MATH225

Abstract Algebra
Course Notes∗
Jan E. Grabowski
Department of Mathematics and Statistics
Lancaster University

[email protected]
http://www.maths.lancs.ac.uk/~grabowsj

6th February 2019

∗
These are course notes, rather than lecture notes. That is, they contain material you
are expected to study that might not be explained in detail in lectures, as well as material
for further reading that is not examinable.

1
 From your previous mathematical studies, you will be aware of examples
of different types of symmetries. As well as geometric symmetries (rota-
tions, reflections, translations), you encountered permutations, which can
be thought of as symmetries of finite sets of objects. You have also studied
linear transformations, as structure-preserving functions of vector spaces -
these too are a type of symmetry.
The notion of a group is designed precisely to capture the common ele-
ments of all of these manifestations of symmetries. It provides an abstract
model, where we can free ourselves from thinking about unimportant inform-
ation that is specific to one particular example, and where we can prove very
general results that we can apply to lots of examples.
In this module, we will look at lots of different examples of groups.
We will begin to study abstract groups, looking at their internal structure
(subgroups) and how groups can be related by structure-preserving functions
(homomorphisms). We will prove many basic results about groups as well as
several more major theorems, which we will apply to our various examples.
You have also encountered several different generalizations of “number
systems” - the integers, the integers modulo n, polynomials and fields. Just
as groups are abstract models of symmetries, so rings are abstract models
for number systems. They have an addition and a multiplication, satisfying
most (but not all) of the familiar properties of the examples we have just
mentioned.
Ring theory also does the same job: stripping away unnecessary com-
plications, making it easier to prove general results that we can apply to
many examples. Again we will study the internal structure of rings, their
structure-preserving maps and some basic results and fundamental theor-
ems.
The examples and results studied here are used and developed further
throughout later pure mathematics modules in algebra, analysis, combinat-
orics and geometry.

2
Syllabus
This module will cover the following topics:

• Permutations, symmetries of two- and three-dimensional shapes and

matrices as examples of symmetries;
• Definition of a group, elementary properties;
• Subgroups, cyclic groups and order of elements; cosets; Lagrange’s
theorem and applications;
• Group homomorphisms and isomorphisms; kernel and image; Cayley’s
theorem;
• Normal subgroups; quotient groups; the First Isomorphism Theorem
for groups;
• Definition of a ring; examples, including polynomial rings; rings with
identity, commutative rings (including the definition of an integral
domain and a field); subrings; ring homomorphisms and isomorphisms;
• Ideals; the ideal structure of the ring of integers;
• Quotient rings; the First Isomorphism Theorem for rings.

3
Aims and Outcomes
The aims of this module are to

• To develop an appreciation of how groups and rings arise as abstract

models for studying symmetries and generalizations of number sys-
tems;

• To become familiar with a range of examples of groups and rings and

to be able to identify important properties of these;

• To study how different groups or rings can be related, through func-

tions between them that preserve the group or ring structure (that is,
homomorphisms);

• To demonstrate how the formal logical theories of groups and of rings

progress, by developing the basic results in these theories and by prov-
ing and applying several fundamental theorems.

On successful completion of the module students should be able to

• Appreciate that algebraic structures such as groups, rings and fields,

are useful as abstract models for studying many natural mathematical
questions, such as symmetries and properties of number systems;

• Give the definitions of a group, a ring and a field (including how these
are related) and give several examples of each of them;

• Understand that groups and group theory allow us to study symmet-

ries of different mathematical objects, and compute with particular
examples, including permutations, integers modulo n and matrices;

• Understand that rings and ring theory allow us to study generalisations

of number systems and compute with particular examples, including
integers, integers modulo n, matrices and polynomials;

• Prove basic results in group theory and ring theory and identify, recall
and apply significant theorems from those theories, including Lag-
range’s Theorem and the First Isomorphism Theorems.

4
Contents
1 Introduction to Abstract Algebra 7
1.1 Binary operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2 Algebraic structures 19

3 The internal structure of groups 23

3.1 Permutations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2 The sign of a permutation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.3 Elementary properties, or, algebraists like to be fussy . . . . . . . . . . . . . 34
3.4 Subgroups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.5 Groups generated by one element . . . . . . . . . . . . . . . . . . . . . . . . 44
3.6 Cosets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.7 Lagrange’s theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

4 Relating different groups 61

4.1 Homomorphisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.2 Kernels and images . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.3 Cayley’s theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.4 Groups of small order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.5 Normal subgroups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.6 Quotient groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.7 The First Isomorphism Theorem for Groups . . . . . . . . . . . . . . . . . . 92
4.8 Looking ahead (NOT EXAMINABLE) . . . . . . . . . . . . . . . . . . . . . 100

5 The internal structure of rings 111

5.1 The definition and some examples . . . . . . . . . . . . . . . . . . . . . . . 111
5.2 Polynomial rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
5.3 Basic properties of rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.4 Subrings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144

6 Relating different rings 152

6.1 Homomorphisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
6.2 Kernels and images . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
6.3 Ideals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
6.4 Ideals generated by one element . . . . . . . . . . . . . . . . . . . . . . . . . 174
6.5 Quotient rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
6.6 The First Isomorphism Theorem for Rings . . . . . . . . . . . . . . . . . . . 187

Index 193

5
References
[AF] Marlow Anderson and Todd Feil. A first course in abstract algebra.
CRC Press, Boca Raton, FL, third edition, 2015. ISBN 978-1-4822-
4552-3.
[Alu] Paolo Aluffi. Algebra: chapter 0, volume 104 of Graduate Studies in
Mathematics. American Mathematical Society, Providence, RI, 2009.
ISBN 978-0-8218-4781-7.
[Cam] Peter J. Cameron. Sets, logic and categories. Springer Undergraduate
Mathematics Series. Springer-Verlag London, Ltd., London, 1999.
ISBN 1-85233-056-2.
[Der] John Derbyshire. Unknown Quantity: A Real and Imaginary History
of Algebra. Atlantic Books, 2008. ISBN 978-1-84354-570-5.
[dS] Marcus du Sautoy. Finding Moonshine: A Mathematician’s Journey
through Symmetry. Fourth Estate, 2008. ISBN 978-0007214617.
[Neu] Peter M. Neumann. A Hundred Years of Finite Group Theory. The
Mathematical Gazette, 80(487):106–118, March 1996.
[NST] Peter M. Neumann, Gabrielle A. Stoy, and Edward C. Thompson.
Groups and geometry. Oxford Science Publications. Oxford Univer-
sity Press, New York, 1994. ISBN 0-19-853452-3; 0-19-853451-5.
[Ron] Mark Ronan. Symmetry and the monster. Oxford University Press,
Oxford, 2006. ISBN 978-0-19-280722-9; 0-19-280722-6.
[ST] Geoff Smith and Olga Tabachnikova. Topics in group theory. Springer
Undergraduate Mathematics Series. Springer-Verlag London Ltd.,
London, 2000. ISBN 1-85233-235-2.

Above, we list the books referred to in these notes and some others
covering the topics herein. There are many books on groups and rings
available in the library or to buy. The books above will provide an alternative
perspective to the recommended texts and you will find others not on this
list that are equally suitable. However, these notes contain all of the material
that you need for this course.

Acknowledgements
This module and its notes owe a great deal to all those who have lectured it
in previous forms, most recently as Groups and Rings. Some parts closely
follow previous iterations of the course notes or are extracted from them;
we thank colleagues for their permission to use these as well as for many
conversations that fed into the development of this module.

6
Lecture plan
1. Overview: what is abstract algebra? Binary operations (pp. 2-17)

2. Binary operations, continued. Algebraic structures (pp. 17-22)

3. Permutations (pp. 23-29)

4. Permutations: signs and An . Basic properties of groups (pp. 29-36)

5. Subgroups (pp. 36-40)

6. Subgroups, continued (pp. 40-44)

7. Groups generated by one element. Cosets (pp. 44-51)

8. Cosets, continued (pp. 51-57)

9. Lagrange’s theorem (pp. 57-60)

10. Homomorphisms (pp. 61-65)

11. Homomorphisms, continued. Kernels and images (pp. 65-70)

12. Kernels and images, continued. Cayley’s theorem (pp. 70-75)

13. Cayley’s theorem, continued. Groups of small order (pp. 75-81)

14. Normal subgroups. Quotient groups (pp. 81-87)

15. Quotient groups, continued (pp. 87-92)

16. The First Isomorphism Theorem (pp. 92-100)

1 Introduction to Abstract Algebra

abstract
adjective
1. existing in thought or as an idea but not having a physical
or concrete existence.
“abstract concepts such as love or beauty”
synonyms: theoretical, conceptual, notional, intellectual, meta-
physical, philosophical, academic;
antonyms: actual, concrete
• dealing with ideas rather than events.
• not based on a particular instance; theoretical.
“we have been discussing the problem in a very abstract
manner”
• (of a noun) denoting an idea, quality, or state rather
than a concrete object.
2. . . .
verb
1. consider something theoretically or separately from (some-
thing else).
2. extract or remove (something).
“applications to abstract more water from streams”
3. make a written summary of (an article or book).
synonyms: summarize, write a summary of, precis, abridge,
condense, compress, shorten, cut down, abbreviate, synop-
size;
algebra
noun
the part of mathematics in which letters and other general sym-
bols are used to represent numbers and quantities in formulae
and equations.
“courses in algebra, geometry, and Newtonian physics”
• a system of algebra based on given axioms.
Origin
late Middle English: from Italian, Spanish, and medieval Latin,
from Arabic al-jabr ‘the reunion of broken parts’, ‘bone-setting’,
from jabara ‘reunite, restore’. The original sense, ‘the surgical
treatment of fractures’, probably came via Spanish, in which
it survives; the mathematical sense comes from the title of a
book, ’ilm al-jabr wa’l-muqābala ‘the science of restoring what
is missing and equating like with like’, by the mathematician
al-Kwārizmı̄ (see algorithm).
¯
Oxford English Dictionary

7
The landscape
Before we embark on our journey into abstract algebra, let us try to get
a sense of the wider landscape. The following diagram over-simplifies, of
course, but it is a starting point.

Pure Applicable Applied

Pure

Algebra Geometry

Arithmetic Analysis

Here we see that the four major topics in pure mathematics all have overlap
with each other. Since we want to understand the place of algebra, let us
zoom in a little further.

8
 Pure

Algebra

Geometry
Groups

Rings
3

2
1

Arithmetic
Analysis

Now we also include two major areas within algebra: groups and rings.
According to this point of view, much of abstract algebra is concerned with
groups, and rings are a particular type of group. This last statement is
technically correct but rather misleading: we will see why, as the module
develops.
Before we move on, we highlight some of the topics where algebra has
particularly important intersection with the other major areas; these are

(1) Number theory: primes modulo n, Fermat, Goldbach, . . .

(2) Cryptography

(3) Algebraic geometry: curves, surfaces, . . .

What is algebra, and how does it work

This module introduces you to a new area of mathematics, algebra, but it
is also trying to do something else. We are encountering a different way of

9
doing mathematics, a place where the sorts of questions we ask are different,
and unsurprisingly the answers are different too. It will help you to recognise
that the newness is on these two levels: it is not just the objects in question
that are new, but the way we approach we are doing is too, and that can
cause you to feel disoriented.
The formal definitions will come soon enough, but first let’s take a bit
of time to try to appreciate what it is that abstract algebra was invented to
study. We will try to make as many links to mathematics you already know
as we can, showing how to build on that, but we will also at various points
look forwards to what the theory we begin to construct in this module can
eventually do, though details will have to be left to future modules.
First and foremost, abstract1 algebra is about mathematical objects,
defined by axioms, and their relationships. The principal examples are
groups, rings, fields and vector spaces, though there are others.
Fundamentally, groups come from considering symmetries. Think of a
square drawn in the plane. What are its symmetries? We would natur-
ally think of the various rotations and reflections, so there is some set of
symmetries. This set includes the (seemingly!) uninteresting symmetry of
“doing nothing”, the identity symmetry. It also, very importantly, has the
property that if we take two symmetries and do one after the other, we get
another symmetry: this puts an operation (of composition, in this case) on
the set. We can also undo each symmetry, and these properties (plus one
other slightly less obvious one) give us the properties of what we now call
groups. Many of the other examples of groups we will meet soon arise in
a similar way, as symmetries of something, but we will make an abstract
definition of a group as a set with an operation satisfying some properties.
Rings are a little more complicated but equally have their origins in
natural and basic mathematical questions. Rings are designed to be a gen-
eralisation of the integers and give us new and different types of “number
systems”. So rings are sets with two operations, satisfying axioms corres-
ponding to natural properties of addition and multiplication. If we just look
at the addition and forget the multiplication, we have a group. So from this
perspective, rings are special sorts of groups.
Fields are special sorts of rings: they are rings—so have addition and
multiplication—with the extra property that they also have division. The
prototype example here is the rationals, where every (non-zero) rational a
has a multiplicative inverse 1/a. Just as subtraction is the inverse operation
for addition, rather than a completely new operation, so fields are sets with
two operations, addition and multiplication, but where the multiplication
1
You might be wondering when you missed out on the “Concrete Algebra” module.
Well, you didn’t, and this is one of those occasions where the adjective—abstract, in
this case—is being used for emphasis, rather than as a qualifier. That is, in modern
mathematics, essentially all algebra is abstract, but we say “abstract algebra” to remind
us that we’re approaching algebra in a particular way.

10
has the extra special property of admitting multiplicative inverses. The reals
and the complex numbers are fields but both of these have extra structure
that is better understood in the context of analysis; from an algebraic point
of view, we consider them as (slightly special) fields.
Vector spaces are also defined in a similar way: they generalise R2 , R3
and so on, and are also sets with two operations. One of these is addition,
so that like rings and fields, vector spaces have a group structure, but what
makes them different is their second operation. Vector spaces do not have
a multiplication like a ring—we can’t multiply a vector by a vector like we
would integers or real numbers—but instead they have scalar multiplication,
where an element of the base field and a vector can be multiplied.
So this is one way to understand the different parts of algebra, namely
group theory, ring theory, linear algebra and so on. But there is another
way to view things, namely by the types of question one asks.
Abstract algebra has several major themes in these terms:

classification: listing or parameterising all of the objects of a particular

type, especially by finding good invariants;

structure theory: looking at how we can decompose an object into simpler

parts, to help understand it, or how to take two objects and combine
them to make another; and

representation theory: many groups can be realised as symmetries in lots

of different ways.

In fact, you are already aware of representation theory without knowing

it: linear algebra is the study of the representation theory of fields. Scalar
multiplication tells us how a field acts on vectors, and one can show that if
a field acts by scalar multiplication in a reasonable way on some set with an
addition operation, then this is necessarily a vector space.
The representation theory of groups has several different aspects too:
some groups naturally act on sets, some on geometric objects and some on
vector spaces (think of the general linear group acting as invertible linear
transformations).
The representation theory of rings is not something you would have come
across so naturally, but it is also an important area of algebra and has very
significant links to geometry.

11
 Abstract algebra also has its own style among the various areas of pure
mathematics. Here are some words commonly associated with algebra:

By the end of this module, hopefully you will appreciate why some of
these words are a good description of algebra, but also hopefully you’ll
disagree with one of them!
The final remarks to be made before we begin on formal definitions are
about how algebra works. The first main theme is “proceeding from the
specific to the general”. That is, take some (relatively) concrete mathemat-
ical situation, such as the symmetries of a square or the arithmetic of the
integers, and try to abstract 2 the situation to give a definition that captures
just the essentials. Then we can study objects that satisfy this definition, in
general, and see what we can learn. Often it is helpful to strip away some of
the clutter of a specific case and by seeing things abstractly we get a better
sense of the essential nature of the question.
The other major theme is “new from old”, taking one or two objects and
producing more related ones. This might be “sub” objects, smaller ones
found inside a given object, or ways to combine two objects to give a larger
one.
In trying to make sense of what will be introduced over the next few
weeks, come back to this introduction and try to see how what we are doing
fits with this philosophy.

2
In the sense of the definition of “abstract” as a verb given at the start of this section.

12
1.1 Binary operations
The starting point for defining any of the algebraic objects referred to above
is the notion of a binary operation. “Binary” means “two”, and “operation”
indicates that we take (two) elements and operate on them, producing an-
other. Formally, however, binary operations are just functions, so we remind
ourselves of the definition of a function first. Before we do, recall that given
sets X and Y , we may form the Cartesian product X × Y of ordered pairs
{(x, y) | x ∈ X, y ∈ Y }.

Definition 1.1. Let X and Y be non-empty sets. A function f : X → Y

is a subset Γ ⊆ X × Y such that for all x ∈ X there exists a unique y ∈ Y
such that (x, y) ∈ Γ.

We write f (x) for the unique element of Y such that (x, f (x)) ∈ Γ and
call f (x) the image of x under f . The set X is called the domain of f , the
set Y is called the codomain of f .
Then the definition of a binary operation is simply the following.

Definition 1.2. Let S be a non-empty set. A binary operation on S is a

function ∗ : S × S → S.

In a similar fashion, we can define a unary operation as a function ∗ : S → S,

a ternary operation as a function ∗ : S × S × S → S and so on. Binary
operations are by far the most commonly encountered, however.
We emphasise that a binary operation being, by definition, a function
has several consequences. Firstly, the binary operation does not exist inde-
pendent of S: the domain and codomain of a function are an integral part
of the data of the function (though later we will abuse notation rather a lot,
which will muddy the waters somewhat).
Secondly, it is required to be well-defined as a single-valued function
(recall from MATH112 Discrete Mathematics that we do not allow multi-
valued functions), so given a, b ∈ S, there exists a unique c ∈ S such that
∗(a, b) = c. That is, given two elements of S, they combine via ∗ in a unique
way to give another element of S.
We usually write a ∗ b for ∗(a, b), as you will see in the examples shortly.
The other important consequence of a binary operation being a function
is that it is necessarily closed. That is, given a, b ∈ S, ∗(a, b) = a ∗ b is
also an element of S. Right now, this might look unimportant but closure
will become relevant as an idea later, when we talk about subobjects. Then
we might have R ⊆ S and an operation ∗ on S and we want to know if
∗ restricts to a binary operation on R (that is, if ∗|R×R : R × R → S has
Im ∗ ⊆ R, so that we obtain a well-defined function ∗R : R × R → R). This
will become clearer in later examples.

13
Examples 1.3.

(a) Let S = N and ∗ : N × N → N, ∗(a, b) = a + b.

This is a function: given a, b ∈ N, there exists a unique c ∈ N such
that c = a + b; the sum of positive integers is again a positive integer.

(b) Let S = N and ∗ : N × N → N, ∗(a, b) = a × b.

Again, since the product of positive integers is a positive integer, this
is a binary operation.

(c) Let S = N and ∗ : N × N → N, ∗(a, b) = a − b.

This is not a binary operation: the pair (1, 2) would have ∗(1, 2) = −1
but −1 ∈ / N, so ∗ does not define a function ((1, 2) has no image). This
is an example where we do not have closure: not all pairs (a, b) have
∗(a, b) in S.

(d) Let S = Z and ∗ : Z × Z → Z, ∗(a, b) = a − b.

This is now a binary operation: the difference of two integers is again
an integer.

You can surely think of many more binary operations, probably mostly
using addition or multiplication. However binary operations can take any
set S as input, not necessarily a set of “numbers”. For example, composition
of functions can give rise to a binary operation: given a non-empty set X
and function f : X → X and g : X → X, their composition g ◦ f : X → X
is again a function. This gives us a binary operation on F(X, X), the set of
functions with domain and codomain X. (Note that if Y �= X, composition
is not a binary operation on F(X, Y ), since we cannot compose two functions
f : X → Y and g : X → Y .)
As a particular example of this, think of the set of rotations and reflec-
tions of a square. More precisely, let S be the unit square centered at the
origin in R2 (thought of as a subset of R2 ) and let Sym(S) be the set of
invertible linear transformations of R2 preserving S (as a set). Composition
is also a binary operation on Sym(S).
In the following diagram, we illustrate the unit square (top left), the
square after applying a rotation clockwise by π about its centre (top right),
after applying a reflection in the line y = x (bottom left) and after first
applying the rotation and then the reflection:

14
 a b c d

d c b a

c b a d

a c
d b

Plain binary operations are not usually very interesting. They become more
so when they have particular properties, such as those that addition or mul-
tiplication of integers satisfy. The most important of these are the following;
from this point onwards, we will use the notation a ∗ b rather than ∗(a, b).
Definition 1.4. Let S be a non-empty set and ∗ : S × S → S be a binary
operation on S. We say ∗ is associative if

(a ∗ b) ∗ c = a ∗ (b ∗ c)

for all a, b, c ∈ S.
This property should be familiar, even if the name is not. Many of
the binary operations you would think of, including all those above, are
associative. There are interesting non-associative binary operations3 but we
won’t encounter them in this course.
There are several different ways to think about associativity, from the
very formal (there are two natural ways to build ternary operations on S
3
The operation [ , ] : Mn (C) × Mn (C) → Mn (C), [A, B] = AB − BA is a very important
example.

15
from ∗ and these two functions should be equal) to the more conceptual
(“we don’t need brackets”).
It is very easy to overlook the extremely important last few words of this
definition: the equation (a ∗ b) ∗ c = a ∗ (b ∗ c) is required to hold for all
a, b, c ∈ S. The operation ∗ is not associative if this equation fails for even a
single triple a, b, c. Hence, to show that an operation is not associative, we
simply need to find a, b, c ∈ S such that the equation fails; we will reiterate
this several times below, but it is crucial that we do so with an explicit
choice of a, b and c.
Class Exercise 1.5. Which of these binary operations is not associative?
(a) S = Z and a ∗ b = a + b;
(b) S = Z and a ∗ b = ab;
(c) S = Z and a ∗ b = a − b.
Find integers a, b, c that show this.
Answer. The answer is (c), since 1 − (2 − 3) = 1 − (−1) = 2 but (1 − 2) − 3 =
−1 − 3 = −4.

Remark 1.6. By an induction argument, one may show that associativity as

defined above implies “all higher associative laws”.
That is, (a ∗ b) ∗ c = a ∗ (b ∗ c) implies
((a ∗ b) ∗ c) ∗ d) = (a ∗ b) ∗ (c ∗ d) = a ∗ (b ∗ c) ∗ d = a ∗ (b ∗ (c ∗ d))
and similarly for expressions involving five or more elements. The effect of
this is that we may insert or remove brackets in these types of expressions
r times
� �� �
as we wish. Also, this tells us that writing a ∗ a ∗ · · · ∗ a makes sense, and
we can call the (unique) element obtained in this way the rth power (or
sometimes, multiple) of a and write ar (or ra) for any natural number r.
The next property of binary operations is also quite natural from the
point of view of addition or multiplication of integers, or for that matter,
symmetries. In the integers we have the number 0 that does not change any
a ∈ Z when we add a and 0. With respect to multiplication, the number
1 does the same thing: 1a = a. In Sym(S), we have the identity linear
transformation. The formal abstract definition is the following.
Definition 1.7. Let S be a non-empty set and let ∗ : S × S → S be a binary
operation on S. We say that an element e ∈ S is an identity for ∗ if
e∗a=a=a∗e
for all a ∈ S. If such an element e exists, we say that ∗ has an identity.

16
 Two important things to notice here are firstly that again we insist that
this equation holds for all a ∈ S, and secondly that both e ∗ a and a ∗ e are
required to be equal to a. There are examples where we might have e ∗ a = a
for all a but where a ∗ e �= a, though it is fair to say that you have to go
looking quite hard in order to find them.
For some binary operations, knowing a “left-handed” statement implies
a “right-handed” one, and vice versa. This can happen if the operation is
rather special4 , in that the order we combine the elements does not matter.
Definition 1.8. Let S be a non-empty set and let ∗ : S × S → S be a binary
operation on S. We say that ∗ is commutative if

a∗b=b∗a

for all a, b ∈ S.
For example, addition and multiplication of integers are both commut-
ative operations but composition of functions is in general not.
Lastly, we need one extra property, but this is of a slightly different
nature. It is a property that elements of S might (or might not) have with
respect to a binary operation on S if that operation has an identity.
Definition 1.9. Let S be a non-empty set and ∗ : S × S → S a binary
operation on S. Assume that there exists e ∈ S such that e is an identity
for ∗. We say that an element a ∈ S has an inverse with respect to ∗ if there
exists b ∈ S such that
a ∗ b = e = b ∗ a.
The prototype example here is the integers: given a ∈ Z, there exists
−a ∈ Z such that a + (−a) = 0 = (−a) + a. In fact, every element of Z has
an inverse with respect to addition, making the integers quite special (hold
this thought for later!).
However this need not always be the case: for example, with respect to
multiplication on the integers, only 1 and −1 have inverses (1 · 1 = 1 = 1 · 1
and (−1)·(−1) = 1 = (−1)·(−1), and in fact these elements are self-inverse).
The number 2 has no inverse in Z because 12 ∈ / Z.
As another (non-)example, most functions are not invertible5 . However,
invertible functions have inverses(!), so by definition the elements of Sym(S)
have inverses.
4
Here your intuition from your previous experience will lead you badly astray: you
might think what follows is typical behaviour, but this is very far from true. And this is
just as well, or else the lecturer’s research area would be even less interesting than most
people already think it is.
5
Recall from MATH112 Discrete Mathematics that a function f : X → X is invertible
if there exists g : X → X such that g ◦ f = IX = f ◦ g. Here IX is the identity function on
X, which is the identity for composition. For composition to be a binary operation, we
need f : X → X rather than f : X → Y , which is why this statement is a slightly special
case of the general definition we gave in MATH112.

17
 You might possibly be wondering about whether a binary operation can
have more than one identity, or if an element can have more than one inverse.
If you are, then congratulations, you are well on the way to becoming an
algebraist. If you are not, then now is a good time to start adopting this
type of healthy questioning when you meet a new definition. Algebraists are
by training sceptical when they see a definition: is the concept well-defined?
is there uniqueness? and so on.
You already know the answer to the second question, at least in the case
of functions (we proved that the inverse of a function is unique if it exists
in MATH112 Discrete Mathematics). It should not therefore be a surprise
that we can prove the following proposition.

Proposition 1.10. Let S be a non-empty set and let ∗ : S × S → S be a

binary operation.

(i) Let e1 and e2 be identities for ∗. Then e1 = e2 .

(ii) Assume that ∗ is associative and has an identity e. Assume that a ∈ S

has two inverses b1 , b2 ∈ S with respect to ∗. Then b1 = b2 .
Proof.

(i) Since e1 is an identity, we have e1 ∗ e2 = e2 . Similarly, e1 ∗ e2 = e1

since e2 is an identity. Hence e1 = e2 .

(ii) We have

b1 = b1 ∗ e
= b1 ∗ (a ∗ b2 )
= (b1 ∗ a) ∗ b2
= e ∗ b2
= b2 .

That is, subject to the assumptions given, both identity elements and
inverses are unique if they exist. The assumptions are necessary, as shown
by the following exercise.
Exercise 1.11. Let S = {a, b, e} and define a binary operation ∗ by the
following table (known as a Cayley table):

∗ e a b
e e a b
a a e e
b b e e

18
 Prove the following:
(a) ∗ has an identity element;
(b) ∗ is not associative;
(c) there exists an element of S having two different inverses.
Notice that we can prove the proposition in complete generality, for
binary operations on arbitrary sets. We now never need to prove either
of these again for particular examples: why would we bother? As we said
before, this idea—proving results in the most general setting possible—is
the “abstract” in Abstract Algebra.

2 Algebraic structures
An algebraic structure is just a set with some operations. We might ask that
the operations have some of the properties discussed above, or that different
operations interact in a particular way. In principle there are many different
types of algebraic structure but in practice some have much richer theories
than others and more (interesting) examples “in nature”. In this section,
we will introduce the main ones we will study in this course.
An algebraic structure consisting of a set S and n operations will be
denoted by (S, ∗1 , ∗2 , . . . , ∗n ). Sometimes we will use different letters for
the set S, to help us remember which type of algebraic structure we are
thinking about, and we might use established or more suggestive notation
for the operations, such as + or ×.
Definition 2.1. A group is an algebraic structure (G, ∗) with ∗ an associ-
ative binary operation, having an identity and such that every element of G
has an inverse with respect to ∗.

Definition 2.2. A group (G, ∗) is called Abelian if ∗ is commutative.

Definition 2.3. A ring is an algebraic structure (R, +, ×) such that (R, +)

is an Abelian group, × is associative and we have

a × (b + c) = (a × b) + (a × c) and (a + b) × c = (a × c) + (b × c)

for all a, b, c ∈ R. The identity element for the group (R, +) will be denoted
0R .
The two equations given here are called the first and second (or left and
right) distributive laws. We could be more general and say that × distributes
over + from the left if a × (b + c) = (a × b) + (a × c), and correspondingly
for the right-handed version. Notice that these are not symmetric in + and
×.

19
Definition 2.4. A field is an algebraic structure (F, +, ×) such that (F, +, ×)
is a ring, × is commutative and has an identity 1F �= 0F and every element
of F \ {0F } has an inverse with respect to ×.

Notice that the definition of a field (F, +, ×) could also be written

equivalently as saying that (F, +) is an Abelian group with identity 0F ,
(F \ {0F }, ×) is an Abelian group and that + and × satisfy the distributive
laws.
We see from these definitions that rings are groups with extra structure,
and fields are rings with extra properties. So every field is in particular a
ring, and every ring is a group. This is why it is sensible to study groups first,
then rings and then fields. Results we prove about groups will automatically
apply to rings and fields, but since rings are special sorts of groups, more
may be true of them.
Groups, rings and fields appear in various different contexts and the
following list of examples is not in any way comprehensive. Rather, they
are examples that should be somewhat familiar to you and these examples
will largely be the ones we will use to illustrate the theory that will follow.

Number systems
• Q, R and C are fields.

• Z is a ring (but not a field); in many ways, Z is the prototype ring and
many ring-theoretic questions are inspired by the structure of Z.

• For each n, the set of integers modulo n, Zn (with addition and mul-
tiplication modulo n) is a ring.

• The set of polynomials in n variables with coefficients from a ring R,

R[x1 , . . . , xn ], is again a ring, with respect to the usual addition and
multiplication of polynomials. This includes examples such as R[x],
C[x, y] and so on.

• Since Q, R and C are fields, as we said in the remarks after the defin-
ition of a field, we have associated groups Q \ {0}, R \ {0} and C \ {0}
with respect to multiplication.
More generally, the set of invertible elements in a ring R forms a group
under multiplication (called the group of units, R∗ ). For a field F , we
have F ∗ = F \ {0F }, because every non-zero element is invertible.

Symmetry groups
• Natural examples of symmetry groups are the set of rotations and
reflections of a square, an n-gon, a circle or other 2-dimensional figures,
as well as rotations and reflections of a cube, tetrahedron, etc.

20
 • The “symmetries of a set” S are given by the set of bijections
def
Bij(S) = {f : S → S | f is a bijection}. We saw in MATH112 Discrete
Mathematics that this set satisfies all the conditions to be a group with
respect to composition of functions.
We also went on to discuss permutations, which are re-orderings of a
set of n objects: a little thought and one realises that permutations
are exactly bijections from the set of objects to itself. These groups
are called the symmetric groups.
They (and the symmetry groups of geometric objects) are usually not
Abelian, in contrast to all the previous examples.

Matrices
• We can add m×n matrices and multiply n×n matrices, we know what
the zero and identity matrices are and we know that sometimes we can
invert matrices. So it should not be a surprise that sets of matrices
can have either group or ring structures, though they are almost never
fields.
More precisely, Mn (Z), Mn (Q), Mn (R) and Mn (C) are all rings; in fact
Mn (R) is one for any ring R. We call these matrix rings. (In linear
algebra, we usually work over a field, because we want to be able to
divide by scalars as well as multiply by them.)

• Inside any matrix ring over a field F , the subset of invertible matrices
(those with non-zero determinant) forms a group, the group of units,
as defined above. We have a special name and notation for the group
of units of matrix rings: we call them general linear groups and write
GLn (F ). The group operation is matrix multiplication.

Sets of functions
• The set of functions from R to R, denoted F(R, R), is a ring with
respect to pointwise addition and multiplication:

(f + g)(x) = f (x) + g(x) and (f g)(x) = f (x)g(x)

(this will be explained in more detail later).

• We can generalise this to functions from any set Y to R (denoted

F(Y, R)) or from any set Y to C (denoted F(Y, C)) or indeed from
any set Y to any ring R (denoted F(Y, R)). Each of these is a ring,
essentially because the codomain (R, C or R) is.

This list might give the (erroneous) impression that there are lot more
rings than groups. This is not the case: since every ring is a group and not

21
every group is a ring, there are definitely more groups than rings. It just so
happens that the examples you are familiar with are mostly rings or fields.
One important observation about the above list is that lots of these ex-
amples, notably the symmetric groups, the matrix rings and general linear
groups, are naturally symmetries of something. The description of sym-
metry groups we gave says this explicitly. One of the main results in Lin-
ear Algebra is that matrices are the same thing as linear transformations
of vector spaces, and hence GLn (F ) is the symmetry group of F n , the n-
dimensional vector space over F . This idea leads to a major area of study
in algebra, called representation theory.
Having mentioned vector spaces, let us briefly say where they fit in terms
of the definitions so far. Every vector space V over a field F is an Abelian
group under addition, (V, +). Vector spaces are not rings (or fields): we do
not multiply vectors by vectors.
However vector spaces have more structure than just addition, namely
they have scalar multiplication. Scalar multiplication is a slightly different
type of operation: it is not a binary operation on V . It takes two inputs, a
scalar λ ∈ F and a vector v ∈ V , and produces a new vector λv ∈ V . We
can formalise this as a function · : F × V → V , λ · v = λv, and say that a
vector space is an algebraic structure (V, +, ·) where + and · satisfy some
compatibility conditions, namely those that give linearity.
This concludes our overview of abstract algebra and we will now concen-
trate on groups and rings in more detail, beginning with groups.

22
3 The internal structure of groups
In this chapter, we will look at the internal structure of groups. That is, we
will have a group—either a specific example or just “some group G”—and
we will ask about properties it has or that its elements or its subsets have.

3.1 Permutations
Firstly, however, we will look in more detail at the symmetric groups, as we
will use these as one of our main classes of examples to illustrate the theory
that follows.
One justification for this is that every finite group can be found inside
a symmetric group. This is called Cayley’s theorem and we will state and
prove it later (Theorem 4.28).

Definition 3.1. A group (G, ∗) is called finite if the set G is finite.

So a group is finite if it has finitely many elements: examples are the

integers modulo 3 and the symmetries of a square. Unsurprisingly, groups
that are not finite are called infinite groups; examples are (Z, +) and GL3 (C)
(with matrix multiplication).

Definition 3.2. The order of a group (G, ∗) is the cardinality of G, |G|.

Recall that the set of bijections from a set S to itself is a group with
respect to composition of functions, (Bij(S), ◦). Let us fix S = {1, 2, . . . , n}
for some n ∈ N.

Definition 3.3. The symmetric group of degree n is defined to be the group

of bijections (Bij({1, 2, . . . , n}), ◦) and is denoted Sn .

In MATH112 Discrete Mathematics, we showed that Sn has n! elements,

that is, the symmetric group of degree n has order |Sn | = n!. We call the
elements of the symmetric group permutations.
The group S1 only has one element, which must be the identity element,
and must be Abelian. The group S2 has two elements and is also Abelian.
However, one very important fact is that Sn is not Abelian if n ≥ 3:
thinking about three objects, first swapping those in positions 1 and 2, then
swapping those in positions 2 and 3 does not give the same arrangement as
swapping the second and third, followed by swapping the first and second.
As the above paragraph illustrates, we need some notation to describe
permutations. Such wordy descriptions are not going to be any good in
general! So we will introduce several notations for permutations, each more
compact than the last, until we arrive at “disjoint cycle notation”, which is
both compact and very useful.

23
 The first approach is called “two-line notation”. Remember that any
function π : {1, 2, . . . , n} → {1, 2, . . . , n} is completely described
� by a list of�
1 2 3 ··· n
its values, π(i) for i ∈ {1, 2, . . . , n}. We can write this as π(1) π(2) π(3) ··· π(n) .
For example, the bijection π sending 1 to 2 (so π(1) = 2), 2 to 3 and 3
to 1 would be written as ( 12 23 31 ).
This helps us see why |Sn | = n!, since there are n choices for π(1), n − 1
for π(2) (π is a bijection so we cannot have π(2) = π(1)) and so on, down
to only one choice left for π(n). That is, n! choices in total.
In fact, the order of the columns does not matter: ( 12 23 31 ) and ( 31 23 12 )
hold the same information.
We can see that ι = ( 11 22 ··· n
··· n ) is the two-line notation for the identity
permutation, which is the identity bijection and so is the identity element
for the composition of permutations.
We can express the composition of permutations as
� a1 a2 ··· an � � 1 2 ··· n � � 1 2 ··· n �
b1 b2 ··· bn ◦ a1 a2 ··· an = b1 b2 ··· bn .

Notice that the order here is in keeping with our convention for composing
functions, writing g ◦ f for first f and then g. So we do the right-hand most
permutation first, and work right to left.
The example in the above paragraph explaining that most symmetric
groups are not Abelian can be written in two-line notation as the results of
the compositions

( 23 11 32 ) ◦ ( 12 21 33 ) = ( 13 21 32 )
and

( 12 33 21 ) ◦ ( 11 23 32 ) = ( 12 23 31 )
not being equal. � �
Finally, the inverse of a11 a22 ··· ··· n
an is ( a11 a22 ··· an
··· n ). For example, the
inverse of ( 12 23 31 ) is ( 21 32 13 ) = ( 13 21 32 ).
However there is a lot of redundancy in two-line notation, and the rule
for composition can be awkward to apply. We will try to simplify things by
using cycle notation; next, we define what we mean by a cycle.

Definition 3.4. Let r ≤ n and let a1 , a2 , . . . , ar be a list of distinct elements

of {1, 2, . . . , n}. Let π ∈ Sn be the permutation defined by

π(ai ) = ai+1 for 1 ≤ i ≤ r − 1

π(ar ) = a1
π(a) = a for a ∈
/ {a1 , a2 , . . . , ar }

We will call π a cycle of length r, or an r-cycle, and we will write (a1 a2 · · · ar ).

24
For example, the permutation that in two-line notation is ( 12 23 31 ) is a 3-
cycle, namely (2 3 1). Similarly, the permutation ( 11 24 35 43 52 ) is a 4-cycle,
namely (2 4 3 5); notice that the numbers in a cycle do not have to come in
increasing order and that 1 is fixed so does not appear in this cycle.
Remark 3.5. We note the following properties of cycles:

(a) An r-cycle can be written in r different ways, since

(a1 a2 · · · ar ) = (a2 a3 · · · ar a1 ) = · · · = (ar a1 · · · ar−1 ).

We usually write cycles with the smallest element of {a1 , a2 , . . . , ar }

first.

(b) The identity permutation ι is represented by any 1-cycle; this looks a

bit odd but the seeming ambiguity will be resolved shortly (by disjoint
cycle notation).

(c) We can compose two cycles by ‘feeding in’ elements from the right
in turn; for example, if π = (1 2 5) and σ = (2 3 4) then π ◦ σ =
(1 2 5)(2 3 4) maps

1 �→ 1 �→ 2, 2 �→ 3 �→ 3, 3 �→ 4 �→ 4, 4 �→ 2 �→ 5, 5 �→ 5 �→ 1.
We often say this as “1 goes to 1, 1 goes to 2; 2 goes to 3, 3 goes to
3” and so on. That is, we start with 1 and write “(1 ”. Then we see
where the first and then the second permutation send it, and write
the partial permutation “(1 2 ”; next it would be “(1 2 3 ” and so on,
until we reach an element that is sent to 1, at which point we close
the cycle. So π ◦ σ = (1 2 3 4 5).

(d) The inverse of a cycle is obtained by simply writing its elements in

reverse order; for example if π is as above then π −1 = (5 2 1) = (1 5 2).
This makes sense because if π sends 2 to 5, then π −1 should send 5 to
2, and so on. (We can also check this by computing (1 2 5)(5 2 1).)

(e) If we compose an r-cycle with itself r times, the result will be the
identity permutation. So every cycle has the property that some power
of it is the identity; we will formalize this idea later.

(f) If π = (a1 a2 · · · ar ) and σ = (b1 b2 · · · bs ) and the sets {ai | 1 ≤ i ≤

r} and {bj | 1 ≤ j ≤ s} are disjoint, then π ◦ σ = σ ◦ π. We can see
this because the elements moved by π and σ are different, so it does
not matter whether π or σ is performed first.
As an example, π = (1 2 3) and σ = (4 5 6 7) are permutations in S8 ,
are disjoint cycles and

25
 π ◦ σ = (1 2 3)(4 5 6 7)
= ( 12 23 31 45 56 67 74 88 )
= (4 5 6 7)(1 2 3) = σ ◦ π
Now not all permutations are cycles. The permutation ( 12 23 31 45 56 67 74 88 ) we
just obtained is not a cycle, for example. However, it is a product of two
cycles which moved disjoint sets of points; this suggests how we may extend
the idea of cycles to cover all permutations.
Definition 3.6. Let π = (a1 a2 · · · ar ) and σ = (b1 b2 · · · bs ) be cycles in
Sn . If the sets {a1 , a2 , . . . , ar } and {b1 , b2 , . . . , bs } are disjoint, we say that π
and σ are disjoint cycles. This terminology is extended in the obvious way
to sets of more than two cycles.
For the permutation α = ( 12 23 31 45 56 67 74 88 ) above, it is easy to recover
the cycles (1 2 3) and (4 5 6 7) that it is a product of. Based on the above
remark, we can write it equally well as any of the following:
α = ( 12 23 31 45 56 67 74 88 )
= (1 2 3)(4 5 6 7)
= (4 5 6 7)(1 2 3)
= (1 2 3)(4 5 6 7)(8)
= (2 3 1)(4 5 6 7)(8)
or in indeed various other equivalent ways. The following theorem shows
that the behaviour observed in this example is typical.
Theorem 3.7. Every permutation in Sn can be written as a product of
disjoint cycles; moreover this expression is unique up to
(i) the order in which the cycles occur,
(ii) the different ways of writing each cycle (rotating the cycle), and
(iii) the presence or absence of 1-cycles.

Proof. Let π ∈ Sn , and choose any a1 ∈ {1, 2, . . . , n}. Consider the sequence
a1 , a2 , a3 , . . . defined by ai+1 = π(ai ) for i ∈ N. Since this sequence is
infinite, and {1, 2, . . . , n} is a finite set, not all the ai can be distinct. Let r
be minimal such that ar+1 = as for some 1 ≤ s ≤ r. We claim that s = 1, in
other words, that the sequence is a1 , a2 , . . . , ar , a1 , . . . where a1 , . . . , ar are
distinct.
To see this, suppose s > 1. Then π(ar ) = ar+1 = as = π(as−1 ). As π is
injective this would force ar = as−1 , contrary to the minimality of r. Thus
we must have s = 1, and we have the cycle γ1 = (a1 a2 · · · ar ).
If r = n then π = γ1 as required, so now assume r < n. We choose
b1 ∈ {1, 2, . . . , n} \ {a1 , a2 , . . . , ar } and proceed as above. This gives another

26
cycle γ2 = (b1 b2 · · · bt ). The cycles γ1 and γ2 are disjoint, because if ai = bj
then
b1 = π t+1−j (bj ) = π t+1−j (ai ) = ai+t+1−j ,
contrary to the choice of b1 . Continuing in this way we obtain π as a product
of disjoint cycles γ1 γ2 · · · γk .
It is clear that two expressions which differ only in ways (i), (ii) and (iii)
represent the same permutation. Conversely, two expressions which differ
in any other way represent different permutations.

From this point onwards, we will often not write π ◦ σ for composition of
permutations, but will just write πσ, as this is more consistent with disjoint
cycle notation. We will also often talk of the product of two permutations,
rather than their composition, although they mean the same thing. Note
however that the order we write the permutations is still important in gen-
eral; only disjoint cycles commute with each other (as in (f) above).
Definition 3.8. A permutation which is written as a product of disjoint
cycles is said to be in disjoint cycle notation; if all 1-cycles are included, it
is said to be in full cycle notation.
In disjoint cycle notation, we often leave out the 1-cycles. Then the
identity permutation ι in disjoint cycle notation has no cycles at all; this is
why we give it a notation, ι, of its own.
Example 3.9. Let π = ( 13 28 36 41 57 64 79 82 95 ) and σ = ( 11 25 36 49 57 62 73 88 94 ); then
the expressions for π, σ, πσ and π −1 in full cycle notation are as fol-
lows:
π = (1 3 6 4)(2 8)(5 7 9),

σ = (2 5 7 3 6)(4 9) = (2 5 7 3 6)(4 9)(1)(8),

πσ = (1 3 6 4)(2 8)(5 7 9)(2 5 7 3 6)(4 9)

= (1 3 4 5 9)(2 7 6 8),

π −1 = (4 6 3 1)(8 2)(9 7 5) = (1 4 6 3)(2 8)(5 9 7).

Recall that composing an r-cycle with itself r times gives the identity
permutation ι. We can extend this idea to all permutations (and later we
will extend it to all elements of a group) as follows.
Definition 3.10. Let π ∈ Sn . The order o(π) of π is the minimal r ∈ N
r times
r
� �� �
such that π = ππ · · · π = ι.

27
 Then one particular advantage of disjoint cycle notation is that it is easy
to calculate the order of a permutation written this way.
Proposition 3.11. If π = γ1 γ2 · · · γk , where the γi are disjoint cycles, then
the order of π is the lowest common multiple of the lengths of the cycles γi .
Proof. Since the γi commute, for n ∈ N we have

π n = (γ1 · · · γk )n = γ1 n · · · γk n .

As the γi move different elements, π n = ι if and only if γi n = ι for all i.

Since by remark (e) above, the order of each cycle γi is its length, the result
follows.

Class Exercise 3.12. What is the order of π = (1 3 6 4)(2 8)(5 7 9)?

Answer. o(π) = lcm(4, 2, 3) = 12.

To conclude this section we focus on the smallest non-trivial cycles.

Definition 3.13. A 2-cycle is called a transposition.
The following result gives some indication of the importance of transposi-
tions.
Theorem 3.14. If n > 1, every permutation in Sn can be written as a
product of transpositions.
Proof. By Theorem 3.7 it suffices to show that every r-cycle for r > 1 can
be written as a product of transpositions. We may write

(a1 a2 a3 . . . ar−1 ar ) = (a1 a2 )(a2 a3 ) · · · (ar−1 ar ).

For looking at the product of transpositions and remembering that we read

right to left, we see that a1 is fixed by all the 2-cycles until (a1 a2 ), which
sends a1 to a2 . Then a2 is fixed by all the 2-cycles until (a2 a3 ), so we
have “(a1 a2 a3 ” so far. Continuing in this way, we see that the product of
transpositions is equal to the r-cycle as claimed.

Class Exercise 3.15. In S8 we have

( 14 25 32 48 53 66 71 87 ) = (1 4 8 7)(2 5 3) = (1 4)(4 8)(8 7)(2 5)(5 3).

28
Notice that in this example the expression (1 4)(4 8)(8 7)(1 2)(1 5)(1 2)(5 3)
gives the same permutation. In general there will be many ways of expressing
a given permutation as a product of transpositions; the essential uniqueness
of Theorem 3.7 occurs because the cycles there are required to be disjoint. In
the next section we will see what can be said about uniqueness of expressions
in terms of transpositions.

3.2 The sign of a permutation

Let π ∈ Sn and suppose that π = γ1 γ2 . . . γk with the γi disjoint cycles. Set
k
�
ν(π) = (|γi | − 1),
i=1

where |γi | denotes the length of the cycle γi . Note that ν(π) is well-defined
because of the uniqueness in Theorem 3.7 (in particular, cycles of length 1
contribute 0 to the sum, and so may be ignored).
Class Exercise 3.16. If π = ( 14 25 32 48 53 66 71 87 ) = (1 4 8 7)(2 5 3)(6), then

ν(π) = (4 − 1) + (3 − 1) + (1 − 1) = 3 + 2 + 0 = 5.
Note that π −1 has the same cycle lengths as π, so we have ν(π −1 ) = ν(π).
Example 3.17. If π is as above, π −1 = (7 8 4 1)(3 5 2)(6), so

ν(π −1 ) = 3 + 2 + 0 = 5 = ν(π).
Clearly ν(π) = 1 if and only if π is a transposition. In fact we may
interpret ν(π) as follows:�if all possible 1-cycles are included in the expression
k
π = γ1 γ2 . . . γk , then i=1 |γi | = n; thus ν(π) = n − k, i.e. ν(π) is the
difference between n and the number of cycles in the full cycle notation.
Example 3.18. If π is as above, then n = 8 and k = 3, so ν(π) = 8 − 3 = 5.
Now ν(π) on its own is not particularly important, but it enables us to
make the following definition.

Definition 3.19. Given π ∈ Sn , the sign of π is defined by

sign(π) = (−1)ν(π) .

Thus sign(π) = ±1; if sign(π) = 1 we call π an even permutation, while if

sign(π) = −1 we call π an odd permutation.

Example 3.20. If π = (1 4 8 7)(2 5 3)(6) as above, we have sign(π) =

(−1)ν(π) = (−1)5 = −1 and so π is odd.
Note that because ν(π −1 ) = ν(π), we have sign(π −1 ) = sign(π).

29
Remark 3.21. It is possible to define the sign of a permutation π in another
way, as (−1)n(π) where n(π) is the number of ‘inversions’ of π, i.e. the
number of pairs (i, j) with i < j and π(i) > π(j); we shall not go into this
here.
We now consider how the sign of a permutation is affected by multiplic-
ation by a transposition.
Lemma 3.22. Let π ∈ Sn and let τ ∈ Sn be a transposition. Then
sign(πτ ) = (−1) · sign(π) = − sign(π).
Proof. Put τ = (a1 a2 ) and consider π as a product of disjoint cycles. We
begin by comparing ν(πτ ) and ν(π) and we deal separately with the two
cases where the cycles in π containing a1 and a2 are different or the same.
If a1 and a2 are in different cycles of π, we have

π = γ1 · · · γk−2 (a1 b1 · · · br )(a2 c1 · · · cs ),

�� �
k−2
where r, s ≥ 0. Then ν(π) = i=1 (|γi | − 1) + r + s. We have

πτ = γ1 · · · γk−2 (a1 b1 · · · br )(a2 c1 · · · cs )(a1 a2 )

= γ1 · · · γk−2 (a1 c1 · · · cs a2 b1 · · · br )
�� �
k−2
and so ν(πτ ) = i=1 (|γ i | − 1) + r + s + 1 = ν(π) + 1.
If on the other hand a1 and a2 are in the same cycle of π, we have

π = γ1 · · · γk−1 (a1 b1 · · · br a2 c1 · · · cs )
�� �
k−1
with r, s ≥ 0. Then ν(π) = i=1 (|γi | − 1) + r + s + 1. This time

πτ = γ1 · · · γk−1 (a1 b1 · · · br a2 c1 · · · cs )(a1 a2 )

= γ1 · · · γk−1 (a1 c1 · · · cs )(a2 b1 · · · br )
�� �
k−1
and so ν(πτ ) = i=1 (|γ i | − 1) + r + s = ν(π) − 1.
Thus in either case we have ν(πτ ) = ν(π) ± 1; so

sign(πτ ) = (−1)ν(πτ ) = (−1)ν(π)±1 = (−1)ν(π) · (−1) = − sign(π).

Example 3.23. Let π = (1 4 8 7)(2 5 3) as above, so that sign(π) = −1. If
τ = (4 5) then
πτ = (1 4 8 7)(2 5 3)(4 5) = (1 4 3 2 5 8 7).

So ν(πτ ) = 6 and sign(πτ ) = (−1)6 = 1 = − sign(π).

On the other hand, if τ = (4 8) then

πτ = (1 4 8 7)(2 5 3)(4 8) = (1 4 7)(2 5 3)(8),

30
So ν(πτ ) = 4 and sign(πτ ) = (−1)4 = 1 = − sign(π).
Corollary 3.24. If τ1 , . . . , τr ∈ Sn are transpositions, then

sign(τ1 · · · τr ) = (−1)r .

Proof. We may use induction on r. The result is true for r = 1, since we have
ν(τ1 ) = 1 and so sign(τ1 ) = (−1)1 . The induction step follows immediately
from Lemma 3.22.

This shows what can be said about an expression for a given permuta-
tion as a product of transpositions: an even (respectively odd) permutation
can only be written as a product of an even (respectively odd) number of
transpositions.
Theorem 3.25. Let π, σ ∈ Sn . Then sign(π) sign(σ) = sign(πσ).
Proof. By Theorem 3.14 we may write

π = τ1 · · · τr , σ = τ1 � · · · τ s �

for some transpositions τ1 , . . . , τr , τ1 � , . . . , τs � . Then πσ = τ1 · · · τr τ1 � · · · τs �

and so
sign(π) sign(σ) = (−1)r (−1)s = (−1)r+s = sign(πσ).

Example 3.26. In S8 , let π = (1 6)(3 7 8)(4 5) and σ = (1 4 8)(2 3 7)(5 6).

Then

πσ = (1 6)(3 7 8)(4 5)(1 4 8)(2 3 7)(5 6) = (1 5)(2 7)(3 8 6 4).

Here

ν(π) = 1 + 2 + 1 = 4 so sign(π) = 1;
ν(σ) = 2 + 2 + 1 = 5 so sign(σ) = −1;
ν(πσ) = 1 + 1 + 3 = 5 so sign(πσ) = −1.
This result shows that if we define a function Σ : Sn → {1, −1} by setting
Σ(π) = sign(π) for all π ∈ Sn , then Σ respects products, in the sense that
Σ(πσ) = Σ(π)Σ(σ) (where the product on the right is the usual product of
integers). Such maps will be very important later on.
Notice too that if we take the product of two even permutations, we
obtain another even permutation (1 · 1 = 1!), so the subset of Sn consisting
of all the even permutations is closed with respect to the group operation.
Also, the identity permutation is even and the inverse of an even permutation
is even. When we move to the general theory shortly, such a subset will be
called a subgroup; a subgroup is in particular a group in its own right.
This particular subgroup of Sn is important enough to have its own
name.

31
Definition 3.27. The subgroup of Sn consisting of the even permutations
is called the alternating group of degree n and is denoted An .

It can be shown that any element of An can be written as a product

of 3-cycles (the ‘smallest’ elements of An , since every transposition is odd);
compare with Theorem 3.14.
Remark 3.28. Another way to describe the sign of a permutation is via
matrices. For a given permutation σ ∈ Sn , let Aσ be the n × n matrix with
i) entry (1 ≤ i ≤ n) and 0 everywhere else. One way to write
1 in the (σ(i),�
this is: Aσ = ni=1 eσ(i),i , where eji is the n × n matrix with 1 in the (j, i)
position and 0 everywhere else.
Then for any row vector (x1 , . . . , xn ) ∈ Rn ,

(x1 , x2 , . . . , xn )Aσ = (xσ(1) , . . . , xσ(n) )

Thus (or by a direct calculation), Aσ Aτ = Aστ .

What we have shown is that σ �→ Aσ defines a function Sn → GLn (R)
that respects multiplication. In this setting, the sign of a permutation σ is
just det Aσ . The matrices Aσ are called permutation matrices or sometimes
monomial matrices.

As an explicit example, let σ = (1 2 3 4) and τ = (1 3 2) be permutations

in S4 . We have
   
0 0 0 1 0 1 0 0
1 0 0 0 0 0 1 0
Aσ =  0 1 0 0
, A τ = 
1 0 0 0 ,


0 0 1 0 0 0 0 1

 
0 0 0 1
0 1 0 0
Aστ = A(1 4) =
0
 and
0 1 0
1 0 0 0
    
0 0 0 1 0 1 0 0 0 0 0 1
1 0 0  
0 0 0 1 
0 0 1 0 0
Aσ Aτ = 
0 =  = Aστ .
1 0 0 1 0 0 0 0 0 1 0
0 0 1 0 0 0 0 1 1 0 0 0

32
 Via permutations, we have actually encountered many of the basic ideas
of group theory: groups, subgroups, functions that preserve the group op-
eration. What we will do next is to turn these into abstract definitions that
can apply to any group.
In doing so, we are following in the footsteps of the founders of the
subject: Galois, Cauchy, Lagrange, Cayley, Abel and others. For groups first
arose as groups of permutations, in the work of Galois on the solvability of
polynomial equations around 1830, and an abstract definition was not given
until work of von Dyck in 1882.

33
3.3 Elementary properties, or, algebraists like to be fussy
We are now going to embark on our systematic study of groups, starting
from the definition and working up. Along the way, there will of course be
many examples drawn from our big list above, but for a while at least things
will become more abstract.
Let us remind ourselves of the definition of a group.

Definition 2.1. A group is an algebraic structure (G, ∗) with ∗ an associ-

ative binary operation, having an identity and such that every element of G
has an inverse with respect to ∗.

By our earlier results on binary operations, we have the following.

Proposition 3.29. Let (G, ∗) be a group.

(i) The identity element of G, e, is unique.

(ii) Every element of G has a unique inverse with respect to ∗.

(iii) (Cancellation) For a, b, c ∈ G, if either a ∗ b = a ∗ c or b ∗ a = c ∗ a

then b = c.

(iv) For a, b ∈ G, the equation a ∗x = b has the unique solution x = a−1 ∗b.

(v) For a, b ∈ G, the equation x ∗a = b has the unique solution x = b∗a−1 .

(vi) If a ∗ b = e then b = a−1 and a = b−1 .

Proof. Parts (i) and (ii) are immediate from Proposition 1.10(i) and (ii),
since ∗ is associative and e is an identity element for ∗.
Then (iii) follows, since if a ∗ b = a ∗ c then

b = e ∗ b = (a−1 ∗ a) ∗ b = a−1 ∗ (a ∗ b) = a−1 ∗ (a ∗ c) = (a−1 ∗ a) ∗ c = e ∗ c = c.

A similar argument, with a−1 on the right, shows that if b ∗ a = c ∗ a then

b = c also.
For (iv),
a ∗ (a−1 ∗ b) = (a ∗ a−1 ) ∗ b = e ∗ b = b
(by associativity, the existence of an identity element and the definition of
inverses—i.e. because G is a group) so that x = a−1 ∗ b is a solution. Then
if x� is another solution, we have a ∗ x = a ∗ x� and so by part (iii), x� = x
and the solution is unique. A similar argument proves (v).
Lastly, for (vi), consider the equation a ∗ x = e. We have that a−1 and
b are solutions, since a−1 is the inverse of a and by our assumption that
a ∗ b = e. Hence by part (iv), a−1 = b. Similarly, using part (v), we may
deduce that b−1 = a.

34
Corollary 3.30. Let (G, ∗) be a group and let a ∈ G. Then there is a
bijection between the set G and the sets {a ∗ g | g ∈ G} and {g ∗ a | g ∈ G}.
Proof. First note that since G is a group, the sets {a ∗ g | g ∈ G} and
{g ∗ a | g ∈ G} are subsets of G.
Define a function La : G → G, La (g) = a ∗ g. Then the image of La is
by definition equal to the set {a ∗ g | g ∈ G} and hence it suffices to show
that La is injective.
To show this, assume that La (b) = La (c). Then

a ∗ b = La (b) = La (c) = a ∗ c.

By part (iii) of the previous proposition, b = c and so La is injective.

Defining Ra : G → G, Ra (g) = g ∗ a, a similar argument completes the
proof.

Remark 3.31. In fact, a little more is true: La is actually surjective, as we

may easily show since we have already done all the work needed. Let b ∈ G.
Then by part (iv), there exists a solution to the equation a ∗ x = b, namely
a−1 ∗ b.
So for any fixed a ∈ G, La : G → G is in fact a bijection, that is, it is a
permutation of the elements of G. Of course, the same is true of Ra .

Proposition 3.32. Let (G, ∗) be a group and let a, b ∈ G. Then

(a ∗ b)−1 = b−1 ∗ a−1 .

Proof. We have

(a ∗ b) ∗ (b−1 ∗ a−1 ) = a ∗ (b ∗ b−1 ) ∗ a−1 = a ∗ e ∗ a−1 = a ∗ a−1 = e

by associativity and the properties of the identity and inverses. Hence by

Proposition 3.29(vi), b−1 ∗ a−1 = (a ∗ b)−1 .

So, we have shown a number of elementary properties of groups, many

of which you may well have thought were “obviously” true, based on your
experience to date. However, algebraists do like to be fussy and to see
proofs of things that are “obvious”; if you stick around in the subject a
little longer, you will encounter plenty of “obvious” claims that are in fact
false, sometimes by very complicated or subtle counterexamples.
All of the above are true, though, and as we said before, now we need
never prove them again: they will be true for any group, no matter what its
underlying set or how complicated its group operation.
We conclude this section with an important construction, used to make
new larger groups from ones we already know. Recall that the Cartesian

35
product of sets X and Y is the set X × Y = {(x, y) | x ∈ X, y ∈ Y } of
ordered pairs of elements from X and Y . If X and Y have more structure,
such as being groups or rings, we may “extend” those structures to X × Y .
Proposition 3.33. Let (G, ∗) and (H, ◦) be groups. Then the Cartesian
product G × H is a group with respect to the binary operation • defined by

(g1 , h1 ) • (g2 , h2 ) = (g1 ∗ g2 , h1 ◦ h2 )

for all g1 , g2 ∈ G, h1 , h2 ∈ H.
Exercise 3.34. Prove Proposition 3.33.
Subsequently, if we refer to the Cartesian product of groups, G × H, we
will mean (G × H, •) with • as given in this Proposition.

3.4 Subgroups
We begin with a comment on notation. If we are working in an arbitrary
group (G, ∗) then we will write e for the identity of the group (sometimes
eG if more than one group is being discussed) and also we will sometimes
suppress the symbol for the binary operation, ∗, and write simply ab for
a ∗ b. This “multiplicative” notation (more properly called concatenation,
meaning “placing side by side”) is used for convenience and does not mean
that the binary operation is necessarily multiplication. We will also talk
about “the group G” rather than writing (G, ∗) every time, although if we
are discussing more than one group or more than one binary operation, or
if we need to explicitly talk about properties of the operation, we might be
more precise again.
If we are working in a particular group and there is an appropriate
symbol for either the identity or the binary operation, such as “0” or “+”,
then we use it.
As we saw with permutations, it is often natural to ask about subsets
of a group that are closed under the binary operation. That is, for any
two elements a, b belonging to the subset, we have a ∗ b also in the subset.
However what we really want is for the subset to itself be a group: then we
can apply all the theory of groups to it too. It turns out that just being
closed under the operation is not enough (shortly we will see what else one
should ask for) and the definition we make next is morally the right one.
Definition 3.35. Let (G, ∗) be a group. A subset H of G is said to be a
subgroup if (H, ∗) is a group. If this is the case, we write H ≤ G.

Remarks 3.36.
(a) From the definition, we see that “H is a subgroup of G” means not
just that “H is a group with respect to some operation”, but it means
that H is a group with respect to the same operation as G.

36
 (b) It is immediate from this definition that every subgroup of an Abelian
group is Abelian: if a∗b = b∗a for all a, b ∈ G, then certainly a∗b = b∗a
for all a, b ∈ H ⊆ G.
(c) If you prefer to be a little bit more formal, remember that ∗ : G × G →
G is a function sending pairs of elements of G to an element in G. To
say that (H, ∗) is a group means that when we restrict to looking at
pairs of elements of H (i.e. looking at the function ∗|H×H : H ×H → G
given by restricting ∗ to H × H), we have Im ∗|H×H ⊆ H so that
there is a well-defined function ∗|H H×H : H × H → H, restricting the
domain to H × H and the codomain to H, and (H, ∗|H H×H ) is a group.
Unsurprisingly, we tend to brush this under the carpet a bit and just
write (H, ∗).
Examples 3.37.
(a) (Z, +) is a subgroup of (Q, +), and both are subgroups of (R, +).
(b) (Q∗ , ×) is a subgroup of (R∗ , ×).
(c) (Q∗ , ×) is not a subgroup of (Q, +), since the two binary operations
are different.
(d) For n ∈ N let ζ = e2πi/n ∈ C. The set Cn = {ζ k | k ∈ Z} is a subgroup
of (C∗ , ×). Notice that |Cn | = n (since e2πi = 1), so that Cn is a finite
group, while C∗ is certainly not.
(e) In the group GLn (R) (where the operation is matrix multiplication),
the set of matrices with determinant 1 is a subgroup; this follows
from properties of determinants you proved in Linear Algebra. This
subgroup has a special name: it is known as the special linear group,
SLn (R). (As before, one may replace R by any field F and obtain
SLn (F ).)
(f) We can consider Sn−1 as a subgroup of Sn : if σ ∈ Sn−1 then we think
of σ as a permutation of {1, 2, . . . , n}, by setting σ(n) = n. More
generally, the set of permutations fixing every element of some subset
of {1, 2, . . . , n} is a subgroup: if two permutations fix some i so does
their product. Also if a permutation fixes i, so does its inverse.
(g) We claimed that the set of even permutations An is a subgroup of Sn ;
this claim will be justified more rigorously very shortly.
(h) In any group G, {e} and G are subgroups.
This last pair of examples is “uninteresting” (in the sense that they don’t
tell us anything we didn’t already know) and on occasion we will want to
exclude these from claims about subgroups. So we introduce the following
terminology.

37
Definition 3.38. Let H be a subgroup of G.
The subgroup {e} is called the trivial subgroup of G. Considered as a
group in its own right, it is called the trivial group. If H �= {e}, we say that
H is a non-trivial subgroup of G.
If H �= G (so that H is a proper subset of G), we say H is a proper
subgroup of G and write H � G or more simply H < G.
In principle, to show that a subset H of a group (G, ∗) is itself a group—
that is, to check the definition of a subgroup above—one should show that
H has a well-defined binary operation that is associative, has an identity
element and where every element has an inverse. However this is a lot to
check and in fact one can check rather less6 .
We will give two lists of properties to check and we will refer to these as
the “subgroup tests”.
Proposition 3.39 (Subgroup test I). Let H be a subset of a group (G, ∗)
and denote the identity element of G by eG . Then H is a subgroup of G if
and only if the following three conditions are satisfied:
(SG1) eG ∈ H;
(SG2) for all h1 , h2 ∈ H, we have h1 ∗ h2 ∈ H;
(SG3) for all h ∈ H, we have h−1 ∈ H.

Proof. Assume that H is a subgroup of G, so that H is a group in its own

right (with respect to the same operation). Then we must have (SG2), or
else the operation on H would not be a binary operation; binary operations
are necessarily closed (see the discussion on page 13).
Secondly, H must have an identity element eH such that eH ∗ h = h =
h ∗ eH for all h ∈ H. Since eG is the identity element of G, we have
eG ∗ h = h = h ∗ eG for all h ∈ H too. But by the uniqueness of identity
elements (Proposition 3.29(i)) eH = eG and so eG ∈ H and we have (SG1).
Finally, each element of H must have an inverse in H, as well as an
inverse in G. By uniqueness of inverses (Proposition 3.29(ii)), these are
again equal, giving (SG3).
Conversely, assume that (SG1), (SG2) and (SG3) hold. Then (SG2)
ensures that ∗ does restrict to a well-defined binary operation on H. Since
∗ is associative for all triples of elements of G, it is associative for all triples
of elements of H, since H is a subset of G.
By (SG1), the element eG belongs to H and since eG ∗ g = g = g ∗ eG for
all g ∈ G, we must have eG ∗ h = h = h ∗ eG for all h ∈ H (again since any
h ∈ H is also an element of G). So H has an identity element, namely eG .

6
The corresponding situation for rings makes this even more desirable: to check that a
set with two binary operations is a ring involves checking nine properties. As we will see,
to show that a subset is a subring, it is possible to get away with checking only four.

38
 It remains to show that every element of H has an inverse in H: we
know that every element of H has an inverse in G but for H to be a group
in its own right, that inverse has to be an element of H. However (SG3) tells
us precisely that the element h−1 inverse to h in G is actually an element of
H, as we need.

Be alert here: it is not enough for a subset to be closed under the group
operation for it to be a subgroup. It is possible to find examples of sub-
sets satisfying (SG2) but not one or other of (SG1) and (SG3), as we will
see shortly. But first, let us start off positively and see some examples of
subgroups and the subgroup test in action.
Examples 3.40.

(a) The set of even integers 2Z = {2r | r ∈ Z} is a subgroup of (Z, +), the
integers under addition.
(SG1) The identity of (Z, +) is 0, which is even.
(SG2) The sum of two even integers is even: 2r +2s = 2(r +s) ∈ 2Z.
(SG3) The negative of an even integer is even: if m = 2r then
−m = −2r = 2(−r) ∈ 2Z.
Similarly, for any m ∈ Z, mZ is a subgroup of (Z, +). If m = 0,
mZ = 0Z is trivial; if m = ±1, mZ = ±1Z = Z so mZ is proper
provided m �= ±1.

(b) For n ∈ N and ζ = e2πi/n ∈ C, the set Cn = {ζ k | k ∈ Z} is a subgroup

of (C∗ , ×).
(SG1) We have 1 = ζ 0 ∈ Cn .
(SG2) By the rules for products of powers, ζ r ζ s = ζ r+s ∈ Cn .
(SG3) Clearly ζ −r is the inverse of ζ r .
We have C1 = {1} trivial and proper, and Cn non-trivial and proper
for n ≥ 2.

(c) As asserted previously, the set of even permutations An is a subgroup

of Sn .
(SG1) The identity permutation ι is even: sign(ι) = (−1)ν(ι) =
(−1)0 = 1.
(SG2) By Theorem 3.25, if π, σ are even, so sign(π) = sign(σ) = 1,
then sign(πσ) = 1 · 1 = 1 and πσ is also even.
(SG3) Since ν(π −1 ) = ν(π), the inverse of an even permutation is
again even.

39
 For n = 1, S1 = A1 = {ι} so A1 is trivial and not proper. For n = 2,
A2 = {ι} < S2 is trivial and proper. For n > 2, An is both proper and
non-trivial.

(d) The special linear group SLn (R) (matrices of determinant 1) is a sub-
group of GLn (R) (invertible matrices, with operation matrix multi-
plication).
(SG1) The identity matrix In has determinant 1.
(SG2) We know that det(AB) = det(A) det(B), so if A, B have
determinant 1, so does AB.
(SG3) We also know that det(A−1 ) = det(A)−1 , so if A has determ-
inant 1, so does A−1 .
For n = 1, GL1 (R) = R∗ = R\{0}; SL1 (R) = {1} is trivial and proper.
For any n > 1, SLn (R) is a proper non-trivial subgroup of GLn (R).

Sometimes it is instructive to see non-examples too, as in this case, to

see why all three conditions in the subgroup test are necessary.
Class Exercise 3.41.

(a) The set of odd integers, 2Z + 1 = {2r + 1 | r ∈ Z}, is not a subgroup

of (Z, +).
(SG1) The identity of (Z, +) is 0 but 0 is not odd.
(SG2) We have that 1 + 3 = 4: 1 and 3 are odd but 1 + 3 is not.
(SG3) 2Z + 1 is closed under negation: −(2r + 1) = 2(−r − 1) + 1.
While it is admittedly somewhat unnatural, we could consider the
subset S = (2Z + 1) ∪ {0}; this subset would satisfy (SG1) and (SG3)
but not (SG2).

(b) The set of positive integers N = {r ∈ Z | r > 0} is not a subgroup of

(Z, +).
(SG1) The identity element 0 is not a positive integer.
(SG2) N is closed under addition.
(SG3) N is not closed under negation.
So N0 = {r ∈ Z | r ≥ 0} = N ∪ {0} satisfies (SG1) and (SG2) but not
(SG3); N0 is not a group with respect to addition.

You might have been expecting an example in which we have (SG2) and
(SG3) but not (SG1). However this is not possible, since h ∗ h−1 = eG would
then be an element of H. So there is some redundancy in the subgroup test
above and we can give a more “efficient” version, as follows.

40
Proposition 3.42 (Subgroup test II). Let H be a subset of a group (G, ∗).
Then H is a subgroup of G if and only if the following two conditions are
satisfied:

(SG1� ) H �= ∅;

(SG2� ) for all h1 , h2 ∈ H, we have h1 ∗ h−1 −1

2 ∈ H, where h2 is the inverse
of h2 in G.
Proof. We prove that (SG1� ) and (SG2� ) hold if and only if (SG1), (SG2)
and (SG3) hold; that is, we show that version II of the subgroup test is
equivalent to version I.
First assume that (SG1), (SG2) and (SG3) hold. Then since by (SG1),
eG ∈ H, H is not empty so (SG1� ) holds. For (SG2� ), let h1 , h2 in H. Then
by (SG3), h−1 −1 �
2 ∈ H and hence by (SG2) h1 ∗ h2 ∈ H, so (SG2 ) also holds.
� �
Conversely, assume that (SG1 ) and (SG2 ) hold. Since H is not empty,
there exists an element h ∈ H. Then by (SG2� ), h ∗ h−1 ∈ H. But h ∗ h−1 =
eG so (SG1) holds. Now eG , h ∈ H so by (SG2� ), eG ∗ h−1 = h−1 ∈ H so
(SG3) is satisfied. Finally, since h1 ∗ (h−1
2 )
−1 = h ∗ h , (SG2) also follows.
1 2

Exercise 3.43. Apply the second subgroup test (Proposition 3.42) to the
examples in Examples 3.40 and 3.41 to confirm what we found out from
using the first subgroup test (Proposition 3.39).
Exercise 3.44. Let (G, ∗) be a group and consider the set

Z(G) = {z ∈ G | z ∗ g = g ∗ z for all g ∈ G}.

Prove that Z(G) is a subgroup of G and that Z(G) is Abelian.

This subgroup is important enough to be given a name, the centre of G.
In order to prove that a given subset is a subgroup we (almost always)
use one of the subgroup tests. Which subgroup test you use is up to you:
they are logically equivalent. The second is slicker but the first is sometimes
clearer, because it breaks the verification down into more and smaller steps.
Having said that, there are other ways to construct subgroups and the
next result is one such.
Recall from MATH112 Discrete Mathematics that if I is a set, we say
that a collection of sets A is an I-indexed family of sets if there is a one-to-
one correspondence between the sets in the collection A and I. Then we may
write A = {Ai | i ∈ I}, i.e. denoting by Ai the member of A corresponding
to I. �
The intersection of the sets in A is i∈I Ai = {x | (∀ i ∈ I)(x ∈ Ai )}.

Proposition 3.45. The intersection of any indexed family of subgroups of

G is again a subgroup of G.

41
Proof. Let H be an I-indexed family of subgroups of (G, ∗), with the ith
subgroup being denoted Hi . That is, for all i ∈ I, Hi is a subgroup of G.
We use the first subgroup test.

� all i ∈ I, by (SG1) for Hi , we have eG ∈ Hi . Then eG ∈

(SG1) For
i∈I Hi .
�
(SG2) Let h1 , h2 ∈ i∈I Hi . Then for all i ∈ I, h1 , h2 ∈ Hi . By �
we have
h1 ∗ h2 ∈ Hi and this is true for all i ∈ I. Hence h1 ∗ h2 ∈ i∈I Hi .
�
(SG3) Let h ∈ i∈I so h ∈ Hi for all � i ∈ I. Then by (SG3) we have
−1 −1
h ∈ Hi , for all i ∈ I. So h ∈ i∈I Hi .

Exercise 3.46 (Not essential, but good practice with abstract arguments).
� the subgroups: for any collection H of subgroups
We do not need to index �
of G, we may define H = {x | (∀ H ∈ H)(x ∈ H)}. Prove that H is a
subgroup of G.
Remarks 3.47.
(a) We used the first subgroup test rather than the second, as�it is not
enough to know that each Hi is non-empty to conclude that i∈I Hi is
non-empty. (Two non-empty but disjoint sets have empty intersection,
e.g. {1}∩{2} = ∅.) We need to know that there is a particular element
that belongs to every Hi , which there is, namely eG .
(b) The union of subgroups need not be a subgroup: let H1 = 2Z, H2 = 3Z
be subgroups of (Z, +). Then 2, 3 ∈ H1 ∪ H2 but 2 + 3 = 5 ∈
/ H 1 ∪ H2 .
Example 3.48. In (Z, +), the intersection of 2Z and 3Z is the set of integers
that are divisible by 2 and divisible by 3, and these are precisely the integers
divisible by 6. So 2Z ∩ 3Z = 6Z.
More generally, mZ ∩ nZ = lcm(m, n)Z and
m1 Z ∩ m2 Z ∩ · · · ∩ mr Z = lcm(m1 , m2 , . . . , mr )Z.
The collection H �= {mZ | m ∈ N} is a N-indexed family of subgroups of
Z. The intersection m∈N mZ consists of all � integers that are divisible by
every m ∈ N. Only 0 has this property, so m∈N mZ = {0} is trivial.
Sometimes we have a subset S of a group G that is not a subgroup, but
where we would like to find a subgroup of G that contains S. Obviously we
could take all of G but this won’t tell us much: we really want the smallest
subgroup of G containing S.
“Smallest” here might mean “smallest size”, but for infinite sets it is
better to think of smallest as “smallest under inclusions of sets”. But we
know how to find the smallest set: it is the intersection. (The intersection is
a subset of every set in the collection and is the unique smallest such under
inclusion.) This leads us to make the following definition.

42
Definition 3.49. Let G be a group and let S ⊆ G be a subset of G. Let
S = {H ≤ G�| S ⊆ H} be the family of subgroups of G that contain S. The
intersection S is a subgroup of�G containing S, called the subgroup of G
generated by S. We write �S� = S.

Notice that by Proposition 3.45 (or rather the slightly more general
version in the exercise that follows) �S� is a subgroup of G. It is also the
unique subgroup having the desired property, namely that it contains S. If
S is actually a subgroup already, then �S� = S.
Example 3.50. In (Z, +), we have �m� = mZ (and in particular �1� = Z)
and �{m, n}� = hcf(m, n)Z, where hcf denotes the highest common factor.
We will discuss this in more detail later.
To simplify notation, if S = {g1 , . . . , gr }, we will write �g1 , . . . , gr � rather
than �{g1 , . . . , gr }� for the subgroup generated by S. So for example, we
would write �m, n� = hcf(m, n)Z.
The above description of the subgroup generated by S as an intersection
is good from the point of view of proving that it is actually a subgroup, but
bad from the point of view of knowing what �S� looks like as a subset of G.
So we are now going to try to tie down a more precise description of
the subgroup generated by S, as the set of elements of G which can be
“expressed in terms of” the elements of S. For example, if x, y ∈ S (and
hence in �S� since S ⊆ �S�) then xy 2 x−1 y 2 x3 y −1 x is expressed in terms of
x and y and is in �S�, by repeated application of the subgroup test.
The following proposition expresses this idea in more formal language.

Proposition 3.51. Let S be a subset of G. Let H be the set of all elements

of G of the form sa11 sa22 · · · sar r , where r is a non-negative integer, s1 , . . . , sr ∈
S and a1 , . . . , ar ∈ Z. (When r = 0, this is the identity element of G.)
Then H is a subgroup of G and H = �S�.
Proof. To show that H is a subgroup of G, we use the first subgroup test.

(SG1) We have e ∈ H by the case r = 0.

b
(SG2) Suppose that we have expressions g = sa11 · · · sar r and h = tb11 · · · tqq ,
where s1 , . . . , sr , t1 , . . . , tq ∈ S and a1 , . . . , ar , b1 , . . . , bq ∈ Z. Then
clearly
b
gh = sa11 · · · sar r tb11 · · · tqq ∈ H.

(SG3) Let g = sa11 · · · sar r ∈ H. Then g −1 = s−a r · · · s−a1 ∈ H.

r 1

So all three conditions are satisfied and hence H is a subgroup of G. It

remains only to show that H = �S�.

43
 We have �S� ⊆ H since H is a subgroup of G which contains S. But
any subgroup of G which contains S must contain each element of H by
(repeated) application of the subgroup test. Thus H ⊆ �S� and so H = �S�.

This gives us (another) natural way to think about the complexity of

groups. The simplest is by increasing order: groups of order 2 are simpler
than groups of order 3 and so on, all of which are simpler than infinite
groups. We will explore this a bit more later, when we examine the possible
groups for various small orders (when we have the technology we will need).
But we could instead organise our groups by the smallest number of
elements needed to generate them. So groups generated by a single element
should be simpler than groups that need two generators, and so on. This
leads to interesting and difficult questions, with the most obvious being
“given a group G, what is the smallest n for which G is generated by a
subset of size n?”.
In the next section, we will see that the class of groups generated by one
element can indeed be described comprehensively; these groups are called
cyclic. (It is beyond the scope of this course to say how and why, but the
class of groups generated by two elements cannot be.)

3.5 Groups generated by one element

We focus on groups and subgroups generated by a single element.

Definition 3.52. Let G be a group. If there exists an element g ∈ G such

that �g� = G, then we say that G is a cyclic group. Such an element g is
called a generator of G.
More generally, a subgroup of the form �g� for g ∈ G is called a cyclic
subgroup of G.

If S ⊆ G, the subgroup �S� generated by S is in particular always a subgroup

of G. So if g ∈ G, �g� is always a subgroup of G; indeed, it is the smallest
subgroup of G containing g.
Examples 3.53.

(a) The integers modulo n under addition, (Zn , +), is a cyclic group.

(b) The integers under addition, Z, is a cyclic group; a generator is 1.

(c) The subgroup {1, −1} of Q∗ is a cyclic subgroup; observe that −1 is a

generator but 1 is not.

(d) The subgroup {2i | i ∈ Z} of Q∗ is cyclic. The set of elements of this

subgroup is {. . . , 18 , 14 , 12 , 1, 2, 4, 8, . . . }. Here both 2 and 12 generate the
subgroup; no other elements do.

44
 (e) The subset C12 of C∗ comprising the roots of x12 −1 is a cyclic subgroup.
More explicitly, C12 = {e2kπi/12 | 0 ≤ k < 12}. Setting ζ = e2πi/12 , we
observe that ζ, ζ 5 , ζ 7 and ζ 11 each �generate
� C12 but no other powers
of ζ do. For example, ζ 6 = −1 and ζ 6 = {ζ 0 , ζ 6 } = {−1, 1}.

ζ3
ζ4 • ζ2
• •
ζ5 ζ1
• •

ζ6
• •
ζ 0 = 1 = ζ 12

• •
ζ7 ζ 11
• •
ζ8 • ζ 10
ζ9

It is not a coincidence that 1, 5, 7 and 11 are the natural numbers

between 0 and 11 that are coprime to 12. The powers of ζ that generate
C12 are called the primitive 12th roots of unity.

Even without many tools to hand, we can completely describe the struc-
ture of cyclic groups, which we shall do in the following results.
First, we generalise Definition 3.10 by giving a definition of the order of
an element of an arbitrary group.

Definition 3.54. Let G be a group and let g ∈ G. Define o(g), the order
of g, to be |�g�|.

The phrasing of this definition may surprise you a little: you might have
been expecting o(g) to be defined to be the least positive integer n such that
g n = e, as for permutations. However for general groups, it is possible that
no such n exists. Rather than have to separate the two cases, this definition
handles them both.

Proposition 3.55. Let G = �g� be a cyclic group. Then exactly one of the
following two statements is true:

(i) 1 ≤ |G| = n < ∞, and the elements e = g 0 , g 1 , g 2 , . . . , g n−1 are dis-

tinct, G = {g 0 , g 1 , g 2 , . . . , g n−1 } and g n = e. For n ≥ 2, so G is non-
trivial, the generic cyclic group will be denoted Cn , the cyclic group of
order n.

45
 (ii) G is (countably) infinite, in which case the elements g i , i ∈ Z, are
distinct. This group will be denoted C∞ .

Proof. First, suppose that there exist s, t ∈ Z such that s �= t but g s = g t .

Without loss of generality, we may assume that s < t. Then t − s > 0 and
g t−s = g t g −s = g t (g s )−1 = g t (g t )−1 = e, so there exists at least one positive
integer M such that g M = e. Let m be the smallest such.
We claim that |�g�| = m. By Proposition 3.51, every element of �g� is
of the form g p for some p ∈ Z. Then there exist integers q, r such that
p = qm + r and 0 ≤ r < m (the division algorithm for integers). So

g p = g qm+r = (g m )q g r = eq g r = g r

since g m = e. Hence every element of �g� is of the form g r for some 0 ≤ r < m
and �g� = {g r | 0 ≤ r < m}.
To complete the proof that |�g�| = m, it remains to show that the
elements g r , 0 ≤ r < m are distinct. But if we had g s = g t for some
0 ≤ s < t < m, then g t−s = e and t − s < m, contradicting the minimality
of m. So |�g�| = m.
Now suppose �g� is finite. The powers of g cannot all be distinct (else
�g� would be infinite) and g s = g t for some s �= t. Then the above proves
all the claims of (i).
Since we have also shown that if the powers of g are not all distinct then
�g� is finite, we conclude that if (i) does not hold, (ii) must.

Remark 3.56. Implicit in the statement of this Proposition is that there

is essentially only one cyclic group of a given order (n ∈ N or countably
infinite). The proof shows that if we fix the order |�g�| = n then we know
all the elements of �g�, namely the powers of g, and of course we also know
the group operation: g r g s = g r+s (replacing r + s with its remainder after
division by n if necessary).
Given �g� and �h� with |�g�| = |�h�|, there is an obvious bijection between
the elements of �g� and �h�, given by g i �→ hi . This bijection respects the
group operation too. From the point of view of group theory, �g� and �h�
are indistinguishable. The only difference between them are the names we
have chosen for the generators. In the near future, we will make this idea
more formal.

Corollary 3.57. Let �g� be a cyclic group. Then �g� is Abelian.

Proof. We have g r g s = g r+s = g s g r for any r, s.

Now we look at the subgroups of the cyclic groups.

46
Remark 3.58. This is a very special case of a general problem, to understand
the subobjects of an algebraic structure. Without any assumptions, it is
hopeless to expect to classify all subgroups of all groups in any meaningful
way and even with assumptions, it is often extremely hard.
However, there are some particular situations where we can give a de-
scription and the case of cyclic groups is one such. Later we will see a closely
related problem for rings, namely the classification of prime ideals of Z.

Proposition 3.59. Let G = �g� be cyclic.

(i) If H is a subgroup of G then H is cyclic. Indeed H = �g n � for some

n ≥ 0.

(ii) If G is finite of order n, then any subgroup of G is of the form �g m �

where m divides n.

(iii) If G is infinite, then any non-trivial subgroup of G is cyclic of infinite

order.

Proof.
(i) If H = {e}, then H is generated by e and certainly cyclic, so assume
H is non-trivial. Then there exists at least one 0 �= k ∈ Z such that
g k ∈ H; without loss of generality, we may assume k > 0 since g k ∈ H
if and only if g −k ∈ H (by (SG3)). Hence we may let n be the smallest
natural number such that g n ∈ H, so that �g n � ≤ H.
We claim that H = �g n �. Let h ∈ H. Since H ≤ G = �g�, we may
write h = g m for some m. Then there exist integers q, r such that
m = qn + r with 0 ≤ r < n. Thus

g r = g qn+r g −qn = g m (g n )−q = h(g n )−q ∈ H;

so since n is minimal positive and 0 ≤ r < n, we must have r = 0. In

other words, h = g m = g qn = (g n )q ∈ �g n � and so H = �g n �. Thus H
is cyclic and generated by g n , as required.

(ii) If G has finite order n then g n = e by Proposition 3.55(i). By the proof

of (i), H = �g m � where 0 ≤ m < n is the smallest
� positive integer such
that g ∈ H. It only remains to prove that m � n.
m

For this, let n = qm + r where 0 ≤ r < m. Then

g r = g n−qm = g n · g −qm = e · (g m )−q ∈ H.

But m is the smallest

� positive integer such that g m ∈ H, so we must
have r = 0, i.e. m � n.

47
(iii) By (i), if G = �g� is of infinite order and H is a subgroup then H =
�g m � for some m ≥ 0. If G is of infinite order then the powers of g are
distinct, so the same is true for the powers of g m , whence H is infinite
by Proposition 3.55.

We conclude our study of cyclic groups with the determination of the

orders of their elements.
Exercise 3.60. Let G be a group and g ∈ G have o(g) = n. Show that
g s =� g t if and only if s ≡ t mod n. Deduce that g s = e if and only if
o(g) � s.
[Hint: Look at the proofs of the results above.]

Proposition 3.61. Let G = �g� be cyclic.

n
(i) If |G| = n < ∞ and j ∈ Z then o(g j ) = (n,j) where (n, j) = hcf(n, j).

(ii) If |G| = ∞ and j ∈ Z \ {0} then o(g j ) = ∞.

Proof.

(i) We have (g j )k = e if and only if n divides jk. Letting µ = n/ hcf(j, n)

and ν = j/ hcf(j, n), so n = µ hcf(j, n) and j = ν hcf(j, n), we have
that n divides jk if and only if µ hcf(j, n) divides νk hcf(j, n), if and
only if µ divides νk. But µ and ν are coprime by construction, so this
is equivalent to µ divides k. Now µ divides k if and only if k is an
integer multiple of µ = n/ hcf(j, n).
So (g j )k = e if and only if k is a multiple of n/ hcf(j, n), and therefore
the smallest positive integer such that g j to that power is equal to the
identity is n/ hcf(j, n), that is, o(g j ) = n/ hcf(j, n).

(ii) This follows immediately from Proposition 3.59(iii).

As an example in the infinite case, 4Z is a cyclic subgroup of Z (under

addition), generated by 4. No non-zero multiple (the additive equivalent of
powers) of 4 is equal to 0, nor is any multiple of any other non-zero element
of 4Z; they all have infinite order.
In a finite cyclic group G = �g� of order n, we can measure how big a
subgroup H = �g r � is relative to G: by Proposition 3.59(ii) and Propos-
ition 3.61(i), r divides n and o(g r ) = n/ hcf(r, n) = n/r, so H has order
n/r.
In infinite cyclic groups—and in infinite groups more generally—such
calculations are less helpful, since all the numbers involved are “infinity”.
For example, how do we measure how big 4Z is relative to Z? It seems
natural that it is “bigger” than 8Z: they have the same number of elements

48
(countably infinitely many) but 8Z ⊆ 4Z so 8Z is smaller. But how would
we compare 4Z and 7Z?
The answer is that we can indeed make mathematical sense of the natural
claims that 4Z covers a quarter of the elements of Z and 7Z a seventh, but
by counting how many “copies” or “translations” of 4Z we need to cover Z
rather than counting the number of elements.
This leads us to the penultimate topic in our study of the internal struc-
ture of groups, namely cosets.

3.6 Cosets
Throughout this section, we will consider a group G and a subgroup H
of G. As previously, we will mostly write the group operation in G by
concatenation: if g, h ∈ G, gh will denote the element obtained from g and
h (in that order) under the binary operation of the group. The exception
will be in groups where the binary operation is addition (such as (Z, +)),
where we will write g + h.

Definition 3.62. For any g ∈ G, the set gH = {gh | h ∈ H} will be called

the left coset of H in G determined by g.

Example 3.63. In S3 , let

H = �(1 2 3)� = {ι, (1 2 3), (1 3 2)}.

The coset (1 2)H in S3 is

{ (1 2)ι = (1 2),
(1 2)(1 2 3) = (1)(2 3) = (2 3),
(1 2)(1 3 2) = (1 3)(2) = (1 3) }
First of all, we see that gH ⊆ G. Since e ∈ H, we have g = ge ∈ gH.
Indeed, every element of gH is, by definition, a “multiple” of g by an element
of H; the left coset gH is precisely all the elements of G that can be written
as gh for some h ∈ H.
If H = {e, h1 , h2 , . . . , hr } is finite, this is clear: gH = {g, gh1 , gh2 , . . . , ghr }
and by Proposition 3.29(iii) (the cancellation property for groups) ghi = ghj
would imply hi = hj , so the elements g, gh1 , . . . , ghr are distinct. In partic-
ular, |gH| = |H|. It is natural to think of gH as the “translations by g” of
the elements of H.
In fact, the claims of the above paragraph still hold if H is infinite, if we
work a little more abstractly.

Lemma 3.64. Let H be a subgroup of G. For any g ∈ G, there is a bijection

Lg : H → gH given by Lg (h) = gh.

49
Proof. This follows immediately from Corollary 3.30 and Remark 3.31, but
it is more instructive to write out the proof directly, so we shall.
We need to show that the function Lg is injective and surjective. As-
sume first that Lg (h) = Lg (k) for some h, k ∈ H. Then gh = gk and by
Proposition 3.29(iii), h = k so Lg is injective.
It is surjective since for gh ∈ gH, Lg (h) = gh.

So any two cosets of H have the same size. Notice too that eH = H so
H is a coset of itself (the “translation” of H that does nothing).
However it is very much not true that if g1 �= g2 then g1 H �= g2 H;
shortly we will see the correct condition for when two cosets are equal. This
condition will also tell us when a coset can be a subgroup (sneak preview:
only the “trivial” coset eH = H is).
First, let us look at a very important example.
Example 3.65. We consider (Z, +) and its (cyclic) subgroup 4Z. Since we
are working in an additive group, we write the group operation in our coset
notation, so the cosets of 4Z are r + 4Z = {r + m | m ∈ 4Z} = {r + 4n | n ∈
Z}.
Explicitly,

0 + 4Z = {0 + m | m ∈ 4Z} = {m | m ∈ 4Z} = {a | a ≡ 0 mod 4} = 4Z

1 + 4Z = {1 + m | m ∈ 4Z} = {1 + 4n | n ∈ Z} = {a | a ≡ 1 mod 4}
2 + 4Z = {2 + m | m ∈ 4Z} = {2 + 4n | n ∈ Z} = {a | a ≡ 2 mod 4}
3 + 4Z = {3 + m | m ∈ 4Z} = {3 + 4n | n ∈ Z} = {a | a ≡ 3 mod 4}
4 + 4Z = {4 + m | m ∈ 4Z} = {4 + 4n | n ∈ Z} = {a | a ≡ 4 mod 4}
= {a | a ≡ 0 mod 4} = 4Z
5 + 4Z = {5 + m | m ∈ 4Z} = {5 + 4n | n ∈ Z} = {a | a ≡ 5 mod 4}
= {a | a ≡ 1 mod 4} = 1 + 4Z
6 + 4Z = {6 + m | m ∈ 4Z} = {6 + 4n | n ∈ Z} = {a | a ≡ 6 mod 4}
= {a | a ≡ 2 mod 4} = 2 + 4Z
..
.

50
 and

−1 + 4Z = {−1 + m | m ∈ 4Z} = {−1 + 4n | n ∈ Z} = {a | a ≡ −1 mod 4}

= {a | a ≡ 3 mod 4} = 3 + 4Z
−2 + 4Z = {−2 + m | m ∈ 4Z} = {−2 + 4n | n ∈ Z} = {a | a ≡ −2 mod 4}
= {a | a ≡ 2 mod 4} = 2 + 4Z
−3 + 4Z = {−3 + m | m ∈ 4Z} = {−3 + 4n | n ∈ Z} = {a | a ≡ −3 mod 4}
= {a | a ≡ 1 mod 4} = 1 + 4Z
−4 + 4Z = {−4 + m | m ∈ 4Z} = {−4 + 4n | n ∈ Z} = {a | a ≡ −4 mod 4}
= {a | a ≡ 0 mod 4} = 4Z
..
.

3 + 4Z -17 -13 -9 -5 -1 3 7 11 15

2 + 4Z -14 -10 -6 -2 2 6 10 14

1 + 4Z -15 -11 -7 -3 1 5 9 13 17

0 + 4Z -16 -12 -8 -4 0 4 8 12 16

-17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

We see that the left cosets of 4Z are precisely the congruence classes of
integers modulo 4. The congruence class modulo 4 containing 3, for example,
is the set of integers {. . . , −9, −5, −1, 3, 7, 11, . . . }, which is exactly the set
of integers of the form 3 + 4n, namely {3 + 4n | n ∈ Z} = 3 + 4Z. So cosets
in particular describe congruence classes of integers.
Notice that every integer belongs to some left coset and that no integer
belongs to more than one left coset. This is because the congruence classes
modulo n for some fixed n partition Z, since congruence modulo n is an
equivalence relation. The following theorem is a vast generalisation of this
result, which corresponds to the specific group (Z, +).

Theorem 3.66. Let G be a group and let H be a subgroup of G.

(i) The relation RH on G defined by

g RH k ⇐⇒ g −1 k ∈ H

is an equivalence relation.

(ii) The equivalence classes for RH are precisely the left cosets gH, g ∈ G.

51
Proof.

(i) For g ∈ G, g −1 g = e ∈ H by (SG1), so g RH g and RH is reflexive.

Given g, k ∈ G, we have g −1 k ∈ H if and only if (g −1 k)−1 = k −1 g ∈ H,
by (SG3), so g RH k if and only if k RH g, and RH is symmetric.
Finally, given g, k, l ∈ G, if g RH k and k RH l, then g −1 k ∈ H and
k −1 l ∈ H. Then by (SG2) (g −1 k)(k −1 l) = g −1 l ∈ H, so g RH l and
RH is transitive. Hence RH is an equivalence relation.

(ii) For k ∈ G, we have

g RH k ⇐⇒ g −1 k ∈ H
⇐⇒ g −1 k = h for some h ∈ H
⇐⇒ k = gh for some h ∈ H
⇐⇒ k ∈ gH.

That is, the equivalence class containing g is

{k | g RH k} = {k | k ∈ gH} = gH.

Corollary 3.67. For H a subgroup of G, the set of left cosets

LH = {gH | g ∈ G} is a partition of G.

That is, for g1 , g2 ∈ G, g1 H ∩ g2 H is either empty or g1 H �

= g2 H (and the
latter happens if and only if g1−1 g2 ∈ H), and furthermore g∈G gH = G.
Remark 3.68. From the proof of Theorem 3.66(i), we see that (g RH k ⇐⇒
g −1 k ∈ H) is an equivalence relation if and only if H is a subgroup of G.
For we required each of (SG1), (SG2) and (SG3) in the proof.

Corollary 3.69. For a subgroup H of G, we have gH = kH if and only if

g −1 k ∈ H.

Examples 3.70.

(a) In S3 , for
H = �(1 2 3)� = {ι, (1 2 3), (1 3 2)}
we saw that the coset (1 2)H in S3 is

{(1 2), (2 3), (1 3)}.

In this example, H has precisely two cosets, namely H itself and

(1 2)H. We see that every element of S3 (there are 3! = 6) lies in
either H or (1 2)H, and no element belongs to both, in accordance
with the theorem.

52
 Notice too that (1 2) RH (1 3) since

(1 2)−1 (1 3) = (1 2)(1 3) = (1 3 2) ∈ H

and (1 3) ∈ (1 2)H as expected.

(b) Recall that in (Z, +), nZ is a subgroup and its cosets are the sets of
the form r + nZ = {r + mn | m ∈ Z}. The corollary tells us that

r + nZ = s + nZ ⇐⇒ r − s ∈ nZ
�
⇐⇒ n � r − s
⇐⇒ r ≡ s mod n,

in line with our observation that r + nZ is the congruence class of

r modulo n. The underlying equivalence relation RnZ is precisely
congruence modulo n, as discussed previously. We can see this visually
by looking again at our picture of the cosets of 4Z:

3 + 4Z -17 -13 -9 -5 -1 3 7 11 15

2 + 4Z -14 -10 -6 -2 2 6 10 14

1 + 4Z -15 -11 -7 -3 1 5 9 13 17

0 + 4Z -16 -12 -8 -4 0 4 8 12 16

-17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

(c) Let us consider Z2 = Z × Z = {(a, b) | a, b ∈ Z} as an Abelian

group under addition: (a1 , b1 ) + (a2 , b2 ) = (a1 + a2 , b1 + b2 ). Let
(3, 2) ∈ Z2 . Let H = �(3, 2)�, the cyclic subgroup generated by (3, 2).
Geometrically, �(3, 2)� is the intersection of the line y = 23 x with Z2 in
the plane.

53
 We see that (1, 2) + H = (7, 6) + H since (7, 6) − (1, 2) = (6, 4) =
2(3, 2) ∈ H. (In the diagram, H = �(3, 2)� is indicated by red circles
and (1, 2) + H is indicated in blue.)

The number of cosets of a given subgroup H of a group G then provides

some measure of the relative size of H in G.

Definition 3.71. Let G be a group and let H be a subgroup of G. The

index of H in G is defined to be the cardinality of the set of left cosets of H
in G. We denote the index of H in G by |G : H|.

This answers our question about the relative size of 4Z and 7Z in (Z, +).
We see from the example above that 4Z has index 4 and a similar set of
calculations shows that 7Z has index 7. So 4Z and its three translations
1 + 4Z, 2 + 4Z and 3 + 4Z cover Z; 7Z needs seven translations to cover Z.

54
Example 3.72. For n ≥ 2, the set of even permutations An in Sn has index
|Sn : An | = 2. By Theorem 3.25, if π ∈
/ An and τ = (1 2) then

sign(τ π) = sign(τ ) sign(π) = (−1)(−1)

so π = (1 2)(1 2)π ∈ (1 2)An . That is, for any π ∈ Sn , either π ∈ An or

π ∈ (1 2)An and the two cosets An and (1 2)An of An in Sn partition Sn .
Therefore the index of An in Sn is at most 2. But transpositions are odd,
so (1 2) ∈
/ An and so there are at least two cosets of An and the index is
exactly 2.
For n = 1, A1 = S1 = {ι} so the index is 1.
When we think about the set of left cosets of a subgroup, it is very
natural to want a complete list of the cosets, without repetitions. But we
have already seen that a coset of H in G has |H| representatives, so clearly
such a list will not be unique in general. Still, we can simply make some
choice.
Definition 3.73. Let H be a subgroup of a group G. A transversal for H
in G is a set TH ⊆ G such that for every coset gH, there exists a unique
element tg ∈ TH such that gH = tg H.
That is, a transversal is a complete set of representatives of the left cosets
of H in G, such that any two distinct cosets have distinct representatives.
If we are looking at a specific group and subgroup, there might be one
or more “natural” choices of transversal. This happens in particular for the
integers under addition and the subgroup nZ. As we saw for 4Z above, the
cosets correspond to the different integers modulo n, so a natural transversal
for nZ in (Z, +) is
TnZ = {r | 0 ≤ r ≤ n − 1}.
Explicitly, for n = 4, we can choose T4Z = {0, 1, 2, 3} as the set of left cosets
is L4Z = {0 + 4Z, 1 + 4Z, 2 + 4Z, 3 + 4Z}.
Other choices are also perfectly valid: {1, 2, 3, 4}, {0, 5, 18, 1003} and
{−3, −2, −1, 0} are all transversals for 4Z. However, it is natural to choose
{0, 1, 2, 3} as these are the minimal positive representatives, and we will
usually make this choice and take TnZ as above.
Note however that {−2, −1, 0, 1, 2} is another natural choice for a trans-
versal of 5Z, and a similar choice can be made for (2m + 1)Z (i.e. odd n).
Since this works less well for even n, it is less commonly used, until one
wants to consider pZ for p prime, when in all but one case p is odd.
You will probably have noticed that we have been saying “left cosets”
and that the definition of gH is asymmetric. Unsurprisingly, one can make
the corresponding definition of a right coset.
Definition 3.74. For any g ∈ G, the set Hg = {hg | h ∈ H} will be called
the right coset of H in G determined by g.

55
 Analogously to Definition 3.71, define the right index of H in G to be
the number of right cosets of H in G and denote it by |H : G|.
Example 3.75. In S4 , let

K = �(1 2 3 4)� = {ι, (1 2 3 4), (1 3)(2 4), (1 4 3 2)}.

The coset K(1 2) in S4 is

{ ι(1 2) = (1 2),
(1 2 3 4)(1 2) = (1 3 4),
(1 3)(2 4)(1 2) = (1 4 2 3)
(1 4 3 2)(1 2) = (2 4 3) }
Notice that this is not the same example as before: this is an example
in S4 and we have used K for the subgroup to emphasise that this is differ-
ent. The reason is that the previous example was too small to illustrate an
important point about left and right cosets.
Namely, in general the left cosets of a subgroup are not the same as the
right cosets; certain special conditions have to hold for this to be the case.
In the example, the left coset (1 2)K is {(1 2), (2 3 4), (1 3 2 4), (1 4 3)}
which we see is not equal to K(1 2).
So we must be careful to specify whether we mean left cosets or right
cosets. Fortunately, while we need to make a choice, some key properties
are independent of which choice we make. The size of a left or a right coset
is the same: every coset, whether left or right, has the same size as H (there
is a natural right-handed version of Lemma 3.64).
A little more work also shows that the number of left cosets of H in G is
equal to the number of right cosets of H in G; that is, |G : H| = |H : G|. In
fact, this is why we just say “index” and not “left index” or “right index”.
This is justified by the following exercise.
Exercise 3.76. Let H be a subgroup of G. Let LH = {gH | g ∈ G} and
RH = {Hg | g ∈ G} be the sets of left and right cosets of G respectively.
Prove that the function ψ : LH → RH , ψ(gH) = Hg −1 is a bijection.
This function ψ is a function between two sets: in general, the set of
cosets of a subgroup has no algebraic structure.
Remark 3.77. Subgroups such that gH = Hg for all g ∈ G are very im-
portant, so much so that they have a special name. They are called normal
subgroups and we will examine them and their importance in detail in the
following sections.
But, as a glimpse of what is to come, we will see that the set of (left)
cosets of a normal subgroup can be given an algebraic structure. Specifically,

56
the group operation descends7 to a group operation on the set of cosets: if
gH = Hg for all g ∈ G, we can define a binary operation on cosets by

(gH)(kH) = g(Hk)H = g(kH)H = gkH 2 = gkH.

If this seems mysterious, don’t worry about it now8 .

3.7 Lagrange’s theorem

In this section, we meet our first “proper” group-theoretic theorem, which
is important enough to be given a name: Lagrange’s theorem. Sometimes
theorems are significant in part because they are complicated and their
proofs involve new and difficult ideas. But sometimes the significance comes
from their elegance or their status as a gateway to new areas of study;
Lagrange’s theorem is in these latter categories.

Theorem 3.78 (Lagrange’s theorem). Let G be a group and H a subgroup

of G. Then |G| = |G : H||H|.
Proof. Since the cosets of H partition G, every element of G is contained in
exactly one coset. There are |G : H| cosets, each of size |H|, so the result
follows.

Corollary 3.79. For any subgroup H of G, both |H| and |G : H| divide

|G|.

Conversely, if m does not divide |G|, G cannot have a subgroup of order m.

Example 3.80. Recall from Example 3.72 that the set of even permutations
An has index 2 in Sn . From Lagrange’s theorem we deduce that 2|An | =
|Sn | = n! so that |An | = n!/2.
Remark 3.81. Strictly speaking, the above proof only works for finite groups.
However the result is true as stated for infinite groups, provided the claim
is interpreted properly:
Exercise 3.82. Let LH denote the set of left cosets of H in G. Choose TH a
transversal for H, that is, TH ⊆ G such that for all gH ∈ LH , there exists
a unique tg ∈ TH such that gH = tg H (this was Definition 3.73). Then for
all g ∈ G, there exists a unique hg ∈ H such that g = tg hg .
Prove that the function ϕ : G → LH ×H, ϕ(g) = (tg H, hg ) is well-defined
and a bijection.
7
This almost certainly seems an odd choice of word; there are reasons why we use it
and this will become clearer later. For now, think that the original group is “bigger” and
the set of cosets is “smaller” and the operation “goes down” from the former to the latter.
8
There will be plenty of opportunity to do so later.

57
 Our mental picture should be the elements of G laid out in a grid, with
each row being a list of the elements of H or a coset of it, and the columns
indexed by (some fixed choice of) representatives for the cosets � (the30 trans-
�
versal� TH
� ). As a specific example, let us take G = C 30 = g|g =e ,
H = g 5 and TH = {g 0 = e, g, g 2 , g 3 , g 4 }, as follows.

TH

g4H g4 g9 g 14 g 19 g 24 g 29

g3H g3 g8 g 13 g 18 g 23 g 28

g2H g2 g7 g 12 g 17 g 22 g 27

g1H g1 g6 g 11 g 16 g 21 g 26

g0H g0 g5 g 10 g 15 g 20 g 25

We reserve a detailed analysis of this picture until Example 4.69(a).

Notice that this picture suggests something interesting: if |G| cannot
be written as a product in a non-trivial way, we must have either one row
or one column. Remarkably, this visual aid is actually telling us a genuine
result, which we will get to in two steps.
Corollary 3.83.
� Let G be a finite group of order |G| = n. Then for all
�
g ∈ G, o(g) n and g n = e.
Proof. Since o(g) =� |�g�| and �g� is a subgroup of G, by Lagrange’s theorem
we must have o(g) � n. Thus if o(g) = r we may write n = qr for some q ∈ N
and then g n = (g r )q = e.

Corollary 3.84. A group of prime order is cyclic and has no proper non-
trivial subgroups. Any non-identity element generates the group.
Proof. If |G| = p with p prime, then any subgroup must have order either
1 or p. Thus if e �= g ∈ G, then |�g�| = p and �g� = G. Hence G is cyclic
and generated by any non-identity element. Any subgroup of such a group
is either trivial (order 1) or it contains �g� for some g �= e and is equal to G.

Groups with no proper non-trivial normal subgroups are called simple groups.
At the end of our work on groups, we will see in what sense these groups are
“simple” (they can be extremely large and complicated by other measures)
and why they are the building blocks for all groups. Cyclic groups of prime
order are very simple: they have no proper non-trivial subgroups at all!

58
Proposition 3.85. If H, K ≤ G and hcf(|H|, |K|) = 1, then H ∩ K = {e}.
Proof. Let |H ∩ K| = r. Then r must divide both |H| and |K|, as H ∩ K is
a subgroup of both H and K, and so r = 1.

Example 3.86. Consider G = S6 , H = �(1 2)�, K1 = �(1 2 3 4 5)� and

K2 = �(1 2 3 4 5 6)�.
We have hcf(|H|, |K1 |) = hcf(2, 5) = 1 so H ∩ K1 = {ι}. (You might like
to write out the elements of H and K1 to check this explicitly.)
However, note also that hcf(|H|, |K2 |) = hcf(2, 6) = 2 but H ∩ K2 = {ι}.
That is, the condition in Proposition 3.85 is sufficient (if hcf(|H|, |K|) = 1
then H ∩ K = {e}) but not necessary (H and K2 show that we can have
H ∩ K2 trivial without hcf(|H|, |K2 |) necessarily being equal to 1).
The previous result can be used to help compute the order of the product
of two elements in certain cases. Firstly we establish the following lemma.

Lemma 3.87. If g, h ∈ G satisfy gh = hg then ghn = hn g and (gh)n = g n hn

for all n ∈ N.
Proof. We use induction on n: both statements are trivially true for n = 1,
so assume they are true for n = k. Then

ghk+1 = ghhk = hghk = hhk g = hk+1 g

and

(gh)k+1 = (gh)k (gh) = g k hk .gh = g k ghk h = g k+1 hk+1 ,

as required.

Proposition 3.88. Let g, h ∈ G with gh = hg. If o(g) = m and o(h) = n

with hcf(m, n) = 1, then o(gh) = mn.
Proof. Set k = o(gh). Now

(gh)mn = g mn hmn = (g m )n (hn )m = en em = e

�
by Lemma 3.87, so k � mn by Exercise 3.60.
On the other hand, e = (gh)k = g k hk by Lemma 3.87, giving g k = h−k .
This is an element of �g� ∩ �h�, which equals
� {e} by Proposition
� 3.85.
It follows that g k = hk = �e, so m � k and n � k by Exercise 3.60
again. Thus mn = lcm(m, n) � k. Combined with the above we obtain
o(gh) = k = mn as required.

59
Example 3.89. If G = S5 , the elements g = (1 2 3) and h = (4 5) satisfy
gh = hg (they are disjoint cycles), and their orders are 3 and 2, which are
coprime; we have gh = (1 2 3)(4 5) = hg, whose order is 6 since

(gh)1 = (1 2 3)(4 5) �= ι
(gh)2 = g 2 h2 = g 2 = (1 3 2) �= ι
(gh)3 = g 3 h3 = h3 = h = (4 5) �= ι
(gh)4 = g 4 h4 = g = (1 2 3) �= ι
(gh)5 = g 5 h5 = g 2 h = (1 3 2)(4 5) �= ι
(gh)6 = (g 3 )2 (h2 )3 = ι
If gh = hg we say that the elements g and h commute. This condition is
certainly needed for the previous result: if we take G = S3 with g = (1 2 3),
h = (2 3), then o(g) = 3 and o(h) = 2, but gh = (1 2) so o(gh) = 2 �= 2 · 3.
(You can easily check for yourself that gh �= hg.)
This concludes our examination of the “internal” structure of groups.
What we are missing is a way to compare two different groups, by some
suitable functions between them that respect the group structure. This is
the topic of the next chapter.

60
4 Relating different groups
4.1 Homomorphisms
When we have two groups G and H and we want to relate the group struc-
ture on G to that on H, to compare properties between them, the right
mathematical thing to do is to start with a function ϕ : G → H. This gives
us a relationship between the underlying sets but this need not relate the
group structures: we need ϕ to be compatible with the binary operations
on G and H.
This leads us to the following definition.
Definition 4.1. Let (G, ∗) and (H, ◦) be groups. A function ϕ : G → H is
called a group homomorphism if
ϕ(g1 ∗ g2 ) = ϕ(g1 ) ◦ ϕ(g2 )
for all g1 , g2 ∈ G.
Notice in particular that the operation in G, ∗, is being used on the left-hand
side of this equation (on g1 , g2 ∈ G), and that the operation in H, ◦ is being
used on the right-hand side (on ϕ(g1 ), ϕ(g2 ) ∈ H). This is an instance when
it is helpful to write the group operations explicitly.
Remark 4.2. This idea of a structure-preserving function is one you have
seen before, in MATH112 Discrete Mathematics. There we discussed iso-
morphisms of graphs, as bijective functions between the sets of vertices that
preserve the edges.
We will come to isomorphisms of groups shortly, but just for now, ignore
the “bijective” bit and concentrate on the “function preserving structure”
bit. In fact, this same idea is used to define homomorphisms for any algebraic
structure, as we will see again later when we talk more about rings.
First, let us deal with some elementary properties of homomorphisms.
Proposition 4.3. Let ϕ : G → H be a group homomorphism. Then
(i) ϕ(eG ) = eH ;
(ii) ϕ(g −1 ) = ϕ(g)−1 for all g ∈ G;
(iii) ϕ(g n ) = ϕ(g)n for all g ∈ G, n ∈ Z and
(iv) o(ϕ(g)) | o(g) for all g ∈ G.

Proof.

(i) Let ϕ(eG ) = h for some h ∈ H. Then since eG eG = eG ,

h2 = ϕ(eG )ϕ(eG ) = ϕ(eG eG ) = ϕ(eG ) = h = heH

and hence by Proposition 3.29(iii), h = eH .

61
 (ii) Since gg −1 = eG = g −1 g, we have

ϕ(g)ϕ(g −1 ) = ϕ(gg −1 ) = ϕ(eG ) = eH = ϕ(eG ) = ϕ(g −1 g) = ϕ(g −1 )ϕ(g).

By uniqueness of inverses (3.29(ii)) ϕ(g −1 ) = ϕ(g)−1 .

(iii) By (ii), it is enough to show ϕ(g n ) = ϕ(g)n for all g ∈ G, n ∈ N.

We do this by induction and note that the claim is certainly true for
n = 1. Assume the claim for n = k. Then

ϕ(g k+1 ) = ϕ(g k g) = ϕ(g k )ϕ(g) = ϕ(g)k ϕ(g) = ϕ(g k+1 )

and the inductive step is verified.

(iv) We have g o(g) = eG so by (iii),

ϕ(g)o(g) = ϕ(g o(g) ) = ϕ(eG ) = eH .

Hence by Exercise 3.60, o(ϕ(g)) | o(g).

The composition of two functions is again a function and we would expect

that if two composable functions preserved group structure, their composi-
tion would too. Indeed this is the case.

Proposition 4.4. Let ϕ : G → H and σ : H → K be group homomorphisms.

Then σ ◦ ϕ : G → K is a group homomorphism.
Proof. Certainly σ ◦ ϕ is a function with domain G and codomain K. Now
if g1 , g2 ∈ G we have

(σ ◦ ϕ)(g1 g2 ) = σ (ϕ(g1 g2 ))
= σ (ϕ(g1 )ϕ(g2 ))
= σ (ϕ(g1 )) σ (ϕ(g2 ))
= (σ ◦ ϕ)(g1 )(σ ◦ ϕ)(g2 )

so that σ ◦ ϕ satisfies the condition to be a group homomorphism.

Examples 4.5.

(a) The function ϕ : Z → Z, ϕ(n) = 2n for all n ∈ Z is a group homo-

morphism from (Z, +) to itself:

ϕ(m + n) = 2(m + n) = 2m + 2n = ϕ(m) + ϕ(n)

for all m, n ∈ Z. (Note the additive notation.)

62
 (b) The function ϕ : R∗ → R∗ , ϕ(x) = x2 for all x ∈ R∗ = (R \ {0}, ×) is
a group homomorphism since

ϕ(xy) = (xy)2 = x2 y 2 = ϕ(x)ϕ(y)

for all x, y ∈ R∗ .

(c) The function ϕ : R → C∗ , ϕ(x) = e2πix for all x ∈ R is a group

homomorphism from (R, +) to (C \ {0}, ×) since

ϕ(x + y) = e2πi(x+y) = e2πix+2πiy = e2πix e2πiy = ϕ(x)ϕ(y)

for all x, y ∈ R. (Note the two different notations for the group oper-
ations!)

(d) Recall from Remark 3.28 that we have a function ϕ : Sn → GLn (R),
ϕ(σ) = Aσ where �
1 if σ(j) = i
(Aσ )ij =
0 otherwise
is the permutation matrix associated to σ.
Then Aστ = Aσ Aτ , where the left-hand side involves the composition
of permutations and the right-hand side is the usual multiplication of
matrices. (The definition of Aσ , with a 1 when σ(j) = i might seem
to be the “wrong way round”; however if we made the more natural-
looking definition with σ(i) = j, we would have an order reversal in
the products.) It follows that ϕ : Sn → GLn (R), ϕ(σ) = Aσ is a group
homomorphism. For an explicit example, look back to Remark 3.28.

Exercise 4.6. Let (G, ∗) and (H, ◦) be groups with identity elements eG and
eH respectively. Show that the function ϕ : G → H defined by ϕ(g) = eH
for all g ∈ G is a group homomorphism. We call this the trivial group
homomorphism.
Recall that certain sorts of functions are special, namely injective, sur-
jective and bijective functions; also, a function is bijective if and only if
it is invertible. If a group homomorphism has one of these properties, we
(sometimes) use a special name for it. So an injective group homomorphism
is also known as a monomorphism and a surjective group homomorphism is
called an epimorphism.

63
Examples 4.7.

(a) The homomorphism ϕ : Z → Z, ϕ(n) = 2n for all n ∈ Z is injective

but not surjective:

ϕ(m) = ϕ(n) ⇒ 2m = 2n ⇒ m=n

but 3 ∈
/ Im ϕ.

(b) The homomorphism ϕ : R∗ → R∗ , ϕ(x) = x2 for all x ∈ R∗ is not

injective or surjective: ϕ(−1) = 1 = ϕ(1) and −1 ∈
/ Im ϕ.

(c) The homomorphism ϕ : R → C∗ , ϕ(x) = e2πix for all x ∈ R is injective

but not surjective:

ϕ(x) = ϕ(y) ⇒ e2πix = e2πiy ⇒ x=y

and Im ϕ = {z ∈ C | �z� = 1} � C∗ .

(d) The homomorphism ϕ : Sn → GLn (R), ϕ(σ) = Aσ is injective but not

surjective: given Aσ , we can easily recover σ (uniquely) but 2In ∈
/
Im ϕ.
(e) The function Σ : Sn → {1, −1}, Σ(π) = sign(π) for all π ∈ Sn is
a homomorphism by Theorem 3.25, where {−1, 1} is a group under
multiplication.
Then Σ is not injective (there are many different permutations have
sign 1; indeed these are the even permutations belonging to An , of
which there are n!/2) but Σ is surjective, since there exist both even
and odd permutations.
We will not use those names and will just say “injective homomorphism”
or “surjective homomorphism”, as this will help us remember what to check.
We will however use the special name for bijective homomorphisms.
Definition 4.8. A group homomorphism ϕ : G → H that is bijective is
called a group isomorphism.
Note that the identity function IG : G → G, IG (g) = g is a group homo-
morphism9 and is a bijection, so is a group isomorphism. This is no surprise:
every group should have the same group structure as itself!
A group isomorphism is a particular kind of homomorphism—a function
with certain properties. We can use these to talk about how two groups
might be related, as follows.
9
This, along with properties proved above, shows that the collection of all groups
together with all group homomorphisms between them forms a category ([Alu, Chapter I,
§3],[Cam, Chapter 7]).

64
Definition 4.9. We say that two groups G and H are isomorphic if and only
if there exists a group isomorphism ϕ : G → H. If G and H are isomorphic,
we often write G ∼ = H, for short.

Lemma 4.10. Let G and H be groups. Then the following are are equival-
ent:
(i) G and H are isomorphic;

(ii) there exists a bijective homomorphism ϕ : G → H; and

(iii) there exist group homomorphisms ϕ : G → H and ψ : H → G such that

ψ ◦ ϕ = IG and ϕ ◦ ψ = IH .
Proof. Parts (i) and (ii) are equivalent by definition (Definition 4.9). To
see that these are equivalent to (iii), assume ϕ : G → H is a bijective ho-
momorphism. Then there exists an inverse function ϕ−1 : H → G, with
ϕ−1 ◦ ϕ = IG and ϕ ◦ ϕ−1 = IH . So we need to prove that ϕ−1 is a group
homomorphism, for then we may take ψ = ϕ−1 .
Now, for h1 , h2 ∈ H, let g1 , g2 ∈ G be the unique elements such that
ϕ−1 (h1 ) = g1 and ϕ−1 (h2 ) = g2 , or equivalently ϕ(g1 ) = h1 and ϕ(g2 ) = h2 .
Then since ϕ is a homomorphism, ϕ(g1 g2 ) = ϕ(g1 )ϕ(g2 ) = h1 h2 and so

ϕ−1 (h1 h2 ) = g1 g2 = ϕ−1 (h1 )ϕ−1 (h2 ).

Hence ϕ−1 is also a homomorphism, as required.

Conversely, assuming (iii), since a function is bijective if and only if it
is invertible, ϕ is a bijection. (Since inverses are unique, ψ is equal to the
inverse of ϕ.) Since ϕ is a homomorphism by assumption, we have (ii).

The point of this lemma is that, depending on the situation at hand, it

might be easier to check (ii) or (iii). Indeed, we see from a careful examin-
ation of the proof that (iii) could be weakened further to
(iv) there exist functions ϕ : G → H and ψ : H → G such that ψ ◦ ϕ = IG
and ϕ ◦ ψ = IH and either ϕ or ψ is a group homomorphism.
That is, if we know that a homomorphism is invertible, it is not necessary
to check separately that the inverse is also a group homomorphism: this
holds automatically. In particular, the inverse of an isomorphism is an
isomorphism.
Since a group isomorphism ϕ : G → H is a bijection, we must have that
|G| = |H|, i.e. isomorphic groups have the same order. The converse is very
definitely false in general but the contrapositive tells us that if two groups
have different orders, they cannot be isomorphic.
As for graphs, this leads us to seek invariants that help us identify when
two given groups are isomorphic or not. If we can prove that a particular

65
property is preserved under isomorphism, then if one group has the property
but the other does not, they cannot be isomorphic.
Examples of invariants are the order of the group (as above), the set of
natural numbers that are the orders of the elements of the group (see the
next example), being Abelian and being cyclic.
Example 4.11. The groups C6 and S3 are not isomorphic. They have the
same order (|C6 | = 6 = 3! = |S3 |) but C6 is Abelian and S3 is not. Indeed,
C6 is cyclic and S3 is not (if it were, every permutation in S3 would have to
be an r-cycle for some r, but S3 has both transpositions and 3-cycles).
However, sharing some properties is not (usually) enough to prove that
the groups are isomorphic: almost always you will need to find an explicit
isomorphism between them. That said, if two groups share lots of properties,
this might reasonably lead you to conjecture that they are isomorphic; this
would not be a proof, though.
Example 4.12. Recall that the subset C12 = {e2kπi/12 | 0 ≤ k < 12} of C∗ is
a cyclic subgroup, so is a cyclic group in its own right. We may write it as
C12 = �ζ� for ζ = e2πi/12 .
Let G = �g� be the (generic) cyclic group of order 12, as defined in
Proposition 3.55. So G = {e, g, g 2 , . . . , g 11 } with group operation g r g s =
g r+s . We claim that C12 is isomorphic to G.
Let ϕ : C12 → G be defined by ζ k = g k for all 0 ≤ k < 12. Clearly ϕ is a
bijection and

ϕ(ζ k ζ l ) = ϕ(ζ k+l ) = g k+l = g k g l = ϕ(ζ k )ϕ(ζ l )

so ϕ is a homomorphism.
Now consider (Z12 , +), the integers modulo 12, with addition (modulo
12). So Z12 = {�
0, � a + �b = �
� with �
1, . . . , 11} c if and only if a + b ≡ c mod 12.
We claim that this group is also isomorphic to G, and hence is isomorphic
to C12 too.
The function we need is ψ : Z12 → G, ψ(� a) = g a . Then ψ is a bijec-
a b
tion: g = g if and only if a ≡ b mod 12 by Exercise 3.60. It is a group
homomorphism since

a + �b) = g a+b = g a g b = ψ(�

ψ(� a)ψ(�b).

So ψ is an isomorphism.
The observant among you will have noticed that the number 12 is irrel-
evant in this example: everything works for a general n ∈ N. We can extend
it further, as follows, obtaining a more formal version of the discussion in
Remark 3.56.
Proposition 4.13. Let G and H be cyclic groups. Then G is isomorphic
to H if and only if they have the same order, |G| = |H|.

66
Proof. As discussed above, if G is isomorphic to H, then G and H have the
same order.
Now assume that |G| = |H|. Let G = �g� and H = �h� and define a
function ϕ : G → H by ϕ(g r ) = hr for all r ∈ Z. By the same argument as
in the previous example, ϕ is a homomorphism.
That ϕ is a bijection follows from Proposition 3.55.

Note that neither the statement or the proof assume that the cyclic
groups are finite.
As we said before, two groups having the same order is far from being
enough to tell us that the groups are isomorphic. The message we take from
this proposition is that being cyclic is a very strong condition: same order
plus cyclic is enough to give us isomorphic.
Previously we noted that the identity map provides an isomorphism of
a group with itself. In fact, a group can have other self-isomorphisms and
it is often an important question to know how many. In some sense, these
are the symmetries of the group (which itself may be the symmetries of
something!).

Definition 4.14. Let G be a group. A group isomorphism ϕ : G → G is

called an automorphism of G.

Lemma 4.15. Let G be a group. The set of automorphisms of G forms a

group under composition.

Proof. Since the composition of bijections is a bijection and since the com-
position of homomorphisms is a homomorphism (Proposition 4.4), we see
that the composition of two automorphisms of G is again an automorphism
of G.
Composition of functions is associative and the identity function is an
automorphism. The inverse of an isomorphism is also an isomorphism
(Lemma 4.10 and the discussion after it) so in particular this is true for
an automorphism. Hence we have a group.

This justifies the following definition.

Definition 4.16. Let G be a group. The group of automorphisms of G,

Aut(G) = {ϕ : G → G | ϕ is an isomorphism} is called the automorphism
group of G.

Example 4.17. One can show that Aut(Cn ) ∼ = ((Zn )∗ , ×), the group of units
of the ring of integers modulo n. In particular, we have Aut(C2 ) = {e} and
Aut(C3 ) ∼= C2 .
We will not justify these claims here: finding the automorphism group of
even a small group usually requires more sophisticated technology than we

67
currently have to hand. We will simply note that the above has an important
relationship with number theory: n �→ |Aut(Cn )| is called the Euler totient
function10 and is a very important function about which much more is said
in MATH328 Number Theory.
Instead, we will move on and look at some important subsets of a group
associated to homomorphisms.

4.2 Kernels and images

We already said that injective and surjective homomorphisms are particu-
larly important; those that are both are very special, being isomorphisms.
However most homomorphisms are neither injective nor surjective, so we
would like a way to measure how far from being injective or surjective a
given homomorphism is.
This is done by means of the kernel and image of the homomorphism.
The image is just the image of the function: recall that being surjective
precisely means that the image is all of the codomain. So the size of the
image measures how close the map is to being surjective—the larger, the
better.
The kernel is a subset of the domain whose size measures how close to
being injective the map is: if the kernel is as small as possible, then the map
is injective.
You have met both of these ideas before, in Linear Algebra. There
you defined the kernel and the image of a linear transformation and also
had a theorem, the Dimension Theorem, relating their dimensions to the
dimension of the domain vector space. The definitions in this section and
the theorems in those that follow are precisely the group theory analogues
of these; the order of a (sub)group will be our replacement for dimension.
Definition 4.18. Let ϕ : G → H be a group homomorphism. The kernel
of ϕ is defined to be the subset of G given by
Ker ϕ = {g ∈ G | ϕ(g) = eH }.

Definition 4.19. Let ϕ : G → H be a group homomorphism. The image of

ϕ is defined to be the subset of H given by
Im ϕ = {h ∈ H | (∃ g ∈ G)(ϕ(g) = h)}.
Notice that the definition of the kernel would not be possible without the
identity element eH and that the image is exactly the image of ϕ as a
function. Be very clear in your mind that Ker ϕ is a subset of the domain,
G, and Im ϕ is a subset of the codomain, H.
You might find the following picture helpful:

10
Also known as Euler’s phi function, since it is often denoted ϕ(n).

68
 Ker ϕ Im ϕ
eG eH

ϕ
G H

Very shortly we will come to examples but first we will prove some basic
facts about kernels and images that will be very helpful in working out the
examples.

69
Proposition 4.20. Let ϕ : G → H be a group homomorphism.

(i) The kernel of ϕ, Ker ϕ, is a subgroup of G.

(ii) The image of ϕ, Im ϕ, is a subgroup of H.

(iii) The homomorphism ϕ is injective if and only if Ker ϕ = {eG }.

(iv) The homomorphism ϕ is surjective if and only if Im ϕ = H.

(v) The homomorphism ϕ is an isomorphism if and only if Ker ϕ = {eG }

and Im ϕ = H.

Proof.

(i) We use the subgroup test:

(SG1) Since by Proposition 4.3(i) we have ϕ(eG ) = eH , eG ∈ Ker ϕ.

(SG2) Let g1 , g2 ∈ Ker ϕ. Then ϕ(g1 ) = ϕ(g2 ) = eH and since ϕ is
a homomorphism, we have

ϕ(g1 g2 ) = ϕ(g1 )ϕ(g2 ) = eH eH = eH

and hence g1 g2 ∈ Ker ϕ.

(SG3) Let g ∈ Ker ϕ, so ϕ(g) = eH . Now by Proposition 4.3(ii),

ϕ(g −1 ) = ϕ(g)−1 = e−1

H = eH

and so g −1 ∈ Ker ϕ.

(ii) Similarly:

(SG1) By Proposition 4.3(i) we have ϕ(eG ) = eH , so eH ∈ Im ϕ.

(SG2) Let h1 , h2 ∈ Im ϕ so that there exist g1 , g2 ∈ G such that
ϕ(g1 ) = h1 and ϕ(g2 ) = h2 . Then since ϕ is a homomorph-
ism,
ϕ(g1 g2 ) = ϕ(g1 )ϕ(g2 ) = h1 h2
and hence h1 h2 ∈ Im ϕ.
(SG3) Let h ∈ Im ϕ so there exists g ∈ G such that ϕ(g) = h. Then
by Proposition 4.3(ii) again,

ϕ(g −1 ) = ϕ(g)−1 = h−1

and so h−1 ∈ Im ϕ.

70
(iii) Assume that ϕ is injective and let g ∈ Ker ϕ, so ϕ(g) = eH . By (i),
eG ∈ Ker ϕ also. Then ϕ(g) = ϕ(eG ), but ϕ is injective, so we must
have g = eG . That is, if ϕ is injective, Ker ϕ = {eG }.
Conversely, assume that Ker ϕ = {eG } and let g1 , g2 ∈ G be such
that ϕ(g1 ) = ϕ(g2 ). Then since ϕ is a homomorphism, using Proposi-
tion 4.3(ii),

ϕ(g1 ) = ϕ(g2 ) ⇐⇒ ϕ(g1 )ϕ(g2 )−1 = eH

⇐⇒ ϕ(g1 )ϕ(g2−1 ) = eH
⇐⇒ ϕ(g1 g2−1 ) = eH .

But then g1 g2−1 ∈ Ker ϕ = {eG }, so g1 g2−1 = eG . Hence g1 = g2 and

we deduce that ϕ is injective.

(iv) By the definition of surjectivity, the homomorphism ϕ is surjective if

and only if Im ϕ = H.

(v) This follows immediately from the previous two parts, since an iso-
morphism is required to be a bijection, i.e. injective and surjective.

Notice the similarities between the proofs of parts (i) and (ii), and what
ingredients we need for these.
Remark 4.21. Let ϕ : G → H be an injective homomorphism. Then we can
consider the function ψ = ϕ|Im ϕ : G → Im ϕ, the codomain restriction to
Im ϕ of ϕ. (This is the function that takes the same values as ϕ but where
we “throw away” any elements of H not in the image of ϕ, and so just take
Im ϕ as the codomain.) Since Ker ψ = Ker ϕ = {eG } and Im ψ = Im ϕ, ψ is
both injective and surjective. Hence ψ is an isomorphism of G with Im ϕ.
Since Im ϕ is a subgroup of H, in this situation we say that G is iso-
morphic to a subgroup of H.
That is, to show that a group G is isomorphic to a subgroup of a group
H, we show that there exists an injective homomorphism from G to H. If
that homomorphism is also surjective, we have that G is isomorphic to H
itself.
Let us revisit the examples from Examples 4.5 and 4.7.
Examples 4.22.

(a) The homomorphism ϕ : Z → Z, ϕ(n) = 2n for all n ∈ Z is injective so

Ker ϕ = {eZ } = {0}.
The image of ϕ is all integers of the form 2n, so Im ϕ = 2Z.

(b) The homomorphism ϕ : R∗ → R∗ , ϕ(x) = x2 for all x ∈ R∗ is not

injective or surjective. Its kernel is all non-zero real numbers whose

71
 square is equal to 1 (which is the identity element in R∗ = (R\{0}, ×)).
So Ker ϕ = {1, −1}.
The image of ϕ is the set of all non-zero real numbers that are squares;
there is no “nice” description of this, so we simply have

Im ϕ = {y ∈ R∗ | y = x2 for some x ∈ R∗ }.

(c) The homomorphism ϕ : R → C∗ , ϕ(x) = e2πix for all x ∈ R is injective,

so Ker ϕ = {eR } = {0} (0 is the identity in (R, +).
As we saw before, the image of ϕ is

Im ϕ = {z ∈ C | �z� = 1} � C∗ .

(d) The homomorphism ϕ : Sn → GLn (R), ϕ(σ) = Aσ is injective, so

Ker ϕ = {eSn } = {ι}. Again the image of ϕ has no particularly nice
description: it is just

Im ϕ = {A ∈ GLn (R) | A = Aσ for some σ ∈ Sn }.

(e) The homomorphism Σ : Sn → {1, −1}, Σ(π) = sign(π) for all π ∈ Sn

is not injective. Its kernel is

Ker Σ = {σ ∈ Sn | sign(σ) = 1}

since 1 is the identity element in ({1, −1}, ×). By definition, these are
the even permutations, An , so Ker Σ = An . Since there exist both
even and odd permutations, Σ is surjective and Im Σ = {1, −1}.

In the picture above, we saw that given a homomorphism ϕ : G → H,

we can split G up into pieces labelled by the element of H that the elements
of that piece map to under ϕ. The next result tells us that this partition of
G, coming from the equivalence relation Rϕ defined by g1 Rϕ g2 if and only
if ϕ(g1 ) = ϕ(g2 ), is a particularly nice partition: it is the same one as we
obtain from RKer ϕ as described in Theorem 3.66.

Proposition 4.23. Let ϕ : G → H be a group homomorphism and let K =

Ker ϕ. Then for all g ∈ G,

gK = {l ∈ G | g Rϕ l} = {l ∈ G | ϕ(g) = ϕ(l)}.

Hence the function ψ : LK → Im ϕ, ψ(gK) = ϕ(g) is a bijection, where

LK = {gK | g ∈ G} is the set of left cosets of K = Ker ϕ.

72
Proof. Given Theorem 3.66 and the above discussion, it suffices to see that
(similarly to the proof of Proposition 4.20(iii))

g 1 Rϕ g 2 ⇐⇒ ϕ(g1 ) = ϕ(g2 )
⇐⇒ eH = ϕ(g1−1 g2 )
⇐⇒ g1−1 g2 ∈ Ker ϕ = K
⇐⇒ g 1 RK g 2 .

Since by Theorem 3.66(ii), the equivalence classes for RK are precisely the
left cosets of K = Ker ϕ, we have the first assertion.
So g1 K = g2 K if and only if ϕ(g1 ) = ϕ(g2 ), so ψ is well-defined and
injective. Since if h ∈ Im ϕ, there exists g ∈ G such that ϕ(g) = h, we have
ψ(gK) = ϕ(g) = h and ψ is surjective and hence a bijection.

Since the index |G : Ker ϕ| is defined to be the cardinality of the set of

left cosets of Ker ϕ, as an immediate corollary of this result and Lagrange’s
theorem we have:

Corollary 4.24. Let ϕ : G → H be a group homomorphism. We have that

|G : Ker ϕ| = |Im ϕ| and hence |G| = |Ker ϕ||Im ϕ|.

This is the analogous result for groups to the Dimension Theorem for vector
spaces, which asserts that for a linear transformation T : V → W we have
dim V = dim Ker T + dim Im T .
In fact, both of these are “numerical shadows” of stronger statements:
shortly we will introduce quotient groups and prove the stronger version for
groups.
Since Ker ϕ ≤ G and Im ϕ ≤ H, we already knew that |Ker ϕ| divides |G|
and |Im ϕ| divides |H|, by Lagrange’s theorem. However this corollary tells
us that |Im ϕ| also divides |G|, which can be useful to know, in applications
such as the following.
Example 4.25. Let G be a group of order 16 and H a group of order 9.
Then the only homomorphism ϕ : G → H is the trivial homomorphism with
ϕ(g) = eH for all g ∈ G. For if ϕ : G → H is a homomorphism, Im ϕ must
divide |G| and |H|, but hcf(16, 9) = 1 so |Im ϕ| = 1 and so Im ϕ = {eH }.
To conclude this section, we recall that properties such as being cyclic
are preserved by isomorphisms and also that (as in Proposition 4.13) if we
have a cyclic group around, we can often say much more than for arbitrary
groups.
In this spirit, we have the following proposition, which tells us that being
cyclic is preserved by surjective homomorphisms (not just isomorphisms,
which are also required to be injective).

73
Proposition 4.26. Let G = �g� be a cyclic group and let H be any group.
If ϕ : G → H is a homomorphism, then Im ϕ is cyclic.
Proof. For any h ∈ Im ϕ, there exists r ∈ Z such that g r ∈ G = �g� and
ϕ(g r ) = h. But by Proposition 4.3(iii), ϕ(g r ) = ϕ(g)r so h ∈ �ϕ(g)� and
hence Im ϕ ⊆ �ϕ(g)�. Since ϕ(g) ∈ Im ϕ, �ϕ(g)� ⊆ Im ϕ and hence Im ϕ =
�ϕ(g)� is cyclic, generated by ϕ(g).

Note again that we make no distinction between finite and infinite cyclic
groups, although if G = �g� is a finite cyclic group, Im ϕ must be finite also.
However, we can certainly have G infinite cyclic and Im ϕ finite, as we can
see from the case G = Z and ϕ : Z → Zn , ϕ(m) = m mod n (which we will
study in more detail soon).

4.3 Cayley’s theorem

Although the philosophy of this course is to try to work as abstractly as pos-
sible, sometimes one needs to calculate in particular explicit examples. This
might be done by hand or, more commonly now, by a computer package.
But in order to do so, we need a “nice” way to represent the elements of our
group, to store them and to calculate the result of the binary operation on
them. In this section, we will see two theoretical results that can be used to
provide such representations. The first is called Cayley’s theorem.

Definition 4.27. Let X be a set. The symmetric group on X is defined to

be the group SX = (Bij(X), ◦) of bijections between X and itself.
Theorem 4.28 (Cayley’s theorem). Let G be a group. There is an injective
homomorphism λ : G → SG .
Proof. For each a ∈ G, recall that we defined La : G → G by La (g) = ag for
all g ∈ G. In Corollary 3.30 and Remark 3.31, we saw that La is a bijection,
so La ∈ SG = Bij(G). So we can define a function λ : G → SG by λ(a) = La .

We claim that λ is an injective homomorphism. Let a, b ∈ G. Then

λ(ab) = Lab . Now for all g ∈ G,

Lab (g) = (ab)g = a(bg) = La (Lb (g)) = (La ◦ Lb )(g)

so Lab = La ◦ Lb and so λ(ab) = λ(a) ◦ λ(b) and λ is a homomorphism.

Now assume we have a, b ∈ G such that λ(a) = λ(b). Then La = Lb , as
functions from G to itself, so in particular La (eG ) = Lb (eG ). But then

a = aeG = La (eG ) = Lb (eG ) = beG = b

so a = b and λ is injective.

74
Corollary 4.29. Every group G is isomorphic to a subgroup of SG .
In particular, every finite group G of order n is isomorphic to a subgroup
of the symmetric group of degree n, Sn .

Proof. This is immediate from the Proposition and Remark 4.21.

Corollary 4.30. Let n ∈ N. There are finitely many non-isomorphic groups

of order n.

Proof. Since Sn is a finite set (of cardinality n!), it has only finitely many
subsets (|P(Sn )| = 2n! ). Therefore Sn has only finitely many subgroups. By
Corollary 4.29, every group of order n is isomorphic to a subgroup of Sn , so
there can be only finitely many groups of order n up to isomorphism.

The phrase “up to isomorphism” at the end here is very important: there
are infinitely many groups with one element, for example, but they are all
isomorphic11 .
Actually, the estimate in this proof is not very good: 2n! is much larger
than the number of subgroups of Sn . This is not surprising: there are lots
of subsets of Sn that are not subgroups (all those that do not contain ι, for
starters).
We do not know exact formulæ for the number of subgroups of Sn , either
up to equality or up to isomorphism, though we do have the following:

• the number of subgroups12 of Sn (not up to isomorphism) is bounded

2 1 2
below by 2n /16 and bounded above by 24( 6 +o(1))n .
2
Since 2n /16 � n! � en ln n−n , this number grows very quickly (super-
exponentially) in n.

• the number of subgroups of Sn up to isomorphism13 is much smaller:

the first thirteen values of this sequence are 1, 2, 4, 9, 16, 29, 55, 137,
241, 453, 894, 2,065 and 3,845; for comparison, the thirteenth value of
the previous sequence is 10,594,925,360.

Cayley’s theorem uses the bijection La : G → G given by La (g) = ag

and the essence of that theorem is that we can think of La as a permutation
of the elements of G. When G is a finite group, we can think of La as a
permutation in two-line notation, such as in the following example.

11
For each x ∈ R, let Ex = ({x}, ∗) be the group with binary operation x ∗ x = x (note
that ∗ is neither + or ×!). Then if x �= y, Ex �= Ey but Ex ∼ = Ey for all x, y ∈ R via
x �→ y.
12
Online Encyclopedia of Integer Sequences, A005432, https://oeis.org/A005432.
13
Online Encyclopedia of Integer Sequences, A174511, https://oeis.org/A174511.

75
Example 4.31. Let G = C6 , generated by g such that g 6 = e. Let a = g 2 .
Then � �
e g g2 g3 g4 g5
La = .
g2 g3 g4 g5 e g
We could then put all of the La ’s together, into a table:
Example 4.32. For G = C6 as above, we have

e g g2 g3 g4 g5
e e g g2 g3 g4 g5
g g g2 g3 g4 g5 e
g2 g2 g3 g4 g5 e g
g3 g3 g4 g5 e g g2
g4 g4 g5 e g g2 g3
g5 g5 e g g2 g3 g4

A table such as this is known as a Cayley table. Although they were used
as ways of explicitly describing groups around the time that group theory
was being developed, the more abstract approaches we have been taking
soon meant that Cayley tables fell out of fashion (except in primary school
teaching, where they are better known as addition and multiplication tables,
or “times tables”). For while Cayley tables are very explicit, it is not hard
to see that for groups larger than about ten elements, they quickly become
unwieldy. Even checking basic properties such as associativity is not always
straightforward, for example.
Cayley’s theorem also gives us the first of our methods for calculating in
groups. Namely, we can take our group and represent it as a group of per-
mutations: G is isomorphic to its image under the injective homomorphism
λ : G → SG . Since computers can compose and invert permutations quickly,
this is effective, even for very large finite groups.
From a theoretical point of view, Cayley’s theorem tells us that per-
mutation representations exist, that is that every group can be thought of
as acting on a set S of cardinality |G| by permuting S. Other permuta-
tion representations might exist too, i.e. G might act on other sets. Group
actions will be studied in more detail in MATH321 Groups and Symmetry.
We can also join two previous results together: Cayley’s theorem and the
homomorphism from the symmetric group to the general linear group given
by taking permutation matrices. This gives us, for free, a matrix version of
Cayley’s theorem.

Theorem 4.33. Let G be a finite group. There is an injective homomorph-

ism µ : G → GL|G| (R).

76
Proof. Firstly, by Cayley’s theorem, there is an injective homomorphism
λ : G → SG . For G finite, there is an isomorphism κ : SG → S|G| so that
composing these, we have an injective homomorphism λ� = κ ◦ λ : G → S|G| .
(Remember from MATH112 Discrete Mathematics that the composition of
two injective functions is again injective.)
Recall from Example 4.5(d) and Example 4.7(d) that there is an injective
homomorphism ϕ : Sn → GLn (R), given by ϕ(σ) = Aσ , the permutation
matrix associated to σ.
Setting n = |G|, we obtain ϕ ◦ λ� : G → GL|G| (R) an injective homo-
morphism as required.

This gives us another approach to computing in large groups: represent

the group elements as matrices, and compute with these. While you might
think that this is not very effective (as matrix multiplication of large matrices
by hand is not much fun), in fact it is, as computers are built precisely to
be extremely good at (integer) addition and multiplication. Again, from a
theoretical perspective, this result tells us that matrix representations exist;
this is explored further in MATH325/425 Representation Theory of Finite
Groups.

4.4 Groups of small order

From all the results we have at our disposal now, we can determine, up to
isomorphism, all of the groups of some given (small) orders. For us, “small”
will mean up to ten, not because one cannot go further but because it is not
especially instructive to go much further.
Actually, the results we already have deal with infinite families of groups,
such as those of prime order, and in fact this suggests a reasonable measure
of the complexity of a group: the number (including multiplicity) of prime
factors in its order. This is why the behaviour can change radically from
order n to order n + 1. For example, order 16 = 24 is quite challenging but
order 17 is not.

Lemma 4.34. There is a unique group of order 1, up to isomorphism.

Proof. Let E1 = ({e1 }, ∗1 ) and E2 = ({e2 }, ∗2 ) be groups of order 1. Then
e1 (respectively e2 ) must be the identity element of E1 (respectively E2 ) and
we must have e1 ∗1 e1 = e1 (respectively e2 ∗2 e2 = e2 ).
Then ϕ : {e1 } → {e2 }, ϕ(e1 ) = e2 is easily seen to be a bijective homo-
morphism and hence an isomorphism.

Proposition 4.35. Let p be a prime. There is a unique group of order p,

up to isomorphism.

77
Proof. Let G and H be groups of order p. By Corollary 3.84, G and H are
cyclic groups (of the same order). Then by Proposition 4.13, G and H are
isomorphic.

This deals with orders 1, 2, 3, 5 and 7: up to isomorphism, there is a

unique group having these orders.
For any order, there is a cyclic group of that order; the above asserts
that if the order is prime, this is all that can happen.
The next most complicated groups are those of order p2 and pq, for p
and q prime. There are some nice general results about these groups but as
they would take us a bit too far afield, we will content ourselves with some
more direct arguments.
In fact, all but one of the non-cyclic groups we will find on our list belong
to a family of groups called the dihedral groups.

Definition 4.36. Let Pn denote a regular n-gon in R2 . The dihedral group

of order 2n, D2n , is the group of isometries of R2 preserving Pn .

Here, “isometry” means a bijection of R2 with itself that preserves all

lengths between pairs of points and preserves all angles. The set of all
isometries of R2 forms a group under composition, with the identity being
the isometry that fixes every point, and the dihedral group preserving Pn
is a subgroup of this (the composition of any two isometries preserving Pn
also preserves Pn and similarly for inverses).
The isometries of R2 can be classified into four types: rotations, reflec-
tions, translations and glide reflections. However, non-identity translations
and glide reflections do not preserve Pn (they move its centre, for example)
so do not appear in the dihedral groups, which therefore consist of rotations
and reflections.
In fact, there is one natural rotation to consider: the anti-clockwise
rotation α by 2π/n about the centre of the n-gon. Taking α and any one of
the reflections preserving Pn , β say, and composing these, we may make any
element of the dihedral group. In line with our terminology for cyclic groups,
we say that these generate the dihedral group: the smallest subgroup of the
dihedral group containing {α, β} is the whole group.

78
 B A C

A α α C α α B

C B A

β β β

C B A

β β β
A α2 α2 C α2 α2 B

B A C

Indeed a little more work shows the following.

Lemma 4.37. Let D2n be the dihedral group of order 2n. Then D2n is
generated by two elements α and β subject to the following relations:

• αn = e,

• β 2 = e, and

• αi β = βα−i for all i ∈ Z.

It follows that D2n = {αi β j | 0 ≤ i ≤ n − 1, 0 ≤ j ≤ 1} and hence that

D2n has 2n elements.

We are quietly overlooking a few minor issues, such as when we say “the”
dihedral group. It is not hard to show that the dihedral group is unique to
up to isomorphism, with suitable technical preparation, but we will omit
this. We have also not carefully defined what we mean by relations but we
need not worry about this either.
Notice that the dihedral groups are not Abelian in general: α and β do
not commute (unless n = 2; see below), since αβ = βα−1 �= βα.
The “base case” D4 needs a little extra comment, as the “regular 2-gon”
does not quite make sense. Rather, D4 is the group of isometries of R2

79
preserving a rectangle. (A rectangle is less symmetric than a square (P4 ),
which is why it has a smaller symmetry group.)
Looking at the explicit description above, D4 = {e, α, β, αβ = βα} and
in fact D4 ∼= C2 × C2 , where the first cyclic group of order 2 is {e1 , a} and
the second is {e2 , b} and α �→ (a, e2 ), β �→ (e1 , b).
Here C2 × C2 is the Cartesian product (Proposition 3.33) of two iso-
morphic copies of C2 . This group is also known as the Viergruppe (German
for “four group”) and notice that D4 is Abelian, since C2 is.
In fact, this and the cyclic group of order 4 are the only possibilities.

Proposition 4.38. Every group of order 4 is isomorphic to C4 or D4 .

Proof. Let G be a group of order 4. Then if G contains an element of order
4, G is cyclic, and vice versa. So assume that G is not cyclic. Then since by
Corollary 3.83, the order of every element of G divides |G| = 4 and only the
identity element has order 1, every non-identity element of G has order 2.
Let G = {e, a, b, c}. Then ab �= a (since b �= e) and ab �= b (since a �= e),
so ab = c. Similarly, ba = c. So G = {e, a, b, ab = ba} with a2 = b2 = e and
ab = ba, so that G is isomorphic to D4 .

We may “improve” on the last part of this proof a little, as follows.

Lemma 4.39. Let a, b be two distinct commuting elements of order 2 in a

� b, a2 = b2 = e and ab = ba. Then |G| is divisible by
group G. That is, a =
4.

Proof. Let H be the subgroup of G generated by {a, b}. Since a and b have
order 2 and commute, any product of a and b will be equal to one of the
elements of the set {e, a, b, ab}. (For example, abbaababa = a5 b4 = a.) This
set is also closed under inverses (indeed, every element is self-inverse) so
H = {e, a, b, ab}. Since by Lagrange’s theorem, the order of a subgroup
divides the order of the group, we are done.

We may use this to establish the groups of order 6.

Proposition 4.40. Every group of order 6 is isomorphic to C6 or D6 .

Proof. Let |G| = 6, and assume G is not cyclic; then the possible orders of
elements of G are 1, 2 and 3 (as it cannot have an element of order 6).
Since |G| is even, there must be some element of order 2 (else G would
consist of distinct pairs of elements {x, x−1 } together with e, which makes
an odd number).
If all elements of G had order 1 or 2, then choose x, y ∈ G with x �= y,
x, y �= e. Then xyxy = (xy)2 = e = x2 y 2 = xxyy. Multiplying the equation
xyxy = xxyy on the left by x−1 and on the right by y −1 , we obtain yx = xy,

80
but then G contains two distinct commuting elements of order 2, which
contradicts Lemma 4.39 as 4 � 6.
Thus there must be x, y ∈ G with o(x) = 3, o(y) = 2; then the elements
e, x, x2 , y, yx, yx2
are all distinct, because if y i xj = y k x� then xj−� = y k−i ∈ �x� ∩ �y� = {e},
and so xj = x� and y i = y k .
We now consider the element xy; this must be equal to yx or yx2 . How-
ever, if xy = yx then o(xy) = 6 by Proposition 3.88 but we are assuming
that G is not cyclic, so this is a contradiction.
So we must have xy = yx2 . Thus we see that x and y satisfy the relations
x = y 2 = e and xy = yx−1 , which are the same as the relations in D6 . One
3

may then check that x �→ α, y �→ β gives an isomorphism of G with D6 .

Notice that as S3 is a non-Abelian group of order 3! = 6, it is isomorphic
to D6 (since the property of being Abelian is preserved under isomorphism,
so S3 cannot be isomorphic to C6 ).
For the remaining two orders less than ten, we will state the classification
without proof. (You may find details in [ST, §4.1], noting that the more
advanced technology listed at the start of that section is only necessary
once one gets to order 12; orders 8 and 9 are treated via “elementary”
considerations.)
Proposition 4.41. Every group of order 8 is isomorphic to C2 × C2 × C2 ,
C2 × C4 , C8 , D8 or H∗ , where H∗ is the quaternion group.
The quaternion group is defined to be the group of units of the ring of
quaternions, H. (Recall that the group of units of a ring is the set of elements
that have a multiplicative inverse, with the operation being multiplication.)
The ring of quaternions has several different descriptions but the simplest
for us to state is the following:
�� � �
w z
H= : w, z ∈ C ⊆ M2 (C).
−z w
Its group of units has 8 elements and can be expressed as
� �
H∗ = i, j, k | i2 = j2 = k2 = ijk = 1 .
In particular, H∗ is non-Abelian.
Proposition 4.42. Every group of order 9 is isomorphic to C3 × C3 or
C9 .
Do not be misled by the apparent simplicity of these results: as men-
tioned previously, for order 16 the classification is much more complicated
and there are 14 isomorphism classes14 , for order 32 there are 51 and for
order 64 there are 267.
14
Online Encyclopedia of Integer Sequences, A000001, https://oeis.org/A000001.

81
4.5 Normal subgroups
In Section 3.6, we defined the left and rights cosets of a subgroup H in a
group G:

gH = {gh | h ∈ H}
Hg = {hg | h ∈ H}

We also saw that in general gH �= Hg, but commented that the class of
subgroups for which this is true for all g ∈ G is important. In this section,
we will begin to see why.

Definition 4.43. Let N be a subgroup of a group G. We say that N is a

normal subgroup if gN = N g for all g ∈ G. We write N � G, or N � G if
N is a proper normal subgroup.

Proposition 4.44 (Normal subgroup test). Let N be a subgroup of a group

G. The following are equivalent:

(NSG1) N is a normal subgroup, so gN = N g for all g ∈ G.

(NSG2) N is closed under conjugation: for all g ∈ G and for all n ∈ N ,

g −1 ng ∈ N .
Proof. Assume (NSG1) holds, so gN = N g for all g ∈ G. Let g ∈ G and
n ∈ N . Then ng ∈ N g = gN , so there exists n� ∈ N such that ng = gn� .
Hence g −1 ng = n� ∈ N as required.
Conversely, assume (NSG2) holds. Let g ∈ G and n ∈ N and consider
ng ∈ N g. By (NSG2), g −1 ng ∈ N so there exists n� such that g −1 ng = n� .
But then ng = gn� ∈ gN , so we see that gN ⊆ N g. Since |gN | = |N g| = |N |,
we conclude that gN = N g.

In practice, we usually check (NSG2), taking an arbitrary element g of

G and an arbitrary element n of N and showing that g −1 ng ∈ N .
As part of this proposition, we have introduced an extra piece of ter-
minology: conjugation. Although we will not use this very much in this
course, the theory of finite groups (as studied further in MATH321 Groups
and Symmetry) makes heavy use of conjugation and its properties.

Definition 4.45. Let g, h ∈ G. The conjugate of h by g is the element

g −1 hg.

Proposition 4.44 tells us that there are two characterizations of normal

subgroups. One is that they are the subgroups for which left and right cosets
coincides. The other is that they are the subgroups that are closed under
conjugation by arbitrary elements of the group. (Note that any subgroup is

82
automatically closed under conjugation by elements of the subgroup, since
subgroups are closed under taking inverses and products.)
You need not worry too much about the apparent asymmetry in (NSG2):
notice that gng −1 = (g −1 )−1 n(g −1 ). Since N is normal and g −1 ∈ G and
n ∈ N , we must have gng −1 = (g −1 )−1 n(g −1 ) ∈ N . So it is equally valid to
show gng −1 ∈ N for all g ∈ G and n ∈ N .
Examples 4.46.

(a) The trivial subgroup {e} and G itself are normal subgroups of G.
To see this for {e},

(NSG1) g{e} = {ge | e ∈ {e}} = {g} = {eg | e ∈ {e}} = {e}g

or equivalently,
(NSG2) for all g ∈ G and e ∈ {e}, g −1 eg = g −1 g = e ∈ {e}.
For G itself,

(NSG1) gG = {gh | h ∈ G} = G = {hg | h ∈ G} = Gg

or equivalently,
(NSG2) for all g ∈ G and h ∈ G, g −1 hg ∈ G, since G is a group.

(b) Every subgroup of an Abelian group is normal.

To see this, let N be a subgroup of an Abelian group G and let g ∈ G.

(NSG1) gN = {gn | n ∈ N } = {ng | n ∈ N } = N g

or equivalently,
(NSG2) for all g ∈ G and n ∈ N , g −1 ng = ng −1 g = ne = n ∈ N .
You should look carefully to see where the assumption that G is
Abelian is used (specifically, that every pair of elements of G com-
mute).
� �
(c) In D6 = α, β | α3 = β 2 = e, αβ = βα2 , the subgroup generated by α
is normal.
The elements of D6 are {e, α, α2 , α3 , β, αβ, α2 β} and the elements of
�α� are {e, α, α2 , α3 }. We use (NSG2) and notice that we need only
check g ∈ D6 \ �α� and n ∈ �α�, since if g ∈ �α�, g −1 ng ∈ �α� automat-
ically because �α� is a subgroup. Similarly we do not need to check
n = e, since g −1 eg = e ∈ �α� for any g.

83
 So first let β ∈ D6 \ �α�. We have

β −1 αβ = β(αβ) = β(βα2 ) = β 2 α2 = α2
and

β −1 α2 β = β(α(αβ)) = β(α(βα2 )) = β 2 α4 = α.
A similar check for αβ and α2 β confirms that �α� is normal.
In fact, there is a much quicker, more general and more theoretical
way to prove this, as we will see shortly.

(d) As we saw in Example 3.75, the subgroup

K = �(1 2 3 4)� = {ι, (1 2 3 4), (1 3)(2 4), (1 4 3 2)}

in S4 is not normal.
Then we saw that the left coset (1 2)K is

{(1 2), (2 3 4), (1 3 2 4), (1 4 3)}

but the right coset K(1 2) is

{(1 2), (1 3 4), (1 4 2 3), (2 4 3)}.

Alternatively, we have that

(1 2 3)−1 (1 2 3 4)(1 2 3) = (1 3 2)(1 2 3 4)(1 2 3) = (1 2 4 3) ∈

/ K.

Exercise 4.47. Recall from Exercise 3.44 that the centre of a group G,

Z(G) = {z ∈ G | zg = gz for all g ∈ G},

is an Abelian subgroup of G. Show that Z(G) is a normal subgroup of G.

We see, especially from (c), that checking by hand that a subgroup is
normal can be somewhat tedious. Fortunately, there are some alternative
approaches.

Proposition 4.48. Let G be a group and N a subgroup of index 2. Then

N is a normal subgroup of G.

Proof. Recall from Theorem 3.66 that the cosets of N are the equivalence
classes for the equivalence relation RN defined in that theorem. In particu-
lar, the cosets are a partition of G.
Now since N is a subgroup of index 2, its left cosets must be eN = N
and everything else, i.e. G \ N , so that G = N ∪ (G \ N ). But its right cosets
must also be N e = N and G \ N . So we see that N has the same left and
right cosets and hence is normal, by (NSG1).
84
Examples 4.49.

(a) Recall that |D2n | = 2n and notice that |�α�| = n, so that by Lagrange’s
theorem |D2n : �α�| = 2. Hence �α� is normal, as we saw explicitly for
n = 3 in (c) above.

(b) The alternating group An has index 2 in Sn (Example 3.72) so is

normal.

Unfortunately, index 2 subgroups are relatively rare. On the other hand,

the following result gives a ready supply of normal subgroups.

Proposition 4.50. Let ϕ : G → H be a group homomorphism. Then the

kernel of ϕ, Ker ϕ, is a normal subgroup of G.
Proof. Recall that Ker ϕ = {g ∈ G | ϕ(g) = eH } and that Ker ϕ is a
subgroup of G (Proposition 4.20(i)). To prove that the kernel is normal, we
use (NSG2).
Let g ∈ G and k ∈ Ker ϕ, so ϕ(k) = eH . Then

ϕ(g −1 kg) = ϕ(g −1 )ϕ(k)ϕ(g) = ϕ(g)−1 eH ϕ(g) = eH

and so g −1 kg ∈ Ker ϕ as required. Here, we use that ϕ is a homomorphism

and Proposition 4.3.

Remark 4.51. Although the image Im ϕ of a group homomorphism ϕ : G →

H is a subgroup of H (Proposition 4.20(ii)), it is not guaranteed to be
normal. If ϕ is surjective, Im ϕ = H, which is normal in H (trivially),
but the following gives an example of a homomorphism whose image is not
normal. � �
Let C4 = g | g 4 = e be cyclic of order 4 and let λ : C4 → S4 be the
injective homomorphism in Cayley’s theorem (Theorem 4.28), which is given
by g �→ (1 2 3 4). Then Im λ = �(1 2 3 4)� and we showed in Example 3.75
and Examples 4.46(d) that this subgroup of S4 is not normal.
Example 4.52. The homomorphism Σ : Sn → {1, −1}, Σ(π) = sign(π) for all
π ∈ Sn has kernel

Ker Σ = {σ ∈ Sn | sign(σ) = 1}

and as noted before these are the even permutations, An , so Ker Σ = An

(see Example 4.22(e)). This gives an alternative proof that An is a normal
subgroup of Sn .
Nevertheless, some groups have very few normal subgroups. Recall from
Examples 4.46(a) that every group has at least the trivial subgroup and
itself as normal subgroups, but it is possible that there are no others.

85
Definition 4.53. A group G is said to be simple if it has no non-trivial
proper normal subgroups.

Examples 4.54. The cyclic groups of prime order, Cp , p prime, are simple
(in fact, by Proposition 3.59(ii), they have no non-trivial proper subgroups
at all). The alternating group A5 is also simple and is in fact the smallest
non-Abelian simple group.

4.6 Quotient groups

Shortly we will see that the above result about kernels of homomorphisms
being normal subgroups is only half of the story: in fact, every normal
subgroup is the kernel of some homomorphism. To prove this will take some
more preparation, however. Firstly, we will, as promised, show that if we
have a normal subgroup N of a group G, we can put an algebraic structure
on the set of cosets of N in G. Specifically, the set of cosets inherits a group
structure from G and this new group is called the quotient group.
Although the construction may seem abstract, we will soon see that a
well-known group is most properly understood as a quotient, namely the
integers modulo n.
The ingredients for the construction are:

• a group (G, ∗) (to try to avoid confusion, we will revert to using an

explicit symbol to denote the binary operation in G);

• a normal subgroup N of G; and

• the set of (left) cosets LN = {gN | g ∈ G} of N in G.

Proposition 4.55. Let G be a group and N a normal subgroup of G. Then

• : LN × LN → LN given by gN • hN = (g ∗ h)N defines a binary operation
on LN .

Before we prove this, let us say something about what the issue is here.
It is that the formula we have written down—•(gN, hN ) = (g ∗ h)N —might
not be a well-defined function. That is, for any pair (gN, hN ) there should
exist a unique coset kN in LN such that •(gN, hN ) = kN . However, this
could depend on the choices of representatives g and h for gN and hN
respectively.
Here is an illustration of what can go wrong. Let G = S4 and let
K = �(1 2 3 4)�.

86
 Consider the following cosets of K:

(1 2)K = {(1 2), (2 3 4), (1 3 2 4), (1 4 3)}

(2 3 4)K = {(2 3 4), (1 3 2 4), (1 4 3), (1 2)} = (1 2)K
(1 3)K = {(1 3), (1 2)(3 4), (2 4), (1 4 2)}
(1 3 2)K = {(1 3 2), (3 4), (1 2 4), (1 4)}
((1 2)(1 3)) K = (1 3 2)K
((2 3 4)(1 3)) K = (1 4 3 2)K = K

Then we see that if we tried to define (1 2)K • (1 3)K = ((1 2)(1 3))K,
on the one hand this would equal (1 3 2)K (from the penultimate line).
On the other hand, (1 2)K = (2 3 4)K, so we should get the same
answer if we calculated ((2 3 4)(1 3))K. But from the last line, this is equal
to K which is certainly not equal to (1 3 2)K. So we cannot define the
operation consistently: we get a different answer for the output depending
on which representative of the coset we choose, so this fails the uniqueness
requirement for a function.
The claim of the proposition is that everything is OK if N is a normal
subgroup, as we prove now.
Proof of Proposition 4.55. We need to prove that •(gN, hN ) = (g ∗ h)N is
a function. For each pair (gN, hN ) ∈ LN × LN , we certainly have a coset
(g ∗ h)N ∈ LN , so (as indicated by the counterexample) the issue lies in the
uniqueness.
Assume that g1 N = g2 N and h1 N = h2 N . We need to prove that
(g1 ∗ h1 )N = (g2 ∗ h2 )N . By Corollary 3.69, this is equivalent to saying
that we have that g1−1 ∗ g2 ∈ N and h−1 1 ∗ h2 ∈ N and want to show that
−1
(g1 ∗ h1 ) ∗ (g2 ∗ h2 ) ∈ N .
Now

(g1 ∗ h1 )−1 ∗ (g2 ∗ h2 ) = h−1 −1 −1 −1 −1

1 ∗ g1 ∗ g2 ∗ h2 = h1 ∗ h2 ∗ h2 ∗ g1 ∗ g2 ∗ h2 .

But h2 ∈ G, g1−1 ∗ g2 ∈ N and N is normal, so h−1 −1

2 ∗ (g1 ∗ g2 ) ∗ h2 ∈ N .
Also h−1
1 ∗ h2 ∈ N so

(g1 ∗ h1 )−1 ∗ (g2 ∗ h2 ) = (h−1 −1 −1

1 ∗ h2 ) ∗ (h2 ∗ (g1 ∗ g2 ) ∗ h2 ) ∈ N

as required.

Indeed, a careful examination of this proof and a little more work would
show that • is a binary operation if and only if N is a normal subgroup.
This is how the above counterexample of K = �(1 2 3 4)� in S4 was chosen:
we showed that K is not normal.

87
Definition 4.56. Let (G, ∗) be a group and let N be a normal subgroup
of G. The quotient group of G by N , denoted G/N , is the group (LN , •)
where LN is the set {gN | g ∈ G} of left cosets of N in G and • is the binary
operation gN • hN = (g ∗ h)N .

Strictly speaking, this definition contains some claims we have not veri-
fied, so we had better do so now.

Proposition 4.57. The binary operation gN • hN = (g ∗ h)N defines a

group structure on LN = {gN | g ∈ G}.

Proof. We need to show that • is associative, that we have an identity

element and that we have inverses. As we will see, all of these properties
are inherited from the corresponding properties of ∗, the original operation
on G.
For associativity, let gN, hN, kN ∈ LN . Then

gN • (hN • kN ) = gN • (h ∗ k)N
= (g ∗ (h ∗ k))N
= ((g ∗ h) ∗ k)N
= (g ∗ h)N • kN
= (gN • hN ) • kN,

as required.
There is a natural candidate for the identity element of G/N , namely
eN = N . Since for all g ∈ G we have

gN • eN = (g ∗ e)N = gN = (e ∗ g)N = eN • gN,

we see that eN does indeed satisfy the required properties.

Again, there is a natural candidate for the inverse of gN , namely g −1 N .
For all g ∈ G,

gN • g −1 N = (g ∗ g −1 )N = eN = (g −1 ∗ g)N = g −1 N • gN

so that (by the uniqueness of inverses) g −1 N is the inverse of gN .

From now on, we will use the notation G/N in preference to (LN , •), so
that when we write G/N it is understood that G/N as a set is the set of
left cosets of a normal subgroup N of G and that the binary operation is •.

Corollary 4.58. We have |G/N | = |G|/|N |.

Proof. The group G/N has as its underlying set LN = {gN | g ∈ G}, the set
of left cosets of N in G, so |G/N | = |G : N | by definition. Since Lagrange’s
theorem (3.78) tells us that |G| = |G : N ||N |, the claim follows immediately.

88
Example 4.59. We will now look carefully at the quotient group construction
for an example we discussed earlier (Example 3.65), that of (G, ∗) = (Z, +)
and N = 4Z = {4m | m ∈ Z}. Note that N = 4Z is normal because G is
Abelian (Example 4.46(b)).
We saw that a natural choice of transversal for 4Z in Z is T4Z = {0, 1, 2, 3},
so that we can write

L4Z = {0 + 4Z, 1 + 4Z, 2 + 4Z, 3 + 4Z}.

This gives us the set underlying the quotient group Z/4Z, so the next step
is to understand how the operation • works.
The results above tell us that

(r + 4Z) • (s + 4Z) = (r + s) + 4Z.

That is, when we apply the binary operation to two cosets r +4Z and s+4Z,
represented by r and s respectively, the result is the coset represented by
r + s. Let us compute this explicitly in a Cayley table for •:

• 0 + 4Z 1 + 4Z 2 + 4Z 3 + 4Z
0 + 4Z 0 + 4Z 1 + 4Z 2 + 4Z 3 + 4Z
1 + 4Z 1 + 4Z 2 + 4Z 3 + 4Z 4 + 4Z
2 + 4Z 2 + 4Z 3 + 4Z 4 + 4Z 5 + 4Z
3 + 4Z 3 + 4Z 4 + 4Z 5 + 4Z 6 + 4Z

But 4+4Z, 5+4Z and 6+4Z, while they are valid cosets, are not written
as elements of L4Z using the transversal we chose. Each of them is a coset
in L4Z and is equal to one of 0 + 4Z, 1 + 4Z, 2 + 4Z and 3 + 4Z, but we
should work out which one. (This is so that we can understand • as a binary
operation on L4Z more clearly, by rewriting our cosets to use our preferred
set of representatives, although strictly speaking it is unnecessary.)
Fortunately, we already saw how to tell when two cosets of 4Z are equal
(Example 3.70(b)): r + 4Z = s + 4Z if and only if r ≡ s mod 4. For example,

(2 + 4Z) • (3 + 4Z) = 5 + 4Z = 1 + 4Z

since 5 ≡ 1 mod 4. So we can rewrite the table as follows:

• 0 + 4Z 1 + 4Z 2 + 4Z 3 + 4Z
0 + 4Z 0 + 4Z 1 + 4Z 2 + 4Z 3 + 4Z
1 + 4Z 1 + 4Z 2 + 4Z 3 + 4Z 0 + 4Z
2 + 4Z 2 + 4Z 3 + 4Z 0 + 4Z 1 + 4Z
3 + 4Z 3 + 4Z 0 + 4Z 1 + 4Z 2 + 4Z

This is a complete description of the group structure of Z/4Z = (L4Z , •).

We see that 0 + 4Z is the identity element, that 3 + 4Z is the inverse to
1 + 4Z and so on.

89
 Indeed, if we look carefully, we deduce that Z/4Z is a cyclic group,
generated by 1 + 4Z. For Z/4Z has 4 elements (|L4Z | = |{0, 1, 2, 3}| = 4)
and 1 + 4Z is a generator, since it has order 4:

(1 + 4Z) • (1 + 4Z) = 2 + 4Z (�= 0 + 4Z)

((1 + 4Z) • (1 + 4Z)) • (1 + 4Z) = 3 + 4Z (�= 0 + 4Z)
(((1 + 4Z) • (1 + 4Z)) • (1 + 4Z)) • (1 + 4Z) = 4 + 4Z = 0 + 4Z

So Z4 ∼ = C4 , as any two cyclic groups of the same order are isomorphic

(Proposition 4.13).
Since the operation • is precisely given by (r + 4Z) • (s + 4Z) = t + 4Z
for t ≡ r + s mod 4, • “is” addition modulo 4 and we will denote it by +4 :

(r + 4Z) +4 (s + 4Z) = t + 4Z

for t ≡ r + s mod 4.
Notice that nowhere in this example did we use any special properties
of 4. Indeed the above discussion works perfectly well for any n ∈ Z. The
definition below formalises our somewhat ad-hoc construction of the integers
modulo n, and has the advantage that the integers modulo n are a group
by definition (so we do not need to prove this).

Definition 4.60. Let n ∈ Z. The group of integers modulo n is defined

to be the quotient group (Z/nZ, +n ). We also often denote this group by
(Zn , +n ).

By the same argument given above for n = 4, Zn ∼

= Cn is cyclic of order
n, with one choice of generator being 1 + nZ.
Example 4.61. We give another example, this time involving the dihedral
group D8 . Concretely, this is the group of isometries of the square but here
we will treat it more abstractly, as the group generated by elements α and
β, satisfying the relations given in Lemma 4.37:

• α4 = e,

• β 2 = e, and

• αi β = βα−i for all i ∈ Z.

From this, we see that D8 = {αi β j | 0 ≤ i ≤ 3, 0 ≤ j ≤ 1}.

Now D8 is not Abelian (since αβ = βα−1 = βα3 ) so we need
� � to do a little
work to identify a normal subgroup. We claim that N = α2 = {e, α2 } is
one such (notice that (α2 )2 = e so indeed N has just these two elements).
To see this, we need to know that N is a subgroup, which we get “for free”
as it is is the subgroup generated by something, and that it satisfies (NSG2).

90
Notice that by Lagrange’s theorem, |G : N | = |G|/|N | = 8/2 = 4 �= 2 so
that we cannot simply use Proposition 4.48.
The condition (NSG2) requires that for all g ∈ G and n ∈ N , we have
g −1 ng ∈ N . Now since (gh)−1 n(gh) = h−1 (g −1 ng)h, we see that it is enough
to check (NSG2) for a generating set of G, since then any element of G is
a product of generators and if N is closed under conjugation by generators,
the aforementioned equation shows that it is closed under conjugation by
any product of generators.
In this case, we have α−1 α2 α = α2 and

β −1 α2 β = βα2 β = (βα2 )β = (α−2 β)β = α2 β 2 = α2

and so N is indeed normal.

A little calculating shows that the cosets of N in G can be written as
follows:

eN = N = {e, α2 }
αN = {α, α3
βN = {β, βα2 = α2 β}
αβN = {αβ, α3 β}

so that a transversal is TN = {e, α, β, αβ}.

Let us again work out the Cayley table15 for •:

• eN αN βN αβN
eN eN αN βN αβN
αN αN eN αβN βN
βN βN αβN eN αN
αβN αβN βN αN eN

Here we are using throughout that gN • hN = (gh)N and the relations in

D8 listed above; this time we have simplified as much as possible first time,
to obtain cosets represented by elements of our chosen transversal TN . For
example, αN • αN = α2 N = N = eN since α2 ∈ N .
We see that every element of G/N has order two: gN • gN = eN for all
g. To identify G/N , we could either make some further direct arguments or
do as we shall, and simply appeal to Proposition 4.38: G/N is not
� cyclic
� of
order 4, due to the lack of an element of order 4, so G/N = D8 / α2 ∼= D4 .
One can try to “see” this geometrically: as an isometry of the square,
α2 is rotation by π about the centre. Taking the quotient of D8 by the
subgroup generated by α2 is forcing the square and its rotation by π to be
15
In general, this would be a terrible idea. But, for practical and pedagogical reasons,
we have chosen a very small example, where the use of more advanced technology is not
merited.

91
identified, breaking the fourfold symmetry. Then a square whose opposite
pairs of edges are no longer “the same” is “like” having a rectangle, whose
isometry group is D4 .
There is also a more algebraic approach. For we have from Exercise 4.47
that the centre of a group is always a normal subgroup. Then we may show
that Z(D4n ) = �αn � and then argue as above to see that D4n /Z(D4n ) ∼
= D2n .

4.7 The First Isomorphism Theorem for Groups

In this penultimate part of our study of group theory, we prove another of
the first main theorems of the subject, the First16 Isomorphism Theorem for
Groups. As the name suggests, there are Second and Third Isomorphism
Theorems, though we will not either state or prove these (you will have to
wait until MATH321 Groups and Symmetry).
First we will establish the aforementioned claim that every normal sub-
group is the kernel of a homomorphism. Then in the First Isomorphism
Theorem we will relate kernels, images and quotients.
Lemma 4.62. Let G be a group and N a normal subgroup of G. The
function π : G → G/N defined by π(g) = gN is a group homomorphism.
Proof. As above, let us denote the group operation in G by ∗ and that in
G/N by •. Let g, h ∈ G. Then

π(g ∗ h) = (g ∗ h)N = gN • hN = π(g) • π(h)

so that π is a homomorphism.

Definition 4.63. Let G be a group and N a normal subgroup of G. The

homomorphism π : G → G/N defined by π(g) = gN is called the quotient
homomorphism associated to G and N .
Proposition 4.64. Let G be a group and N a normal subgroup of G. The
quotient homomorphism π : G → G/N is surjective and its kernel is Ker π =
N.

Proof. Let gN ∈ G/N . Then π(g) = gN so π is certainly surjective.

Recall that eN is the identity element in G/N . We have

Ker π = {g ∈ G | π(g) = eN }
= {g ∈ G | gN = eN }
= {g ∈ G | g ∈ N }
= N.

16
In previous years, this has been known as the Fundamental Isomorphism Theorem
and you will see that name in previous exams.
92
 Then we have shown the following.
Corollary 4.65. Let G be a group and N a normal subgroup of G. Then N
is the kernel of a group homomorphism with domain G, namely the quotient
homomorphism π : G → G/N .
Examples 4.66.
(a) Consider G = (Z, +) and N = nZ for n ∈ Z. Recall that N is
normal since G is Abelian. Then nZ is the kernel of the quotient
homomorphism π : Z → Z/nZ given by π(r) = r + nZ.
But the cosets a + nZ are precisely the sets of integers congruent to
a modulo n. So π is the map that takes an integer to its congruence
class modulo n.
This is a homomorphism, as you already know in a disguised fashion:
in the notation of MATH111 Numbers and Relations, writing � a for
a + nZ,
a�+b=� a + �b.
(On the right hand side here, it would be consistent with the above to
write +n to emphasise that this is addition modulo n in Zn .)
(b) We saw in Example 4.49(b) that the subgroup of even permutations,
An , is a normal subgroup of Sn (since it has index 2). So An is the
kernel of π : Sn → Sn /An . Since An has index 2, there are precisely
two cosets in Sn /An and, as in the proof of Proposition 4.48, Sn =
An ∪ Sn \ An . Now (1 2) is odd so (1 2) ∈
/ An and hence a transversal
for An in Sn is {ι, (1 2)}.
We see that the two cosets are precisely the even elements and the odd
elements in Sn . Then via the isomorphism sending the coset ιAn to 1
and (1 2)An to −1, we see that π in this case is essentially the same
as the sign homomorphism Σ : Sn → {1, −1}, Σ(τ ) = sign(τ ) for all
τ ∈ Sn from Example 4.7(e) and Example 4.52.
Next, we make some preparations for the First Isomorphism Theorem.
Most likely, it will not be obvious to you why we do what we are about
to, but do not worry about this. It is part of a much larger picture that
would eventually become clearer after more advanced (undergraduate and
postgraduate) study in algebra; here it can simply be seen as a technical
result that we need to be able to prove what we want.
Theorem 4.67 (Universal property of the quotient group). Let G and H
be groups. Let N be a normal subgroup of G and let π : G → G/N be the
associated quotient homomorphism.
Then for every homomorphism ϕ : G → H such that N ⊆ Ker ϕ, there
exists a unique homomorphism ϕ̄ : G/N → H such that ϕ = ϕ̄ ◦ π.
Furthermore, Ker ϕ̄ = (Ker ϕ)/N and Im ϕ̄ = Im ϕ.

93
 We can illustrate the maps in this theorem via the following diagram,
known as a “commutative diagram” (the “commutative” refers to the fact
that following along the arrows in either possible way gives the same result).
ϕ
G H
π
∃!ϕ̄

G/N

Proof. Let ϕ̄ : G/N → H be a homomorphism such that ϕ = ϕ̄ ◦ π. Then

we have
ϕ̄(gN ) = ϕ̄(π(g)) = (ϕ̄ ◦ π)(g) = ϕ(g).
Since the right-hand side is completely determined by the original homo-
morphism ϕ, it follows that there is at most one such homomorphism. That
is, if such a homomorphism ϕ̄ exists, it is unique.
So let us define ϕ̄ : G/N → H by ϕ̄(gN ) = ϕ(g). We check that it is
well-defined. Let g1 , g2 ∈ G be such that g1 N = g2 N , so that g1−1 g2 ∈ N .
Then
ϕ̄(g2 N ) = ϕ(g2 )
= ϕ(g1 g1−1 g2 )
= ϕ(g1 )ϕ(g1−1 g2 )
= ϕ̄(g1 N )ϕ̄(g1−1 g2 N )
= ϕ̄(g1 N )ϕ̄(eG N )
= ϕ̄(g1 N )ϕ(eG )
= ϕ̄(g1 N )eH
= ϕ̄(g1 N ).

Now ϕ̄ is a group homomorphism because

ϕ̄(gN • hN ) = ϕ̄(ghN ) = ϕ(gh) = ϕ(g)ϕ(h) = ϕ̄(gN )ϕ̄(hN ).

The kernel of ϕ̄ is

Ker ϕ̄ = {gN ∈ G/N | ϕ̄(gN ) = eH }

= {gN ∈ G/N | ϕ(g) = eH }
= {gN ∈ G/N | g ∈ Ker ϕ}
= {kN ∈ (Ker ϕ)/N }
= (Ker ϕ)/N

where we remark that the quotient (Ker ϕ)/N makes sense because N being
normal in G implies that N is normal in any subgroup of G containing N ,
such as Ker ϕ in this case.

94
 The image of ϕ̄ is

Im ϕ̄ = {h ∈ H | (∃ gN ∈ G/N )(ϕ̄(gN ) = h)}

= {h ∈ H | (∃ gN ∈ G/N )(ϕ(g) = h)}
= {h ∈ H | (∃ g ∈ G)(ϕ(g) = h)}
= Im ϕ.

Hence all the claims of the theorem are proved.

A consequence of this theorem is that there is a one-to-one correspond-

ence between homomorphisms ϕ : G → H with N ⊆ Ker ϕ and homomorph-
isms ϕ̄ : G/N → H, the correspondence being given by composing with π.
Since quotient groups can be complicated to understand, this tells us that
to find homomorphisms whose domain is G/N , we can instead look for ho-
momorphisms whose domain is G and just check if N is a subgroup of the
kernel of these. (Depending on the situation, the converse direction can be
very helpful too, of course.)
The above theorem has done much of the heavy lifting to enable us to
state and prove in a nice, clean fashion the First Isomorphism Theorem.
(Arguably, it should be called the “First Isomorphism Corollary” since it
follows essentially immediately from the previous theorem, but often the-
orems are called theorems for their significance rather than their outright
difficulty.)
Theorem 4.68 (First Isomorphism Theorem for Groups). Let G and H be
groups and let ϕ : G → H be a group homomorphism. Then

G/ Ker ϕ ∼
= Im ϕ
Proof. We apply Theorem 4.67 to G, Im ϕ (instead of H), ϕ|Im ϕ : G → Im ϕ
(instead of ϕ) and N = Ker ϕ, all of which satisfy the necessary assumptions.
Then from that theorem, we conclude that there exists a homomorphism
ϕ̄ : G/ Ker ϕ → Im ϕ with Ker ϕ̄ = (Ker ϕ)/ Ker ϕ = {e Ker ϕ} trivial and
Im ϕ̄ = Im ϕ. That is, ϕ̄ is an injective and surjective homomorphism, hence
an isomorphism ϕ̄ : G/ Ker ϕ → Im ϕ.

Recall from Corollary 4.58 that |G/N | = |G|/|N |, so it follows from the
First Isomorphism Theorem that |G| = |Ker ϕ||Im ϕ|. In fact, we already
knew this, and showed it directly (without the need for quotient groups) in
Corollary 4.24. So at first sight, all we have done is provided a much more
complicated proof.
It would be a mistake to think this way, though. Actually, the First
Isomorphism Theorem is a much stronger claim: it asserts that the group
structures of the groups G/ Ker ϕ and Im ϕ are the same, not just the car-
dinalities of their underlying sets. The latter is an immediate consequence of

95
the former, but we already know that there are two non-isomorphic groups
of order 4, so to say that the groups are isomorphic is a much more signific-
ant claim. This is why the proof needed is more complex: stronger claims
need more justification.
When we originally proved Corollary 4.24, we remarked that the formula
|G| = |Ker ϕ||Im ϕ| is a group-theoretic analogue of the Dimension Theorem
from Linear Algebra, dim V = dim Ker T + dim Im T for a linear transform-
ation T : V → W . With just a little more work than was done in MATH220
Linear Algebra II, one can define quotient vector spaces and prove a First
Isomorphism Theorem for Vector Spaces, which in turn implies the Dimen-
sion Theorem. The existence of bases for vector spaces simplifies matters
somewhat, and allows us to think of quotients in terms of complementary
subspaces (orthogonal subspaces, for real and complex vector spaces).
We conclude this section with some examples of applications of the First
Isomorphism Theorem. Largely, they fall into the class “Trying To Avoid
Using Quotients”.
Examples 4.69.

(a) Recall from Proposition 3.59(ii) that if G = �g | g n = e� is a finite

cyclic group of order n, then any subgroup N of G is of the form
N = �g r � where r divides n. Note that N is cyclic of order n/r. Since
cyclic groups are Abelian, N is normal so it is natural to ask whether
we can identify the group G/N explicitly.
Note first that |G/N | = |G|/|N | = n/(n/r) = r. Let s be such that
n = rs, so (g s )r = e. Consider the coset g s N ∈ G/N . Then

(g s N )r = g rs N = eN

but (g s )t N �= eN for 0 ≤ t < r. So g s N has order r in G/N and G/N

is cyclic of order r, generated by g s N .
We have therefore shown that every quotient of a cyclic group is cyclic,
just as every subgroup of a cyclic group is cyclic, as we proved before.
Moreover, Crs /Cs ∼= Cr .

96
 ζ3
ζ4 ζ2

ζ5 ζ1

ζ6
ζ 0 = 1 = ζ 12

ζ7 ζ 11

ζ8 ζ 10
ζ9

Here we have illustrated the case r = 4, s = 3 using the picture of a

cyclic�group
� as roots of unity (12th roots, in this case). The subgroup
N = ζ 4 is represented by the red nodes and triangle (N has order
s = 3, hence is a triangle). The elements of the quotient C4 are the
four triangles and one sees that rotating the red triangle one position
anti-clockwise cycles us through the yellow, green and blue triangles,
returning to the red triangle after four rotations. Another way to
visualise this result (for r = 5, s = 6) was given on page 58, when we
discussed Lagrange’s theorem.
To relate this to the First Isomorphism Theorem, let ϕ : Cn → Cn be
the homomorphism given by ϕ(g i ) = (g i )s . It is straightforward to
check that this is a homomorphism. Then

Ker ϕ = {e, g r , g 2r , g 3r , . . . , g (s−1)r } = �g r � = N

and

Im ϕ = {(g i )s | 0 ≤ i ≤ n − 1} = {(g s )i | 0 ≤ i ≤ n − 1} = �g s � ∼
= Cr .

From the First Isomorphism Theorem, we conclude that Crs /Cs ∼

= Cr .

(b) Consider ϕ : C∗ → R∗ , ϕ(z) = �z�, where F∗ = (F \ {0}, ×) for F =

R, C. This is a homomorphism since

ϕ(z1 z2 ) = �z1 z2 � = �z1 ��z2 � = ϕ(z1 )ϕ(z2 )

for all z1 , z2 ∈ C∗ .
So the First Isomorphism Theorem tells us that C∗ / Ker ϕ ∼
= Im ϕ.
The kernel of ϕ is

Ker ϕ = {z ∈ C∗ | �z� = 1} = {eiθ | 0 ≤ θ < 2π} = S 1

97
 where S 1 is common notation for the circle of radius 1 centered at 0
in the complex plane. The image of ϕ is

Im ϕ = {r ∈ R∗ | (∃ z ∈ C∗ )(�z� = r} = {r ∈ R∗ | r > 0} = R+

since every non-zero positive real number is the modulus of some non-
zero complex number: �r + 0i� = r for r ∈ R+ , for example. It is also
straightforward to check directly that R+ is a subgroup of R∗ (remem-
bering that the operation being used is multiplication), although we
get this automatically from it being the image of a homomorphism.
Hence C∗ /S 1 ∼= R+ . Geometrically, this corresponds to viewing a
point in the complex plane (without the origin) as being determined
completely by an angle θ and a positive real number r: this is precisely
the polar form z = reiθ . Below we give an illustration of this.

(r, θ) = (3, 2π/10)

S 1 = {z ∈ C∗ | �z� = 1}

θ = 2π/10

R+ = {r ∈ R∗ | r > 0} r=3

(c) Recall Example 3.70(c), namely Z2 = Z × Z = {(a, b) | a, b ∈ Z} as an

Abelian group under addition: (a1 , b1 )+(a2 , b2 ) = (a1 +a2 , b1 +b2 ). Let
(3, 2) ∈ Z2 . Since Z2 is Abelian, the cyclic subgroup �(3, 2)� is normal;
we would like to know what the quotient Z2 / �(3, 2)� is. Geometrically,
�(3, 2)� is the intersection of the line y = 23 x with Z2 in the plane, as
we saw before.

98
Let us construct a homomorphism ϕ : Z2 → Z such that �(3, 2)� =
Ker ϕ. (One might reasonably ask why I am “guessing” that I can
put Z as the codomain here. A heuristic would be that Z2 is “2-
dimensional”, �(3, 2)� is a line and hence “1-dimensional”, so that we
would expect the quotient to be “1-dimensional”, so look like Z. Fortu-
nately, this somewhat casual reasoning turns out to be correct. Indeed,
we see that there are |Z| cosets, indicated by the dotted lines parallel
to the line spanned by (3, 2).)
We want ϕ to respect addition in Z2 so it had better be linear in
both arguments: ϕ(a, b) = am + bn for some m, n. Since we want
ϕ(3, 2) = 0, choosing m = 2, n = −3 looks plausible.

99
 We double-check that this does define a homomorphism:

ϕ((a1 , b1 ) + (a2 , b2 )) = ϕ(a1 + a2 , b1 + b2 )

= 2(a1 + a2 ) − 3(b1 + b2 )
= (2a1 − 3b1 ) + (2a2 − 3b2 )
= ϕ(a1 , b1 ) + ϕ(a2 , b2 )
So by the First Isomorphism Theorem, Z2 / Ker ϕ ∼
= Im ϕ and it re-
mains to identify Ker ϕ and Im ϕ.
Although we chose ϕ so that �(3, 2)� ⊆ Ker ϕ, potentially Ker ϕ could
be larger. So let us consider (a, b) ∈ Ker ϕ and see what we can say.
We have ϕ(a, b) = 2a − 3b = 0 so 2a = 3b. The only solutions to this
equation in integers are (3m, 2m) for m ∈ Z, so Ker ϕ = �(3, 2)�, as
we wanted.
Now since hcf(3, 2) = 1, for any n ∈ Z, there exist a, b ∈ Z such
that ϕ(a, b) = 2a − 3b = n (see MATH111 Numbers and Relations,
for example). Geometrically, this corresponds to the observation that
the blue line spanned by (2, −3) (that is, the line orthogonal to that
spanned by (3, 2)) intersects every coset (exactly once). So Im ϕ = Z
and ϕ is surjective.
Hence Z2 / �(3, 2)� ∼
= Z.

4.8 Looking ahead (NOT EXAMINABLE)

In this final section on group theory, before we turn our attention to rings,
we will take a look ahead at the broader landscape of finite group theory
and briefly survey the major results and questions that drove that subject
area forwards over the twentieth century. We will restrict ourselves to finite
groups in part because infinite group theory has a rather different flavour
and requires rather more advanced material to even discuss, and partly
because the story of finite group theory from 1900 to 2000 is very compelling.
Indeed, whole books have been written on the topic. You might like to take
some light reading on your summer holiday in the form of [Ron] or [dS], for
example.
We begin by recalling the classification of groups of order less than 10, up
to isomorphism, in Table 1, which we explained in some detail in Section 4.4.
As we saw, it is not reasonable to try to understand every group or even
to classify the groups of a given order up to isomorphism, so one tries to
get a handle on the finite simple groups, or rather, for which orders there
are simple groups and what those groups are when they do exist. Recall
that a simple group is one with no non-trivial proper normal subgroups
(Definition 4.53).

100
 n Isomorphism types of groups of order n
1 {e}
2 C2
3 C3
4 C4 , C2 × C2
5 C5
6 C6 , D 6
7 C7
8 C8 , C2 × C4 , C2 × C2 × C2 , D 8 , H∗
9 C9 , C3 × C3

Table 1: List of groups of order less than 10, up to isomorphism.

This has been a huge endeavour for group theorists and it is now com-
monly held to have been completed. The whole classification is well beyond
our reach—it would take long enough just to define all of the groups ap-
pearing in the classification—but we can make a start.
One might legitimately ask why we single out the finite simple groups for
classification. This is because if we knew all finite simple groups and if we
knew all the possible ways to “glue” simple groups together to make bigger
groups, then we would know all finite groups. As hard as this approach may
appear, it is considerably easier than the order-by-order approach.

4.8.1 Composition series and the Jordan–Hölder theorem

Informally, a series for a group G is a chain of subgroups of G. In general,
a group has many different series. However, if we ask for certain special
properties of our series then perhaps we will be lucky and find that there is
a property that is invariant among all such series. For composition series,
a theorem called the Jordan–Hölder theorem tells us we are lucky. This
theorem should be thought of as a “unique factorization” theorem.

Definition 4.70. Suppose that G is a group with subgroups

{e} = G0 � G1 � G2 � · · · � Gn = G

for some n ∈ N. We say that such a list of inclusions is a series for G. A

composition series for a group G is a series

{e} � G1 � · · · � Gn−1 � Gn = G

101
with Gi a maximal proper normal subgroup of Gi+1 (normal in Gi+1 but
not necessarily in G). That is, for all i, there does not exist Ni � Gi+1 with
Gi � Ni � Gi+1 .
The factors Gi+1 /Gi are called composition factors.

Composition factors are simple: N is a maximal normal subgroup of G if

and only if the only proper normal subgroup of G/N is N/N , i.e. is trivial.
Examples 4.71.

(a) C12 has several composition series, for example

{e} � C2 � C4 � C12 with factors C2 , C2 and C3

{e} � C2 � C6 � C12 with factors C2 , C3 and C2
{e} � C3 � C6 � C12 with factors C3 , C2 and C2

(b) A4 has three composition series, all of the form

{e} � C2 � C2 × C2 � A4

with factors C2 , C2 and C3 . (There are three choices of subgroups of

C2 × C2 isomorphic to C2 .)
Note that C2 � C2 × C2 and C2 × C2 � A4 but C2 � A4 .

Example (a) is suggestive of a special property of composition series: the

set of composition factors of C12 is the same for all three series. Another
group-theoretic miracle is that this is true not just for cyclic groups, or even
just for finite Abelian groups but in fact for any finite group. This is the
celebrated Jordan–Hölder theorem.

Theorem 4.72 (Jordan–Hölder theorem). Let G be a finite group. Then

all composition series of G have the same composition factors including
multiplicities, up to isomorphism.

Once one has proved this theorem, we can talk about the composition
factors of G, rather than of a particular composition series for G. What
this means from a classification point of view is that it justifies our focus
on simple groups: the composition factors of a group are simple, and every
group is iteratively built up out of these factors.
A vector space analogue says that if V is an n-dimensional vector space
then a descending sequence of subspaces V � V � � V �� � · · · , each of
dimension exactly one less than the last, terminates in the zero subspace
and has length n + 1. The factors are all copies of the field (so here there is
one factor, up to isomorphism, but it occurs with maximum multiplicity).

102
4.8.2 Finite Abelian groups
Abelian groups are considerably easier to understand than groups in general:
here we will sketch the (full!) classification of finite Abelian groups, up to
isomorphism. By “classify”, we mean that we will state a theorem that says
that every finite Abelian group is isomorphic to one of a list of explicitly
described groups. The list will be infinite but this is not surprising—there
is a cyclic group having any order you care to name and all of these are
non-isomorphic. On the other hand, the explicit description is really very
explicit and given an Abelian group, one can find out which group on the
list it is isomorphic to relatively easily.
The strategy is the following:
(a) show that a finite Abelian group A has a Cartesian product decom-
position into subgroups having prime-power order, for different primes
dividing |A|, then

(b) show that each of these subgroups is a direct product of cyclic groups.
Hence A will be a direct product of cyclic groups. We see here a hint of how
important prime numbers are in finite group theory.
We begin with a general definition, even though we will only be interested
in the Abelian case for now.
Definition 4.73. Let G be a group and p a prime. We say G is a p-group
if the order of every element of G is a power of p. That is, for all g ∈ G
o(g) = pk for some k depending on g.
Clearly a non-trivial group G can be a p-group for at most one p but lots
of groups are not p-groups (e.g. C2 × C3 , S3 ). A very useful characterisation
of finite Abelian p-groups is the following.
Lemma 4.74. Let A be a non-trivial finite Abelian group and let p be a
prime. Then A is a p-group if and only if A has p-power order, i.e. |A| = pr
for some r ≥ 1.
Remark 4.75. This result is true for non-Abelian groups but is much harder
to prove in generality (it relies in part on a result known as Cauchy’s theorem
[ST, Theorem 3.33]).
Another name for Abelian p-groups is primary groups, and so the follow-
ing theorem is often called the Primary Decomposition Theorem for finite
Abelian groups.
Theorem 4.76. Let A be a finite Abelian group and let P = {p1 , . . . , pn } be
the set of primes occurring as orders of elements of A. Then the elements
of A of pi -power order form a subgroup A(pi ) and

A∼
= A(p1 ) × A(p2 ) × · · · × A(pn ).

103
 This establishes part (a) of the strategy, the reduction to the case of
finite Abelian p-groups. Now we look for part (b).
Theorem 4.77. A finite Abelian p-group A is isomorphic to a direct product
of cyclic p-groups.
The two theorems of this section then yield the classification of finite
Abelian groups.
Theorem 4.78. Every finite Abelian group A is a direct product of cyclic
groups.
Proof. Theorem 4.76 tells us that a finite Abelian group is the direct product
of its primary components, each of which is a finite Abelian p-group (for some
p) and thus a direct product of cyclic groups by Theorem 4.77.

This theorem may be improved, to give a recipe for identifying a “stand-

ard” decomposition of A into cyclic groups.
Theorem 4.79. Let A be a finite Abelian group. Then

A∼
= C d1 × C d2 × · · · × C dt

such that 1 < di | di+1 for 1 ≤ i < t and t and the sequence (di ) are uniquely
determined by A.
This standard decomposition might not be exactly what you expected:
C2 × C3 would have the associated data t = 1, d1 = 6 (corresponding to its
isomorphic form C6 ), rather than t = 2, (d1 , d2 ) = (2, 3), since 2 � 3.
Example 4.80. If A is Abelian of order 8 then A is isomorphic to exactly
one of C8 , C2 × C4 and C2 × C2 × C2 .
Exercise 4.81. Classify the finite Abelian groups having order 100, order 128
and order 74,088.

4.8.3 Non-Abelian finite simple groups

Rather than work one possible group order at a time, one tries to use the
machinery of finite group theory to strike out many possibilities in one go,
based on the prime factorisation of the order. It is of course sufficient to
restrict our attention to non-Abelian finite simple groups, as we have already
classified the Abelian ones, these being cyclic of prime order.
If the order of our group has a special form, we can often see whether it
can be simple. Results such as the following are typical and the interested
reader can find many of them given in detail in [ST].
Proposition 4.82. There is no non-Abelian simple group of order
(i) 2p with p prime, or

104
 (ii) pq with p �= q primes, or

(iii) pn with p prime, n ≥ 0.

By proving results like this, one obtains theorems such as the following.

Theorem 4.83. There is no non-Abelian simple group G having order

|G| < 60 or 60 < |G| ≤ 167.

You should be wondering about the exclusion of |G| = 60 from the

statement of the above theorem. This is because there is a non-Abelian
simple group of order 60. Similarly, there is a good reason for stopping at
167, but we will have to wait for that too.

Proposition 4.84. The alternating group of degree 5, A5 , is the unique

simple group of order 60, up to isomorphism.

Using this as the base case for an induction, one can then tackle the
inductive step and hence prove the following theorem.

Theorem 4.85. Let n ≥ 5. The alternating group of degree n, An , is a

simple group, of order n!/2.

One may also show that there is a unique non-Abelian simple group of
order 168, and there are none having intermediate order. This simple group
is not A6 , of course: |A6 | = 360. We will show that it is a group called
PSL(2, 7); however, the next simple group after PSL(2, 7) has order 360 and
is indeed A6 .

4.8.4 The projective special linear groups

Next we recall the definitions of the general and special linear groups and
introduce two more types of linear group, the projective general and special
linear groups.
Recall from Exercise 4.47 that the centre of a group G, Z(G), is the
normal subgroup {z ∈ G | zg = gz for all g ∈ G} (the subgroup of elements
of G that commute with every element of G).

Definition 4.86. Let F be a field.

(a) The general linear group of degree n over F , GLn (F ), is the group of
n × n matrices with entries from F and non-zero determinant.

(b) The special linear group of degree n over F , SLn (F ), is the subgroup
of GLn (F ) of matrices of determinant one.

(c) The projective general linear group of degree n over F , PGL(n, F ), is

the quotient group of GLn (F ) by its centre Z(GLn (F )).

105
 (d) The projective special linear group of degree n over F , PSL(n, F ), is
the quotient group of SLn (F ) by its centre Z(SLn (F )).

The definition of the special linear group is concrete enough for us to

know what its elements are, but this is not the case with the projective
versions. So we start by identifying Z(GLn (F )) and Z(SLn (F )).

106
Lemma 4.87. Let F be a field.

(i) Z(GLn (F )) = {λ · In | λ ∈ F \ {0}} where In is the n × n identity

matrix.

(ii) Z(SLn (F )) = Z(GLn (F )) ∩ SLn (F ) = {λ · In | λ ∈ F \ {0}, λn = 1}.

Proof. The first part is linear algebra; the second follows immediately.

Notice that although SLn (F ) is a subgroup of GLn (F ), PSL(n, F ) is not

a subgroup of PGL(n, F ). However, there is an obvious natural isomorphism
of PSL(n, F ) with a normal subgroup of PGL(n, F ).
Of particular importance to us will be the case of F a finite field. In
MATH220 Linear Algebra II, it was mentioned that the integers modulo 2
and the integers modulo 5 are fields. We will see later that Zp for p prime is
a field and it is in fact the case that any finite field has prime power order
and that any two finite fields of the same order are isomorphic. These fields
are denoted Fq .
If |F | = q, we will write GLn (q) instead of GLn (F ), and similarly for the
other matrix groups above. If F is a finite field of order q then GLn (q) is
finite (and hence the other matrix groups are too) and we can give formulæ
for the orders of these groups.

Lemma 4.88. Let F be a field of order q and let t be the number of roots
of the polynomial xn − 1 over F . Then t = (n, q − 1), the highest common
factor of n and q − 1. Furthermore,
�
(i) |GLn (q)| = n−1 n i
i=0 (q − q ),

(ii) |SLn (q)| = |GLn (q)|/(q − 1),

(iii) |PGL(n, q)| = |GLn (q)|/(q − 1) and

1 n(n−1)/2
�n i
(iv) |PSL(n, q)| = |SLn (q)|/t = (n,q−1) q i=2 (q − 1).

These groups are more familiar to you than they might appear at first
sight.
Examples 4.89.

(a) GL1 (F ) ∼
= F ∗ = F \ {0} and so

= SL1 (F ) ∼
PGL(1, F ) ∼ = PSL(1, F ) ∼
= {e}.

(b) Z(SL2 (C)) = {I2 , −I2 }, since the only two square roots of unity in
C (or indeed, any field) are 1 and −1. In fact, one can show that
PSL(2, C) ∼= PGL(2, C). This group is known by a special name: it
is the group of Möbius transformations of the Riemann sphere (the
extended complex plane).

107
 (c) Taking F = F2 to be the field of two elements (i.e. Z2 with addition
and multiplication modulo 2), we have the finite group PSL(2, 2) of
order ((22 − 1)(22 − 21 )/(2 − 1))/(1) = 6. One can easily show that
this group is not Abelian and hence, by our previous classification, is
isomorphic to D6 (∼= S3 ).

(d) With F = F2 again, q − 1 = 1 and (n, q − 1) = (n, 1) = 1, so all four

of GLn (2), SLn (2), PGL(n, 2) and PSL(n, 2) have the same order. In
fact, we see that Z(GLn (2)) = Z(SLn (2)) = {e} so GLn (2) = SLn (2),
PGL(n, 2) = PSL(n, 2) and there is a canonical isomorphism between
these pairs (the isomorphism G/{e} ∼= G).

(e) Taking F = F3 , the field of three elements, |PSL(2, 3)| = 12 and one
may show that PSL(2, 3) is isomorphic to A4 .

The terminology “projective . . . ” refers to the fact that these groups act
on projective spaces. Spaces such as Cn and Rn are called affine spaces and
projective spaces are quotients of these. One major geometric reason for
studying projective space rather than affine space is that projective spaces
are compact topological spaces and are thus better behaved. The general
definition of projective n-space is as follows.

Definition 4.90. Let F be a field. Then the projective n-space over F ,

F Pn , is the quotient of F n+1 \ {0} by the equivalence relation ∼ given by

v ∼ w ⇐⇒ ∃ λ ∈ F \ {0} such that v = λw.

Equivalently, projective n-space is the space of lines in F n+1 passing through

the origin.

Then, just as GLn (F ) is the group of automorphisms of F n , PGL(n +

1, F ) is the group of automorphisms of F Pn . In particular, PGL(2, C) is the
group of automorphisms of CP1 .
The smallest projective plane is F2 P2 and, from the above, its auto-
morphism group is PGL(3, 2) (= GL3 (2) = . . . ). The projective plane F2 P2
is well-known and usually goes by the name of the Fano plane. Its points
are in bijection with the elements of F32 \ {0} so there are seven of them. The
points and lines are as in Figure 1, where we have written the coordinates as
row vectors as this is more convenient for drawing the picture. The action of
GL3 (2) on the points of the Fano plane is simply by matrix multiplication
on the vector indexing the point. We also note that by a “coincidence”,
GL3 (2) ∼ = PSL(2, 7) but the presence of the “7” should now not be such a
surprise!
For us, the primary interest is in the projective special linear groups
over finite fields as finite groups. We note that the above isomorphisms
PSL(2, 2) ∼ = S3 and PSL(2, 3) ∼
= A4 show that these groups are not simple.

108
 (0, 1, 0)

(0, 1, 1) (1, 1, 0)

(1, 1, 1)

(0, 0, 1) (1, 0, 1) (1, 0, 0)

Figure 1: The Fano plane.

However, the following isomorphisms suggest that this is not typical:

PSL(2, 4) ∼
= PSL(2, 5) ∼= A5 , PSL(2, 9) ∼= A6 and PSL(4, 2) ∼ = A8 , so these
groups are all simple.
We now look at the first PSL(2, q) not already discussed: this is PSL(2, 7).
There is no PSL(2, 6), as 6 is not a prime power so no field of order 6 exists.

Theorem 4.91. The group PSL(2, 7) of order 168 is simple.

As noted at the end of the previous section, the list of non-Abelian finite
simple groups begins A5 , PSL(2, 7), A6 (where we prefer to label the first
and third of these by their alternating group name, rather than (one of)
their projective special linear group name(s)). Their orders are 60, 168 and
360 respectively. We know that the alternating groups of degree at least
5 continue to contribute to this sequence, but what about the projective
special linear groups? Do these stop being simple at some point, or are
they all alternating groups? Well, no and no (the latter by looking at their
orders).

Theorem 4.92. For any n ≥ 2 and q a prime power, the group PSL(n, q) is
simple, of order given by the formula in Lemma 4.88, unless (n, q) = (2, 2)
or (2, 3).

We remark that the proof does not proceed in the way we did for
PSL(2, 7). Instead, one identifies certain elements with nice properties and
uses linear algebra and geometrical arguments.

109
 The projective special linear groups are part of a family called the groups
of Lie type and the finite simple members of this family make up the third
and final infinite class of finite simple groups. Groups of Lie type are matrix
groups over finite fields (including orthogonal and unitary groups, among
others) or “twisted” versions of these. We conclude this discussion by looking
at the result we have been aiming towards, the classification of finite simple
groups.

4.8.5 The Classification of Finite Simple Groups

The full classification theorem would take a significant amount of time and
space to state in detail but one gets an idea of the flavour of the theorem
from the following summary.
Theorem 4.93. Every finite simple group is isomorphic to one of the fol-
lowing:
(a) a cyclic group of prime order,

(b) an alternating group An , for n ≥ 5,

(c) a simple group of Lie type over Fq , or

(d) the 26 sporadic simple groups.

The first two parts of this theorem have been covered in detail already
and the third was introduced in the previous section. In particular, the
groups PSL(n, q) are simple groups of Lie type. Some other Lie type groups
can be constructed as matrix groups preserving inner products of various
sorts (symmetric inner products give orthogonal groups, skew-symmetric
ones symplectic groups, Hermitian ones unitary groups).
The 26 sporadic groups are:
• the Mathieu groups, M11 , M12 , M22 , M23 and M24 ;

• the Janko groups, J1 , J2 , J3 and J4 ;

• the Conway groups, Co1 , Co2 and Co3 ;

• the Fischer groups, Fi22 , Fi23 and Fi24 ;

• the Higman–Sims group, HS;

• the McLaughlin group, McL;

• the Held group, He;

• the Rudvalis group, Ru;

• the Suzuki sporadic group, Suz;

110
 • the O’Nan group, O� N;

• the Harada–Norton group, HN;

• the Lyons group, Ly;

• the Thompson group, Th;

• the Baby Monster group, B; and

• the Fischer–Griess Monster group, M.

The group M11 is the smallest sporadic simple group, of order 7920, and the
group M, the largest of the sporadic simple groups, has order

246 · 320 · 59 · 76 · 112 · 133 · 17 · 19 · 23 · 29 · 31 · 41 · 47 · 59 · 71

= 808017424794512875886459904961710757005754368000000000.

The finite simple groups still have much to tell us, particularly with
regard to their representation theory. The study of finite simple groups is
by no means over, and certainly the study of finite groups is not. However,
our study of them is at an end.
For further reading, a good starting place is [Neu]. Recent books on the
subject include [Ron] and [dS], both “popular” books but with the “profes-
sional” reader in mind, also. Those interested in the developmental history
of modern abstract algebra more generally (as sketched in the introduction
to these notes), i.e. encompassing fields, rings, algebraic geometry and more,
should consider [Der].
Now, however, we turn our attention to rings.

111