapter
https://ebookmeta.com/product/abstract-algebra-2nd-edition-i-h-
sheth/
https://ebookmeta.com/product/contemporary-abstract-algebra-
joseph-a-gallian/
https://ebookmeta.com/product/concrete-algebra-with-a-view-
toward-abstract-algebra-benjamin-mckay/
https://ebookmeta.com/product/concrete-algebra-with-a-view-
toward-abstract-algebra-benjamin-mckay-2/
Abstract Algebra: An Inquiry-Based Approach, 2nd
Edition Jonathan K. Hodge & Steven Schlicker & Ted
Sundstrom
https://ebookmeta.com/product/abstract-algebra-an-inquiry-based-
approach-2nd-edition-jonathan-k-hodge-steven-schlicker-ted-
sundstrom/
https://ebookmeta.com/product/contemporary-abstract-algebra-9th-
edition-joseph-a-gallian-2/
https://ebookmeta.com/product/contemporary-abstract-algebra-10th-
edition-joseph-a-gallian/
https://ebookmeta.com/product/contemporary-abstract-algebra-9th-
edition-joseph-a-gallian/
https://ebookmeta.com/product/how-to-think-about-abstract-
algebra-1st-edition-lara-alcock/
AMS / MAA TEXTBOOKS VOL 67
D30
2
{ρ2 , δ1 } 8 ρ 9 {ρ , 1 }
2
ρ
3 5
4 5 ρ ρ 6 7
3
6
1 ρ ρ15 ρ10 2
{ι}
Discovering Abstract Algebra
AMS/MAA TEXTBOOKS
VOL 67
Discovering Abstract
Algebra
Copying and reprinting. Individual readers of this publication, and nonprofit libraries acting for them,
are permitted to make fair use of the material, such as to copy select pages for use in teaching or research.
Permission is granted to quote brief passages from this publication in reviews, provided the customary ac-
knowledgment of the source is given.
Republication, systematic copying, or multiple reproduction of any material in this publication is permit-
ted only under license from the American Mathematical Society. Requests for permission to reuse portions
of AMS publication content are handled by the Copyright Clearance Center. For more information, please
visit www.ams.org/publications/pubpermissions.
Send requests for translation rights and licensed reprints to [email protected].
© 2021 by the American Mathematical Society. All rights reserved.
The American Mathematical Society retains all rights
except those granted to the United States Government.
Printed in the United States of America.
⃝
∞ The paper used in this book is acid-free and falls within the guidelines
established to ensure permanence and durability.
Visit the AMS home page at https://www.ams.org/
10 9 8 7 6 5 4 3 2 1 26 25 24 23 22 21
This book is dedicated to my wife, Carolyn,
whose patience and support made this possible.
Contents
Acknowledgments xi
1 Introduction 3
1.1 A brief backstory 3
1.2 Properties of the integers 4
2 Binary operations 7
2.1 Closure 7
2.2 Binary tables 10
2.3 Isomorphic structures 11
3 Groups 15
3.1 Basic properties of groups 15
3.2 Group notation 18
3.3 Group tables and the order of a group 20
5 Applications of subgroups 31
5.1 Cosets 31
5.2 Lagrange’s theorem 33
5.3 Conjugation 34
6 Quotient groups 39
6.1 Homomorphisms and kernel 39
6.2 Normal subgroups 41
6.3 The natural projection homomorphism 44
vii
viii Contents
7 Cyclic groups 47
7.1 Properties of cyclic groups 47
7.2 Infinite cyclic groups 48
7.3 Finite cyclic groups 49
8 Direct products 53
8.1 External direct products 53
8.2 Finitely generated abelian groups 55
11 Alternating groups 71
11.1 Orbits and cycles 71
11.2 Transpositions and the parity of a permutation 73
11.3 The alternating group 74
11.4 Generating sets for symmetric groups 75
11.5 The simplicity of 𝐴5 78
12 Rings 83
12.1 Basic properties of rings 83
12.2 Homomorphisms 87
12.3 Polynomials 89
13 Commutative rings 93
13.1 Integral domains 93
13.2 The Ring ℤ𝑛 95
13.3 Polynomials over integral domains 96
14 Fields 99
14.1 The field of quotients 99
14.2 The characteristic of a ring 101
14.3 Polynomials over a field 102
Appendices 169
B Matrices 179
B.1 Matrix algebra 179
B.2 Matrix inverses 182
x Contents
Index 197
Acknowledgments
This book represents the fruits of many years of labor, starting with my arrival at the
University of Dallas in the Fall of 2009. The mathematics department was (and still is)
steeped in inquiry-based approaches to teaching, beginning with Dr. Charles Coppin,
who “brought” these methods to the university early on. Soon after my arrival, I recog-
nized the need for an inquiry-based approach in my own abstract algebra course, and
hence I began my journey to write a theorem sequence appropriate for undergraduates.
Consequently, this book would not have been possible for me to write and use
without Dr. Coppin’s legacy, nor without the overt support of my colleagues at the
University of Dallas, Paul Phillips, Rob Hochberg, and David Andrews. I am partic-
ularly grateful to my longtime friend Paul, who offered many suggestions (and found
many typos) in the latter half of the book.
I also wish to thank several other colleagues of mine, who used this book in their
own classes during the various stages of its writing. Their testing of the book provided
essential feedback for improvement, pointing out where the scaffolding could be re-
fined or clarified, the need for examples and appendices, and mistakes in the theorem
sequence or exercises. My deep gratitude for being testers of this book goes to:
Heidi Andersen (American International University – Kuwait)
Sean Bailey (Texas A&M University – Texarkana)
Suzanne Caulk (Regis University)
Doug Dailey (Christendom College)
Paul Phillips (University of Dallas)
Josephine Yu (Georgia Tech University)
In addition, the MAA Textbook editorial board provided essential comments and cri-
tiques of the book which vastly improved the style and clarity of the manuscript. I
would like to thank Steve Kennedy especially, who paitently shepherded and guided
me through the majority of the process.
Yet my students at the University of Dallas bore the majority of the testing of this
book. Only because of their diligence in working the theorems and exercises, and shar-
ing with me their struggles on various topics, was I able to construct a theorem se-
quence that guides students through this seemingly intangible subject. Indeed, many
of the exercises I included, along with some motivating theorems, are a direct result of
student feedback and their need for just a bit more intuition or insight. I am indebted
to their hard work on these theorems.
Finally, I cannot begin to thank the members of my family enough for their sup-
port. The countless hours spent writing and revising this book were only possible with
their patience and understanding. I owe my wife, Carolyn, my deepest thanks, for
xi
xii Acknowledgments
without her gift of time for me to write, this book never would have been completed.
It is to her that I dedicate this book.
John Osoinach
April, 2021
To the instructor
Class structure
The premise in using a theorem sequence is for the students to derive the proofs on
their own. That doesn’t mean that the instructor doesn’t have a role, but it does mean
that the instructor’s role isn’t that of showing students “how it’s done.” Rather, it’s the
students who present their ideas to each other, while the instructor makes sure that
the ensuing conversation is adequate to address the issues raised. Consequently, the
only real way to circumvent the intention of the theorem sequence is for the instructor
to prove the theorems for the students. Yes, there will always be a couple of theorems
that students just can’t get, or perhaps the students will be given a break one day, so
xiii
xiv To the instructor
that the instructor will prove a small number of theorems for them. But by and large,
the instructor’s job isn’t to show the students how to prove the theorems. That’s their
job.
Hence, there are several very reasonable structures an instructor can employ to be
faithful to the inquiry-based approach of this book. Any structure taken, though, relies
on student engagement with the theorems at hand, along with student contributions
of their results to the class. There is no one single class structure that an instructor
needs to use for this book to be an effective tool. As a sample of reasonable approaches
an instructor might take, the following are all faithful ways to structure a course using
this theorem sequence.
• Assign a small number of theorems or exercises for the students to prepare for class
each day. Students work independently before class, and during class students
discuss the topics and present their theorems to each other for analysis.
• Have students work in groups outside class to develop proofs of all or most of the
theorems from a section. During class, individuals from different groups take turns
presenting their proofs to the class. Groups can post their proofs beforehand to a
discussion board or other shared medium.
• Have students work together in class on some of the more straightforward the-
orems and exercises. Outside class, have students prove additional theorems or
work additional exercises, with the more central theorems presented the next class
period.
• For an independent or directed study, you and the student share the task of prepar-
ing proofs of the theorems.
Whatever approach is taken, the role of the instructor is not that of a presenter.
Rather, depending on how the class is structured, the instructor’s role is that of a mod-
erator, or a judge, or a guide, or a resource for students. After all, the students’ ideas are
what drive the course; what the instructor does to help students grow is to emphasize
or highlight particularly salient aspects of what the students themselves came up with
and presented.
you feel it’s important for students to grapple with skipped theorems, assigning them
as homework problems is a really good idea.
On the other hand, if your institution has a second course in abstract algebra, then
the remaining chapters (Parts IV – VI) along with anything you skipped make for a
very nice and full second course. You’ll notice that there are fewer chapters in these
last three parts, and that’s by design. This part of the book was written with two goals
in mind: allow students to continue their investigations into abstract algebra, and to
support the MAA’s CUPM guide for recommended topics in the subject. Indeed, the
topics included in Parts IV – VI are not intended to limit what instructors can have their
students discover, but rather to act as a platform for deeper investigations by students,
either as projects, instructor interests, etc.
When I teach from Parts I – III, I’ve found that a reasonable way to keep students on
track is to set a goal of 12 theorems each week, with the expectation that students will
likely discuss only 8-10 of those theorems during the week. After each class meeting, I
assign the next few theorems to the class as a whole, based on where in the sequence the
class ended. For a class that meets three times a week, I assign three or four theorems
or exercises for presentation at the next class meeting; for a class that meets twice a
week, I assign five or six theorems or exercises for the next class meeting. For a second
course, the theorems tend to be a bit more involved, and presenting and discussing full
proofs of those theorems can take a bit longer. Hence, for a second course, I moderate
my expectations slightly, assigning fewer theorems each week, but having a deeper and
richer discussion in class.
Exercises
Finally, I need to mention the role that the exercises play. The ideas and tools in ab-
stract algebra are almost always foreign to students, and students struggle seeing the
connections between disparate objects that the theorems of the subject make. Hence,
the exercises are crucial in giving students tangible applications and expositions of the
theorems, as well as developing their intuition to make more of these connections on
their own. There are too many exercises to warrant assigning them all, but their com-
pletion is fundamental for the development of student understanding of the theory.
Some exercises are excellent motivators for the theory, occurring just before a defini-
tion or theorem; those that follow definitions and theorems are good applications of
what the students just encountered.
Consequently, while the theorems take center stage in this book, including exer-
cises as part of the class routine is necessary for students to check their understanding
of what a theorem really does for them. A good way to use them is as prompts for in-
class discussion, either as preparation for an upcoming theorem, or as a follow-up from
the previous theorem(s). This low-stakes approach gives students the freedom to make
mistakes and instructors the ability to check how well the class as a whole understands
the recent material.
To the student
I have a question for you. Given that the title of this book is, “Discovering Abstract
Algebra,” do you think this book intends to show you how abstract algebra (whatever
that is) was discovered by others, or that this book intends for you to discover abstract
algebra (whatever that is) for yourself?
If your response to that question is the former, then it might be because that’s
the way many of the math courses you’ve taken have been taught. You are shown a
particular topic or idea, chosen by the instructor (or the book the instructor is using),
after which the instructor might give several examples of what to do with the ideas.
Between classes, you are asked to practice what you saw by doing homework problems
that repeat the examples you saw in class. Every so often, you’re examined on how
well you can remember what you were shown. When you can recite how the ideas
were used, you are evaluated as knowing the material.
In contrast, this book’s intent is precisely the latter: for you to discover for your-
self the main themes, ideas, and consequences of the subject we call abstract algebra.
Indeed, the book intends to lead you through a journey, introducing you to interest-
ing objects, suggesting things for you to investigate about them, and prompting you to
think about where your conclusions might lead you. Consequently, the book provides
you with the definitions of the objects and the theorems you are to prove concerning
them, but there are no proofs and only a few examples. The development of abstract
algebra is yours to claim for yourself.
Think about what that means the book asks of you. You cannot passively read the
book, expecting the book to do the work of explaining the material to you for you to
simply believe and repeat. No, the book requires you to be active in your engagement
with the material. You will have to prove the theorems, you will have to work out
examples, you will have to practice with the exercises.
Each chapter of this book represents a new object to study, or a known object to
analyze more deeply. The most important aspects in each chapter with which to en-
gage are the theorems, of course. But usually there’s an example or exercise that helps
motivate a theorem or to which you can apply the theorem you just proved. You can’t
really claim to understand the subject simply by doing computations, and unless you
can apply the theorems to specific examples, you don’t really know what’s going on
either. Yes, the theorems take center stage, but it’s not a solo act. The exercises and
examples help support and describe what’s going on.
Now, please don’t feel like you have to prove every single theorem and work every
single exercise correctly to learn the material. Part of actively engaging the subject
includes the expectation that you will not always succeed: you won’t see the insight of
a proof, or how an example motivates a theorem, or how to solve an exercise using the
xvii
xviii To the student
concepts developed. What you can expect, though, is that your diligent, active attempts
to resolve the issues at hand are exactly what it takes to learn the subject. Being able
to learn why something doesn’t work is just as valuable and important as seeing why
something does work.
As you work through this sequence of theorems, then, avoid giving in to the temp-
tation to give up just because you don’t immediately know how to prove the theorem
before you. Indeed, another part of what actively engaging the subject includes is a
devotion to play. Yes, that’s right, play. Play with examples, ideas, definitions, the-
orems – really, anything that you can which you think might help you along with a
proof. But unlike homework problems that can be solved in a couple of lines by us-
ing a formula, proving theorems takes perseverance. Mathematical play gives you the
time, space, and freedom to try different things, good and bad, so that you can work
towards a proof or solution. And while not everything you try will work, do have faith
in yourself that your mathematical background has provided you with enough tools to
engage in productive play.
Finally, but certainly not least, don’t expect your ideas to be succinct, refined, or
sophisticated. When actively engaging a subject for the first time, the first proofs of
theorems tend be, well, rough. But again, that’s not the point of working through the
theorems. Any correct proof of a theorem is no more or less valid than any other one.
That’s why you can’t really talk about “the” proof of a theorem at all (there are over 350
different proofs of the Pythagorean Theorem alone). Each proof reflects one person’s
insights into the problem, and the more proofs you see and understand, the more you
grow and the more tools you have with which to play. Hence, be prepared to share
your ideas with your classmates, so that everyone can see what you see, but also so
that others can suggest improvements to make your ideas clearer.
So, go into this process expecting things to get messy. Mathematical play is sup-
posed to be messy. The objects you’re about to meet took decades to become as clear
and precise as they are now. Give yourself that same freedom to be messy, with the
goal of becoming more and more clear and precise as you proceed through the theo-
rem sequence. And when you’re done, you can say confidently that you did, in fact,
discover abstract algebra.
Part 1
Group theory
Introduction
1
1.1 A brief backstory
Algebra is an old subject. I mean, really, really old. But exactly how old the subject
is depends on what we mean by algebra.1 For instance, consider Diophantus, a 3rd
century Greek mathematician of Alexandria. His 13-volume treatise, the Arithmetica,
includes what many view as a first attempt at algebra. In it, Diophantus articulates a
process for finding a number that solves a particular computational problem. To aid
his exposition of solving these problems, he develops a system to notate the unknown
solution and the operations that describe that process. Is this where algebra started?
Perhaps instead we should consider Muhammad ibn Musa al-Khwarizmi, a 9th
century Arabic mathematician of Baghdad. One of his great works, Hisab al-jabr w’al-
muqabala, is also considered a treatise on algebra. Indeed, it is this book from which
our term “algebra” is derived, coming from the Latin translation of the Arabic word “al-
jabr” as “algebrae.” In his book, al-Khwarizmi poses problems to find a number that
satisfies a computational procedure, much like Diophantus. But al-Khwarizmi groups
his problems into different types, and for each type of problem he provides a process to
find the desired number. Is this really the beginnings of algebra?
To be fair, this brief introduction misses the point. What’s relevant is not the who,
where, or when, but rather the what: what topic are these two individuals interested in
understanding? From the brief summary above, it’s clear that both of them are inter-
ested in developing a process (if one exists) for identifying one or more numbers that
satisfy an initial description. In our modern terminology, what they’re both doing is
solving equations. Hence, this subject which we now call algebra concerns itself with
solving equations, and that will be our primary goal in this book as well.
That said, we do not intend to start with Diophantus or with al-Khwarizmi and pro-
ceed through nearly two millennia of mathematics. After all, during that time, mathe-
maticians developed and advanced what we now recognize as our notions of equation
1 For an excellent exposition of the mathematicians at the heart of algebra’s origins, read Chapter 5 of
3
4 Chapter 1. Introduction
and solution, along with a variety of notations, objects, and operations to which those
same concepts apply. Furthermore, due to the wide variety of objects to which these
concepts apply, mathematicians have done what they do best: identify common ele-
ments from the disparate objects and create a body of knowledge around them. This
tendency of mathematics to abstract common themes is why this course is commonly
referred to as abstract algebra. Therefore, while the topics you’re about to encounter
may seem far, far removed from these humble origins, the quest to solve equations is
nonetheless the driving force behind everything you will learn.
Property 2. If 𝑥 and 𝑦 are integers, then so is 𝑥𝑦; if 𝑥 and 𝑦 are positive integers, then
so is 𝑥𝑦.
This last property is particularly important and useful in algebra. It says that if
you create a set of positive integers and can prove that it’s not empty, then your set
has a smallest element. Consequently, this property is often used when you want to
prove that a smallest number exists in a set that you’ve created. For practice, read the
following definition, then use the well-ordering property to prove your first theorems.
This first theorem is a good application of the well-ordering property, but its sim-
plicity is due mostly to the fact that the subset of ℤ+ was (nearly) given to you. More
often, you’ll have to construct your own subset of positive integers, then apply the well-
ordering property to the set you created. The following very useful theorem illustrates
this point.
Theorem 1.4. Let 𝑆 be a nonempty set of integers. If there exists an integer 𝑚 such that
𝑚 < 𝑠 for all 𝑠 ∈ 𝑆, then 𝑆 has a least element.
1.2. Properties of the integers 5
Let’s use what we’ve learned to prove our first major theorem, called the Division
Algorithm. As an aid, first prove this little lemma, then use that lemma to prove the
theorem.
Lemma 1.5. Let 𝑚 and 𝑛 be integers, with 𝑛 > 0. Then the set {𝑚 − 𝑥 | 𝑥 ∈ 𝑛ℤ} has a
least nonnegative element.
Theorem 1.6 (The Division Algorithm). Let 𝑚 and 𝑛 be integers, with 𝑛 > 0. Then
there exist unique integers 𝑞 and 𝑟 such that 𝑚 = 𝑛𝑞 + 𝑟, where 0 ≤ 𝑟 < 𝑛.
Definition 1.7. Let 𝑚 and 𝑛 be integers, with 𝑛 > 0. Then the unique integers 𝑞 and
𝑟 such that 𝑚 = 𝑛𝑞 + 𝑟, where 0 ≤ 𝑟 < 𝑛, are called the quotient and remainder,
respectively, of 𝑚 divided by 𝑛.
Corollary 1.8. Let 𝑚 and 𝑛 be integers, with 𝑛 > 0. Then 𝑚 is a multiple of 𝑛 if and only
if the remainder of 𝑚 divided by 𝑛 in the Division Algorithm is 0.
Theorem 1.9. Let 𝑎, 𝑏, 𝑛 ∈ ℤ with 𝑛 > 0. If 𝑟𝑎 and 𝑟 𝑏 are the remainders of 𝑎 and 𝑏
divided by 𝑛, respectively, then the remainder of 𝑎 + 𝑏 divided by 𝑛 equals the remainder
of 𝑟𝑎 + 𝑟 𝑏 divided by 𝑛, and the remainder of 𝑎𝑏 divided by 𝑛 equals the remainder of 𝑟𝑎 𝑟 𝑏
divided by 𝑛.
Binary operations
2
2.1 Closure
When one begins to think on the variety of equations one tries to solve, it doesn’t take
long to realize that each subject has its own objects, upon which are defined a variety
of processes that you can use to manipulate them: addition on numbers, matrix multi-
plication on matrices, etc. But despite the apparent differences between them, they all
share one fundamental property: whenever you combine two or more of the objects,
you get another object of the same type: the sum of two numbers is a number, the prod-
uct of two 𝑛 × 𝑛 matrices is another matrix, etc. Hence, that’s where we’ll start our
study of abstract algebra proper.
To begin, let’s reflect on what we’re really doing when we “combine” objects. We
take two objects from a set and apply some kind of rule or process to this pair of objects,
and the result produces an object from that same set. Yet what we’ve described here
is nothing more than a function whose domain is ordered pairs of elements from a set
and whose range is contained back in that same set. This basic observation is codified
in the following definition.
(1) If ∗ is a binary operation on a set 𝐺, then the element 𝑎 ∗ 𝑏 cannot be undefined for
𝑎 𝑐 𝑎 𝑐
any elements 𝑎, 𝑏 ∈ 𝐺. For instance, if we let 𝐺 = ℚ and define 𝑏 ∗ 𝑑 = 𝑏 / 𝑑 , then
this rule isn’t defined when 𝑐 = 0, since division by zero isn’t defined.
7
8 Chapter 2. Binary operations
(2) Likewise, if ∗ is a binary operation on a set 𝐺, then the element 𝑎 ∗ 𝑏 must be well-
defined for each pair of elements 𝑎, 𝑏 ∈ 𝐺. For instance, if we let 𝐺 = ℚ and define
𝑎 𝑐 1 2
𝑏
∗ 𝑑 = 𝑎 + 𝑐, then this rule isn’t well-defined, since 2 and 4 represent the same
𝑎 1 𝑎 2
number, but 𝑏
∗ 2
= 𝑎 + 1, and 𝑏
∗ 4
= 𝑎 + 2, which are always different.
Aside. Students often fail to appreciate the import of this example. The issue of a well-
defined function always arises whenever the objects in the set have more than one de-
scription or form. The standard way to verify that a binary operation ∗ is well-defined is
to take two equivalent forms of the objects in your set and prove that the result does not
depend on the form of the objects. Functions dealing with rational numbers frequently
fall in this category, since every rational number has infinitely many equivalent frac-
tional forms. Hence, to check that a function 𝑓 is well-defined in this case, you would
𝑎 𝑐 𝑎 𝑐
need to suppose that 𝑏 = 𝑑 and prove that 𝑓 ( 𝑏 ) = 𝑓 ( 𝑑 ).
This idea of closure isn’t new – go back and look at the first two properties of the
integers and positive integers. Those two properties simply state that addition and
multiplication are both binary operations on ℤ and ℤ+ , so that ℤ and ℤ+ are closed
under both addition and multiplication. In general, it’s pretty easy to define a binary
operation on a set – simply make sure that it’s well-defined, it’s defined everywhere,
and that it’s closed, and you’re done!
Example 2.2. Most operations you can think of are binary operations. For instance:
(1) Addition, subtraction, and multiplication of integers, real numbers, and complex
numbers are all binary operations on those sets, but not division, since division by
zero isn’t defined.
(2) Addition and subtraction of 𝑚 × 𝑛 matrices with either integer, real, or complex
entries are binary operations on those sets, but not multiplication, since multipli-
cation isn’t defined when 𝑚 ≠ 𝑛. However, if your set is all 𝑛 × 𝑛 matrices, then
multiplication is a binary operation on that set of matrices.
(3) Addition, subtraction, and multiplication of functions from ℤ, ℝ, or ℂ to ℤ, ℝ, or ℂ
are all binary operations on those sets of functions, which brings us to an important
notational issue. We define the sum (or difference, or product) of two functions 𝑓
and 𝑔 by defining (𝑓 + 𝑔)(𝑥) = 𝑓(𝑥) + 𝑔(𝑥) (likewise for subtraction and multi-
plication). In other words, the symbol 𝑓 + 𝑔 is the name of the sum of 𝑓 and 𝑔,
and the image of the “sum” function 𝑓 + 𝑔 is the sum of the individual images of
𝑓 and 𝑔.
(4) Given any set 𝐴, then composition of functions from 𝐴 to itself is a binary operation
on the set of all functions from 𝐴 to itself.
Some less usual ones for you to consider are:
(1) For any nonempty set 𝐴, pick any element 𝑐 ∈ 𝐴, and define 𝑎 ∗ 𝑏 = 𝑐 for all
𝑎, 𝑏 ∈ 𝐴. In other words, you always get 𝑐.
(2) For any nonempty set 𝐴, define 𝑎∗𝑏 = 𝑎. In other words, forget the second element
entirely.
(3) On the set ℤ+ , the operation 𝑎 ∗ 𝑏 = 𝑎𝑏 is a perfectly fine binary operation.
2.1. Closure 9
𝑎
(4) The set { 𝑏 ∈ ℚ | 𝑏 = 2𝑘 , 𝑘 ∈ ℤ} is called the set of dyadic rationals, and multipli-
cation is a binary operation on that set.
Exercise 2.3. For each set below, determine if the proposed operation on the set 𝐺 is
a binary operation. If not, explain if the operation isn’t well-defined, if it’s undefined,
or if it’s not closed.
(1) 𝐺 = ℤ; 𝑎 ∗ 𝑏 = 𝑎𝑏
𝑎 𝑐 𝑎+𝑐
(2) 𝐺 = ℚ; 𝑏 ∗ 𝑑
= 𝑏𝑑
(3) 𝐺 = ℝ; 𝑎 ∗ 𝑏 = |𝑏 − 𝑎|.
(4) 𝐺 = {𝑎 + 𝑏𝑖 ∈ ℂ | 𝑎, 𝑏 ∈ ℚ, 𝑎2 + 𝑏2 ≠ 0}; 𝑧 ∗ 𝑤 = 𝑧/𝑤
(5) 𝐺 is the set of dyadic rationals; 𝑎 ∗ 𝑏 = 𝑎 + 𝑏.
(6) 𝐺 = {𝑓 ∶ ℝ → ℝ | 𝑓(𝑥) = 𝑎𝑥 + 𝑏, 𝑎, 𝑏 ∈ ℝ}; 𝑓 ∗ 𝑔 = 𝑓 ∘ 𝑔
𝑎 𝑏 |
(7) 𝐺 = {[ ] | 𝑎, 𝑏, 𝑐, 𝑑 are odd integers} ; 𝐴 ∗ 𝐵 = 𝐴𝐵
𝑐 𝑑 |
𝑎 𝑏 |
(8) 𝐺 = {[ ] | 𝑎, 𝑏 ∈ ℝ, 𝑎𝑑 ≠ 𝑏𝑐} ; 𝐴 ∗ 𝐵 = 𝐴𝐵
𝑐 𝑑 |
Notice that in the above examples and exercises, the order in which the elements
appear in the operation sometimes matters; that is, computing 𝑎 ∗ 𝑏 might be different
from 𝑏 ∗ 𝑎. A very good example of this is matrix multiplication: given two 𝑛 × 𝑛
matrices 𝐴 and 𝐵, the product 𝐴𝐵 need not equal 𝐵𝐴 (in fact, it’s rarely true that those
two are equal). Likewise, the order in which you compose functions also matters, since
𝑓 ∘ 𝑔 and 𝑔 ∘ 𝑓 are typically different functions. Hence, we should probably give this
concept a name.
Definition 2.4. Let ⟨𝐺, ∗⟩ be a binary structure. Then ∗ is commutative if and only
if 𝑎 ∗ 𝑏 = 𝑏 ∗ 𝑎 for all 𝑎, 𝑏 ∈ 𝐺.
Exercise 2.5. Identify which of the binary operations in Example 2.2 and Exercise 2.3
are commutative and which are not.
Our next concept needs a bit of motivation before we define it. To begin, if ⟨𝐺, ∗⟩
is a binary structure, then for any 𝑎, 𝑏 ∈ 𝐺, the object 𝑎 ∗ 𝑏 is an element back in 𝐺.
That means we can pick a third element, 𝑐 ∈ 𝐺, and we can compute (𝑎 ∗ 𝑏) ∗ 𝑐. On
the other hand, suppose we compute 𝑏 ∗ 𝑐 first, and then compute 𝑎 ∗ (𝑏 ∗ 𝑐) second.
Should we expect (or even want) those to be the same?
This doesn’t always happen, unfortunately, which is something each grade-school
child learns. Specifically, consider the operation of subtraction on the integers. If you
take three numbers, 𝑎, 𝑏, 𝑐 ∈ ℤ, and compare (𝑎 − 𝑏) − 𝑐 and 𝑎 − (𝑏 − 𝑐), you don’t
always get the same answer. That’s why we can’t just write down the symbol 𝑎 − 𝑏 − 𝑐
and have it make sense – depending on how you group the numbers, you’ll possibly
get different results. On the other hand, addition and multiplication of integers (and
real and complex numbers too, for that matter) can be grouped differently without
10 Chapter 2. Binary operations
changing the result: (𝑎 + 𝑏) + 𝑐 = 𝑎 + (𝑏 + 𝑐) and (𝑎𝑏)𝑐 = 𝑎(𝑏𝑐). That’s why the symbols
𝑎 + 𝑏 + 𝑐 and 𝑎𝑏𝑐 are unambiguous. It has nothing to do with the order in which the
numbers appear – that’s the commutative property. It’s all about how the numbers are
grouped, and that’s our next property to define.
Definition 2.6. Let ⟨𝐺, ∗⟩ be a binary structure. Then ∗ is associative if and only if
(𝑎 ∗ 𝑏) ∗ 𝑐 = 𝑎 ∗ (𝑏 ∗ 𝑐) for all 𝑎, 𝑏, 𝑐 ∈ 𝐺.
Example 2.7. Most of the common examples given previously are associative opera-
tions: addition and multiplication of objects such as numbers, functions, and matrices
are all associative operations, as well as function composition.
Exercise 2.8. Go back to the less usual examples from Example 2.2 as well as Exercise
2.5, and determine which of those operations are associative.
When a binary structure ⟨𝐺, ∗⟩ has an associative binary operation, we simply write
𝑎 ∗ 𝑏 ∗ 𝑐 to indicate either (𝑎 ∗ 𝑏) ∗ 𝑐 or 𝑎 ∗ (𝑏 ∗ 𝑐), since there is no ambiguity. We
can then extend this notation to any finite number of elements in 𝐺; that is, we write
𝑎1 ∗ 𝑎2 ∗ ⋯ ∗ 𝑎𝑛 to indicate any of the possible groupings of the 𝑎𝑖 , all of which yield the
same result due to associativity. Technically, this fact needs a proof, but the induction
is needlessly messy, and it really doesn’t shed any light on the concept.
Exercise 2.9. Let 𝐺 = {𝑎, 𝑏, 𝑐}. Suppose we define a binary operation on 𝐺 with the
following table:
∗ 𝑎 𝑏 𝑐
𝑎 𝑎 𝑐 𝑐
(So, for instance, 𝑐 ∗ 𝑎 = 𝑎.)
𝑏 𝑏 𝑏 𝑏
𝑐 𝑎 𝑎 𝑏
(1) Compute 𝑎 ∗ 𝑏, 𝑏 ∗ 𝑎, 𝑏 ∗ 𝑏, (𝑎 ∗ 𝑐) ∗ 𝑏, 𝑎 ∗ (𝑐 ∗ 𝑏), 𝑐 ∗ (𝑐 ∗ 𝑐), and (𝑐 ∗ 𝑐) ∗ 𝑐.
(2) Determine if the operation is commutative, associative, neither, or both.
2.3. Isomorphic structures 11
There’s really no theory about binary tables, but there are several observations that
are useful to have. First, since you can put any of the 𝑛 elements of 𝐺 into any of the
2
𝑛2 entries, there are a total of 𝑛𝑛 possible ways to construct a binary operation on 𝐺.
Second, checking to see if a binary table’s operation is commutative is easy: reflect
the table along the “main diagonal” and compare with the original table. If they’re
the same, then it’s a commutative operation; otherwise, you’ll have at least one pair of
elements 𝑎𝑖 , 𝑎𝑗 such that 𝑎𝑖 ∗ 𝑎𝑗 ≠ 𝑎𝑗 ∗ 𝑎𝑖 .
Associativity, on the other hand, is never easy to check by looking at the table.
That’s because checking associativity deals with using the table sequentially. It also
means verifying that (𝑎 ∗ 𝑏) ∗ 𝑐 = 𝑎 ∗ (𝑏 ∗ 𝑐) for all choices of 𝑎, 𝑏, 𝑐 ∈ 𝐺, which means
you’ve got 𝑛3 different pairs of triples to compare. That’s just too tedious to do by hand;
a computer is almost a necessity if you need to know if your operation is associative.
Exercise 2.10. Several entries in the binary table below have been left blank. It turns
out that there’s one and only one way to fill in the table so that the resulting binary
operation is associative. Fill in the missing entries to create an associative binary op-
eration. Is the operation commutative?
∗ 𝑎 𝑏 𝑐
𝑎 𝑎
𝑏 𝑐
𝑐 𝑏
(Hint: You need 𝑏 ∗ (𝑏 ∗ 𝑐) = (𝑏 ∗ 𝑏) ∗ 𝑐 and 𝑏 ∗ (𝑐 ∗ 𝑐) = (𝑏 ∗ 𝑐) ∗ 𝑐. What is the
only element that 𝑏 ∗ 𝑐 can equal?)
So, an injection is a function such that no two different elements in the domain
are mapped to the same element in the range; a surjection is a function whose range is
the entire codomain. But bijections are the functions for which every element in the
codomain is the image of one and only one element in the domain. That’s exactly what
we’re looking for when we want to match up elements in two sets, and that’s what we’ll
use in our key term of this section.
Exercise 2.13. Which, if any, of the following are isomorphisms? Are any automor-
phisms?
(1) 𝜙 ∶ ⟨ℝ, +⟩ → ⟨ℝ+ , ⋅⟩ is given by 𝜙(𝑥) = 𝑒𝑥 .
(2) 𝜙 ∶ ⟨ℤ, ⋅⟩ → ⟨2ℤ, ⋅⟩ is given by 𝜙(𝑥) = 2𝑥.
1
(3) 𝜙 ∶ ⟨ℚ+ , ⋅⟩ → ⟨ℚ+ , ⋅⟩ is given by 𝜙(𝑥) = 𝑥 .
(4) 𝐺 is the set of all 𝑛 × 𝑛 invertible matrices, and 𝜙 ∶ ⟨𝐺, ⋅⟩ → ⟨𝐺, ⋅⟩ is given by
𝜙(𝐴) = 𝐴−1 .
Of course, that this means that the relation of “isomorphic” is an equivalence re-
lation on the set of binary structures (see Section A.1 in Appendix A). In this way, we
can simply say that ⟨𝐺, ∗⟩ and ⟨𝐺 ′ , ∗′ ⟩ are isomorphic.
It’s worthwhile to highlight both what it takes to use the definition directly to show
that two structures are isomorphic, and what it takes to show that two structures are not
isomorphic. For the former, you must first find a function 𝜙 ∶ 𝐺 → 𝐺 ′ , then prove that
your function is one-to-one, onto, and that for all 𝑎, 𝑏 ∈ 𝐺, that 𝜙(𝑎 ∗ 𝑏) = 𝜙(𝑎) ∗′ 𝜙(𝑏).
On the other hand, to show that two structures are not isomorphic, you must show that
2.3. Isomorphic structures 13
it is impossible to find such a function. Showing that one or two functions don’t work
isn’t enough – you’ve got to prove that no isomorphism exists. Although that may seem
hard, there is a classic way to do this.
Identifying algebraic properties is exactly what’s used to identify when two struc-
tures are not isomorphic. After all, if one structure has a particular algebraic property
that another one doesn’t, then it’s impossible for them to be isomorphic: if they were
isomorphic, then they’d both share the same property! Hence, identifying those prop-
erties that are algebraic will be a continuing theme in this book. Let’s build some now
to help develop this idea.
Theorem 2.19. If a binary structure ⟨𝐺, ∗⟩ has an identity element, then that identity
element is unique.
Thus, when a binary structure has an identity element, we can talk about the iden-
tity element. Because of this, when you want to make a table of a binary structure that
has an identity, there is a convention we use: list the identity element first.
Exercise 2.20. The same theorem isn’t true for left or right identity elements. Create
a table to define a binary structure which has two left identity elements.
Theorem 2.21. If ⟨𝐺, ∗⟩ and ⟨𝐺 ′ , ∗′ ⟩ are isomorphic binary structures, and ⟨𝐺, ∗⟩ has an
identity element 𝑒 ∈ 𝐺, then 𝜙(𝑒) is the identity element for 𝐺 ′ , where 𝜙 ∶ 𝐺 → 𝐺 ′ is any
isomorphism from 𝐺 to 𝐺 ′ . Consequently, the property “⟨𝐺, ∗⟩ has an identity element” is
an algebraic property.
Exercise 2.22. In each of the following standard binary structures, identify the identity
element, if it exists.
(1) ⟨ℤ, +⟩ (likewise for ℚ, ℝ, ℂ).
(2) ⟨ℤ, ⋅⟩ (likewise for ℚ, ℝ, ℂ).
14 Chapter 2. Binary operations
(3) ⟨𝐺, ⋅⟩, where 𝐺 is the set of all 𝑛 × 𝑛 matrices over ℤ (likewise for matrices over
ℚ, ℝ, ℂ).
(4) ⟨𝐺, ∘⟩, where 𝐺 is the set of all functions 𝑓 ∶ 𝐴 → 𝐴 for a given nonempty set 𝐴.
Definition 2.23. Let ⟨𝐺, ∗⟩ be a binary structure with an identity element 𝑒. A left
inverse for an element 𝑎 ∈ 𝐺 is any element 𝑎′ ∈ 𝐺 such that 𝑎′ ∗ 𝑎 = 𝑒. Likewise, a
right inverse for an element 𝑎 ∈ 𝐺 is any element 𝑎′ ∈ 𝐺 such that 𝑎 ∗ 𝑎′ = 𝑒. An
element 𝑎′ ∈ 𝐺 that is both a left and right inverse for 𝑎 ∈ 𝐺 is called an inverse for
𝑎 ∈ 𝐺.
Theorem 2.24. Let ⟨𝐺, ∗⟩ be a binary structure with an identity element 𝑒. If 𝑎′ is a left
inverse for 𝑎, then 𝑎 is a right inverse for 𝑎′ ; if 𝑎′ is a right inverse for 𝑎, then 𝑎 is a left
inverse for 𝑎′ ; and if 𝑎′ is an inverse for 𝑎, then 𝑎 is an inverse for 𝑎′ .
Theorem 2.25. If ⟨𝐺, ∗⟩ and ⟨𝐺 ′ , ∗′ ⟩ are isomorphic binary structures with an identity
such that every element in ⟨𝐺, ∗⟩ has an inverse, then whenever 𝑎′ ∈ 𝐺 is an inverse
of 𝑎 ∈ 𝐺, the element 𝜙(𝑎′ ) ∈ 𝐺 ′ is an inverse for 𝜙(𝑎) ∈ 𝐺 ′ for any isomorphism
𝜙 ∶ 𝐺 → 𝐺 ′ . Consequently, the property “Every element in ⟨𝐺, ∗⟩ has an inverse” is an
algebraic property.
Exercise 2.26. Create a table to define a binary structure which has an element with
two different inverses.
Exercise 2.27. In each of the following standard binary structures, identify those ele-
ments that have an inverse.
(1) ⟨ℤ, +⟩ (likewise for ℚ, ℝ, ℂ).
(2) ⟨ℤ, ⋅⟩.
(3) ⟨ℚ, ⋅⟩ (likewise for ℝ, ℂ).
(4) ⟨𝐺, ⋅⟩, where 𝐺 is the set of all 𝑛 × 𝑛 matrices over ℤ (likewise for matrices over
ℚ, ℝ, ℂ).
(5) ⟨𝐺, ∘⟩, where 𝐺 is the set of all functions 𝑓 ∶ 𝐴 → 𝐴 for a given nonempty set 𝐴.
Groups
3
3.1 Basic properties of groups
It’s now time for us to define our first algebraic structure. As we said in the introduc-
tion, the underlying theme of this subject is that of solving equations. In the context of
binary structures, that means that we want to answer the following question:
Given a binary structure ⟨𝐺, ∗⟩ and any two elements 𝑎, 𝑏 ∈ 𝐺, does there exist an
element 𝑥 ∈ 𝐺 such that 𝑎 ∗ 𝑥 = 𝑏? If so, how many such elements exist?
This simple question is the prototype for all questions concerning the solution of
equations: does a solution exist, and if so, how many?
Let’s try to play around with the equation 𝑎 ∗ 𝑥 = 𝑏 to see what we might desire
in order to solve this equation. Intuitively, what we want to do is “move” the 𝑎 to the
other side somehow. For instance, if 𝐺 were the real numbers and ∗ was addition, we
could subtract 𝑎 from both sides. If 𝐺 were the set of 𝑛 × 𝑛 matrices and ∗ was matrix
multiplication, we could try to multiply by 𝐴−1 , if 𝐴 were invertible. In both cases,
we are using the idea of an inverse of 𝑎: either adding −𝑎 to both sides (that’s what
subtraction really is), or multiplying by the inverse matrix of 𝐴. So, it looks like we
really, really need every element of 𝐺 to have an inverse. However, we can’t talk about
an inverse of an element without first having an identity element. So, it looks like at
the very least we really, really need for 𝐺 to have an identity element as well.
Let’s press this further and posit that 𝐺 has an identity, 𝑒, and that every element
𝑎 ∈ 𝐺 does, in fact, have an inverse 𝑎′ . Let’s play around and see if this is what we
need!
Suppose 𝑎 ∗ 𝑥 = 𝑏. Then we can write 𝑎′ ∗ (𝑎 ∗ 𝑥) = 𝑎′ ∗ 𝑏, which results in
an expression (on the left hand side) with three elements of 𝐺 grouped together. And
since we really would like to group the 𝑎′ with the 𝑎, we really, really need for the
operation ∗ to be associative. Fine. Let’s posit that property too. That would mean that
𝑎′ ∗ (𝑎 ∗ 𝑥) = (𝑎′ ∗ 𝑎) ∗ 𝑥 = 𝑒 ∗ 𝑥 = 𝑥, and thus 𝑥 = 𝑎′ ∗ 𝑏. And since the operation ∗
is closed, the element 𝑎′ ∗ 𝑏 is an element of 𝐺. Thus, if the equation 𝑎 ∗ 𝑥 = 𝑏 has a
solution, it must have the form 𝑥 = 𝑎′ ∗ 𝑏. The only thing left to check is to verify that
15
16 Chapter 3. Groups
Definition 3.1. Let ⟨𝐺, ∗⟩ be a binary structure. We say that ⟨𝐺, ∗⟩ is a group if the
following three properties hold:
Example 3.2. The following list contains the notation of the standard groups you’ll
need to know as you proceed through the book.
(1) The integers, rationals, reals, and complex numbers are all groups under addition.
We notate these groups simply by ℤ, ℚ, ℝ, and ℂ, respectively, without explicit ref-
erence to addition as the operation. The additive identity element is 0, of course,
and the (additive) inverse of a number 𝑥 is −𝑥.
(2) The set of all nonzero rationals, reals, and complex numbers forms a group under
multiplication. We denote these groups by ℚ∗ , ℝ∗ , and ℂ∗ , respectively, without
explicit reference to multiplication as the operation. The multiplicative identity
element is 1, of course, and the (multiplicative) inverse of a number 𝑥 is its recip-
rocal, 1/𝑥.
(3) The set of all invertible 𝑛 × 𝑛 matrices over the integers, rationals, reals, or complex
numbers forms a group under matrix multiplication. This group is called the gen-
eral linear group and is denoted 𝐺𝐿𝑛 (ℤ) (or ℚ, ℝ, ℂ). The identity element here is
the identity matrix 𝐼𝑛 , and the inverse of a matrix 𝐴 is denoted 𝐴−1 (see Definitions
B.9 and B.13).
(4) The group of all bijections from a set 𝐴 to itself forms a group under function
composition. This group is denoted 𝑆𝐴 , and its identity element is the function
𝜄 ∶ 𝐴 → 𝐴 given by 𝜄(𝑎) = 𝑎 for all 𝑎 ∈ 𝐴 (see Exercise 2.22 number 4). The inverse
of a bijection 𝑓 ∶ 𝐴 → 𝐴 is denoted 𝑓−1 (see Definition A.23).
Aside. Part of the definition of a group requires that we begin with a binary structure
⟨𝐺, ∗⟩. Hence, before we check the three properties listed in the definition, we must
first verify that we are, in fact, dealing with a binary operation on the set. For instance,
if we let 𝐺 = {0, ±1} and define 𝑎 ∗ 𝑏 = 𝑎 + 𝑏, then this operation satisfies all three
of the properties. But ⟨𝐺, ∗⟩ isn’t a binary structure, since 1 + 1 ∉ 𝐺. Having a binary
structure is a necessary precondition to apply the definition!
3.1. Basic properties of groups 17
Exercise 3.4. Which, if any, of the following are groups? For those that are, give the
identity element of the group. (Hint: check associativity last.)
(1) 𝐺 = the nonnegative integers; 𝑚 ∗ 𝑛 = 𝑚 + 𝑛.
(2) 𝐺 = ℤ; 𝑥 ∗ 𝑦 = 𝑥 + 𝑦 − 2.
(3) 𝐺 = ℚ; 𝑟 ∗ 𝑠 = 𝑟𝑠.
(4) 𝐺 = {1, 2, 3, 4}, 𝑎 ∗ 𝑏 = 𝑟, where 𝑟 is the unique remainder of 𝑎𝑏 divided by 5.
(5) 𝐺 = the set of all 𝑛 × 𝑛 diagonal matrices over ℝ; 𝐴 ∗ 𝐵 = 𝐴𝐵.
(6) 𝐺 = {𝑓 ∶ ℝ → ℝ | 𝑓(𝑥) ≥ 0}; (𝑓 ∗ 𝑔)(𝑥) = |𝑓(𝑥) − 𝑔(𝑥)|.
(7) 𝐺 = {𝑎 + 𝑏𝑖 ∈ ℂ | 𝑎2 + 𝑏2 = 1}; 𝑧 ∗ 𝑤 = 𝑧𝑤.
(8) Let 𝑋 be a set, and let 𝐺 = 𝒫(𝑋), the power set of 𝑋; 𝑆 ∗ 𝑇 = (𝑆 ∪ 𝑇) − (𝑆 ∩ 𝑇).
18 Chapter 3. Groups
Based on our previous discussion, we have that the equation 𝑎 ∗ 𝑥 = 𝑏 always has a
solution in a group. But how many solutions does it have? We know that every solution
has the form 𝑥 = 𝑎′ ∗ 𝑏, so if we knew how many inverses each element has, then we
could answer the question. Hence, that should be our first theorem of the section.
Theorem 3.6. Let ⟨𝐺, ∗⟩ be a group, and let 𝑔 ∈ 𝐺. Then there exists exactly one inverse
for 𝑔.
Aside. The three group properties were posited as useful for solving simple equations,
as this corollary shows. It’s worth asking if they’re necessary; that is, do you need a
group structure to solve these equations. As it turns out, the answer is no; you can find
a binary operation on a set with three elements that is not associative but for which
there are unique solutions to the equations 𝑎 ∗ 𝑥 = 𝑏 and 𝑦 ∗ 𝑎 = 𝑏. However, if you
require an associative binary operation, then a group structure is, in fact, necessary to
have unique solutions to those equations.
Theorem 3.8 (The Cancellation Laws). Let ⟨𝐺, ∗⟩ be a group, and let 𝑎, 𝑏, 𝑐 ∈ 𝐺. Then
the following hold:
(1) (Left Cancellation Law) If 𝑐 ∗ 𝑎 = 𝑐 ∗ 𝑏, then 𝑎 = 𝑏.
(2) (Right Cancellation Law) If 𝑎 ∗ 𝑐 = 𝑏 ∗ 𝑐, then 𝑎 = 𝑏.
Definition 3.9. Let ⟨𝐺, ∗⟩ be a group. We say that ⟨𝐺, ∗⟩ is an abelian group if and
only if the operation ∗ is commutative.
So here’s our first convention: If we use the symbol + to denote our group’s oper-
ation, we are implying that the group is an abelian group.
On the other hand, the notation for multiplication is used both for abelian and
nonabelian group operations. Since that’s the more “generic” symbol, we’ll use that
instead of ∗, and we now simply write that 𝐺 is a group, without including the “name”
of the operation. Hence, our convention is the following:
When 𝐺 is a group, to denote the element 𝑎 ∗ 𝑏, we write 𝑎𝑏 for an abstract group,
and we call 𝑎𝑏 the product of 𝑎 with 𝑏; if we write 𝑎 + 𝑏, we imply that the group is
abelian, and we call 𝑎 + 𝑏 the sum of 𝑎 and 𝑏.
This notation has proved so useful that we adopt exponential notation for a group
𝐺:
𝑛 times
(1) For any element 𝑎 ∈ 𝐺 and any positive integer 𝑛, we write 𝑎𝑛 = 𝑎
⏞⎴⋅ 𝑎⎴⏞⎴
⋅ ⋯⎴⏞
⋅ 𝑎.
(2) If 𝑒 is the identity element of 𝐺, then we define 𝑎0 = 𝑒 for any 𝑎 ∈ 𝐺.
(3) For any 𝑎 ∈ 𝐺, we notate the inverse of 𝑎 as 𝑎−1 .
If we use the symbol + instead (so that 𝐺 is an abelian group), then we adopt the
corresponding additive notation:
𝑛 times
(1) For any positive integer 𝑛, we write 𝑛𝑎 = ⏞⎴⎴⎴⏞⎴⎴⎴⏞
𝑎 + 𝑎 + ⋯ + 𝑎.
(2) If 𝑒 is the identity element of 𝐺, then we define 0𝑎 = 𝑒.
(3) For any 𝑎 ∈ 𝐺, we notate the inverse of 𝑎 as −𝑎.
To practice this notation, let’s prove some facts about inverses.
Theorem 3.10. Let 𝐺 be a group. Then for all 𝑎, 𝑏 ∈ 𝐺, we have (𝑎−1 )−1 = 𝑎 and
(𝑎𝑏)−1 = 𝑏−1 𝑎−1 .
Please notice the importance of the order of the elements used in writing the in-
verse of 𝑎𝑏: it’s the inverse of 𝑏, followed by the inverse of 𝑎. Since the operation need
not be commutative, then we must be very careful when computing (𝑎𝑏)−1 .
Exercise 3.11. Prove that if 𝐺 is a group, then (𝑎𝑏)−1 = 𝑎−1 𝑏−1 for all 𝑎, 𝑏 ∈ 𝐺 if and
only if 𝐺 is abelian.
How well do you understand the definitions and notations so far? Try the next
theorem.
Theorem 3.12. Let 𝐺 be a group and 𝑎 ∈ 𝐺. Then for any nonnegative integer 𝑛, we
have (𝑎−1 )𝑛 = (𝑎𝑛 )−1 .
This suggests that, for any element 𝑎 in a group and any positive integer 𝑛, we
define 𝑎−𝑛 to be 𝑎−𝑛 = (𝑎−1 )𝑛 = (𝑎𝑛 )−1 . Likewise, if we use additive notation instead,
we define −𝑛𝑎 to be −𝑛𝑎 = 𝑛(−𝑎) = −(𝑛𝑎). When we do that, we finally have a nice
pair of summary theorems to prove.
20 Chapter 3. Groups
Theorem 3.13 (Laws of Exponents). Let 𝐺 be a group and 𝑎 ∈ 𝐺. Then for any integers
𝑛, 𝑚, we have 𝑎𝑛 𝑎𝑚 = 𝑎𝑛+𝑚 and (𝑎𝑚 )𝑛 = 𝑎𝑛𝑚 .
Corollary 3.14 (Laws of Coefficients). Let 𝐺 be an abelian group under addition, and
let 𝑎 ∈ 𝐺. Then for any integers 𝑛, 𝑚, we have 𝑛𝑎 + 𝑚𝑎 = (𝑛 + 𝑚)𝑎 and 𝑛(𝑚𝑎) = (𝑛𝑚)𝑎.
One word of warning is appropriate here, however. Keep in mind that although the
laws of exponents hold, we don’t have a distributive law of exponents, simply because
our group need not be abelian! Countless beginning algebra students have made the
mistake of assuming that (𝑎𝑏)𝑛 = 𝑎𝑛 𝑏𝑛 , just like they learned in school. Once again,
𝑛 times
𝑎𝑛 𝑏𝑛 = 𝑎
⏞⎴⋅ 𝑎⎴⏞⎴
⋅ ⋯⎴⏞ ⏞⎴
⋅ 𝑎⋅𝑏 ⋅ 𝑏⎴⏞⎴
⋅ ⋯⎴⏞
⋅ 𝑏. In general, those two expressions need not be equal,
so be careful and use only those laws you’ve proved.
Exercise 3.15. Prove that if 𝐺 is a group, then (𝑎𝑏)2 = 𝑎2 𝑏2 for all 𝑎, 𝑏 ∈ 𝐺 if and only
if 𝐺 is abelian.
Theorem 3.16. Let 𝐺 be a group. Then for every 𝑔 ∈ 𝐺, the functions 𝜆𝑔 ∶ 𝐺 → 𝐺 and
𝜌𝑔 ∶ 𝐺 → 𝐺 defined by 𝜆𝑔 (𝑥) = 𝑔𝑥 and 𝜌𝑔 (𝑥) = 𝑥𝑔 are bijections.
This simple theorem is the key to understanding the form that a group table must
have. Consider a particular element 𝑎 in a group 𝐺. The row headed by 𝑎 consists of all
elements of the form 𝑎𝑔, where 𝑔 ∈ 𝐺. However, the theorem says that each element
of 𝐺 occurs once and only once in that form! Thus, we conclude that every row must
have each element of 𝐺 occur once and only once. Likewise, each column must have
each element of 𝐺 occur once and only once.
Armed with this theorem, and recalling that we will always list the identity ele-
ment first in a table, we can begin to construct various tables of small groups. To aid
us, let’s introduce a convenient term dealing with the size of a group. You might find
a quick review of what cardinality means in Section A.3 of Appendix A useful.
Definition 3.17. Let 𝐺 be a group. Then the cardinality of 𝐺 is called the order of 𝐺,
written |𝐺|. In particular, when the cardinality of 𝐺 is finite, we say that 𝐺 is a finite
group and that 𝐺 has finite order. Likewise, if the cardinality of 𝐺 is not finite, we
say that 𝐺 is an infinite group and that 𝐺 has infinite order.
Exercise 3.18. Show that there is only one way (up to isomorphism) to create a table
for a group of order 1.
3.3. Group tables and the order of a group 21
Definition 3.19. Any group with exactly one element is called the trivial group.
Exercise 3.20. Show that there is only one way (up to isomorphism) to create a table
for a group of order 2.
Exercise 3.21. Show that there is only one way (up to isomorphism) to create a table
for a group of order 3.
What the above three exercises show is that there is a unique group of orders 1,
2, and 3, in the sense that any two groups of order 𝑘 are isomorphic, for 𝑘 = 1, 2, 3.
Thus, every group of order 1 is isomorphic to the trivial group; every group of order 2
is isomorphic to ℤ2 , and every group of order 3 is isomorphic to ℤ3 . Based on this, you
might think that there’s only one way to create a table for a group of any finite order, but
you’d be wrong. In general, the higher the order, the more options you have. Hence,
in order to describe most groups with large order, we need to develop some new ideas.
However, we can at least show that there are only two groups (up to isomorphism) of
order 4.
Exercise 3.22. In this exercise, you will create the only two groups of order 4 (up to
isomorphism). One of them must be isomorphic to ℤ4 , so the other will be a new
group, called the Klein four-group, denoted 𝑉. Use the set {𝑒, 𝑎, 𝑏, 𝑐}, with 𝑒 denoting
the identity element, to create two nonisomorphic group tables in the following way.
First, create the unique group table for which every element is its own inverse. Since
not every element in ℤ4 is its own inverse, this will be the Klein four-group 𝑉. Then
create a group table for which some element is not its own inverse. In so doing, explain
why any choices you make are either forced or arbitrary. This shows that every group of
order 4 is either isomorphic to ℤ4 or to 𝑉. Are either, both, or neither groups abelian?
Subgroups and generating sets
4
4.1 Subgroups
As we have seen in Chapter 3, examples of groups are quite varied: sets of numbers
or matrices under addition or multiplication, sets of functions under composition, and
even finite sets with a suitably constructed table are groups. However, for many of these
examples, it’s not the operation that’s different, but rather only the set that changes.
Example 4.1. Consider the groups 𝐺 = ℝ and 𝐻 = ℚ. Although they are both groups
in their own right, there’s a natural relationship between them: not only is ℚ a subset of
ℝ, but their operations are identical: addition of elements of ℚ is identical to addition
of those same elements in ℝ. When this special relationship of being both a subset and
agreement on the operation occurs, we tend to think not of two different groups, but
of one “large” group inside of which lies the “smaller” group as a subset.
There is a technical issue we must first raise. The binary operation of a group ⟨𝐺, ⋅⟩
is a function with domain 𝐺 × 𝐺. Given a subset 𝐻 ⊂ 𝐺, we can restrict the domain of
⋅ to 𝐻 × 𝐻 and see if this restricted operation – the “same” one used to define a valid
operation on 𝐺 – yields a group (or a binary operation, at least). Hence, we need a
preliminary definition to deal with this technicality.
Definition 4.2. Let ⟨𝐺, ⋅⟩ be a binary structure and let 𝐻 ⊂ 𝐺. Then the function
∗ ∶ 𝐻 × 𝐻 → 𝐺 given by ∗(𝑎, 𝑏) = 𝑎 ⋅ 𝑏 is called the operation induced by ⋅.
With this terminology, we can now define the key term precisely.
Given a group 𝐺, there are always two (not necessarily different) subgroups of 𝐺:
the group 𝐺 itself, and the subgroup consisting of the identity element alone. Since
23
24 Chapter 4. Subgroups and generating sets
we’re often interested in subgroups other than those two, let’s name them for future
reference.
Example 4.5. Before we turn to examples of subgroups of groups, let’s see what can
go wrong with some “non-examples.”
(1) The odd integers are not a subgroup of the integers ℤ, since the odd integers are
not closed under addition.
(2) The set of positive rational numbers is not a subgroup of ℚ, since the positive ra-
tionals lacks an additive identity element.
(3) The set of 𝑛×𝑛 invertible matrices with integer entries is not a subgroup of 𝐺𝐿𝑛 (ℝ),
since the inverse of many such matrices requires non-integer entries.
(4) The set ℤ𝑛 is not a subgroup of ℤ. It fails primarily because the operation on ℤ𝑛 is
not addition – it’s addition modulo 𝑛. Even if you use the induced addition opera-
tion on ℤ𝑛 instead, then the set fails to be closed under addition.
Example 4.6. As with groups before, the following list contains the notation of the
standard subgroups you’ll need to know as you proceed through the book.
(2) Although we considered ℤ, ℚ, ℝ, and ℂ as groups under addition in their own right,
there is a natural and obvious subgroup relationship ℤ < ℚ < ℝ < ℂ.
(3) The set of 𝑛 × 𝑛 matrices with integer entries with determinant 1 is a subgroup of
𝐺𝐿𝑛 (ℤ), called the special linear group and is denoted 𝑆𝐿𝑛 (ℤ) (or ℚ, ℝ, ℂ).
(4) When 𝐺 is any binary structure, the set of all automorphisms of 𝐺 is a subgroup of
𝑆 𝐺 and is denoted 𝐴𝑢𝑡(𝐺).
(6) The set of all complex numbers 𝑎 + 𝑏𝑖 that are distance 1 away from the origin (that
is, all complex numbers 𝑎 + 𝑏𝑖 such that 𝑎2 + 𝑏2 = 1) is a subgroup of ℂ∗ and is
simply denoted 𝑈 (see Section C.2 in Appendix C).
(7) For any positive integer 𝑛, the set of all complex numbers 𝑧 such that 𝑧𝑛 = 1 is a
subgroup of ℂ∗ and of 𝑈, and is denoted 𝑈𝑛 . This subgroup has 𝑛 elements, called
the 𝑛th roots of unity, and we notate those 𝑛 elements by 𝜁0 , 𝜁1 , . . . , 𝜁𝑛−1 (see Section
C.3 in Appendix C). Specifically, 𝜁𝑘 = cos(2𝜋𝑘/𝑛) + 𝑖 sin(2𝜋𝑘/𝑛) for 0 ≤ 𝑘 < 𝑛.
There are two things to note in this important theorem. First, you don’t have to
check to see if the induced operation is associative. After all, the operation is associa-
tive on all elements in 𝐺, and 𝐻 is a subset of 𝐺. Thus, it’s not necessary to check
associativity for 𝐻.
But the second issue is a bit more subtle. You do not check to see if there is an
identity element, nor do you need to check to see if each element of 𝐻 has an inverse.
Again, those two facts are given to you by the fact that 𝐺 is itself a group. What the
theorem says you must do is verify that the identity element – which already exists – is a
member of 𝐻, and that for each element 𝑎 ∈ 𝐻, its inverse element 𝑎−1 – which already
exists – is also a member of 𝐻. Hence, the theorem can be thought of as answering three
questions:
(1) Is 𝐻 closed?
(2) Is 𝑒 ∈ 𝐻?
(3) If 𝑎 ∈ 𝐻, is 𝑎−1 ∈ 𝐻?
If the answer is yes for all three questions, then the theorem allows you to conclude
that 𝐻 is a subgroup of 𝐺. Should the answer to even one of the three questions be no,
then the theorem allows you to conclude that 𝐻 is not a subgroup of 𝐺. This three-
pronged test is the easiest way to check if a subset 𝐻 is a subgroup of a group 𝐺.
Example 4.8. Let’s use this 3-pronged test in an example. Let 𝐴 be a set, and pick any
non-empty subset 𝐵 ⊂ 𝐴. Recall from Example 3.2 that 𝑆𝐴 is the group of bijections
from 𝐴 to itself. Let 𝐻 be those elements 𝜎 ∈ 𝑆𝐴 such that 𝜎(𝑥) = 𝑥 for all 𝑥 ∉ 𝐵. Is 𝐻
a subgroup of 𝑆𝐴 ?
We have three things to check:
(1) Is 𝐻 closed? To check this, we pick two elements 𝜎, 𝜏 ∈ 𝐻. This means we know
that 𝜎(𝑥) = 𝑥 and 𝜏(𝑥) = 𝑥 for all 𝑥 ∉ 𝐵. Our task is to determine if 𝜎 ∘ 𝜏 ∈ 𝐻. In
order to find that out, we ask: does (𝜎 ∘ 𝜏)(𝑥) = 𝑥 for all 𝑥 ∉ 𝐵? Let’s pick such an
element and find out. So, let 𝑥 ∉ 𝐵. Then (𝜎 ∘ 𝜏)(𝑥) = 𝜎(𝜏(𝑥)). But 𝜏(𝑥) = 𝑥, since
𝜏 ∈ 𝐻, and when we apply 𝜎 to 𝜏(𝑥) = 𝑥, we get 𝜎(𝜏(𝑥)) = 𝜎(𝑥) = 𝑥, since 𝜎 ∈ 𝐻.
Thus, 𝜎 ∘ 𝜏 ∈ 𝐻.
(2) Is the identity element of 𝑆𝐴 in 𝐻? Remember: we already know that the function
𝜄 ∶ 𝐴 → 𝐴 given by 𝜄(𝑥) = 𝑥 for all 𝑥 ∈ 𝐴 is the identity element of 𝑆𝐴 ; what we
don’t know is if 𝜄 ∈ 𝐻. Again, how do we check? We pick 𝑥 ∉ 𝐵 and compute:
𝜄(𝑥) = 𝑥, which is exactly what it means for 𝜄 ∈ 𝐻. Thus, 𝐻 contains the identity
element of 𝑆𝐴 .
(3) If 𝜎 ∈ 𝐻, is 𝜎−1 ∈ 𝐻? Remember: we already know that any 𝜎 ∈ 𝐻 has an inverse
𝜎−1 in 𝑆𝐴 ; what we don’t know is if 𝜎−1 is in 𝐻. Again, pick 𝑥 ∉ 𝐵 and ask: does
26 Chapter 4. Subgroups and generating sets
The reason for this rather lengthy example is because this is how we can actually
think of 𝑆𝐴 as a subgroup of 𝑆 𝐵 whenever 𝐴 ⊂ 𝐵. Since 𝑆𝐴 is the set of all bijections
of 𝐴 to itself, we can pick any 𝜎 ∈ 𝐴 and extend it to a bijection of all of 𝐵 be defining
𝜎(𝑏) = 𝑏 whenever 𝑏 ∈ (𝐵 − 𝐴). In other words, think of 𝜎 as moving the elements of
𝐴 around, and leaving everything else in 𝐵 alone.
Now you try out this test on the next theorems.
Exercise 4.11. It’s tempting to try to prove that the union of a collection of subgroups
is also a subgroup, but it’s rarely the case. Prove that the union of two subgroups is a
subgroup if and only if one is a subset of the other. (Interestingly enough, it is possible
to find three subgroups whose union is a subgroup, no two of which are subsets of the
other.)
Exercise 4.12. Determine if the following subsets 𝐻 are subgroups of the given group
𝐺 under the operation induced from 𝐺. Provide either a proof or a counterexample to
justify your answers.
(1) 𝐺 = ℚ; 𝐻 = the set of dyadic rationals.
(2) 𝐺 = ℝ; 𝐻 = {2𝛼 | 𝛼 ∈ ℚ}.
(3) 𝐺 = ℤ9 ; 𝐻 = {0, 2, 4, 6, 8}.
𝑎 𝑏 |
(4) 𝐺 = 𝐺𝐿2 (ℝ); 𝐻 = {[ ] | 𝑎, 𝑏, 𝑑 ∈ ℝ, 𝑎𝑑 ≠ 0}.
0 𝑑 |
(5) 𝐺 = 𝑆 ℤ ; 𝐻 = {𝑓 ∈ 𝑆 ℤ | 𝑓(ℤ+ ) ⊂ ℤ+ }.
(6) 𝐺 = ℂ∗ ; 𝐻 = {𝑎 + 𝑏𝑖 ∈ ℂ∗ | 𝑎, 𝑏 ∈ ℚ}.
Theorem 4.13. Let 𝐺 be a group. Then the subset 𝑍(𝐺) = {𝑔 ∈ 𝐺 | 𝑥𝑔 = 𝑔𝑥 for all 𝑥 ∈ 𝐺}
is a subgroup of 𝐺.
Another random document with
no related content on Scribd:
perhaps a people that can dance may dispense with them. They
impart a pensive, wayward pleasure to the mind, and are a kind of
chronology of happy events, often serious in the retrospect—births,
marriages, and so forth. Coleridge calls them ‘the poor man’s only
music.’ A village-spire in England peeping from its cluster of trees is
always associated in imagination with this cheerful accompaniment,
and may be expected to pour its joyous tidings on the gale. In
Catholic countries, you are stunned with the everlasting tolling of
bells to prayers or for the dead. In the Apennines, and other wild and
mountainous districts of Italy, the little chapel-bell with its simple
tinkling sound has a romantic and charming effect. The Monks in
former times appear to have taken a pride in the construction of bells
as well as churches; and some of those of the great cathedrals abroad
(as at Cologne and Rouen) may be fairly said to be hoarse with
counting the flight of ages. The chimes in Holland are a nuisance.
They dance in the hours and the quarters. They leave no respite to
the imagination. Before one set has done ringing in your ears,
another begins. You do not know whether the hours move or stand
still, go backwards or forwards, so fantastical and perplexing are
their accompaniments. Time is a more staid personage, and not so
full of gambols. It puts you in mind of a tune with variations, or of an
embroidered dress. Surely, nothing is more simple than time. His
march is straightforward; but we should have leisure allowed us to
look back upon the distance we have come, and not be counting his
steps every moment. Time in Holland is a foolish old fellow with all
the antics of a youth, who ‘goes to church in a coranto, and lights his
pipe in a cinque-pace.’ The chimes with us, on the contrary, as they
come in every three or four hours, are like stages in the journey of
the day. They give a fillip to the lazy, creeping hours, and relieve the
lassitude of country places. At noon, their desultory, trivial song is
diffused through the hamlet with the odour of rashers of bacon; at
the close of day they send the toil-worn sleepers to their beds. Their
discontinuance would be a great loss to the thinking or unthinking
public. Mr. Wordsworth has painted their effect on the mind when
he makes his friend Matthew, in a fit of inspired dotage,
‘Sing those witty rhymes
About the crazy old church-clock
And the bewilder’d chimes.’
The tolling of the bell for deaths and executions is a fearful
summons, though, as it announces, not the advance of time but the
approach of fate, it happily makes no part of our subject. Otherwise,
the ‘sound of the bell’ for Macheath’s execution in the ‘Beggar’s
Opera,’ or for that of the Conspirators in ‘Venice Preserved,’ with the
roll of the drum at a soldier’s funeral, and a digression to that of my
Uncle Toby, as it is so finely described by Sterne, would furnish
ample topics to descant upon. If I were a moralist, I might
disapprove the ringing in the new and ringing out the old year.
‘Why dance ye, mortals, o’er the grave of Time?’
St Paul’s bell tolls only for the death of our English kings, or a
distinguished personage or two, with long intervals between.[15]
Those who have no artificial means of ascertaining the progress of
time, are in general the most acute in discerning its immediate signs,
and are most retentive of individual dates. The mechanical aids to
knowledge are not sharpeners of the wits. The understanding of a
savage is a kind of natural almanac, and more true in its
prognostication of the future. In his mind’s eye he sees what has
happened or what is likely to happen to him, ‘as in a map the voyager
his course.’ Those who read the times and seasons in the aspect of
the heavens and the configurations of the stars, who count by moons
and know when the sun rises and sets, are by no means ignorant of
their own affairs or of the common concatenation of events. People
in such situations have not their faculties distracted by any
multiplicity of inquiries beyond what befalls themselves, and the
outward appearances that mark the change. There is, therefore, a
simplicity and clearness in the knowledge they possess, which often
puzzles the more learned. I am sometimes surprised at a shepherd-
boy by the road-side, who sees nothing but the earth and sky, asking
me the time of day—he ought to know so much better than any one
how far the sun is above the horizon. I suppose he wants to ask a
question of a passenger, or to see if he has a watch. Robinson Crusoe
lost his reckoning in the monotony of his life and that bewildering
dream of solitude, and was fain to have recourse to the notches in a
piece of wood. What a diary was his! And how time must have spread
its circuit round him, vast and pathless as the ocean!
For myself, I have never had a watch nor any other mode of
keeping time in my possession, nor ever wish to learn how time goes.
It is a sign I have had little to do, few avocations, few engagements.
When I am in a town, I can hear the clock; and when I am in the
country, I can listen to the silence. What I like best is to lie whole
mornings on a sunny bank on Salisbury Plain, without any object
before me, neither knowing nor caring how time passes, and thus
‘with light-winged toys of feathered Idleness’ to melt down hours to
moments. Perhaps some such thoughts as I have here set down float
before me like motes before my half-shut eyes, or some vivid image
of the past by forcible contrast rushes by me—‘Diana and her fawn,
and all the glories of the antique world;’ then I start away to prevent
the iron from entering my soul, and let fall some tears into that
stream of time which separates me farther and farther from all I once
loved! At length I rouse myself from my reverie, and home to dinner,
proud of killing time with thought, nay even without thinking.
Somewhat of this idle humour I inherit from my father, though he
had not the same freedom from ennui, for he was not a
metaphysician; and there were stops and vacant intervals in his
being which he did not know how to fill up. He used in these cases,
and as an obvious resource, carefully to wind up his watch at night,
and ‘with lack-lustre eye’ more than once in the course of the day
look to see what o’clock it was. Yet he had nothing else in his
character in common with the elder Mr. Shandy. Were I to attempt a
sketch of him, for my own or the reader’s satisfaction, it would be
after the following manner:——but now I recollect, I have done
something of the kind once before, and were I to resume the subject
here, some bat or owl of a critic, with spectacled gravity, might swear
I had stolen the whole of this Essay from myself—or (what is worse)
from him! So I had better let it go as it is.
WHY THE HEROES OF ROMANCE ARE
INSIPID