ABSTRACT

ALGEBRA

MTS 303 & 306

PROF. OLUSEGUN OLOTU

Course Content

 Definition and Examples (including permutation groups).

 Subgroups, Cosets.
 Lagrange’s Theorem & Applications.
 Euclidean Algorithm for Polynomials.
 Normal Subgroups & Quotient Groups.
 Homomorphism, Monomorphism and Isomorphism Theorems.
 Symmetric Groups, Automorphisms.
 Conjugate Classes Normalizers.
 The Sylow’s Theorems.
 Normal and Composition Series.
 The Jordan-Holder Theorem.
 Direct Products.
 Solvable Groups.

2|Page
 CHAPTER ONE

1 BASIC TERMINOLOGIES

1.1 GROUPOID

This is a set on which a binary operation 𝛼 is defined, e.g. (𝑍 + , +), (𝑍, −) not (𝑍 + , −).

1.2 SEMIGROUP

This is a groupoid which is associative, e.g. (𝑍, +), (𝑄\{0},×) not (𝑍, −).

1.3 MONOID

This is a semigroup with an identity element, e.g. (𝑄\{0},×).

1.4 GROUP

This is a monoid with element invertible, e.g. (𝑍, +), (𝑄\{0},×), (𝑄, +), (𝑅, +), (𝑅\{0},×).

1.5 𝐸𝑥𝑎𝑚𝑝𝑙𝑒𝑠

(a) 𝑎 ∗ 𝑏 = 𝑎2 + 𝑏 2 + 1 is a binary operation on 𝑍, 𝑄, 𝑅, 𝑍 +, 𝑄 + or 𝑅 +.

(b) Let 𝑆 be some set of all subsets of 𝑋. For example, let 𝑋 = {1}, then 𝑆 = {∅, [1]}. If 𝑋 =
{1,2}, 𝑆 = {∅, [1], [2], [1,2]}. The operation ∩ (intersection) is a binary operation on 𝑆. Also
∪ (union) is another binary operation on 𝑆 since if 𝐴, 𝐵 ∈ 𝑆, then 𝐴 ∗ 𝐵 = (𝐴 ∩ 𝐵) ∈ 𝑆. Also,
(𝐴 ∪ 𝐵) ∈ 𝑆, for 𝐴, 𝐵 ∈ 𝑆.
𝑎 𝑏
(c) Let 𝑆 = {𝐺𝐿(2, 𝑅)} = {( ) |𝑎, 𝑏, 𝑐, 𝑑 ∈ 𝑅}.
𝑐 𝑑
𝑎 𝑏 𝑒 𝑓 𝑎𝑒 + 𝑏𝑔 𝑎𝑓 + 𝑏ℎ
Define 𝑥 ∗ 𝑦 = ( )∗( )=( ). Then ∗ is a binary operation on
𝑐 𝑑 𝑔 ℎ 𝑐𝑒 + 𝑑𝑔 𝑐𝑓 + 𝑑ℎ
𝑆.
NB: If 𝐴 ∈ 𝑆, then 𝑎𝑑 − 𝑏𝑐 is called the determinant of 𝐴. If 𝑎𝑑 − 𝑏𝑐 ≠ 0, then 𝐴−1 exists.

1.6 𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛

A binary operation on 𝑆 is commutative if 𝑎 ∗ 𝑏 = 𝑏 ∗ 𝑎, ⍱𝑎, 𝑏 ∈ 𝑆.

A binary operation on 𝑆 is associative if (𝑎 ∗ 𝑏) ∗ 𝑐 = 𝑎 ∗ (𝑏 ∗ 𝑐), ⍱𝑎, 𝑏, 𝑐 ∈ 𝑆.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒𝑠

1. (𝑍, −) is neither commutative nor associative [i.e. 1 − 2 ≠ 2 − 1 and (1 − 2) − 3 ≠ 1 −

(2 − 3)].
2. 𝑎 ∗ 𝑏 = 2(𝑎 + 𝑏) on 𝑍 is commutative. However, it is not associative.
3. 𝑎 ∗ 𝑏 = 2𝑎𝑏 on 𝑍 is both commutative and associative.
4. 𝐺𝐿(2, 𝑅) is associative but not commutative.

3|Page
 5. 𝐴∆𝐵 (symmetric difference of A and B) is the set of elements in either 𝐴 or 𝐵 but not both.

Thus, 𝐴∆𝐵 = (𝐴 − 𝐵) ∪ (𝐵\𝐴) or 𝐴∆𝐵 = (𝐴 − 𝐵) − (𝐴 ∩ 𝐵).

Let 𝐴 = {1,2,3,4,5}, 𝐵 = {2,4,6}, then 𝐴∆𝐵 = {1,3,5,6}.

Let 𝐴 = {𝑥|𝑥 ∈ 𝑧 𝑎𝑛𝑑 𝑥 ≥ 0} and 𝐵 = {𝑥|𝑥 ∈ 𝑧 𝑎𝑛𝑑 𝑥 ≤ 0}, then 𝐴∆𝐵 = {𝑥|𝑥 ∈ 𝑧 𝑎𝑛𝑑 𝑥 ≠ 0}.

For any set 𝐴, 𝐴∆𝐴 = ∅, 𝐴∆∅ = 𝐴.

6. Let 𝑆 = {𝑎, 𝑏, 𝑐, 𝑑}, then the following gives a binary operation on 𝑆.

Columns
* a b C d
a a c B d
Rows b c a D b
c b b A c
d d d C a
Is the above binary operation commutative? Is it associative?
7. In each of the following, determine whether or not the given operation ∗ is a binary
operation on the given set 𝑆.
(i) 𝑆 = 𝑍, 𝑎 ∗ 𝑏 = 𝑎 + 𝑏 2.
(ii) 𝑆 = 𝑅, 𝑎 ∗ 𝑏 = 𝑏.
(iii) 𝑆 = {1, −2,3,2, −4}, 𝑎 ∗ 𝑏 = |𝑏|

1.7 FORMAL DEFINITION OF A GROUP

An ordered pair (𝐺,∗), where 𝐺 ≠ ∅ is a group provided the following are satisfied:

(i) ∗ is a binary operation on 𝐺.

(ii) ∗ is associative.
(iii) There is an element (identity) 𝑖 such that 𝑥 ∗ 𝑖 = 𝑖 ∗ 𝑥 = 𝑥, ⍱𝑥 ∈ 𝐺.
(iv) ⍱𝑥 ∈ 𝐺 Ǝ 𝑥̅ such that 𝑥 ∗ 𝑥̅ = 𝑥̅ ∗ 𝑥 = 𝑖.

N.B.:

(a) From (iii), 𝐺 ≠ ∅. Also from (iii), the element 𝑖 is called the identity and it is unique.

(b) From (iv), 𝑥̅ is called the inverse and we shall see later that an element 𝑥 has only one inverse.

(c) We do not assume that ∗ is commutative. However, groups with the special added property that
their operation is commutative are called ABELIAN after the Norwegian mathematician Niels Abel
Henrik (1802-1829) who died of poverty and tuberculosis at the age of 27.

1.8 𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛

Abelian Group

4|Page
A group (𝐺,∗) is abelian when 𝑎 ∗ 𝑏 = 𝑏 ∗ 𝑎, ⍱𝑎, 𝑏 ∈ 𝐺.

Example: (𝑍, +), (𝑄, +), (𝑅, +), (𝑅 + ,×) is group. (𝑄 + ,×) also.

1.9a ADDITIVE GROUP OF INTEGERS MODULO 𝑛

Let 𝑛 be a positive integer and consider the set 𝑍𝑛 = {0,1,2,3, … , 𝑛 − 1}.

Then (𝑍𝑛 ,⊕) where ⊕ is the binary operation on 𝑍 gotten by adding two integers in 𝑍 and then
disregarding the multiples of 𝑛 so as to obtain an answer that is one of 0,1,2,3, … , 𝑛 − 1. There is a
lemma which allows this to work. [A lemma in mathematics is a preliminary or auxiliary result one is
aiming at.]

1.9b 𝐿𝑒𝑚𝑚𝑎

The Division Algorithm

If 𝑎 and 𝑛 are integers and 𝑛 is a positive integer, then Ǝ 𝑞, 𝑟 ∈ 𝑍 such that 𝑎 = 𝑛𝑞 + 𝑟 and 0 ≤ 𝑟 <
𝑛, where 𝑞 = 𝑞𝑢𝑜𝑡𝑖𝑒𝑛𝑡 and 𝑟 = 𝑟𝑒𝑚𝑎𝑖𝑛𝑑𝑒𝑟.

𝑃𝑟𝑜𝑜𝑓

We prove the existence and uniqueness of 𝑞 and 𝑟.

Existence

Let 𝑞𝑛 be the greatest multiple of 𝑛 less than or equal to 𝑎 (𝑞𝑛 ≤ 𝑎).

If 𝑟 = 𝑎 − 𝑞𝑛, then 𝑎 − 𝑞𝑛 ≥ 0 and 𝑎 = 𝑞𝑛 + 𝑟, so all we need to check is that 𝑟 < 𝑛.

But if 𝑟 ≥ 𝑛, then

𝑟 − 𝑛 = 𝑎 − 𝑞𝑛 − 𝑛

𝑟 − 𝑛 = 𝑎 − (𝑞 + 1)𝑛 ≥ 0

𝑟 − 𝑛 = 𝑎 − |𝑞 + 1|𝑛

So, (𝑞 + 1)𝑛 is a multiple of 𝑛 that is less than or equal to 𝑎, contradicting the choice of 𝑞𝑛 as the
greatest such multiple.

Uniqueness

If 𝑞1 𝑛 + 𝑟1 = 𝑞2 𝑛 + 𝑟2 and 0 ≤ 𝑟1 < 𝑛, 0 ≤ 𝑟2 < 𝑛, then 𝑞1 𝑛 − 𝑞2 𝑛 = 𝑟2 − 𝑟1 , so, 𝑟2 − 𝑟1 is a

multiple of 𝑛, but −𝑛 ≤ 𝑟2 − 𝑟1 < 𝑛. So, 𝑟2 − 𝑟1 = 0 and 𝑟2 = 𝑟1 .

Thus, 𝑞1 𝑛 = 𝑞2 𝑛. ∴ 𝑞1 = 𝑞2 ∎

N.B.: Denote by 𝑎̅ the unique 𝑟 guaranteed by the lemma, and call it the remainder of mod𝑛.

5|Page
𝐸𝑥𝑎𝑚𝑝𝑙𝑒𝑠

Let 𝑛 = 7.

If 𝑎 = 16, then 16 = 2.7 + 2, so 𝑞 = 2, 𝑟 = 2 and ̅16

̅̅̅ = 𝑟 = 2.

̅̅̅̅ = 0.
If 𝑎 = 35, then 35 = 5.7 + 0, so 𝑞 = 5, 𝑟 = 0 and 35

With the lemma, we can unambiguously define operation ⊕ on 𝑍 by setting 𝑥 ⊕ 𝑦 = ̅̅̅̅̅̅̅

𝑥 + 𝑦.

For example:

If 𝑛 = 5, then 𝑍𝑛 = 𝑍5 = {0,1,2,3,4,5} and

3 ⊕ 2 = 3̅ + 2̅ = 5̅ = 0

3 ⊕ 4 = 3̅ + 4̅ = 7̅ = 2

2 ⊕ 4 = 2̅ + 4̅ = 6̅ = 1

We claim that (𝑍𝑛 ,⊕) is a group called the additive group of integers mod𝑛.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒

For Klein’s 4-group

Let 𝐺 = 𝐾𝑙𝑒𝑖𝑛′ 𝑠 4 − 𝑔𝑟𝑜𝑢𝑝 = {𝑒, 𝑎, 𝑏, 𝑐}, where 𝑎2 = 𝑏 2 = 𝑐 2 = 𝑒.

Let 𝐻 =< 𝑎 >, then 𝐻 is normal since 𝐺 is abelian and 𝐺\𝐻 is a group with two elements namely
𝐻 =< 𝑎 > and 𝐻𝑏 = 𝐻𝑐.

METABELIAN GROUP

A group 𝐺 is called metabelian if 𝐺 has an abelian normal subgroup 𝐻 such that 𝐺\𝐻 is also abelian.

6|Page
 CHAPTER TWO

ELEMENTARY PROPERTIES OF A GROUP

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 1

If 𝐺 is a group, then 𝑎 ∗ 𝑏 = 𝑐 ∗ 𝑏 → 𝑎 = 𝑐.

Also, 𝑎 ∗ 𝑏 = 𝑎 ∗ 𝑐 → 𝑏 = 𝑎.

These are called the right and left cancellation laws.

𝑃𝑟𝑜𝑜𝑓

Assume 𝑎 ∗ 𝑏 = 𝑎 ∗ 𝑐.

Then,

(𝑎 ∗ 𝑏) ∗ 𝑏 ′ = (𝑐 ∗ 𝑏) ∗ 𝑏 ′

𝑎 ∗ (𝑏 ∗ 𝑏 ′ ) = 𝑐 ∗ (𝑏 ∗ 𝑏 ′ ) (Associative)

𝑎 ∗ 𝑒 = 𝑐 ∗ 𝑒 (Inverse)

𝑎 = 𝑐 (Identity)

Similarly, the secondary association can be proved.

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 2

Unique Inverse

An element 𝑔 in a group 𝐺 has a unique inverse.

𝑃𝑟𝑜𝑜𝑓

Let 𝑔1 and 𝑔2 be two inverses of 𝑔. Show 𝑔1 = 𝑔2 .

Now,

𝑔1 = 𝑔1 ∗ (𝑔 ∗ 𝑔2+ )

𝑔1 = (𝑔1 ∗ 𝑔) ∗ 𝑔2+

𝑔1 = 𝑒 ∗ 𝑔2+

∴ 𝑔1 = 𝑔2

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 3

7|Page
Identity is unique.

𝑃𝑟𝑜𝑜𝑓

Suppose 𝑒1 and 𝑒2 are two identities of a group 𝐺. Show 𝑒1 = 𝑒2 .

Now,

𝑔 ∗ 𝑒1 = 𝑔, ⍱𝑔 ∈ 𝐺 (1)

and

𝑒2 ∗ 𝑦 = 𝑦, ⍱𝑦 ∈ 𝐺 (2)

So,

𝑒1 = 𝑒2 ∗ 𝑒1 by (2)

𝑒1 = 𝑒2 by (1)

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 4

If 𝐺 is a group, then 𝑥 = (𝑥 −1 )−1 = 𝑥 for any 𝑥 ∈ 𝐺.

𝑃𝑟𝑜𝑜𝑓

Since 𝑥 −1 is the inverse of 𝑥, we know that 𝑥 −1 ∗ 𝑥 = 𝑥 ∗ 𝑥 −1 = 𝑒.

By these equations, 𝑥 satisfies the definition of (𝑥 −1 )−1.

So, 𝑥 = (𝑥 −1 )−1 by the uniqueness of inverse.

PERMUTATION GROUP

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛

A function 𝑓: 𝐴 → 𝐴 is a permutation on 𝐴 if 𝑓 is one-to-one and onto.

The set of all permutations on 𝐴 with composition of functions as the operation is called the
symmetric group on 𝐴 denoted by 𝑆𝑦𝑚(𝐴).

Any subgroup of 𝑆𝑦𝑚(𝐴) is a permutation group on 𝐴.

8|Page
We shall consider a case where 𝐴 is finite, say 𝐴 = {1,2,3, … , 𝑛} and we accordingly denote 𝑆𝑦𝑚(𝐴)
𝑆𝑛 called the symmetric group of degree 𝑛.

Let 𝑓 ∈ 𝑆𝑛 . Then 𝑓 shuffles the elements 1,2,3,4,5,6, … , 𝑛 − 1, 𝑛 and we can represent 𝑓 explicitly by

1 2 3 4 5 … 𝑛
𝑓=( ) where 𝑓(𝑘) is placed under 𝑘 for each 𝑘 between
𝑓(1) 𝑓(2) 𝑓(3) 𝑓(4) 𝑓(5) … 𝑓(𝑛)
𝑖 and 𝑛.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒

Let 𝑛 = 4 and consider 𝑔 and 𝑓 in 𝑆4 , where

1 2 3 4 1 2 3 4
𝑔=( ), 𝑓 = ( )
2 4 1 3 3 2 4 1

Then it is easy to calculate products in 𝑆4 . To see what goes under 1 in the product, we just recall
that the product is the composition of the two given permutations, the one on the right being
performed first.

So,

(𝑔°𝑓)(1) = (1 2 3 4 1 2 3 4
)°( ) (1)
2 4 1 3 3 2 4 1

(𝑔°𝑓)(1) = (1 2 3 4 (3 )
)
2 4 1 3
(𝑔°𝑓)(1) = 1

Similarly,

(𝑔°𝑓)(1) = (1 2 3 4 1 2 3 4
)°( ) (2)
2 4 1 3 3 2 4 1

(𝑔°𝑓)(1) = (1 2 3 4
) (2)
2 4 1 3
(𝑔°𝑓)(1) = 4

So,

1 2 3 4
𝑔°𝑓 = ( )
1 4 3 2

9|Page
However, this notation is rather uneconomical. Hence, a much more efficient notation is necessary
for permutations. Also, in order to introduce an important subgroup of 𝑆𝑛 , we consider special
permutations called CYCLES.

A permutation of the form (𝑎1 , 𝑎2 , 𝑎3 , … , 𝑎𝑘 ) is called a cycle or k-cycle. A 2-cycle is called a

transposition.

Two cycle 𝛼 = (𝑎1 , 𝑎2 , 𝑎3 , … , 𝑎𝑛 ) and 𝛽 = (𝑏1 , 𝑏2 , 𝑏3 , … , 𝑏𝑚 ) are disjoint if 𝛼 ∩ 𝛽 = ∅. A set Γ of

cycles is disjoint if any two disjoint cycles in Γ are disjoint.

The new insight here is that a permutation like 𝛼 = (1 3 5)(2 4 6 8) can be viewed as a product of
two permutations in 𝑆𝑦𝑚(8). That is 𝛼 = 𝑎1 𝑎2, where 𝑎1 = (1 3 5) where 𝑎1 fixes 246, 7 and 8 and
𝑎2 = (2468) where 𝑎2 fixes 1,3,5,7.

1 2 3 4 5 6 7
Again suppose 𝛼 = ( ).
7 1 2 6 4 4 3

Since the top row is almost wasted, we suspect a more efficient way to write 𝛼 without losing any
information. Since 𝛼 takes 1 to 7, we could write 𝛼 = (17 … ). Instead of now seeing where 2 goes,
let us note that 𝛼 takes 7 to 3. 𝛼 = (173 … ). Next, 𝛼 takes 3 to 2 and 2 back to 1. So, we can write
𝛼 = (1732) with the understanding that 𝛼 takes each entry to the following entry except that the
last entry in parenthesis is taken back to the first entry in parenthesis. This is all well except that we
have ignored 4, 5 and 6. We repeat the process starting with 4, and find that 𝛼 = (1732)(46). Now
the only entry left out is 5. Since 𝛼 takes 5 to 5, the complete situation is described by 𝛼 =
(1732)(46)(5). This is called cycle notation for 𝛼.

We can shorten this a little more by dropping (5) and having an understanding that 𝛼 takes each
omitted entry to itself.

𝐸𝑥𝑒𝑟𝑐𝑖𝑠𝑒

1. Rewrite the following permutations in the cycle notation:

1 2 3 4 5
(a) ( )
2 3 4 5 1
1 2 3 4 5 6
(b) ( )
2 1 4 3 6 5
1 2 3 4 5 6 7
(c) ( )
2 3 1 4 5 7 6
(d) The following permutations in 𝑆𝑦𝑚(8) are written in the cycle notation. Rewrite in the
old notation.
(a) (1 3 5 7 2 4 6 8)
(b) (1 2 3)(7 6)(4 5)
(c) (1 8)(2 7)(3 6)(4 5)
(d) What is (1 3 5 7 2 4 6 8)−1 ?
𝑛!
N.B.: |𝑆𝑛 | = 𝑛! and |𝐴𝑛 | =
2

10 | P a g e
𝐸𝑥𝑎𝑚𝑝𝑙𝑒

Let us look at 𝑆𝑦𝑚(3), the group of all permutations of 3 elements. If 𝑓(1) = 2, 𝑓(2) = 3 and
1 2 3
𝑓(3) = 1, then we write this as ( ).
2 3 1
1 2 3 1 2 3
So, if 𝑓 = ( ), 𝑔 = ( ), then
2 3 1 3 2 1

1 2 3 1 2 3
(𝑓°𝑔)(1) = (( )°( ) ) (1)
2 3 1 3 2 1

(𝑓°𝑔)(1) = (1 2 3
) ((
1 2 3
) (1))
2 3 1 3 2 1

(𝑓°𝑔)(1) = (1 2 3
) (3)
2 3 1
(𝑓°𝑔)(1) = 1

Also,

(𝑓°𝑔)(2) = ((1 2 3
)°(
1 2 3
) ) (2)
2 3 1 3 2 1

(𝑓°𝑔)(2) = (1 2 3
) (2)
2 3 1

(𝑓°𝑔)(2) = 3

and

(𝑓°𝑔)(3) = ((1 2 3
)°(
1 2 3
) ) (3)
2 3 1 3 2 1

1 2 3
(𝑓°𝑔)(3) = ( ) (1)
2 3 1
(𝑓°𝑔)(3) = 2

Putting these three computations together, we see that

1 2 3 1 2 3 1 2 3
( )°( )=( )
2 3 1 3 2 1 1 3 2

11 | P a g e
Remember that we are dealing with functions, apply the function on the right first, then the one on
the left and over through the elements 1,2,3, … , 𝑛 to determine the product permutation.

GEOMETRIC INTERPRETATION OF 𝑆𝑌𝑀(3) = 𝑆3

Before writing down a multiplication table and subgroup lattice for 𝑆𝑦𝑚(3), we give it a geometric
interpretation. Suppose we have a fixed hollow equilateral triangular frame with vertices labelled as
shown:

1 2

Assume also we have a triangle which fits into this frame and which can be taken out and fitted back
in. Each such movement is called symmetry. For instance, taking the triangle out and rotating it 120𝑜
clockwise will move the vertex at 1 to 3, 3 to 2 and 2 to 1. So, this rotation corresponds to
1 2 3
( ) = (132) (Cycle Rotation)
3 1 2

Let us see what happens if we take out the triangle, flip it about the line passing through 2 that
bisects the triangle and replace it in the frame. The vertices at 1 and 3 are interchanged and the
1 2 3
vertex at 2 is not moved. Therefore, this is written as ( ) = (13)
3 2 1
1 2 3 1 2 3 1 2 3
Multiplying as before ( )( )=( ) = (12)
3 1 2 3 2 1 2 1 3

The product corresponds to flipping the triangle about the bisector through 3. Thus, rotating the
triangle 120𝑜 and then flipping about 2 is the same as first flipping the triangle about 3.

1 2 3 1 2 3
If we allow 𝑎 = ( ) = 240𝑜 𝑟𝑜𝑡𝑎𝑡𝑖𝑜𝑛 clockwise and 𝑏 = ( ), then 𝑎2 = 120𝑜
2 3 1 3 2 1
1 2 3 1 2 3 1 2 3 1 2 3
rotation clockwise = ( ), 𝑎𝑏 = ( )( )=( ) and
3 1 2 2 3 1 3 2 1 1 3 2
1 2 3 1 2 3 1 2 3
𝑎2 𝑏 = ( )( )=( ).
3 1 2 3 2 1 2 1 3

We obtain the following multiplication table:

12 | P a g e
 * 𝒆 𝒂 𝒂𝟐 𝒃 𝒂𝒃 𝒂𝟐 𝒃
𝒆 𝑒 𝑎 𝑎2 𝑏 𝑎𝑏 𝑎2 𝑏
𝒂 𝑎 2 𝑒 𝑎𝑏 𝑎2 𝑏 𝑏
𝑎
𝟐 2 𝑒 𝑎 2 𝑏 𝑎𝑏
𝒂 𝑎 𝑎 𝑏
𝒃 𝑏 𝑎2 𝑏 𝑎𝑏 𝑒 𝑎2 𝑎
2
𝒂𝒃 𝑎𝑏 𝑏 𝑎 𝑏 𝑎 𝑒 𝑎2
𝒂𝟐 𝒃 𝑎2 𝑏 𝑎𝑏 𝑏 𝑎2 𝑎 𝑒
Figure: Multiplication Table for 𝑆3 .

Several items are worth noting.

First,

1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
𝑎𝑏 = ( )( )=( )≠( )=( )( ) = 𝑏𝑎
2 3 1 3 2 1 1 3 2 2 1 3 3 2 1 2 3 1
{, , 2, ,
So, 𝑆𝑦𝑚(3) is non abelian.

In fact, 𝑆𝑦𝑚(3) is the least non abelian group of order 6. Note, every group of order 5 or less is
abelian. The proper subgroups of 𝑆3 are 𝐻1 =< 𝑏 >, 𝐻2 =< 𝑎𝑏 >, 𝐻3 =< 𝑎2 𝑏 > each with two
elements where 𝑏, 𝑎𝑏, 𝑎2 𝑏 are the generators of the subgroups respectively and 𝐴3 =< 𝑎 > with 3
elements.

So, a subgroup lattice for 𝑆3 is given as

𝑆3 = {𝑒, 𝑎, 𝑎2 , 𝑏, 𝑎𝑏, 𝑎2 𝑏 }

𝐴3 =< 𝑎 > 𝐻1 =< 𝑏 > 𝐻2 =< 𝑎𝑏 > 𝐻3 =< 𝑎2 𝑏 >

{𝑒}
Figure: Subgroup Lattice for 𝑆3 .

In 𝑆3 = {𝑒, 𝑎, 𝑎2 , 𝑏, 𝑎𝑏, 𝑎2 𝑏 }, we have 𝑏𝑎 = 𝑎2 𝑏 and 𝑏𝑎2 = 𝑎𝑏. These two equations tell us how to
multiply the elements of 𝑆3 without writing down the permutation notation.

For instance, (𝑎𝑏)(𝑎𝑏) = 𝑎(𝑏𝑎)𝑏 = 𝑎(𝑎2 𝑏)𝑏 = 𝑎3 𝑏 2 = 𝑒

and

(𝑎2 𝑏)(𝑎2 𝑏) = (𝑎2 )(𝑏𝑎2 )𝑏 = 𝑎2 (𝑎𝑏)𝑏 = 𝑎3 𝑏 2 = 𝑒

13 | P a g e
So, both have order 2.

Observe that in these calculations, we simply use the equations 𝑏𝑎 = 𝑎2 𝑏 and 𝑏𝑎2 = 𝑎𝑏 to “move
𝑓𝑠 and 𝑔𝑠” and thus get all the 𝑓𝑠 together and all the 𝑔𝑠 together.

Let us figure out the subgroups of 𝑆3 . Of course, we have cyclic subgroups < 𝑒 >, < 𝑎 >, < 𝑎2 >, <
𝑏 >, < 𝑎𝑏 > and < 𝑎2 𝑏 > since < 𝑎 >=< 𝑎2 > and 𝑏, 𝑎𝑏, 𝑎2 𝑏 are each of order 2, which is only
one cyclic subgroup of order 3 and there are three cyclic subgroups of order 2.

These subgroups have been produced by choosing elements arbitrarily and considering their powers
which form the cyclic subgroups.

Note that 𝑆3 can now be written as 𝑆3 = {𝑒, 𝑎, 𝑎2 , 𝑏, 𝑎𝑏, 𝑎2 𝑏 } where 𝑒 is the identity element.

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛

SUBGROUP

Suppose 𝐺 is a group and 𝐵 is non−∅ subset of 𝐺, then 𝐵 is a subgroup of 𝐺 provided 𝐵 inherits the
characteristics of the group 𝐺.

OR

A subset 𝐵 of a group (𝐺,∗) is a subgroup if the elements of 𝐵 form a group under ∗.

𝐿𝑒𝑚𝑚𝑎

If 𝑓: 𝐴 → 𝐵 and 𝑔: 𝐵 → 𝐶 are functions, then

a. If 𝑓 and 𝑔 are one-to-one, then 𝑔°𝑓 is one-to-one.

b. If 𝑓 and 𝑔 are onto, then 𝑔°𝑓 is onto.

𝑃𝑟𝑜𝑜𝑓

a. Let (𝑔°𝑓)(𝑥1 ) = (𝑔°𝑓)(𝑥2 ). We wish to show that 𝑥1 = 𝑥2 [This is sufficient to show 𝑔°𝑓 is one-
to-one].

Now, (𝑔°𝑓)(𝑥1 ) = (𝑔°𝑓)(𝑥2 ) → 𝑔(𝑓(𝑥1 )) = 𝑔(𝑓(𝑥2 )) → 𝑔(𝑥1 ) = 𝑔(𝑥2 ) since 𝑓 is 1 − 1.

Also, 𝑔(𝑥1 ) = 𝑔(𝑥2 ) → 𝑥1 = 𝑥2 since 𝑔 is 1 − 1.

Therefore, 𝑔°𝑓 is 1 − 1.

b. Pick an arbitrary element 𝑐 ∈ 𝐶.

Informally, we need to fill in the blank so that (𝑔°𝑓)(. ) = 𝑐 is satisfied.

14 | P a g e
 𝒈°𝒇
A B C

We again work back to 𝐵 and then to 𝐴. Since 𝑔 is onto, there is a 𝑏 ∈ 𝐵 such that 𝑔(𝑏) = 𝑐.

Since 𝑓 is onto, there is an 𝑎 ∈ 𝐴 such that 𝑓(𝑎) = 𝑏.

Thus, (𝑔°𝑓)(𝑎) = 𝑔(𝑓(𝑎)) = 𝑔(𝑏) = 𝑐

So, we can proceed in this way for any 𝑐 ∈ 𝐶, and 𝑔°𝑓 is onto ∎

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛

If 𝐴 is a set, then the set 𝑆𝑦𝑚(𝐴) of all permutations on 𝐴 is a group, where the group operation is
composition of functions.

𝑃𝑟𝑜𝑜𝑓

If 𝑓 and 𝑔: 𝐴 → 𝐴 are functions, then 𝑓°𝑔 is also: 𝐴 → 𝐴 a function.

(i) Closure

By the lemma, 𝑓°𝑔 is one-to-one and onto, since both 𝑓 and 𝑔 are one-to-one and onto.

(ii) Associativity

Clearly, for 𝑓, 𝑔 and ℎ: 𝐴 → 𝐴, then (𝑓°𝑔°ℎ): 𝐴 → 𝐴 and

(𝑓°(𝑔°ℎ))(𝑎) = ((𝑓°𝑔)°ℎ)(𝑎)

(𝑓°(𝑔ℎ))(𝑎) = ((𝑓𝑔)°ℎ)(𝑎)

𝑓° (𝑔(ℎ(𝑎))) = 𝑓°(𝑔(𝑎))

𝑓°(𝑔(𝑎)) = 𝑓(𝑔(𝑎))

𝑓(𝑔(𝑎)) = 𝑓(𝑎)

𝑓(𝑎) = 𝑎

(iii) Identity Element

If 𝐴 is any set, then 𝐼𝐴 : 𝐴 → 𝐴 is the identity function such that 𝐼𝐴 (𝑎) = 𝑎, ⍱𝑎 ∈ 𝐴.

15 | P a g e
 (iv) Inverse Element

If 𝑓: 𝐴 → 𝐴 is a function, then 𝑓 −1 : 𝐴 → 𝐴 is also a function.

This is true since (𝑓 −1 °𝑓)(𝑎) = 𝑓 −1 (𝑓(𝑎)) = 𝑎 = 𝐼𝐴 (𝑎), ⍱𝑎 ∈ 𝐴.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒

Another important group is 𝐷8, the dihedral group of order 8. Consider the vertices of a square
as follows.

1 4

2 3

Again, consider this to be a fixed square frame to use in labelling the movements of a square fitted
inside the frame.

1 2 3 4
90𝑜 = 𝑎 = ( ) counter clockwise
2 3 4 1
1 2 3 4
180𝑜 = 𝑎2 = ( )
3 4 1 2

Checking, we have

1 2 3 4 1 2 3 4 1 2 3 4
𝑎𝑎 = 𝑎2 = ( )( )=( )
2 3 4 1 2 3 4 1 3 4 1 2

Also,

1 2 3 4 1 2 3 4
𝑎3 = ( ) and 𝑎4 = ( )=𝐼
4 1 2 3 1 2 3 4

16 | P a g e
 1 2 3 4
Let 𝑏 be the flip about the diagonal through the vertices at 1 to 3; then 𝑏 = ( ); 𝑏𝑎 =
1 4 3 2
1 2 3 4 1 2 3 4
( ) = 𝑡ℎ𝑒 "𝑣𝑒𝑟𝑡𝑖𝑐𝑎𝑙 𝑓𝑙𝑖𝑝"; 𝑏𝑎2 = ( ) = "𝑓𝑙𝑖𝑝 𝑎𝑏𝑜𝑢𝑡 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑑𝑖𝑎𝑔𝑜𝑛𝑎𝑙";
4 3 2 1 3 2 1 4
1 2 3 4
𝑏𝑎3 = ( ) = "𝑓𝑙𝑖𝑝 𝑡ℎ𝑟𝑜𝑢𝑔ℎ 𝑡ℎ𝑒 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠"
2 1 4 3

We note that 𝐼 = 𝑎4 = 360𝑜 , 𝑏 2 = 𝐼 and 𝑎𝑏 = 𝑏𝑎3 and 𝑏𝑎 ≠ 𝑏𝑎3 .

Therefore, 𝐷8 is non-abelian using 𝑎4 = 𝐼 = 𝑏 2 and 𝑏𝑎3 = 𝑎𝑏, we can write the multiplication table
for 𝐷8 as follows:

* 𝑰 𝒂 𝒂𝟐 𝒂𝟑 𝒃 𝒃𝒂 𝒃𝒂𝟐 𝒃𝒂𝟑
𝑰 𝐼 𝑎 𝑎2 𝑎3 𝑏 𝑏𝑎 𝑏𝑎2 𝑏𝑎3
𝒂 𝑎 𝑎2 𝑎3 𝐼 𝑏𝑎3 𝑏 𝑏𝑎 𝑏𝑎2
𝒂𝟐 𝑎2 𝑎3 𝐼 𝑎 𝑏𝑎2 𝑏𝑎3 𝑏 𝑏𝑎
𝒂𝟑 𝑎3 𝐼 𝑎 𝑎2 𝑏𝑎 𝑏𝑎2 𝑏𝑎3 𝑏
𝒃 𝑏 𝑏𝑎 𝑏𝑎2 𝑏𝑎3 𝐼 𝑎 𝑎2 𝑎3
𝒃𝒂 𝑏𝑎 𝑏𝑎2 𝑏𝑎3 𝑏 𝑎3 𝐼 𝑎 𝑎2
𝒃𝒂𝟐 𝑏𝑎2 𝑏𝑎3 𝑏 𝑏𝑎 𝑎2 𝑎3 𝐼 𝑎
𝒃𝒂𝟑 𝑏𝑎3 𝑏 𝑏𝑎 𝑏𝑎2 𝑎 𝑎2 𝑎3 𝐼

The following is a group lattice for 𝐷8 where 𝐼 = 𝑒

DIAGRAM

𝐸𝑥𝑒𝑟𝑐𝑖𝑠𝑒

1. Verify that {𝑒, 𝑎2 , 𝑏, 𝑏𝑎2 } is a subgroup of 𝐷8.

2. Write down a multiplication table for a group given only that 𝑎5 = 𝑒 = 𝑏 2 and 𝑎𝑏 = 𝑏𝑎4 .
3. Show that if 𝑎 is the only element of order 2 then 𝑥𝑎 = 𝑎𝑥, ⍱𝑥 ∈ 𝐺 where 𝐺 is a group.
(Hint: Use 𝑥 𝑛 = 𝑒 iff (𝑦 −1 × 𝑦)𝑛 = 𝑒).
4. Let 𝐻 = {𝑒, 𝑎2 , 𝑏𝑎, 𝑏𝑎3 }.
(i) Verify that 𝐻 is a subgroup of 𝐷8.
(ii) Verify that 𝐻 is abelian.
(iii) Find a subgroup 𝐾 of 𝐻.
5. (a) 𝐺 is an abelian group iff (𝑎𝑏)2 = 𝑎2 𝑏 2 , ⍱𝑎, 𝑏 ∈ 𝐺.
(b) If 𝐺 is abelian such that |𝐺| is an odd integer, then show that the product of all elements
in 𝐺 is 𝑒.

6. If 𝐺 is a group of even order, then show that there is an element 𝑥 ∈ 𝐺 other than 𝑒 such that
𝑥 2 = 𝑒. [In fact, there are odd numbers of such elements called involutions].

7. Let (𝐺,∗) be a group and 𝑥. 𝑥 = 𝑥, then 𝑥 = 𝑒.

8. If 𝐺 is a finite group and 𝑥 ∈ 𝐺, then 𝑥 𝑛 = 𝑒 for some positive integer 𝑛.

17 | P a g e
 CHAPTER THREE

COSETS

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛

Let 𝐻 be a subgroup of 𝐺, then by a right coset of 𝐻 in 𝐺, we mean a subset of 𝐺 of the form 𝐻𝑎 =

{ℎ𝑎|ℎ ∈ 𝐻}, 𝑎 ∈ 𝐺. So, a coset of 𝐻 is a “translate” of 𝐻 by some element of 𝐺.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒

1. Let 𝐺 = (𝑅\{0}, . ) and 𝐻 = {1,2,5} where 𝐻 is a subset of 𝐺 and 𝐾 = {−1, 3, 3\2}. Then,
3 15
𝐻𝐾 = {−1, −2, −5, 3, 6, 15, 2
, 3, 2 }. Hence, 𝐻𝐾 is just a subset of 𝐺. However, if 𝐺 =
1
(𝑅, +) and 𝑍 a subgroup of 𝐺, then 𝑍 + 2 consists of the points we get by shifting all the
integers one-half units to the right.

1 3 5 7
𝑍 = {0, ±1, ±2, ±3, … }, 𝑍 = {± , ± , ± , ± , … }
2 2 2 2

2. Show 𝑥𝑦 −1 𝑎 = 𝑎𝑥𝑦 −1 .
Since 𝑥, 𝑦 ∈ 𝐻𝑎 = {𝑥 ∈ 𝐺|𝑥𝑎 = 𝑥𝑎}, then 𝑦𝑎 = 𝑎𝑦 → 𝑎 = 𝑦 −1 𝑎𝑦 → 𝑦𝑎𝑦 −1 = 𝑎 or 𝑎𝑦 −1 =
𝑦 −1 𝑎 → 𝑦 −1 ∈ 𝐻𝑎. So, 𝑥𝑦 −1 𝑎 = 𝑥𝑎𝑦 −1 = 𝑎𝑥𝑦 −1 ∈ 𝐻𝑎.

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛

Relations

A relation 𝑅: 𝑆 → 𝑇 is a subset of 𝑆 × 𝑇, the set of ordered pairs (𝑠, 𝑡) with 𝑠 ∈ 𝑆 and 𝑡 ∈ 𝑇. If

(𝑠, 𝑡) ∈ 𝑅, we write this as 𝑠𝑅𝑡 or 𝑠~𝑡.

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛

Equivalence Relations

A relation 𝑅: 𝑆 → 𝑆 is called an equivalence relation on 𝑆 if

(i) 𝑠~𝑠 ⍱ 𝑠 ∈ 𝑆 (reflexive law).

(ii) 𝑠~𝑡, then 𝑡~𝑠 (symmetric law).
(iii) 𝑠~𝑡 and 𝑡~𝑢, then 𝑠~𝑢 (transitive law).

N.B.: If ~ is an equivalence relation, 𝑠~𝑡 is read as “𝑠 is equivalent to 𝑡”.

18 | P a g e
𝐸𝑥𝑎𝑚𝑝𝑙𝑒 1

If 𝑠 is a set and "~” is “=”, then = is an equivalence relation on 𝑆. Thus, equality is an equivalence
relation and one way to view equivalence relations is a generalization of equality.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒 2

Let 𝑆 = 𝑍 and define 𝑎~𝑏 iff 𝑎 − 𝑏 = 2𝑘 where 𝑘 ∈ 𝑍, claim ~ is an equivalence relation

(i) since 𝑎 − 𝑎 = 0 = 2 − 0 ⍱𝑎 ∈ 𝑍, we have 𝑎~𝑎.

(ii) if 𝑎 − 𝑏 = 2𝑘, then 𝑏 − 𝑎 = −2𝑘 = 2(−𝑘), so 𝑎~𝑏 implies 𝑏~𝑎.
(iii) if 𝑎 − 𝑏 = 2𝑘 and 𝑏 − 𝑐 = 2𝑚, then (𝑎 − 𝑏) + (𝑏 − 𝑐) = (𝑎 − 𝑐) = 2(𝑘 + 𝑚). So 𝑎~𝑏,
𝑏~𝑐 → 𝑎~𝑐.

∴ This is an equivalence relation ∎

𝐸𝑥𝑒𝑟𝑐𝑖𝑠𝑒𝑠

Which of the following are equivalence relations?

(i) 𝑆 = 𝑍 and 𝑎~𝑏 if 𝑎 − 𝑏 = 𝑠𝑘, 𝑘 ∈ 𝑍.

(ii) 𝑆 = 𝑅 and 𝑎~𝑏 if 𝑎 − 𝑏 < 0.
(iii) 𝑆 = 𝑡ℎ𝑒 𝑠𝑒𝑡 𝑜𝑓 𝑎𝑙𝑙 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛𝑠: 𝑅 → 𝑅 where 𝑓~𝑔 if 𝑓(𝑥) − 𝑔(𝑥) = 𝑐 for some constant
𝑐.

EQUIVALENCE CLASS

If ~ is an equivalence relation on a set 𝑠, then for 𝑡 ∈ 𝑆, 𝑡̅ = {𝑢|𝑡~𝑢} is the equivalence class

containing 𝑡.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒

Let 𝑆 = 𝑍, 𝑎~𝑏 if 𝑎 − 𝑏 = 2𝑘, 𝑘 ∈ 𝑍.

1̅ = {𝑛|1~𝑛} = {𝑛|1 − 𝑛 = 2𝑘} = {𝑛|𝑛 = 1 − 2𝑘} = 𝑡ℎ𝑒 𝑜𝑑𝑑 𝑖𝑛𝑡𝑒𝑔𝑒𝑟𝑠 = 3̅ = 5̅ = 0̅ =

𝑒𝑣𝑒𝑛 𝑖𝑛𝑡𝑒𝑔𝑒𝑟𝑠

𝐿𝑒𝑚𝑚𝑎

If 𝑆 is a set, and ~ is an equivalence relation on 𝑆, then 𝑠 ∈ 𝜏̅ → 𝑠̅ = 𝑡̅.

19 | P a g e
𝑃𝑟𝑜𝑜𝑓

Show 𝑆̅ ⊆ 𝜏̅ and 𝜏̅ ⊆ 𝑆̅. If 𝑥 ∈ 𝑆̅, then 𝑥~𝑠. Since 𝑠 ∈ ô̅ by assumption 𝑠~𝜏.

Using transitivity from 𝑥~𝑠 and 𝑠~𝑡, we have 𝑥~𝑡. So 𝑥 ∈ 𝜏̅. ∴ 𝑠̅ ⊆ 𝜏̅ .

Conversely, let 𝑦 ∈ 𝜏̅, then 𝑦~𝜏 and we still have 𝑡~𝑠.

By symmetry from 𝑡~𝑠, we have 𝑠~𝑡.

So by transitivity we have 𝑠~𝑦.

Hence 𝑦 ∈ 𝑆̅. So, 𝜏̅ ⊆ 𝑆̅. ∴ 𝜏̅ = 𝑆̅ ∎

PARTITION

If 𝑆 is a set and {𝑃𝛼 } is a collection of non-empty subsets of 𝑆, then {𝑃𝛼 } is a partition of 𝑆 if

(i) ⋃𝛼 𝑃𝛼 = 𝑆
(ii) 𝑃𝛼 ∩ 𝑃𝛽 = ∅ for 𝛼 ≠ 𝛽.

N.B.: This is stated less formally as every element of 𝑆 is in one of the 𝑃𝛼 and in only one of the 𝑃𝛼 .

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛

If 𝑆 is a set and ~ is an equivalence relation on 𝑆, then the equivalence classes form a partition of 𝑆.

𝑃𝑟𝑜𝑜𝑓

Since 𝑡 ∈ 𝜏̅ ⍱𝑡 ∈ 𝑆, we see that every element 𝑆 is in at least one equivalence class. If we assume
̅, then by lemma 𝑡̅ = 𝑢̅ = 𝑚
𝑡 ∈ 𝑢̅ ∩ 𝑚 ̅ . So 𝑡 is in but one equivalence class (going by many names).
So the equivalences form a partition of 𝑆 ∎

20 | P a g e
 CHAPTER FOUR

LANGRANGE’S THEOREM

This is the first important theorem in group theory. Before, we present the theorem, we shall offer
some definitions and propositions.

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛 4.1

Refer to the definition of COSET.

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛 4.2

(a) Order of a group 𝐺: If 𝐺 is a finite group, then |𝐺| is the number of elements in 𝐺, called the
order of 𝐺.
(b) Order of an element in a group 𝐺: If 𝑔 ∈ 𝐺, then the order of 𝑔 is the order of the subgroup
< 𝑔 >, where < 𝑔 >= {𝑒, 𝑔, 𝑔2 , … , 𝑔−1 , 𝑔−2 , … }. If < 𝑔 > is finite, then < 𝑔 >=
{𝑔, 𝑔2 , 𝑔3 , … , 𝑔𝑚 = 𝑒}. So, the order of 𝑔 is the least positive integer 𝑚 such that 𝑔𝑚 = 𝑒

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 4.3

If 𝐻 is a finite subgroup of a group 𝐺, then |𝐻| = |𝐻𝑎| where 𝐻𝑎 is the right coset of 𝐻 in 𝐺.

𝑃𝑟𝑜𝑜𝑓

Define 𝑓: 𝐻 → 𝐻𝑎 by 𝑓𝑎 (ℎ) = ℎ𝑎. Show 𝑓 is 1 − 1 and onto.

Let 𝑓𝑎 (ℎ1 ) = 𝑓𝑎 (ℎ2 ). Show ℎ1 = ℎ2 .

𝑓𝑎 (ℎ1 ) = 𝑓𝑎 (ℎ2 ) → ℎ1 𝑎 = ℎ2 𝑎

By the right cancellation law, ℎ1 = ℎ2 .

∴ 𝑓 𝑖𝑠 1 − 1

Show onto:

⍱ℎ𝑎 ∈ 𝐻𝑎, find ℎ ∈ 𝐻 such that 𝑓𝑎 (ℎ) = ℎ𝑎.

Obviously, 𝑓𝑎 (ℎ) = ℎ𝑎, so 𝑓 is onto.

∴ |𝐻| = |𝐻𝑎| ∎

𝑇ℎ𝑒𝑜𝑟𝑒𝑚 4.4

21 | P a g e
If 𝐺 is a finite group and 𝐻 is a subgroup of 𝐺, then |𝐻|/|𝐺| [the divider here means ‘divides evenly
with zero remainder’].

𝑃𝑟𝑜𝑜𝑓

First Proof

Let 𝐻 = {ℎ1 , ℎ2 , … , ℎ𝑘 }. Start writing the elements of 𝐺 as follows:

ℎ1 , ℎ2 , … , ℎ𝑘

ℎ1 𝑎, ℎ2 𝑎, … , ℎ𝑘 𝑎

ℎ1 𝑏, ℎ2 𝑏, … , ℎ𝑘 𝑏

Where 𝑎 is picked to be an element not in the first row where 𝑏 is not in the first two rows.

Continuing in this fashion until all the elements of 𝐺 are listed as shown:

𝐻 = {ℎ1 , ℎ2 , … , ℎ𝑘 }

𝐻𝑎 = {ℎ1 𝑎, ℎ2 𝑎, … , ℎ𝑘 𝑎}

𝐻𝑏 = {ℎ1 𝑏, ℎ2 𝑏, … , ℎ𝑘 𝑏}

𝐻𝑥 = {ℎ1 𝑥, ℎ2 𝑥, … , ℎ𝑘 𝑥}

N.B.: Each row is a right coset with |𝐻| = |𝐻𝑎| = |𝐻𝑏| = ⋯ = |𝐻𝑥| by the proposition above.

What happens if two rows contain an element in common?

Say 𝐻𝑦 contains ℎ𝑖 𝑤 where 𝐻𝑤 ≠ 𝐻𝑦. Then ℎ𝑖 𝑤 = ℎ𝑗 𝑦 for some ℎ𝑗 ∈ 𝐻. Then 𝑦 = ℎ𝑗 −1 ℎ𝑖 𝑤 ∈ 𝐻𝑤.

So, 𝑦 would have been included in an earlier row and would not have selected 𝑦 to start a row. Thus,
no two rows have a common element. If 𝑚 is the number of distinct right cosets, we see that 𝐺 can
be divided into 𝑚 rows of |𝐻| elements each.

So, |𝐺| = |𝐻|𝑚

∴ |𝐻|/|𝐺|

INDEX OF 𝑯 IN 𝑮

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛 4.5

If 𝐻 is a subgroup of a finite group 𝐺, and |𝐺| = 𝑚|𝐻|, then 𝑚 is the index of 𝐻 in 𝐺.

22 | P a g e
𝐶𝑜𝑟𝑜𝑙𝑙𝑎𝑟𝑦 4.6

If 𝑔 is an element of a finite group 𝐺, then the order of 𝑔 divides |𝐺|.

𝑃𝑟𝑜𝑜𝑓

| < 𝑔 > | is the order of the subgroup generated by 𝑔. By Lagrange’s theorem, | < 𝑔 > |/|𝐺|∎

𝐶𝑜𝑟𝑜𝑙𝑙𝑎𝑟𝑦 4.7

If 𝐺 is a group of prime order 𝑝, then 𝐺 is cyclic.

𝑃𝑟𝑜𝑜𝑓

Let 𝑔 ≠ 𝑒 ∈ 𝐺.

Then, | < 𝑔 > | = 1 or 𝑝, since |𝐺| = 𝑝 = 𝑝𝑟𝑖𝑚𝑒

Since 𝑒, 𝑔 ∈< 𝑔 >, we must have < 𝑔 >= 𝑝 = |𝐺|. So, 𝐺 =< 𝑔 > and cyclic.

𝐸𝑥𝑒𝑟𝑐𝑖𝑠𝑒𝑠

1. If |𝐺| = 2𝑝 where 𝑝 is prime, show that every proper subgroup of 𝐺 is cyclic.

2. Show that every proper subgroup of 𝑆3 is cyclic.
3. Suppose 𝐺 is a group and 𝑥 has order 𝑛 in 𝐺, and 𝑑/𝑛 where 𝑑 is a positive integer, then 𝐺
has an element of order 𝑑.

N.B.: Lagrange died in 1812, 20 years before Galois first defined group more or less as we know it
and 40 years before groups were discussed widely.

1. Let 𝐺 = (𝑍, 4) and 𝐻 = 6𝑍, the right cosets of 𝐻 in 𝐺 are 6𝑍, 6𝑍 + 1, 6𝑍 + 2, 6𝑍 + 3, 6𝑍 + 4,

6𝑍 + 5

2. Let 𝐺 ′ = (𝑍12 ,⊕). 𝐻 =< 4 >= {4,8,0}. The right cosets of 𝐻 have 𝐻, 𝐻 ⊕ 1, 𝐻 ⊕ 2, 𝐻 ⊕ 3.

𝐻 ⊕ 1 = {5,9,1}; 𝐻 ⊕ 2 = {6,10,2}; 𝐻 ⊕ 3 = {7,11,3}.

23 | P a g e
 CHAPTER FIVE

NORMAL SUBGROUPS

Let 𝐺 be a group and 𝐻 ⊂ 𝐺. Then we say 𝐻 is a normal subgroup of 𝐺 (denoted by 𝐻∆𝐺).

If (i) ⍱ℎ ∈ 𝐻; (ii) 𝑔 ∈ 𝐺; then 𝑔ℎ𝑔−1 ∈ 𝐻

OR

𝐻 is normal in 𝐺 iff 𝑔𝐻𝑔−1 ⊆ 𝐻 where 𝑔𝐻𝑔−1 = {𝑔ℎ𝑔−1 |ℎ ∈ 𝐻}.

N.B.: The definition does not say that 𝑔ℎ𝑔−1 = ℎ, ⍱𝑔 ∈ 𝐺, ℎ ∈ 𝐻, rather it says that 𝑔ℎ𝑔−1 will be
some element of 𝐻 not necessarily ℎ. In fact, 𝑔ℎ𝑔−1 = ℎ → 𝑔ℎ = ℎ𝑔 saying that 𝑔 and ℎ commute
with each other.

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 5.1

Let 𝐻 ⊂ 𝐺 where 𝐺 is a group and 𝐻 is a subgroup of 𝐺, then the following are equivalent:

(i) 𝐻 is normal.
(ii) 𝑔ℎ𝑔−1 = 𝐻, ⍱𝑔 ∈ 𝐺
(iii) 𝑔𝐻 = 𝐻𝑔, ⍱𝑔 ∈ 𝐺 [i.e. the left cosets are the same as the right cosets]

𝑅𝑒𝑚𝑎𝑟𝑘

(iii) does not say that ⍱ℎ ∈ 𝐻, 𝑔ℎ = ℎ𝑔. It implies that 𝑔ℎ will be ℎ′ 𝑔 for some ℎ′ , but it is not
necessary for ℎ′ to be ℎ.

𝑃𝑟𝑜𝑜𝑓

(𝑖) ↔ (𝑖𝑖). Clearly, (𝑖𝑖) → (𝑖). Show (𝑖) → (𝑖𝑖).

Assume 𝑔𝐻𝑔−1 ⊆ 𝐻, ⍱𝑔 ∈ 𝐺. Then, for any 𝑔, 𝑔−1 𝐻(𝑔−1 )−1 ⊆ 𝐻. Thus, 𝑔−1 𝐻 ⊆ 𝐻𝑔−1 and 𝐻 ⊆
𝑔𝐻𝑔−1.

So, with the assumption 𝑔𝐻𝑔−1 ⊆ 𝐻 we have 𝑔𝐻𝑔−1 = 𝐻 ∎

(𝑖) ↔ (𝑖𝑖) [i.e. 𝑔𝐻𝑔−1 = 𝐻 ↔ 𝑔𝐻 = 𝐻𝑔]. This is immediate since (𝑖𝑖) and (𝑖𝑖𝑖) are obtained from
each other by multiplying on the right by 𝑔 and 𝑔−1 .

𝑃𝑟𝑜𝑜𝑓

Let 𝐺 and 𝐻 be groups and consider 𝐺 × 𝐻. The subgroup {𝑒𝐺 } × 𝐻 is a normal subgroup of 𝐺 × 𝐻.
Let (𝑔, ℎ) ∈ 𝐺 × 𝐻 and (𝑒𝐺 , 𝑥) ∈ {𝑒𝐺 } × 𝐻.

24 | P a g e
Then, (𝑔, ℎ)(𝑒𝐺 , 𝑥)(𝑔, ℎ)−1 = (𝑔𝑒𝐺 𝑔−1 , ℎ𝑥ℎ−1 ) = (𝑒𝐺 , ℎ𝑥ℎ−1 ) ∈ {𝑒𝐺 } × 𝐻.

Similarly, (𝐺 × {𝑒𝐺 })∆(𝐺 × 𝐻).

Let us look at some criteria for the determination of normal subgroups without referring to the
definition of normality. They are as follows:

(i) The first involves the idea of elements commuting with each other.
(ii) The second involves counting arguments.

N.B.: (a) By 𝑍(𝐺), we mean the centre of a group 𝐺 denoted by 𝑍(𝐺) = {𝑦|𝑦𝑔 = 𝑔𝑦, ⍱𝑔 ∈ 𝐺}

(b) 𝑍(𝐺)∆𝐺, ⍱𝐺.

𝑇ℎ𝑒𝑜𝑟𝑒𝑚 [𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛] 5.2

Let 𝐺 be a group. Then any subgroup 𝐻 of 𝑍(𝐺) is a normal subgroup of 𝐺.

𝑃𝑟𝑜𝑜𝑓

Let 𝐻 ⊂ 𝑍(𝐺). Show 𝐻∆𝐺.

One way to show this is to show that 𝑔ℎ𝑔−1 ∈ 𝐻 ∈, ⍱𝑔 ∈ 𝐺, ℎ ∈ 𝐻 but since ℎ ∈ 𝑍(𝐺), 𝑔ℎ𝑔−1 =
ℎ𝑔𝑔−1 = ℎ ∈ 𝐻 where 𝑔ℎ = ℎ𝑔.

𝐶𝑜𝑟𝑜𝑙𝑙𝑎𝑟𝑦 5.3

If 𝐺 is abelian, then every subgroup of 𝐺 is normal.

𝑃𝑟𝑜𝑜𝑓

If 𝐺 is abelian, then 𝑍(𝐺) = 𝐺. So, every subgroup of 𝐺 is a subgroup of 𝑍(𝐺) hence normal by the
previous theorem ∎

𝑇ℎ𝑒𝑜𝑟𝑒𝑚 [𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛] 5.4

Let 𝐻 be a subgroup of 𝐺 such that [𝐺: 𝐻] = 2[𝑖. 𝑒. 𝐼𝑛𝑑𝑒𝑥 𝑜𝑓 𝐻 𝑖𝑛 𝐺 𝑖𝑠 2]. Then 𝐻∆𝐺.

𝑃𝑟𝑜𝑜𝑓

Sufficient to show 𝑔𝐻 = 𝐻𝑔, ⍱𝑔 ∈ 𝐺. By assumption, there exactly two cosets (namely two right
cosets) of 𝐻 in 𝐺, namely 𝐻 and 𝐺\𝐻.

Similarly, the left cosets are 𝐻 and 𝐺\𝐻.

25 | P a g e
Thus, the left cosets are the same as the right cosets and for 𝑔 ∈ 𝐺, the left coset containing 𝑔 is
also the right coset containing 𝑔, i.e. 𝑔𝐻 = 𝐻𝑔 ∎

OR

Let 𝑔 ∈ 𝐺, then 𝑔 ∈ 𝐻 or 𝑔 ∈ 𝐺\𝐻. If 𝑔 ∈ 𝐻, then 𝑔𝐻 = 𝐻 = 𝐻𝑔.

If 𝑔 ∉ 𝐻, then 𝑔 ∈ 𝐺\𝐻 = (𝐺\𝐻)𝑔 since 𝐻 ∩ 𝐻𝑔 = ∅.

Thus, 𝑔𝐻 = 𝐺\𝐻 = 𝐻𝑔.

∴ By part (iii) of proposition 5.1 𝐻∆𝐺 ∎

Another criterion is that the product of two left cosets is a left coset.

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 5.5

A subgroup 𝐻 ⊂ 𝐺 is normal in 𝑥𝐻𝑦𝐻 = 𝑥𝑦𝐻, ⍱𝑥, 𝑦 ∈ 𝐺.

→ 𝑃𝑟𝑜𝑜𝑓

Since 𝑥𝐻 and 𝑦𝐻 are subsets of 𝐺, 𝑥𝐻𝑦𝐻 = {𝑥ℎ𝑦ℎ′ , ℎ′ , ℎ ∈ 𝐻 }.

Thus, if 𝐻∆𝐺, 𝑦 −1 𝐻𝑦 = 𝐻.

So, 𝑥𝐻𝑦𝐻 = 𝑥(𝑦𝑦 −1 )𝐻𝑦𝐻 = 𝑥𝑦(𝑦 −1 𝐻𝑦)𝐻 = 𝑥𝑦𝐻𝐻 = 𝑥𝑦𝐻 ∎

If 𝐻∆𝐺, then Ǝ a 𝑦 in 𝐺 such that 𝑦 −1 𝐻𝑦 ≠ 𝐻, so 𝑦 −1 𝐻𝑦𝐻 ≠ 𝑦 −1 𝑦𝐻 = 𝐻, so 𝑥𝐻𝑦𝐻 ≠ 𝑥𝑦𝐻 if we set

𝑥 = 𝑦 −1 ∎

← 𝑃𝑟𝑜𝑜𝑓

Assuming 𝑥𝐻𝑦𝐻 = 𝑥𝑦𝐻 → 𝐻𝑦 = 𝑦𝐻 ↔ 𝑦 −1 𝐻𝑦 = 𝐻.

DIRECT PRODUCT

Let 𝐺 = 𝑍3 × 𝑍2 and 𝐻 = {(0,0), (0,1)}. 𝐻 is a subgroup of 𝐺 and its right cosets are

̅̅̅̅) = (0,1
𝐻 = 𝐻 + (0,0) = {(0,0), (0,1)} = 𝐻 + (0,1) = (0,0 ̅̅̅̅)

̅̅̅̅) = (1,1
𝐻 + (1,0) = {(1,0), (1,1)} = 𝐻 + (1,1) = (1,0 ̅̅̅̅)

̅̅̅̅) = (2,1
𝐻 + (2,0) = {(2,0), (2,1)} = 𝐻 + (2,1) = (2,0 ̅̅̅̅)

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛

26 | P a g e
If 𝐻∆𝐺 then 𝐺\𝐻, the set of left cosets of 𝐻 in 𝐺 is a group by defining an operation ∗ on 𝐺\𝐻 by
𝑥𝐻 ∗ 𝑦𝑥 = 𝑥𝐻𝑦𝐻. This is true only if 𝐻∆𝐺.

𝑃𝑟𝑜𝑜𝑓

(a) Closure
𝑥𝐻 ∗ 𝑦𝐻 = 𝑥𝐻𝑦𝐻 = 𝑥𝑦𝐻 iff 𝐻∆𝐺 by proposition 5.5.
(b) Associativity
(𝑥𝐻 ∗ 𝑦𝐻) ∗ 𝑧𝐻 = (𝑥𝑦)𝑧𝐻 = 𝑥(𝑦𝑧)𝐻 = 𝑥𝐻 ∗ (𝑦𝐻 ∗ 𝑧𝐻)
So, 𝐺\𝐻 satisfies associative law.
(c) Identity
𝐻 = 𝑒𝐻 is the identity since 𝑒𝐻𝑥𝐻 = 𝑒𝑥𝐻 = 𝑥𝐻 = 𝑥𝐻𝑒𝐻.
(d) Inverse
𝑥 −1 𝐻 is the inverse of 𝑥𝐻 since 𝑥𝐻 ∗ 𝑥 −1 𝐻 = 𝑥𝐻𝑥 −1 𝐻 = 𝑥𝑥 −1 𝐻 = 𝐻 = 𝑥 −1 𝐻𝑥𝐻

To conclude the proof, we must show ∗ is a function on 𝐺\𝐻. Specifically, if 𝑥1 𝐻 = 𝑥2 𝐻 [which does
not mean 𝑥1 = 𝑥2 ], do we have 𝑥1 𝐻 ∗ 𝑦𝐻 = 𝑥2 𝐻 ∗ 𝑦𝐻?

There is no trouble since 𝑥1 𝐻 ∗ 𝑦𝐻 = 𝑥1 𝐻𝑦𝐻 = 𝑥2 𝐻𝑦𝐻 = 𝑥2 𝐻 ∗ 𝑦𝐻.

If 𝑦1 𝐻 = 𝑦2 𝐻, we can similarly show that 𝑥1 𝐻 ∗ 𝑦1 𝐻 = 𝑥2 𝐻 ∗ 𝑦2 𝐻 and ∗ is a well-defined function

∎

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛

If 𝑁∆𝐺 then the group (𝐺\𝑁,∗) of left cosets of 𝑁 with 𝑥𝑁 ∗ 𝑦𝑁 = 𝑥𝑦𝑁 is called the Quotient Group
or Factor Group of 𝐺 over 𝑁.

HOMOMORPHISM

Let 𝑓: 𝐴 → 𝐵 be a function from a group 𝐴 into group 𝐵, then 𝑓 is an homomorphism provided

𝑓(𝑥°𝑦) = 𝑓(𝑥) ∗ 𝑓(𝑦), ⍱𝑥, 𝑦 ∈ 𝐴 where ° and ∗ are the binary operations defined on 𝐴 and 𝐵
respectively.

Monomorphism

This is a homomorphism which is 1 − 1.

Epimorphism

This is a homomorphism which is onto.

Isomorphism

If 𝑓: 𝐻 → 𝐾 is a homomorphism, then 𝑓 is an isomorphism if 𝑓 is both 1 − 1 and onto.

27 | P a g e
𝐸𝑥𝑎𝑚𝑝𝑙𝑒𝑠

1. Let 𝜑: 𝐺 → 𝐺 be defined by 𝜑(𝑥) = 𝑥. Then 𝜑 is an isomorphism.

2. Suppose 𝜑: 𝐺 → 𝐺 is defined by 𝜑(𝑥) = 𝑒, ⍱𝑥 ∈ 𝐺. Then 𝜑 is a homomorphism but neither
a monomorphism nor an epimorphism unless 𝐺 is the trivial group [i.e. 𝐺 = {𝑒}].

𝐸𝑥𝑒𝑟𝑐𝑖𝑠𝑒𝑠

1. Let 𝐺 = (𝑍, +) and 𝜑: (𝑍, +) → (𝑍, +) is defined by 𝜑(𝑛) = 2𝑛, ⍱𝑛 ∈ 𝑍. Show that 𝜑 is a
homomorphism, a monomorphism but not an epimorphism.
2. Let 𝐺 = (𝑍, +) and 𝜑: 𝐺 → (𝑍, +) is defined by 𝜑(𝑛) = −𝑛. Show that 𝜑 is an isomorphism.
3. Let 𝜑: (𝑍, +) → (𝑍𝑛 ,⊕) be defined by 𝜑(𝑚) = 𝑚 ̅, the remainder of 𝑚 (𝑚𝑜𝑑 𝑚). Show that
𝜑 is a homomorphism; 𝜑 is not a monomorphism, 𝜑(𝑜) = 𝜑(𝑛) yet 𝑛 ≠ 𝑜; and that 𝜑 is an
epimorphism, however.
4. Let 𝑓: (𝑅, +) → (𝑅 + , . ) be given by 𝑓(𝑥) = 𝑒 𝑥 , show that is a homomorphism. In fact, show
𝑓 is monomorphism and epimorphism.
1,𝑖𝑓 𝑓 𝑖𝑠 𝑒𝑣𝑒𝑛
5. Let 𝛼: 𝑆𝑛 → (𝑅\{0}, . ) be given by 𝛼(𝑓) = {−1,𝑖𝑓 𝑓 𝑖𝑠 𝑜𝑑𝑑 . Then 𝛼 is homomorphism.

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛

If 𝜑: 𝐺 → 𝐺̅ be a homomorphism. Then

(i) 𝜑(𝑒𝐺 ) = 𝑒𝐺̅

(ii) ⍱𝑥 ∈ 𝐺 and any integer 𝑛, 𝜑(𝑥 𝑛 ) = [𝜑(𝑥)]𝑛
(iii) if 𝑜(𝑥) = 𝑛, then 𝑜(𝜑(𝑥))\𝑛

𝑃𝑟𝑜𝑜𝑓

If 𝑥 ∈ 𝐺, 𝑜(𝑥) = 𝑛, then 𝑥 𝑛 = 𝑒𝐺 . So, by parts (i) and (ii), [𝜑(𝑥)]𝑛 = 𝑒𝐺̅ . Thus, 𝜑(𝑥) has finite order.

𝑇ℎ𝑒𝑜𝑟𝑒𝑚

Let 𝐺 be a group and 𝑔 ∈ 𝐺. If

(i) 𝑜(𝑔) = 𝑜(𝑔−1 )

(ii) 𝑜(𝑔) = 𝑛 and 𝑔𝑚 = 𝑒, then 𝑛\𝑚
(iii) 𝑜(𝑔) = 𝑛 and (𝑚, 𝑛) = 𝑑

Then, 𝑜(𝑔𝑚 ) = 𝑛\𝑑.

𝑃𝑟𝑜𝑜𝑓 (𝑖)

28 | P a g e
Show 𝑜(𝑔) = 𝑜(𝑔−1 ). Suppose 𝑜(𝑔) = 𝑛 and 𝑜(𝑔−1 ) = 𝑚 and suppose 𝑚 ≠ 𝑛. Then 𝑔𝑛 = 𝑒 and
(𝑔−1 )𝑚 = 𝑒. So, 𝑒. 𝑒 = 𝑔𝑛 . (𝑔−1 )𝑚 → 𝑔𝑛 . 𝑔−𝑚 = 𝑔𝑛−𝑚 = 𝑒. Hence, 𝑛\𝑛 − 𝑚 or 𝑚\𝑛 − 𝑚 is
impossible.

𝑃𝑟𝑜𝑜𝑓 (𝑖𝑖)

Given 𝑔𝑚 = 𝑒 and we seek to make something out of the fact that 𝑛 is the smallest positive integer
such that 𝑔𝑛 = 𝑒.

Write 𝑚 = 𝑞𝑛 + 𝑟, where 0 ≤ 𝑟 < 𝑛.

Then, 𝑔𝑚 = 𝑒 becomes 𝑔𝑞𝑛+𝑟 = 𝑒 → 𝑔𝑞𝑛 𝑔𝑟 = 𝑒 → (𝑔𝑛 )𝑞 𝑔𝑟 = 𝑒 but 𝑔𝑛 = 𝑒, so the last equation

becomes 𝑔𝑟 = 𝑒.

But 𝑟 is smaller than 𝑛, so this is impossible unless 𝑟 = 0.

Thus, 𝑚 = 𝑞𝑛 + 𝑟, so 𝑛\𝑚 ∎

𝑃𝑟𝑜𝑜𝑓 (𝑖𝑖𝑖)

Here, we use our information about great circle distance (gcd). We must show that 𝑛\𝑑 is the
smallest positive integer 𝑘 such that (𝑔𝑚 )𝑘 = 𝑒.

First of all, (𝑔𝑚 )𝑛\𝑑 = 𝑔𝑚(𝑛\𝑑) = 𝑔(𝑚\𝑑)𝑛 = (𝑔𝑛 )𝑚\𝑑 = 𝑒 𝑚\𝑑 = 𝑒 since 𝑜(𝑔) = 𝑛.

Now, suppose 𝑘 > 0 is smaller than 𝑛\𝑑 and (𝑔𝑚 )𝑘 = 𝑒.

We show 𝑛\𝑑 divides 𝑘, thus giving a contradiction. We have 𝑔𝑚𝑘 = 𝑒, so by part (ii), we know
𝑛\𝑚𝑘 which implies that 𝑛\𝑑 divides (𝑚\𝑑)𝑘. Since (𝑚\𝑑, 𝑛\𝑑) = 1 → 𝑛\𝑑 divides 𝑘 which is
absurd since 𝑘 is a positive integer smaller than |𝑛\𝑑|.

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛 𝑎𝑛𝑑 𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛

If 𝑁∆𝐺 and 𝛼: 𝐺 → 𝐺\𝑁 is given by 𝛼(𝑔) = 𝑔𝑁, then 𝛼 is called the canonical homomorphism from
𝐺 to 𝐺\𝑁.

𝑃𝑟𝑜𝑜𝑓

Show 𝛼 is a homomorphism and onto.

For homomorphism, show 𝛼(𝑔ℎ) = 𝛼(𝑔)𝛼(ℎ).

⍱𝑔, ℎ ∈ 𝐺, 𝛼(𝑔ℎ) = 𝑔ℎ𝑁 = 𝑔𝑁ℎ𝑁 = 𝛼(𝑔)𝛼(ℎ)

So, 𝛼 is a homomorphism.

Also, for any 𝑥𝑁 in 𝐺\𝑁, 𝛼(𝑥) = 𝑥𝑁 (Onto) ∎

29 | P a g e
N.B.: Starting with any Normal Subgroup of a group 𝐺, we obtain a homomorphism. This is the first
half of the connection between normal subgroups and isomorphisms. After, we prove the
correspondence theorem we can prove that any homomorphism gives us an accompanying normal
subgroup.

𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛

The Correspondence Theorem

If 𝐺 and 𝐺̅ are groups and 𝜑: 𝐺 → 𝐺̅ is an onto homomorphism. Then

(a) 𝜑(𝑒𝐺 ) = 𝑒𝐺̅ where 𝑒𝐺 is the identity in 𝐺. 𝑒𝐺̅ is the identity in 𝐺̅ .

(b) (𝜑(𝑥))−1 = 𝜑(𝑥 −1 )
(c) if 𝐻 is a subgroup of 𝐺 then 𝜑(𝐻) is a subgroup of 𝐺̅ .
(d) ̅ is a subgroup of 𝐺̅ , then 𝜑−1 (𝐻
if 𝐻 ̅ ) is a subgroup of 𝐺 where for any subset 𝑋 of 𝐺̅ ,
𝜑−1 (𝑋) = {𝑔|𝑔 ∈ 𝐺 𝑎𝑛𝑑 𝜑(𝑔) ∈ 𝑋}.
(e) if 𝐻∆𝐺, then 𝜑(𝐻)∆𝐺̅ .
̅ ∆𝐺̅ , then 𝜑−1 (𝐻
(f) if 𝐻 ̅ )∆𝐺.

𝑃𝑟𝑜𝑜𝑓

(a) 𝜑(𝑒𝐺 ) = 𝜑(𝑒𝐺 2 ) = 𝜑(𝑒𝐺 . 𝑒𝐺 ) = 𝜑(𝑒𝐺 )𝜑(𝑒𝐺 ) = (𝜑(𝑒𝐺 ))2

We have a solution to 𝑥 2 = 𝑥 in 𝐺̅ .
So, 𝑥 = 𝑒𝐺̅ = 𝜑(𝑒𝐺 )

N.B.: Since 𝐺̅ is a group, this yields 𝜑(𝑒𝐺 ) = 𝑒𝐺̅ as desired.

(b) Since 𝜑(𝑥)𝜑(𝑥 −1 ) = 𝜑(𝑥𝑥 −1 ) = 𝜑(𝑒𝐺 ) = 𝑒𝐺̅ ,

We have 𝜑(𝑥 −1 ) = (𝜑(𝑥))−1
In general, 𝜑(𝑥 𝑛 ) = (𝜑(𝑥))𝑛 . For 𝑛 = 0 𝑜𝑟 1, this is obvious.
For 𝑛 > 1, we prove by induction.
So, 𝜑(𝑥 𝑛 ) = 𝜑(𝑥(𝑥 𝑛−1 )) = 𝜑(𝑥)𝜑(𝑥 𝑛−1 )) = (𝜑(𝑥))𝑛 .

(c) Let 𝐻 be a subgroup of 𝐺. By (a), 𝑒𝐺̅ ∈ 𝜑(𝐻) since 𝜑(𝑒𝐺 ) = 𝑒𝐺̅ ∈ 𝐻 ̅.

So, 𝜑(𝐻) ≠ ∅.
By (b), if 𝜑(ℎ) ∈ 𝜑(𝐻), then (𝜑(ℎ))−1 ∈ 𝜑(𝐻) since (𝜑(ℎ))−1 = 𝜑(ℎ−1 ).
Similarly, if 𝜑(ℎ1 ), 𝜑(ℎ2 ) ∈ 𝜑(𝐻), then 𝜑(ℎ1 )𝜑(ℎ2 ) = 𝜑(ℎ1 ℎ2 ) ∈ 𝜑(𝐻).
So, 𝜑(𝐻) is closed and 𝜑(𝐻) is a subgroup.

̅ be a subgroup of 𝐺̅ . Then 𝑒𝐺 ∈ 𝜑−1 (𝐻

(d) Let 𝐻 ̅ ) since 𝜑(𝑒𝐺 ) = 𝑒𝐺̅ ∈ 𝐻
̅.
̅ ), then 𝜑(𝑥) ∈ 𝐻
Inverse: If 𝑥 ∈ 𝜑−1 (𝐻 ̅.

30 | P a g e
 So, (𝜑(𝑥))−1 ∈ 𝐻 ̅ since (𝜑(𝑥))−1 = 𝜑(𝑥 −1 ).
We have 𝑥 −1 ∈ 𝜑−1 (𝐻 ̅ ).
Closure: If 𝑔1, 𝑔2 ∈ 𝜑 −1 (𝐻 ̅ ), then 𝜑(𝑔1 ), 𝜑(𝑔2 ) ∈ 𝐻
̅
Since 𝐻̅ is closed being a subgroup of 𝐺̅ , 𝜑(𝑔1 )𝜑(𝑔2 ) = 𝜑(𝑔1 𝑔2 ) ∈ 𝐻
̅.
So, 𝑔1 , 𝑔2 ∈ 𝜑 ̅
−1 (𝐻 ).

(e) Given 𝐻∆𝐺, then 𝜑(𝐻) is a subgroup of 𝐺̅ by (c).

It is sufficient to show 𝜑(𝐻)∆𝐺̅ .
If 𝑔̅ ∈ 𝐺̅ and knowing 𝜑(𝑔) = 𝑔̅ for some 𝑔 ∈ 𝐺, since 𝜑 is onto, therefore
𝑔̅ −1 𝜑(𝐻)𝑔̅ = (𝜑(𝑔))−1 𝜑(𝐻) 𝜑(𝑔) = 𝜑(𝑔−1 )𝜑(𝐻) 𝜑(𝑔) = 𝜑(𝑔−1 𝐻𝑔) = 𝜑(𝐻) since
𝐻∆𝐺.
∴ 𝜑(𝐻)∆𝐺̅ ∎

(f) Suppose 𝐻̅ ∆𝐺̅ , show 𝜑−1 (𝐻̅ )∆𝐺 [i.e. 𝑔−1 (𝜑−1 (𝐻
̅ ))𝑔 ⊆ 𝜑−1 (𝐻
̅ )]
Apply 𝜑 to 𝑔−1 𝜑−1 (𝐻 ̅ )𝑔, to obtain 𝜑(𝑔−1 𝜑−1 (𝐻 ̅ )𝑔) = 𝜑(𝑔−1 )𝐻 ̅ 𝜑(𝑔) = 𝐻
̅.
̅ )𝑔 ⊆ 𝜑−1 (𝐻
So, 𝑔−1 𝜑−1 (𝐻 ̅ ), ⍱𝑔 ∈ 𝐺.
̅ )∆𝐺 ∎
Thus, 𝜑−1 (𝐻

𝑅𝑒𝑚𝑎𝑟𝑘

̅ ∆ 𝜑(𝐺).
We use the hypothesis “𝜑 𝑖𝑠 𝑜𝑛𝑡𝑜” only in (e). In part (f), we need only 𝐻

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛

If 𝜑: 𝐺 → 𝐺 is a homomorphism, then 𝐾𝑒𝑟𝜑 = {𝑔 ∈ 𝐺|𝜑(𝑔) = 𝑒𝐺 } is the Kernel of the

homomorphism 𝜑.

𝐶𝑜𝑟𝑜𝑙𝑙𝑎𝑟𝑦

If 𝜑 is a homomorphism from 𝐺 into 𝐺̅ , then 𝐾𝑒𝑟𝜑∆𝐺.

𝑃𝑟𝑜𝑜𝑓

Note that 𝐾𝑒𝑟𝜑 = 𝜑 −1 (𝑒𝐺̅ ) and 𝑒𝐺̅ 𝐺̅ . Then by (f) in the previous proposition 𝜑−1 (𝑒𝐺̅ ) = 𝐾𝑒𝑟𝜑∆𝐺 ∎

Thus, for any homomorphism 𝜑, we obtain a normal subgroup. This is the second half of the
connection between these two concepts.

𝐸𝑥𝑒𝑟𝑐𝑖𝑠𝑒𝑠

31 | P a g e
 1. Which of the following functions 𝜑 are homomorphisms? Of the homomorphisms, which are
isomorphisms?
(a) 𝐺 = (𝑅{0}, . ) = 𝐺̅ , 𝜑(𝑥) = 3𝑥.
(b) 𝐺 = 𝐺̅ = 𝑎𝑏𝑒𝑙𝑖𝑎𝑛 𝑔𝑟𝑜𝑢𝑝, 𝜑(𝑥) = 𝑥 4 where 𝐺 = (𝑅{0},∗).

2. If 𝐴 is an abelian group of order 𝑛, and 𝜑 is a function from 𝐴 to 𝐴 given by 𝜑(𝑎) = 𝑎𝑘 ,

where 𝑘 is an integer, then show that
(i) 𝜑 is a homomorphism.
(ii) 𝜑 is an isomorphism if gcd(𝑘, 𝑛) = 1.

We now come to an important theorem which makes precise the notion of every homomorphism
leading to a normal subgroup.

𝑇ℎ𝑒𝑜𝑟𝑒𝑚

THE FIRST ISOMORPHISM THEOREM FOR GROUPS or THE FUNDAMENTAL THEOREM OF GROUP
HOMOMORPHISMS

(i) If 𝜑 is a homorphism from a group 𝐺 into a group 𝐺̅ and 𝑁 = ker(𝜑) then G/N is an
isomorphism to 𝐺̅ .
(ii) Further, there is a 1 − 1 correspondence between subgroups of 𝐺̅ and those of G that
contain N.

𝑃𝑟𝑜𝑜𝑓 𝑜𝑓 𝑡ℎ𝑒 𝐹𝑖𝑟𝑠𝑡 𝑃𝑎𝑟𝑡

Since ker(𝜑) ∆𝐺, G/N exists. We have the following picture; where the map 𝜂: G → G/N is the
canonical homomorphism defined earlier on.

𝜑
𝐺 𝐺̅

G/N

We want to construct a map 𝛼: G → G/N → 𝐺̅ which has some relationship to , and then show that
𝛼 is an isomorphism.

𝐺̅
𝛼

32 | P a g e
Let 𝑁𝑔 ∈ G/N. WhatG/N
is the logical for 𝛼(𝑁𝑔)?

We need an element in 𝐺̅ related to both 𝜑 and 𝑁𝑔. So we try 𝛼(𝑁𝑔) = 𝜑(𝑔).

We must first show that while 𝑁𝑔 = 𝑁ℎ does not imply 𝑔 = ℎ. This ambiquity does not make 𝛼
multiple valued.

𝑁𝑔 = 𝑁ℎ

↔ 𝑁𝑔ℎ+ = 𝑁

↔ 𝑔ℎ−1 ∈ 𝑁 = ker(𝜑)

↔ 𝜑(𝑔)(𝜑(ℎ))−1 = 𝜑(𝑔)𝜑(ℎ−1 ) = 𝜑(𝑔ℎ−1 ) = 𝑒𝐺̅

↔ 𝛼(𝑁𝑔 ) = 𝜑(𝑔) = 𝜑(ℎ) = 𝛼(𝑁ℎ)

Thus 𝛼 is a function.

Since every step is reversible, reading from bottom to top yields that 𝛼 is one-to-one.

If 𝑔̅ ∈ 𝐺̅ and 𝜑 is onto, so 𝑔̅ = 𝜑(𝑔) for some 𝑔 ∈ 𝐺.

Thus, 𝛼(𝑁𝑔 ) = 𝜑(𝑔) = 𝑔̅ and 𝛼 is onto.

Since 𝑁∆𝐺, we have 𝑁𝑔ℎ = 𝑁𝑔𝑁ℎ.

Thus 𝛼(𝑁𝑔 )𝛼(𝑁ℎ ) = 𝜑(𝑔)𝜑(ℎ) = 𝜑(𝑔ℎ) = 𝛼(𝑁𝑔ℎ) = 𝛼(𝑁𝑔 𝑁ℎ )

So 𝛼 is a homorphism and thus an isomorphism ∎

To prove the second part, we start with a lemma.

𝐿𝑒𝑚𝑚𝑎

If 𝜑: 𝐺 → 𝐺̅ and 𝐻 ≥ 𝑁 = 𝐾𝑒𝑟𝜑 is a subgroup of G, then 𝜑−1 (𝜑(𝐻)) = 𝐻

𝑃𝑟𝑜𝑜𝑓 𝑜𝑓 𝐿𝑒𝑚𝑚𝑎

If 𝑥 ∈ 𝜑−1 (𝜑(𝐻)) then 𝜑(𝑥) ∈ 𝜑(𝐻).

So 𝜑(𝑥) = 𝜑(ℎ) for some ℎ ∈ 𝐻.

Thus, 𝜑(𝑥)𝜑(ℎ)−1 = 𝜑(𝑥ℎ−1 ) = 𝑒𝐺̅ ,

And 𝑥ℎ−1 ∈ ker(𝜑) = 𝑁.

33 | P a g e
Thus 𝑥 ∈ 𝑁ℎ ⊆ 𝐻ℎ = 𝐻

And 𝜑−1 (𝜑(𝐻)) ⊆ 𝐻.

By definition of 𝜑−1 we must have 𝜑−1 (𝜑(𝐻)) ≥ 𝐻

Therefore 𝜑−1 (𝜑(𝐻)) = 𝐻 ∎

𝑃𝑟𝑜𝑜𝑓 𝑜𝑓 𝑡ℎ𝑒 𝑆𝑒𝑐𝑜𝑛𝑑 𝑃𝑎𝑟𝑡

If 𝐻 and 𝐾 are two distinct subgroups of 𝐺 both containing 𝑁, then 𝜑(𝐻) and 𝜑(𝐾) are distinct in 𝐺̅ .

This follows from the lemma since 𝜑(𝐻) = 𝜑(𝐾) then 𝜑−1 (𝜑(𝐻)) = 𝜑−1 (𝜑(𝐻)) = 𝐻.

Thus 𝜑, considered as a function from the subgroups of 𝐺 containing 𝑁 to subgroups of 𝐺̅ , is one to

one.

̅ ⊂ 𝐺̅ , then 𝜑−1 (𝐻) is a subgroup of 𝐺 containing 𝑁 and 𝜑(𝜑−1 (𝐻

It is onto because if 𝐻 ̅ )) = 𝐻
̅.

KLEIN’S FOUR GROUP

𝐾𝑙𝑒𝑖𝑛’𝑠 𝐹𝑜𝑢𝑟 𝐺𝑟𝑜𝑢𝑝 is named after the German mathematician Felix Klein (1849-1925). The
German word for 4 − 𝐺𝑟𝑜𝑢𝑝 is 𝑉𝑒𝑟𝑔𝑟𝑢𝑝𝑝𝑒 denoted by 𝑉.

Let 𝐷8 → 𝑉 be given by 𝜑(𝑎) = 𝑎̅ and 𝜑(𝑏) = 𝑏̅ where 𝑏 4 = 𝑏 2 = 𝑒 and 𝑏 −1 𝑎𝑏 = 𝑎−1 in 𝐷8and

𝑎̅2 = 𝑏̅ 2 = (𝑎𝑏
̅̅̅)2 = 𝑒̅ ∈ 𝑉.

Klein’s Four Group 𝑉 = {𝑒, 𝑎, 𝑏, 𝑐} is not cyclic.

Then 𝑁 = 𝐾𝑒𝑟(𝜑) = {𝑒, 𝑎2 } ∈ 𝐷8 where 𝐷8 = {𝑒, 𝑎, 𝑎2 , 𝑎3 , 𝑏, 𝑎𝑏, 𝑎2 𝑏, 𝑎3 𝑏}

𝑎2 =< 𝑎 >= {𝑎, 𝑒}

𝑏 2 =< 𝑏 >= {𝑏, 𝑒}

𝑐 2 =< 𝑐 >= {𝑐, 𝑒}

𝑎2 = 𝑏 2 = 𝑐 2 = 𝑒, 𝑏𝑐 = 𝑐𝑏 = 𝑎, 𝑎𝑏 = 𝑏𝑎 = 𝑐, 𝑎𝑐 = 𝑐𝑎 = 𝑏

𝑉 = {𝑒, 𝑎, 𝑏, 𝑐}

Note: The subgroup lattice above 𝑁 corresponds with the subgroup lattice of 𝑉 = {𝑒, 𝑎, 𝑏, 𝑐}, is
abelian Klein’s four group but not cyclic.

34 | P a g e
𝐸𝑥𝑒𝑟𝑐𝑖𝑠𝑒

1. If 𝜑: 𝐺 → 𝐴 is a homorphism where 𝐺 is any group and 𝐴 is abelian, show that any subgroup
of 𝐺 containing 𝑁 is normal in 𝐺.
𝑜𝑛𝑡𝑜
2. If 𝐺 = (𝑅\{0}, ∙) is a group and 𝑔(𝑥) = 𝑒 𝑥 ; 𝜑:→ 𝑅 + is a homomorphism
(i) Compute the ker(𝜑).
(ii) Find the quotient group.
(iii) What is it isomorphic to?

N.B.: Know the second fundamental theorem for groups.

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛

Centralizer of an Element in a Group 𝐺

If 𝐺 is a group and 𝑦 ∈ 𝐺, then centralizer of 𝑦 is denoted by 𝑍(𝑦) = {𝑥 ∈ 𝐺|𝑥𝑦 = 𝑦𝑥}.

N.B.:

(i) 𝑍(𝑦) is always a subgroup of 𝐺.

(ii) 𝑍(𝐺) ⊆ 𝑍(𝑦) where 𝑍(𝐺) is the centre of 𝐺: 𝑍(𝐺) = {𝑥|𝑥𝑦 = 𝑦𝑥 ⩝ 𝑦 ∈ 𝐺}

Generally, 𝑍(𝑦)may be larger than 𝑍(𝐺).

𝐸𝑥𝑎𝑚𝑝𝑙𝑒𝑠

(i) If 𝐺 is abelian, then 𝑍(𝐺) = 𝐺, ⩝ 𝑦 ∈ 𝐺.

(ii) If 𝐺 = 𝑆З , then 𝑍(𝑎) = [𝑒, 𝑎, 𝑎2 ]. Also 𝑍(𝑏) = [𝑒, 𝑏].

𝐿𝑒𝑚𝑚𝑎

Let 𝐺 be a group, define a relation 𝑅 on 𝐺 by 𝑎𝑅𝑏 ↔ 𝑎 is conjugate to 𝑏 [i.e. 𝑎𝑅𝑏 ↔ 𝑎 ⁆ some 𝑥 ∈ 𝐺,

such that 𝑎 = 𝑥𝑏𝑥 −1 , then 𝑅 is an equivalence relation.

𝑃𝑟𝑜𝑜𝑓

(i) Reflexivity

If 𝑎 ∈ 𝐺, then 𝑎 = 𝑒𝑎𝑒 −1. So, 𝑎𝑅𝑎

35 | P a g e
 (ii) Symmetricity
If 𝑎𝑅𝑏, show 𝑏𝑅𝑎 by 𝑎𝑅𝑏 ↔ 𝑎 = 𝑥𝑏𝑥 −1 for some 𝑥.
Hence, 𝑏 = 𝑥 −1 𝑎𝑥 = 𝑥 −1 𝑎(𝑥 −1 )−1. So, 𝑏𝑅𝑎.

(iii) Transitivity
If 𝑎𝑅𝑏 and 𝑏𝑅𝑐, show 𝑎𝑅𝑐.
By assumption, 𝑎 = 𝑥𝑏𝑥 −1 and 𝑏 = 𝑦𝑐𝑦 −1 for some 𝑥, 𝑦 ∈ 𝐺.
Thus 𝑎 = 𝑥(𝑦𝑐𝑦 −1 )𝑥 −1 = (𝑥𝑦)𝑐(𝑥𝑦)−1
∴ 𝑎𝑅𝑐 and 𝑅 is an Equivalence Relation.

𝐷𝑒𝑓𝑖𝑛𝑖𝑡𝑖𝑜𝑛

Conjugate Class of an Element 𝑎 in Group 𝐺

The equivalence class 𝑎̅ of an element 𝑎 ∈ 𝐺 under the relation 𝑅 is called the conjugate class of a
and consists of all the conjugates of 𝑎 [i.e. 𝑎̅ = {𝑥𝑎𝑥 −1 |𝑥 ∈ 𝐺} = {𝑥 ∈ 𝐺|𝑥𝑅𝑎}. 𝑥 = 𝑦𝑎𝑦 −1.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒

In 𝑆З = {𝑒, 𝑎, 𝑎2 , 𝑏, 𝑎𝑏, 𝑎2 𝑏}, 𝑒̅ = {𝑒} = {𝑥𝑒𝑥 −1 |𝑥 ∈ 𝐺}

Also ̅𝑎 = {𝑎, 𝑎2 }

𝑏̅ = {𝑏, 𝑎𝑏, 𝑎2 𝑏}

Thus, 𝑅 breaks 𝑆З into classes i.e. 𝑆З = {𝑒} ∪ {𝑎, 𝑎2 } ∪ {𝑏, 𝑎𝑏, 𝑎2 𝑏}.

Our general aim is to count an arbitrary finite group 𝐺 by counting each conjugate class separately,
so, we need to find an expression for the number of conjugates of an element 𝑎.

In 𝑆З, 𝑎 has two conjugates and two is the index of 𝑍(𝑎) = {𝑎, 𝑎2 } in 𝑆З

𝑏 ∈ 𝑆З , has three conjugates and again this number is the index of 𝑍(𝑏).

𝑒 ∈ 𝑆З has only one conjugate and [𝑆З : 𝑍(𝑒)] = 1.

𝐿𝑒𝑚𝑚𝑎

Let 𝐺 be a finite group and 𝑎 ∈ 𝐺. Then the number of distinct conjugates of 𝑎 in 𝐺 is [𝐺: 𝑍(𝑎)].

𝑃𝑟𝑜𝑜𝑓

36 | P a g e
Let 𝑥, 𝑦 ∈ 𝐺.

Then 𝑥ℎ𝑥 −1 = 𝑦𝑎𝑦 −1

iff 𝑎𝑥 −1 𝑦 = 𝑥 −1 𝑦𝑎

iff 𝑥 −1 𝑦 ∈ 𝑍(𝑎)

iff 𝑥𝑍(𝑎) = 𝑦𝑍(𝑎)

In other words, 𝑥 and 𝑦 determines the same conjugate of 𝑎 iff they determine the same left coset
of 𝑍(𝑎), so the mapping 𝑥𝑎𝑥 −1 → 𝑥𝑍(𝑎) establishes a 1-1 correspondence between the conjugates
of 𝑎 and the left cosets of 𝑍(𝑎) in 𝐺.

By theorem, the number of left cosets of 𝑍(𝑎) is [𝐺: 𝑍(𝑎)].

So, the number of conjugates of 𝑎 is [𝐺: 𝑍(𝑎)].

𝑇ℎ𝑒𝑜𝑟𝑒𝑚

CLASS EQUATION

Let 𝐺 be a finite group and let {𝑎1 , 𝑎2 , … , 𝑎𝑘 } consist of one element from each conjugacy class
containing at least two elements. Then |𝐺| = |𝑍(𝐺)| + [𝐺: 𝑍(𝑎1 )] + [𝐺: 𝑍(𝑎2 )] + ⋯ + [𝐺: 𝑍(𝑎𝑘 )].

𝑃𝑟𝑜𝑜𝑓

Let 𝑎𝑘+1 , 𝑎𝑘+2 , … , 𝑎𝑘+𝑠 be the element of the conjugacy classes containing only one element each.

Then, we know that 𝐺 is the disjoint union. 𝐺 = 𝑎̅1 ∪ 𝑎̅2 ∪ … ∪ 𝑎̅𝑘 ∪ 𝑎̅𝑘+1 … ∪ 𝑎̅𝑘+𝑠 .

By lemma, 𝑎̅𝑗 contains exactly [𝐺: 𝑍(𝑎𝑗 )] elements, for 𝑗 = 1,2, … , 𝑘.

Thus, |𝐺| = [𝐺: 𝑍(𝑎1 )] + [𝐺: 𝑍(𝑎2 )] + ⋯ + [𝐺: 𝑍(𝑎𝑘+𝑠 )]

Also by lemma, for any 𝑎 ∈ 𝐺, 𝑎̅ consists of 𝑎 alone

↔ [𝐺: 𝑍(𝑎)] = 1

↔ 𝑍(𝑎) = 𝐺

↔ 𝑎 ∈ 𝑍(𝐺)

and therefore 𝑎𝑘+1 , 𝑎𝑘+2 , … , 𝑎𝑘+𝑠 are precisely the elements of 𝑍(𝐺). Hence, 𝑠 = |𝑍(𝐺)| ∎

N.B.: Perhaps the most noticeable difference between the situation here and that in the proof of
Lagrange’s Theorem is that it need not be the case that all the equivalence classes have the same
number of elements.

37 | P a g e
𝐸𝑥𝑎𝑚𝑝𝑙𝑒

𝑆3 = 𝑒̅ ∪ 𝑓 ̅ ∪ 𝑔̅ where 𝑒̅ = {𝑒}, 𝑓 ̅ = {𝑎, 𝑎2 }, 𝑔̅ = {𝑏, 𝑎𝑏, 𝑎2 }

So, the class equation 𝑆3 is 6 = 1 + 2 + 3.

𝐸𝑥𝑒𝑟𝑐𝑖𝑠𝑒𝑠

1. (a) Let 𝐺 = (𝑍48 ,⊕). Find (i) [𝐺: 𝐻] for 𝐻 =< 32 >
(b) Let 𝐺 = (𝑍54 ,⊕). Find (i) [𝐺: 𝐻] for 𝐻 =< 24 >
2. Let 𝐺 be a group of order 𝑝2 𝑝 prime. Show that 𝐺 must have a subgroup of order 𝑝.
3. Let 𝐺 be a group and suppose 𝑔 ∈ 𝐺 such that 𝑍(𝑔) = 𝑍(𝐺). Show that 𝐺 is abelian.
4. Let G be a finite group. Show that [𝐺: 𝑍(𝐺)] cannot be prime.

𝑅𝑒𝑚𝑎𝑟𝑘

𝑉∆𝐴4 since 𝑉 is the only subgroup of order 4, where 𝑉 = {𝑒, (12)(34), (13)(24), (14)(23)}

𝑉 ≈ 𝑍2 × 𝑍2 and it is abelian.

Let 𝑁1 = 𝐴4

𝑁2 = 𝑉

𝑁3 = {𝑒𝐺 }
𝑁 𝑉
So, there is the series {𝑁𝑖 }3𝑖=1 , with 𝑁1 /𝑁2 having three elements and 𝑁1 /𝑁2 ≈ 𝑍3 and 𝑁2 ≈ {𝑒} ≈ 𝑉.
1

Since both 𝑍3 and 𝑉 are abelian, 𝐴4 is solvable.

From example 3, is 𝑆4 solvable?

N.B.: 𝐴𝑛 for 𝑛 ≥ 5 is not solvable since it has no proper subgroups.

38 | P a g e
 RING THEORY

A ring (𝑅, +,∙) is a set 𝑅 with the binary operations + and ∙ such that

(i) (𝑅, +) is an abelian group.

(ii) ∙ is closed and associative.
(iii) Distributive law holds, i.e. 𝑥(𝑦 + 𝑡) = 𝑥𝑦 + 𝑥𝑡.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒𝑠

(𝑅, +,∙), (𝑄, +,∙), (𝑍, +,∙), (2𝑍, +), (𝑍𝑛 , +), ⍱𝑛 ∈ 𝑍 +

N.B.: If 𝐴 is abelian such that 𝑎𝑏 = 0, then 𝐴 is a trivial ring.

COMMUTATIVE RING

A ring 𝑅 is commutative if 𝑎𝑏 = 𝑏𝑎, ⍱𝑎, 𝑏 ∈ 𝑅.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒𝑠

(𝑍, +,∙), (𝑅, +,∙), (2𝑍, +)

RING WITH UNITY

A ring 𝑅 with an element 1 such that 1𝑎 = 𝑎1 = 𝑎, ⍱𝑎 ∈ 𝑅.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒

(𝑍, +,∙)

A ZERO DIVISOR IN A RING

If 𝑅 is a nonzero element 𝑎 such that a nonzero 𝑏 → 𝑎𝑏 = 0 or 𝑏𝑎 = 0. So, 𝑎 is a left-zero divisor

and 𝑏 is a right-zero divisor.

INTEGRAL DOMAIN

A commutative ring 𝑅 with unity 1 ≠ 0 satisfying the cancellation laws without zero divisor for
(𝑅\{0},∙) is called an integral domain.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒

(𝑍, +), 𝑍𝑝 , 𝑝 prime. Each 𝑟 ∈ 𝑍𝑝 satisfies (𝑟, 𝑝) = 1 and is therefore a unit in 𝑍𝑝 .

39 | P a g e
𝐸𝑥𝑎𝑚𝑝𝑙𝑒

Find all zero divisors in 𝑍6 .

𝑆𝑜𝑙𝑢𝑡𝑖𝑜𝑛

𝑍6 = {0,1,2,3,4,5}, then 23 = 0, 34 = 0. So, 2, 3 and 4 are zero divisors in 𝑍6 .

In 𝑍6 , let 𝑅 = {0̅, 2̅, 4̅}, then 𝑅 is commutative with 4 as unity, e.g. 4.2 = 2.4 = 2.

In 𝑍7 , find the unity element.

RING HOMOMORPHISM

If 𝑅 and 𝑆 are rings, then 𝜑: 𝑅 → 𝑆 is a ring homomorphism. If 𝜑(𝑔 + ℎ) = 𝜑(𝑔) + 𝜑(ℎ) and
𝜑(𝑔ℎ) = 𝜑(𝑔)𝜑(ℎ), ⍱𝑔, ℎ ∈ 𝑅.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒

𝑎 𝑏 𝑎 𝑏
𝜑: 𝐶 → {( ) |𝑎, 𝑏 ∈ 𝑅} defined by 𝜑(𝑎 + 𝑏𝑖) = ( ) is a ring homomorphism.
−𝑏 𝑎 −𝑏 𝑎

DIVISION RING

A division ring is a ring 𝑅 such that (𝑅\{0},∙) forms a group. Example: (𝜑\{0},∙) is a division ring.

FIELD

A field is a commutative division ring.

𝐸𝑥𝑎𝑚𝑝𝑙𝑒𝑠

(𝜑, +,∙), (𝑅, +,∙)

N.B.: 𝑅 is called a division ring if 𝑅 has a unity 1 ≠ 0 and every nonzero element of 𝑅 is a unit. By
unit 𝑎, we mean 𝑎 has an inverse in 𝑅 (multiplicative).

𝑇ℎ𝑒𝑜𝑟𝑒𝑚

Let 𝑅 be a ring with 𝑎, 𝑏 ∈ 𝑅. Suppose 𝑎 is not a zero divisor, then if 𝑎𝑏 = 𝑎𝑐, 𝑏 = 𝑐.

𝑃𝑟𝑜𝑜𝑓

𝑎𝑏 = 𝑎𝑐 → 𝑎𝑏 − 𝑎𝑐 = 0 → 𝑎(𝑏 − 𝑐) = 0. Since 𝑎 is not a zero divisor, then (𝑏 − 𝑐) = 0 → 𝑏 = 𝑐.

40 | P a g e
𝐶𝑜𝑟𝑜𝑙𝑙𝑎𝑟𝑦

Let 𝑅 be a commutative ring with unity 1 ≠ 0. Then, 𝑅 is an Integral Domain iff whenever 𝑎, 𝑏 ∈ 𝑅
satisfy 𝑎𝑏 = 𝑎𝑐 and 𝑎 ≠ 0, then 𝑏 = 𝑐.

𝑃𝑟𝑜𝑜𝑓

Assume 𝑅 is Integral Domain with 𝑎, 𝑏 ∈ 𝑅 and 𝑎 ≠ 0.

Then, 𝑎 is not a zero divisor.

Then by the theorem above 𝑏 = 𝑐 ∎

Suppose 𝑅 is not an Integral Domain, we find 𝑎, 𝑏, 𝑐 ∈ 𝑅, 𝑎 ≠ 0 such that 𝑎𝑏 = 𝑎𝑐 but 𝑏 ≠ 0.

In fact, we know that there is a nonzero divisor 𝑎 ∈ 𝑅.

Let 𝑏 ∈ 𝑅 such that 𝑏 ≠ 0 and 𝑎𝑏 = 0. Then, 𝑎𝑏 = 𝑎0 but 𝑏 ≠ 0 contradiction.

𝑇ℎ𝑒𝑜𝑟𝑒𝑚

Every finite Integral Domain is a field.

𝑃𝑟𝑜𝑜𝑓

Let 𝑅 be finite. Then 𝑅 is commutative and 1 ≠ 0 in 𝑅.

Show that if 𝑟 ≠ 0, 𝑟 ∈ 𝑅.

Then 𝑟 has a multiplicative inverse in 𝑅.

Since 𝑅 is finite, its elements are {𝑟1 , 𝑟2 , … , 𝑟𝑛 }. Consider the elements {𝑟𝑟1 , 𝑟𝑟2 , … , 𝑟𝑟𝑛 }.

Since 𝑟 ≠ 0, these are distinct since they are all in 𝑅. These must account for all elements in 𝑅.

In particular, one of them is 1.

So, 𝑟𝑟1 = 𝑟1 𝑟 = 1. For some 1 and 1 is a unit.

41 | P a g e
TEST

1. Define a group 𝐺 and show that 𝐺 = (𝑍, +) is group.

2. Let 𝐺 = (𝑍, +) and 𝐻 = (2𝑍, +).
(i) Find the cosets of 𝐻 in 𝐺.
(ii) Is 𝐻 a subgroup of 𝐺.
3. Show that the identity in a group is unique.
1 2 3
4. If 𝐺 = 𝑆3 and ( ) ∈ 𝐺. Find
2 3 1
1 2 3 −1
(i) ( ) .
2 3 1
1 2 3 1 2 3
(ii) ( )( ) ∈ 𝑆3 .
2 3 1 3 1 2

42 | P a g e