Math 110B - Group Theory Lecture Notes

MATH 110B - GROUP THEORY
MATTHEW GHERMAN
These notes are based on Hungerford, Abstract Algebra 3rd edition.
Contents
1. Groups 2
1.1. Definition of a group 2
1.2. Basic properties of groups 4
1.3. Subgroups 7
1.4. Isomorphisms and Homomorphisms 12
2. Normal Subgroups and Quotient Groups 17
2.1. Congruence and Lagrange’s Theorem 17
2.2. Normal Subgroups 20
2.3. Quotient Groups 22
2.4. Isomorphism Theorems 25
2.5. The Symmetric and Alternating Groups 30
3. Topics in Group Theory 33
3.1. Direct Products 33
3.2. Finite Abelian Groups 37
3.3. The Sylow Theorems 42
3.4. Conjugacy and the Proof of Sylow’s Theorems 44
3.5. The Structure of Finite Groups 48
4. Applications of Group Theory 53
4.1. Group Actions 53
4.2. Solvable and Nilpotent Groups 58
1
2 MATTHEW GHERMAN
1. Groups
1.1. Definition of a group.
Definition 1.1. A group is a non-empty set G closed under a binary operation · that satisfies the
following three axioms.
(1) (Associativity) For all a, b, c ∈ G, (a · b) · c = a · (b · c).
(2) (Identity) There is an element e ∈ G such that a · e = e · a = a for all a ∈ G.
(3) (Inverses) For each a ∈ G, there is some b ∈ G such that a · b = b · a = e.
Definition 1.2. A group is abelian if the binary operation is commutative. In other words,
a · b = b · a for all a, b ∈ G.
Example 1.1. Every ring under addition is an abelian group. For instance, (Z, +) or (Q, +).
However, (Z, ·) is not a group since there is no multiplicative inverse to 21 . We can make (Q, ·) into
a group by removing zero.
Example 1.2. Let R be a ring. Recall that a unit in R is an element that has a multiplicative
inverse. The set of units R× of R is a group under multiplication.
The quintessential example is (Z/nZ)× . From Math 110A, we know that an equivalence class a
is invertible in Z/nZ if and only if a and n are relatively prime as integers. Thus the number of
elements in (Z/nZ)× is the Euler totient function φ(n).
The multiplication table for (Z/8Z)× is the following.
· 1 3 5 7
1 1 3 5 7
3 3 1 7 5
5 5 7 1 3
7 7 5 3 1
We find that each non-identity element is its own inverse. Further, the symmetry along the left to
right diagonal reveals that (Z/8Z)× is an abelian group.
Since each element has an inverse, each row and column of a group multiplication table will
contain one and only one copy of each element of the group. The group multiplication table is
much like a Sudoku puzzle.
Example 1.3. Let GLn (F ) be invertible n × n matrices with coefficients in a field F . We can
likewise classify GLn (F ) by the n × n matrices A with det(A) ̸= 0. The common choices for F are
R, C, or Z/pZ for p a prime integer. Except for n = 1, these are non-abelian groups.
Example 1.4. Many groups arise as symmetries of shapes. Take an equilateral triangle with
labeled vertices 1, 2, and 3. We can reflect the triangle about any axis through one of the vertices.
Likewise, we can rotate by an angle of 2π3

counterclockwise. The set of these symmetries form
a group under composition called a dihedral group. We can replace the equilateral triangle with
any regular n-gon to obtain a dihedral group denoted Dn . Let r represent rotation by 2π 3
and s
2 2
represent a reflection. The elements of D3 are {e, r, r , s, sr, sr }. This is another example of a
non-abelian group since the element rs is not the same as sr.
MATH 110B - GROUP THEORY 3
Example 1.5. Let Xn = {1, 2, . . . , n}. The set Sn of bijections of Xn under composition is a
group called the symmetric group on n elements. For n ≥ 3, Sn will be non-abelian.
As an example, take n = 3. The permutation
1→2
2→3
3→1
is denoted (123). Each number is sent to the number on its right while the rightmost number is
sent to the first. Another permutation is (12). Reading left to right,
1→2
2→1
3 → 3.
The set S3 is {e, (12), (13), (24), (123), (132)} where e is the identity element that keeps each
number fixed. We compose permutations as functions compose or, in other words, we read right
to left. For example, (123)(12) is the permutation
1→2→3
2→1→2
3→3→1
so (123)(12) = (13). The permutation (12)(123) is
1→2→1
2→3→3
3→1→2
so (12)(123) = (23).
With some more machinery, we will show that S3 is the same as the dihedral group of the
equilateral triangle. For i, j, k ∈ {1, 2, 3}, an element of the form (ijk) is a rotation while an
element of the form (ij) is a reflection about the axis through vertex k. For larger n, however, the
group Sn is much larger than Dn .
End of lecture 1
4 MATTHEW GHERMAN
1.2. Basic properties of groups.

Proposition 1.1. Let G be a group and a, b, c ∈ G.
(1) G has a unique identity element.
(2) If ab = ac or ba = ca, then b = c.
(3) Each element of G has a unique inverse.
Proof. (1) Let e and e′ be two elements such that ae = ea = a and ae′ = e′ a = a for all a ∈ G.
Then e = ee′ = e′ so there is exactly one identity element.
(2) Let ab = ac. Multiply both sides by a−1 on the left to obtain a−1 (ab) = a−1 (ac). By
associativity, (a−1 a)b = (a−1 a)c or eb = ec. We conclude that b = c. The other claim can
be shown via right multiplication.
(3) Suppose d and d′ are inverse of a. Then ad = e = ad′ . By (2), d = d′ as desired.
□
The result of Proposition 1.1 allows us to denote the unique inverse of a ∈ G by a−1 .
Corollary 1.1. Let G be a group with a, b ∈ G.
(1) (ab)−1 = b−1 a−1
(2) (a−1 )−1 = a
Proof. (1) We want to show b−1 a−1 is the inverse of ab. By associativity,
(b−1 a−1 )(ab) = b−1 (a−1 a)b
= b−1 eb
= b−1 b
= e.
A similar argument can be made to show (ab)(b−1 a−1 ) = e. The uniqueness of the inverse
proves (ab)−1 = b−1 a−1 .
(2) We want to show that a is the inverse of a−1 . We have
aa−1 = e
a−1 a = e
so a is the inverse of a−1 .
□
For a positive integer n and a ∈ G, we will denote by an the product of n copies of a. We will
denote a−n to be the product of n copies of a−1 . For convenience, we may write a0 = e.
Remark 1.1. For a non-abelian group G, note that (ab)n might not be equal to an bn .
The following exponent rules do hold, however.
Proposition 1.2. Let G be a group with a ∈ G. For all m, n ∈ Z,
am an = am+n
(am )m = amn .
Proof. We will prove the result for n ≥ 0 inductively. The other cases will be left as an exercise.
As a base case, let n = 0. Then am a0 = am = am+0 . As the inductive step, assume that
a a = am+k for all 0 ≤ k ≤ n − 1. We want to show, am an = am+n . Via associativity and the
m k
inductive hypothesis,
am an = (am an−1 )a
= am+n−1 a
= am+n .
As a base case, let n = 0. Then (am )0 = e = am·0 . Assume that (am )k = amk for all 0 ≤ k ≤ n−1.
By the inductive hypothesis and the above result,
(am )n = (am )n−1 am
= am(n−1) am
= am(n−1)+m
= amn .
□
Remark 1.2. Let G be an abelian group with a ∈ G. The group operation will sometimes be
denoted by + instead of ·. In these cases, the identity element will be denoted 0 and the inverse
of a ∈ G will be −a. Further, na is the sum of n copies of a and −na is the sum of n copies of
a−1 . We will call G an additive group.
Definition 1.3. Let G be a group with a ∈ G. If aM = e for a positive integer M , then the order
of a is the smallest positive integer k for which ak = e. If no such M exists, then a is said to have
infinite order. We denote by |a| the order of a.
Definition 1.4. The size of a group G will also be called the order of G. We will use the notation
|G| to denote the order of the group G.
The connection between order of elements and order of groups will be explored next chapter.
Example 1.6. What is the order of the element (123) in S3 ? The element (123)2 is
1→2→3
2→3→1
3→1→2
so (123)2 = (132). Then (123)3 = (123)(123)2 = (123)(132) is
1→3→1
2→1→2
3 → 2 → 3.
We conclude that (123)3 = e and |(123)| = 3. Further, we can show |(132)| = 3 and |(ij)| = 2 for
any distinct i, j ∈ {1, 2, 3}.
In general, the order of Sn is n! since there are n! different ways of permuting a set of size n.
Example 1.7. In the additive group Z/12Z, the element 8 has order 3.
8+8=4
8+8+8=8+4=0
Proposition 1.3. Let G be a group with a ∈ G.
(1) If ai = aj with i ̸= j, then a has finite order.
(2) If a has infinite order, then the elements ak for k ∈ Z are distinct.
6 MATTHEW GHERMAN
Proof. (1) Assume ai = aj for i ̸= j. Without loss of generality, assume that i > j. Multiply
both sides by a−j . Then ai−j = e and a has finite order.
(2) The statement is the contrapositive of (1) and, thus, equivalent.
□
We can reduce many group theory problems to matters of elementary number theory via the
following result.
Proposition 1.4. Let G be a group with a ∈ G an element of finite order n.
(1) ak = e if and only if n|k
(2) ai = aj if and only if i ≡ j (mod n)
(3) If n = td with d ≥ 1, then at has order d
Proof. (1) (⇒) Suppose ak = e. By the division algorithm, k = nq + r with 0 ≤ r < n. Thus
ak = anq+r
= (an )q ar
= e q ar
= ar .
Since ak = e, we conclude ar = e. The order of a is n so n is the smallest positive integer
for which an = e. Since r is strictly less than n, we have r = 0.
(⇐) Assume k = nt. Then ak = ant = (an )t = et = e. End of lecture 2
(2) Equivalenty, we will prove that ai−j = e if and only if i − j ≡ 0 (mod n). By (1), ai−j = e
if and only if n divides i − j. This is equivalent to i ≡ j (mod n).
(3) We have (at )d = atd = an = e. We want to prove that d is the smallest such positive
exponent. Let k be a positive integer such that (at )k = e. Then atk = e and n divides tk
by (1). We can write tk = nr = (td)r for some integer r. We have k = dr. Since k and d
are positive and d|k, we have d ≤ k.
□
Lemma 1.1. Let G be a group. Let a, b ∈ G be elements of finite order such that ab = ba. If
gcd(|a|, |b|) = 1, then |ab| = |a||b|.
Proof. Let m = |a|, n = |b|, and k = |ab|. We have
(ab)mn = amn bmn = (am )n (bn )m = enG em
G = eG
so k divides mn by Proposition 1.4(1).
Next,
eG = enG = ((ab)k )n = (ab)kn = akn bkn = akn (bn )k = akn eG = akn
so Proposition 1.4(1) implies that m divides kn. Since m and n are relatively prime, m divides k.
The same argument with m and n swapped proves that n divides k as well. Since m and n are
relatively prime, mn divides k and k = mn. □
Lemma 1.2. Let G be a group. Let a, b ∈ G be elements of finite order such that ab = ba. Let
m = |a| and n = |b|. Then |ab| divides lcm(m, n).
Proof. We have ℓ = lcm(m, n) where ℓ = rm and ℓ = sn for some integers r, s ∈ Z. Since a and b
commute,
(ab)ℓ = aℓ bℓ = arm bsn = (am )r (bn )s = eG .
Therefore, |ab| divides ℓ by Proposition 1.4(1). □
Note that the |ab| is not necessarily lcm(|a|, |b|) when b = a−1 for instance.
1.3. Subgroups.
Definition 1.5. A non-empty subset H of G is a subgroup of G if the following hold.
(1) (Closure under group operation) For each a, b ∈ H, ab ∈ H.
(2) (Closure under inverse) For each a ∈ H, a−1 ∈ H.
Every non-trivial group G has at least two subgroups: the trivial subgroup {e} and G itself. All
other groups are called proper subgroups.
Proposition 1.5. Let H be a finite non-empty subset of a group G. If H is closed under the
group operation in G, then H is a subgroup of G.
Proof. We only need to check that H is closed under inverse. If a ∈ H, then closure under the
group operation implies ak ∈ H for all positive integers k. Since H is finite, a is an element of
finite order. Let |a| = n. If n = 1, then a = e and we are done. Assume n > 1. Since n − 1 ≡ −1
(mod n), Proposition 1.4 implies that an−1 = a−1 . □
Example 1.8. The subset
1 a
H= :a∈R
0 1
is a subgroup of GL2 (R). We have

1 a 1 b 1 a+b
= ∈H
0 1 0 1 0 1
−1
1 a 1 −a
= ∈H
0 1 0 1
Example 1.9. Recall the dihedral group D3 has elements {e, r, r2 , s, sr, sr2 } where r represents a
rotation of an equilateral triangle by an angle of 2π 3
counterclockwise and s represents reflection
about an axis through a vertex of the equilateral triangle. The subset H = {e, r, r2 } is the subgroup
of rotations of D3 . Likewise, K = {e, sri } is a subgroup of D3 containing one of the reflections for
each 0 ≤ i ≤ 2.
Example 1.10. Let S4 be the symmetric group containing all permutations of X4 = {1, 2, 3, 4}.
The subset H of all permutations that fix 4 is a subgroup. In fact, this subgroup is basically the
same as S3 since it contains all possible permutations of the subset {1, 2, 3} in X4 .
Definition 1.6. The quaternion group is the set Q8 = {1, −1, i, −i, j, −j, k, −k} where −i is the
inverse of i, −j is the inverse of j, and −k is the inverse of k. Further,
(−1)2 = 1
i2 = j 2 = k 2 = −1
ij = k
ji = −k.
The quaternion group is inspired by the standard unit vectors, {i, j, k}, of R3 where we operate
via the cross-product. Right-hand rule implies that i × j = k and j × i = −k. Q8 is a helpful
example of a small non-abelian group.
Example 1.11. The subsets {1, −1} and {1, i, i2 = −1, i3 = −i} are subgroups of Q8 .
If we want a subgroup H to contain {i, j}, then it must contain the inverses {−i, −j} and be
closed under multiplication. Thus ij = k ∈ H, and we conclude H = Q8 .
8 MATTHEW GHERMAN
1.3.1. Center of a group.
Definition 1.7. The center of a group G is
Z(G) = {h ∈ G : gh = hg for all g ∈ G}.
In other words, an element of G is in the center if and only if it commutes with all other elements
of G. If G is an abelian group, then Z(G) = G. If G is not abelian, then Z(G) will be a proper
subset of G.
End of lecture 3
Proposition 1.6. The center of G is a subgroup of G.
Proof. The identity of G is always in the center so Z(G) is non-empty.

Let h1 , h2 ∈ Z(G). Then (h1 h2 )g = h1 gh2 = g(h1 h2 ) for all g ∈ G. Thus h1 h2 ∈ Z(G).
Let h ∈ G. For each g ∈ G, we have gh−1 = (hg −1 )−1 = (g −1 h)−1 = h−1 g. Thus h−1 ∈ Z(G). □

a b
Example 1.12. We will compute the center of GL2 (R). Let A = be an element of the
c d

0 1
center of GL2 (R). Take B = . Then
1 0

b a
AB =
d c

c d
BA =
a b

1 1
are equal so a = d and b = c. Take C = . Then
0 1

a a+b
AC =
c c+d

a a+b
=
b a+b

a+c b+d
CA =
c d

a+b a+b
=
b a

a 0 ×
are equal so b = c = 0. We conclude that Z(GL2 (R)) = :a∈R is the subset of
0 a
invertible scalar matrices.
Example 1.13. The elements of D4 are symmetries of the square. We can rotate by a multiple
of π2 or reflect about one of the axes pictured below.
Let r be a rotation by π2 counterclockwise and s be a reflection. Then the elements of D4 are

{e, r, r2 , r3 , s, sr, sr2 , sr3 }. We find
rs = sr3
r3 s = sr
so r, r3 , and s are not elements of the center. Further,
r(sri ) = sri+3
(sri )r = sri+1
where i + 3 ̸≡ i + 1 (mod 4). Thus sri is not an element of the center for each 0 ≤ i ≤ 3. We can
show that r2 does commute with each element so Z(D4 ) = {e, r2 }.
Example 1.14. We will show that Z(S3 ) = {e}. Let i, j, k ∈ {1, 2, 3} be distinct. Then
(ij)(ijk) = (jk)
(ijk)(ij) = (ik)
proves that no non-trivial element can be in the center of G. By a similar argument, we can show
that Z(Sn ) = {e} for all n ≥ 3. We can interpret the result as symmetric groups are as far from
abelian as possible.
Example 1.15. Since ij = k ̸= −k = ji and jk = i ̸= −i = kj, the center of the quaternion
group does not contain {±i, ±j, ±k}. We can show that Z(Q8 ) = {1, −1}.
1.3.2. Cyclic subgroups.
Proposition 1.7. If G is a group and a ∈ G, then ⟨a⟩ = {an : n ∈ Z} is a subgroup of G.
Proof. Let H = ⟨a⟩. Each element of H is ak for some k ∈ Z. Then
a−k ∈ H
ak aℓ = ak+ℓ ∈ H.
□
Definition 1.8. The group ⟨a⟩ is the cyclic subgroup generated by a. If G = ⟨a⟩, then G is a cyclic
group. Note that every cyclic group is abelian since ak aℓ = ak+ℓ = aℓ ak .
Some of the most useful examples of groups are cyclic. We will be able to describe a large class
of groups by their cyclic subgroups.
Example 1.16. The group (Z/9Z)× = {1, 2, 4, 5, 7, 8} is a cyclic group generated by 2. However,
the element 8 is order 2 and does not generated the group.
Proposition 1.8. Let G be a group with a ∈ G.
(1) If a has infinite order, then ⟨a⟩ is an infinite subgroup consisting of the distinct elements
ak for k ∈ Z.
(2) If a has finite order n, then ⟨a⟩ is a subgroup of order n and ⟨a⟩ = {e, a, a2 , . . . , an−1 }.
Proof. The result follows from Proposition 1.3. □
Remark 1.3. There is an additive version of Proposition 1.8. Let a have finite order n of an
additive group G. Then ⟨a⟩ = {0, a, 2a, . . . , (n − 1)a}.
Example 1.17. The group Z is an infinite cyclic group generated by 1.
Example 1.18. The group Z/nZ is a finite cyclic group generated by 1 for each n.
10 MATTHEW GHERMAN
Example 1.19. For G = D4 , examples of cyclic subgroups of D4 are ⟨r⟩ or ⟨sri ⟩ for 0 ≤ i ≤ 3.
The subgroup H = ⟨r2 , s⟩ is a non-cyclic subgroup.
Proposition 1.9. Every subgroup of a cyclic group is cyclic.
Proof. Let G = ⟨a⟩ and H a subgroup of G. If H = ⟨e⟩, then H is the cyclic subgroup generated
by e. If H ̸= ⟨e⟩, then there is some non-identity element ai . By possibly taking an inverse, we
may assume i is positive. Let k be the smallest positive integer for which ak ∈ H. We will prove
that H = ⟨ak ⟩. Let h ∈ H so h = am for some m. By the division algorithm, m = kq + r for
0 ≤ r < k. Thus
ar = am−kq = am a−kq = am (ak )−q ∈ H.
Since k is the smallest positive integer, r = 0 and h = am = (ak )q as desired. □
End of lecture 4
1.3.3. Generators of a group. Cyclic subgroups are the easiest groups to describe. It turns out we
can describe groups as products of cyclic subgroups. Often large groups can be described with
only a small set of fundamental elements.
Definition 1.9. Let S be a non-empty subset of a group G. Let ⟨S⟩ be the subset of all possible
finite products of integer powers of elements of S. We will refer to ⟨S⟩ as the subgroup generated
by S.
If S = {a}, then ⟨S⟩ is the cyclic subgroup generated by a. If G = ⟨S⟩, then we say that G
is generated by S. The next result proves that ⟨S⟩ is the smallest subgroup of G containing the
subset S.
Proposition 1.10. (1) ⟨S⟩ is a subgroup of G that contains S.
(2) If H is a subgroup of G that contains the set of S, then H contains the subgroup ⟨S⟩.
Proof. (1) ⟨S⟩ is non-empty because the set S is non-empty and every element of S (considered
as a one-element product) is an element of ⟨S⟩. If a, b ∈ ⟨S⟩, then
a = a1 a2 · · · ak
where k ≥ 1 and each ai is either an element of S or the inverse of an element of S.
Similarly,
b = b1 b2 · · · bt
with t ≥ 1 and each bj is either an element of S or the inverse of an element of S. Therefore,
the product
ab = a1 a2 · · · ak b1 b2 · · · bt
consists of elements of S or inverses of elements of S. Hence, ab ∈ ⟨S⟩ and ⟨S⟩ is closed
under products. Further,
a−1 = a−1 −1 −1
k · · · a2 a1
is in ⟨S⟩ so ⟨S⟩ is closed under inverses.
(2) Any subgroup that contains the set S must include the inverse of every element of S. By
closure, this subgroup must also contain all possible products, in every order, of elements
of S and their inverses. Therefore, every subgroup that contains S must also contain the
entire group ⟨S⟩.
□
2π
Example 1.20. The group Dn is generated by the set S = {r, s} where r is the rotation by n
counterclockwise and s is a reflection.
Example 1.21. As we noted in Example 1.11, Q8 is generated by {i, j}.
Definition 1.10. Let G1 and G2 be two groups. The Cartesian product is the set of ordered pairs
G1 × G2 = {(g1 , g2 ) : g1 ∈ G1 , g2 ∈ G2 }
with the operation (g1 , g2 ) · (h1 , h2 ) = (g1 · h1 , g2 · h2 ).
We will be able to describe many groups as the Cartesian product of the above group examples.
Example 1.22. Let G = ⟨S⟩ and H = ⟨T ⟩ be two groups. Then the Cartesian product G × H is
generated by the set of ordered pairs (S × {eH }) ∪ ({eG } × T ).
Once we determine a set of generators for a group, much of the information of the group can be
determined by checking only on the generators. For instance, the next result shows that we can
check that a group is abelian on its generators.
Proposition 1.11. Let G = ⟨S⟩. If ab = ba for any two a, b ∈ S, then G is abelian.
Proof. Let x, y ∈ G. Then x = a1 a2 · · · ak for ai ∈ S and y = b1 b2 · · · bℓ for bj ∈ S. Since each ai
commutes with each bj ,
ab = (a1 a2 · · · ak )(b1 b2 · · · bℓ )
= (a1 a2 · · · ak−1 )(b1 b2 · · · bℓ )ak
..
.
= (b1 b2 · · · bℓ )(a1 a2 · · · ak )
= ba.
□
12 MATTHEW GHERMAN
1.4. Isomorphisms and Homomorphisms. There are many examples of groups that use dif-
ferent symbols to represent the same underlying structure. These groups will be referred to as
isomorphic groups. In practice, we will treat isomorphic groups as the same. Once we identify
elements properly, there is no discernible difference except in the symbols we use to represent them.
Example 1.23. Below are the multiplication tables of (Z/5Z)× and Z/4Z.
· 1 2 3 4 + 0 1 2 3
1 1 2 3 4 0 0 1 2 3
2 2 4 1 3 1 1 2 3 0
3 3 1 4 2 2 2 3 0 1
4 4 3 2 1 3 3 0 1 2
Via the identification
1 7→ 0
2 7→ 1
3 7→ 3
4 7→ 2,
we can see that the two groups have the same underlying structure since the product of any two
elements in (Z/5Z)× corresponds to the product of the corresponding elements in Z/4Z. In fact,
we will prove in Proposition 1.16 that any two cyclic groups of the same order will be isomorphic
by identifying a generator of one with a generator of the other.
Writing down multiplication tables is not an effective technique for large finite or infinite groups.
Thus we will want a simpler formal process for finding when two groups are the same. In short,
a group homomorphism is a function that preserves the multiplication structure on groups. This
should remind you of the definition of a ring homomorphism except with only one operation
involved.
Definition 1.11. Let G and H be groups. A function f : G → H is a group homomorphism if
f (ab) = f (a)f (b) for all a, b ∈ G.
Proposition 1.12. Let f : G → H be a group homomorphism.
(1) f (eG ) is the identity of H.
(2) f (a−1 ) = f (a)−1 for every a ∈ G.
Proof. (1) For g ∈ G, we have f (g) = f (geG ) = f (g)f (eG ). Multiply on the left by f (g)−1 to
prove that eH = f (eG ).
(2) By (1),
f (a−1 )f (a) = f (a−1 a) = f (eG ) = eH .
Thus f (a−1 ) is the inverse of f (a).
□
The other piece of an isomorphism is that there is a one-to-one correspondence between the
elements of the respective groups.
Definition 1.12. Let A and B be sets. A function f : A → B is injective if f (x) = f (y) implies
x = y. The function is surjective if for each z ∈ B, there is some x ∈ A for which f (x) = z. The
function is a bijection if f is both injective and surjective. Equivalently, a function is bijective if
it has an inverse.
Definition 1.13. Let G and H be groups. A function f : G → H is a group isomorphism if f is a
bijective group homomorphism. We write G ≃ H when there is a group isomorphism f : G → H.
Example 1.24. Let E = 2Z be the additive group of even integers. Define f : Z → E as

f (n) = 2n. We claim that f is a group isomorphism. Note that
f (m + n) = 2(m + n) = 2m + 2n = f (m) + f (n)
so f is a group homomorphism. If f (m) = f (n), then 2m = 2n and m = n by dividing by 2.
Thus f is injective. Since E is the set of even integers, f is surjective. We conclude that f is an
isomorphism and Z ≃ E. This is another example of two isomorphic cyclic groups.
Example 1.25. Let K be the multiplicative group of positive real numbers. Define f : R → K
as f (r) = 10r . We claim that f is a group isomorphism. Then
f (r + s) = 10r+s = 10r 10s = f (r)f (s)
so f is a group homomorphism. If f (r) = f (s), then 10r = 10s and r = s by taking log10 of both
sides. Thus f is injective. For each k ∈ K, we find
f (log10 (k)) = 10log10 (k) = k
so f is surjective. We conclude that f is a group isomorphism and R ≃ K.
Example 1.26. The function f : R× → R× given by f (r) = r2 is a group homomorphism since
f (rs) = (rs)2 = r2 s2 = f (r)f (s).
However, f (−1) = f (1) = 1 so f is not injective. Further, f is not surjective since f (r) = r2 ≥ 0.
Example 1.27. The function f : Z → Z/nZ given by f (a) = a is a homomorphism of additive
groups since
f (a + b) = a + b = a + b = f (a) + f (b).
The function is surjective but not injective. For a ≡ b (mod n), we have f (a) = f (b).
Example 1.28. Let G and H be groups. The function f : G × H → G given by f (g, h) = g is
projection onto the first coordinate. The function f is a surjective group homomorphism that is
not injective for non-trivial H.
Example 1.29. A bijection between finite sets implies that the sets have the same size. Thus two
finite groups of different order cannot be isomorphic. It is not generally true that two finite groups
of the same size are isomorphic, however.
End of lecture 5
Proposition 1.13. If G is abelian and H is non-abelian, then G and H are not isomorphic.
Proof. We will prove the contrapositive. Assume that f : G → H is an isomorphism and G is
abelian. For each h1 , h2 ∈ H, there exist g1 , g2 ∈ G such that f (g1 ) = h1 and f (g2 ) = h2 . Then
h1 h2 = f (g1 )f (g2 ) = f (g1 g2 ) = f (g2 g1 ) = f (g2 )f (g1 ) = h2 h1 .
We conclude that H is abelian. □
Note that in the proof of the above result we only use the surjectivity of f . Thus we can relax
the assumption to any surjective group homomorphism.
Example 1.30. The groups Z/6Z and S3 have the same order but are not isomorphic by Propo-
sition 1.13.
Proposition 1.14. If f : G → H is an isomorphism, then a ∈ G and f (a) ∈ H have the same
order.
14 MATTHEW GHERMAN
Proof. Let n be the order of a and k the order of f (a). Then

f (a)n = f (an ) = f (eG ) = eH
by Proposition 1.12(1) so k ≤ n. Further,
f (a)k = f (ak ) = eH .
Since f (eG ) = eH and f is injective, ak = eG . Thus k ≥ n. □
Note that we only needed f to be injective in the above argument. Thus the same result is true
for any injective group homomorphism. We state and prove a generalization for general group
homomorphisms.
Proposition 1.15. Let a ∈ G be an element of finite order. If f : G → H is a homomorphism,
then |f (a)| divides |a|.
Proof. Let n = |a|. Then f (a)n = f (an ) = f (eG ) = eH by Proposition 1.12(1). Thus Proposition
1.4(1) implies that |f (a)| divides n = |a|. □
Example 1.31. The groups Z/4Z and Z/2Z × Z/2Z are not isomorphic by Proposition 1.14.
Every element of Z/2Z × Z/2Z has order at most 2 while 1 and 3 have order 4 in Z/4Z.
Example 1.32. Let c be a fixed element of a group G. Define the function f : G → G as
f (a) = cac−1 . This is called conjugation by c. For a, b ∈ G,
f (ab) = c(ab)c−1 = ca(c−1 c)bc−1 = (cac−1 )(cbc−1 ) = f (a)f (b)
so f is a group homomorphism. The function g : G → G given by g(a) = c−1 ac is inverse to f so
f is an isomorphism.
An isomorphism from a group to itself is an automorphism. An automorphism defined by
conjugation by an element of G is an inner automorphism.
Proposition 1.16. Let G be a cyclic group.
(1) If G is infinite, then G is isomorphic to the additive group Z.
(2) If G is finite of order n, then G is isomorphic to the additive group Z/nZ.
Proof. (1) Suppose G = ⟨a⟩ is an infinite cyclic group. Define the function f : G → Z as
f (ak ) = k. Then f is a homomorphism since
f (ai aj ) = f (ai+j ) = i + j = f (ai ) + f (aj ).
If f (ai ) = f (aj ), then i = j and f is injective. For each i ∈ Z, the element ai ∈ G satisfies
f (ai ) = i so f is bijective. Therefore, G ≃ Z.
(2) Suppose G = ⟨b⟩. Define the function f : G → Z/nZ as f (bk ) = k. Since there are multiple
ways of writing an element of G, we need to check that f is well-defined. Let bk = bℓ . Then
f (bk ) = k = ℓ = f (bℓ ) by Proposition 1.4(2). Further, f is a homomorphism since
f (bi bj ) = f (bi+j ) = i + j = i + j = f (bi ) + f (bj )
If f (bi ) = f (bj ), then i ≡ j (mod n). By Proposition 1.4, bi = bj when i ≡ j (mod n).
Thus f is an injection and f is a bijection since G and Z/nZ have the same order.
□
Much like Math 110A, we will define the kernel and image of a group homomorphism. In the
best case scenario, knowledge of these two subgroups will give a nearly complete picture of the
group homomorphism.
Definition 1.14. Let f : G → H be a group homomorphism. Define the kernel of f and image
of f respectively as
ker(f ) = {a ∈ G : f (a) = eH }
im(f ) = {b ∈ H : b = f (a) for some a ∈ G}.
Proposition 1.17. The kernel of f is a subgroup of G and the image of f is a subgroup of H.
Proof. The identity of G is an element of ker(f ) so ker(f ) is non-empty.
Let a, b ∈ G be elements of the kernel of f . Then
f (ab) = f (a)f (b) = eH eH = eH
so ab ∈ ker(f ).
By Proposition 1.12(2),
f (a−1 ) = f (a)−1 = e−1
H = eH
so ker(f ) is a subgroup of G.
Since f (eG ) = eH , the identity element of H is always an element of im(f ). Thus im(f ) is
non-empty.
Let c, d ∈ H be elements of the image of f . Then there are a, b ∈ G for which f (a) = c and
f (b) = d. Thus
f (ab) = f (a)f (b) = cd
and cd ∈ im(f ).
Again by Proposition 1.12(2),
f (a−1 ) = f (a)−1 = c−1
so c−1 ∈ im(f ). We conclude that im(f ) is a subgroup of H. □
For now, we will work with the image. We will generalize the next result in the First Isomorphism
Theorem once we have the notion of a quotient group.
Proposition 1.18. Let f : G → H be a group homomorphism. If f is injective, then G ≃ im(f ).
Proof. Any group homomorphism f : G → H can be regarded as a surjective group homomorphism
f : G → im(f ). If f is injective, then f is an isomorphism onto im(f ) and G ≃ im(f ). □
Example 1.33. Define f : Z/2Z → Z/4Z as f (k) = 2k. Then
f (k + ℓ) = f (k + ℓ) = 2(k + ℓ) = 2k + 2ℓ = f (k) + f (ℓ)

and f is a group homomorphism. We can see that f is injective since f (0) = 0 ̸= 2 = f (1).
Proposition 1.18 implies that Z/2Z ≃ im(f ) = ⟨2⟩. In other words, a copy of Z/2Z sits inside
Z/4Z as the subgroup generated by 2.
Example 1.34. Recall Example 1.27 where we define the surjective group homomorphism
f : Z → Z/nZ
as f (k) = k. Since f is surjective, im(f ) = Z/nZ. Let k ∈ ker(f ) so f (k) = 0. Then k ≡ 0 (mod
n) or n|k. In other words, ker(f ) ⊂ nZ. To prove the reverse containment, take a ∈ nZ. Then
a = nk for some k ∈ Z and f (nk) = nk = 0. Therefore, ker(f ) = nZ. We can rephrase this as
multiples of n are the elements that map to 0 under f .
16 MATTHEW GHERMAN
Example 1.35. Let G and H be groups. Recall Example 1.28 where we define projection onto
the first coordinate f : G × H → G as f (g, h) = g. Since f is a surjective group homomorphism,
im(f ) = G. The elements that map to eG will be of the form ker(f ) = {(eG , h) : h ∈ H}. We
can identify this set with H via (eG , h) 7→ h. With some work, we can show that ker(f ) ≃ H.
Therefore, there is an isomorphic copy of H in the Cartesian product G × H.
What happens if we instead project onto the second coordinate?
End of lecture 6
2. Normal Subgroups and Quotient Groups

2.1. Congruence and Lagrange’s Theorem. Recall that two integers are congruent modulo n
if a − b is divisible by n. The set nZ in Z is the subgroup of multiples of n. We can translate
the congruence into multiplicative notation as ab−1 is an element of nZ. The following definition
generalizes the notion of congruence to an arbitrary subgroup of any group.
Definition 2.1. Let K be a subgroup of a group G. Let a, b ∈ G. Then a is congruent to b modulo
K if ab−1 ∈ K.
Much as in the case of Z/nZ, our goal is to partition G so that any two elements in a grouping
are congruent modulo K. Each piece of the partition will then be treated as a single element in
the quotient. For instance, any two integers that are 1 modulo n are elements of the equivalence
class 1 in Z/nZ.
Proposition 2.1. For K a subgroup of G, congruence modulo K is an equivalence relation.
Proof. Let a, b, c ∈ G.
Reflexivity: aa−1 = e ∈ K since K is a subgroup of G. Thus a is congruent to a modulo K.
Symmetry: Assume a is congruent to b modulo K. We want to show that b is congruent to a
modulo K. We have ab−1 ∈ K so (ab−1 )−1 ∈ K since K is a subgroup. Then ba−1 ∈ K and b is
congruent to a modulo K.
Transitivity: Assume that a is congruent to b modulo K and b is congruent to c modulo K. We
want to show that a is congruent to c modulo K. We know that ab−1 ∈ K and bc−1 ∈ K. The
product of two elements of K is another element of K so (ab−1 )(bc−1 ) = a(b−1 b)c−1 = ac−1 ∈ K.
Thus a is congruent to c modulo K. □
For a subgroup K of G and an element a ∈ G, define the congruence class of a modulo K to

be the set of all elements of G that are congruent to a modulo K. In other words,
a = {b ∈ G : ba−1 ∈ K}.
Let k = ba−1 ∈ K for some b ∈ G. Then b = ka so we can rewrite
a = {ka : k ∈ K}.
As a result, we will adopt the notation Ka for the congruence class of a modulo K.
Definition 2.2. A right coset of K in G is a congruence class of some a ∈ G modulo K. We will
denote the right coset Ka.
In the next section, we introduce left cosets aK. We can rephrase everything in this section via
left cosets. For non-abelian groups, the left and right cosets might contain different elements, but
there will always be the same number of left and right cosets of a subgroup.
Example 2.1. Let G = D4 and K = ⟨s⟩ = {e, s}. The subgroup K is a right coset. We also have
the right cosets Kr = {r, sr} and Kr2 = {r2 , sr2 }. The remaining right coset is Kr3 = {r3 , sr3 }.
Note that each element of D4 appears in one and only one right coset of K.
Let G = D4 and N = ⟨r⟩. The subgroup N is a right coset. The only other right coset is
N s = {s, rs = sr3 , r2 s = sr2 , r3 s = sr}. Once again, each element of D4 appears in one and only
one right coset of K.
Proposition 2.2. Let K be a subgroup of G and a, c ∈ G. Then a is congruent to c modulo K if
and only if Ka = Kc.
18 MATTHEW GHERMAN
Proof. (⇒) Assume a is congruent to c modulo K. Then ac−1 ∈ K and a = kc for some k ∈ K.
Then Ka = K(kc) ⊂ Kc. Further, c−1 = a−1 k for some k ∈ K or c = k −1 a by taking the inverse
of both sides. Then Kc = K(k −1 a) ⊂ Ka and Ka = Kc.
(⇐) Assume that Ka = Kc. Then ea = a ∈ Kc and a = kc for some k ∈ K. Thus ac−1 = k ∈ K
by multiplying by c−1 on the right. We conclude that a is congruent to c modulo K. □
Corollary 2.1. Let K be a subgroup of G. Two right cosets of K in G are either disjoint or
identical.
Proof. Let Ka and Kb be two right cosets of K in G. If Ka ∩ Kb is non-trivial, then there are
k1 , k2 ∈ K for which k1 a = k2 b. Then ab−1 = k1−1 k2 ∈ K. By Proposition 2.2, Ka = Kb. □
2.1.1. Lagrange’s Theorem. In this subsection, we will develop our first powerful theorem for deal-
ing with groups. Lagrange’s Theorem will reduce some of our group questions to a question about
divisibility of integers.
Proposition 2.3. Let K be a subgroup of G.
S
(1) G is the union of the right cosets of K, G = a∈G Ka.
(2) For each a ∈ G, there is a bijection of sets f : K → Ka. If K is finite, any two right cosets
contain the same number of elements.
S
Proof. (1)
S Since Ka is a subset
S of G, we have S Ka ⊂ G. Let b ∈ G, then b = eb ∈ Kb and
a∈G
b ∈ a∈G Ka. Thus G ⊂ a∈G Ka and G = a∈G Ka.
(2) Define the set function f : K → Ka as f (k) = ka for each k ∈ K. By definition of Ka,
f is surjective. Assume f (x) = f (y) for x, y ∈ K. Then xa = ya and x = y by right
multiplication by a−1 . Thus f is injective and, hence, bijective.
If K is finite, every coset Ka has the same number of elements as K. Thus Ka and Kb
have the same number of elements for any a, b ∈ G.
□
Definition 2.3. Let H be a subgroup of G. Then the index of H in G is the number of distinct
right cosets of H in G, denoted [G : H].
If G is finite, then the index of any subgroup is finite. If G is infinite, the index of a subgroup
could be finite or infinite.
Example 2.2. Let nZ be the subgroup of Z containing multiples of n. Then [Z : nZ] = n since
there are n distinct congruence classes.
Example 2.3. Let Z be the integral subgroup of the additive group Q. For a, b ∈ Q, the cosets
Z + a and Z + b are equal if and only if a − b ∈ Z. Each coset has a unique representative x ∈ Z + a
for which 0 ≤ x < 1. If 0 ≤ x < y < 1, then Z + x ̸= Z + y. Thus [Q : Z] is infinite.
Theorem 2.1 (Lagrange’s Theorem). Let K be a subgroup of a finite group G. Then the order
of K divides the order of G. Further,
|G| = |K|[G : K].
Proof. Suppose that [G : K] = n and denoteSn the n distinct cosets of K in G by Kc1 , . . . , Kcn for
each ci ∈ G. By Proposition 2.3(1), G = i=1P Kci . The cosets
P are distinct so, by Corollary 2.1,
the cosets are pairwise disjoint. Thus |G| = ni=1 |Kci | = ni=1 |K| by Proposition 2.3(2). We
conclude that |G| = |K|[G : K]. □
Corollary 2.2. Let G be a finite group.
(1) If a ∈ G, then the order of a divides the order of G.
(2) If |G| = k, then ak = e for every a ∈ G.

Proof. (1) The cyclic subgroup ⟨a⟩ of G has order |a| by Proposition 1.8(2). By Lagrange’s
Theorem, |a| divides |G|.
(2) If |G| = k and |a| = n, then n|k by (1). Thus k = nt for some t ∈ Z and
ak = ant = (an )t = et = e.
□
Example 2.4. Let p be a prime integer. Every group G of order p is cyclic and isomorphic to
Z/pZ. Let a ∈ G. Then |a| divides p by Corollary 2.2. Since p is prime, |a| = 1 or |a| = p. If
|a| = 1, then a is the identity of G. If |a| = p, then ⟨a⟩ = G and G is cyclic of order p. By
Proposition 1.16(2), G ≃ Z/pZ.
End of lecture 7
Example 2.5. Every group of order 4 is isomorphic to Z/4Z or Z/2Z × Z/2Z. Let a ∈ G. By
Lagrange’s Theorem, |a| divides 4 so |a| = 1, |a| = 2, or |a| = 4. If G has an element of order 4,
then G ≃ Z/4Z by Proposition 1.16(2).
Suppose that G does not contain an element of order 4. Then each non-identity element of G
has order 2. Let G = {e, a, b, c}. We can fill in the multiplication table as follows.
· e a b c
e e a b c
a a e
b b e
c c e
−1
If ab = e, then b = a = a, a contradiction. If ab = a, then b = e, a contradiction. If ab = b, then
a = e, a contradiction. Thus ab = c. We obtain the multiplication table below by filling in the
grid like a Sudoku puzzle.
· e a b c
e e a b c
a a e c b
b b c e a
c c b a e
This is the multiplication table for Z/2Z × Z/2Z via the following identification
e 7→ (0, 0)
a 7→ (1, 0)
b 7→ (0, 1)
c 7→ (1, 1).
20 MATTHEW GHERMAN
2.2. Normal Subgroups. We would like the set of cosets of a subgroup K in a group G to form
a group. For a, b ∈ G obvious multiplication would be (Ka)(Kb) = K(ab). However, the elements
of (Ka)(Kb) have the form k1 ak2 b while the elements of K(ab) have the form k1 ab for k1 , k2 ∈ K.
For non-abelian groups, these two descriptions can be quite different. A normal subgroup will be
one for which the set of right cosets is a group under the desired operation.
Definition 2.4. Let K be a subgroup of G and a ∈ G. A left coset is aK = {ak : k ∈ K}.
Definition 2.5. A subgroup N of G is normal if aN = N a for all a ∈ G. In other words, the left
coset containing a coincides with the right coset containing a. We write N ◁ G.
Example 2.6. Recall Example 2.1 where G = D4 with subgroups K = ⟨s⟩ and N = ⟨r⟩. The left
coset rK = {r, rs = sr3 } while the right coset Kr = {r, sr = r3 s}. Thus K is not normal.
However, sN = N s and rN = N = N r so N is a normal subgroup. Note that we only need to
check the normality condition on the generators of the group.
Example 2.7. Let G be an abelian group with subgroup N . Since an = na for all a ∈ G and
n ∈ N , N is a normal subgroup of G. We conclude that every subgroup of an abelian group is
normal.
Example 2.8. Let G be a group with subgroup N . Recall that the center Z(G) of G from
Definition 1.7 is the subgroup of elements that commute with all elements of G. Let N ⊆ Z(G).
Take a ∈ G and n ∈ N . Then an = na so any subgroup of Z(G) is always a normal subgroup of
G. In particular, this implies that Z(G) is always normal.
Remark 2.1. It is important to note that the condition aN = N a does not imply that an = na
for each n ∈ N . The equality aN = N a only requires that an1 = n2 a for some possibly distinct
n1 , n2 ∈ N . We can think of this as a relaxed commutativity condition.
The following result is the motivation for normal subgroups. We will be able to define an
intuitive multiplication on the right (or left) cosets of normal subgroups.
Proposition 2.4. Let N be a normal subgroup of G with a, b, c, d ∈ G. If N a = N b and N c = N d,
then N (ac) = N (bd).
Proof. There are elements m, n ∈ N for which ab−1 = m and cd−1 = n. Since N is normal,
an = n1 a for some n1 ∈ N . Then
(ac)(bd)−1 = acd−1 b−1
= a(cd−1 )b−1
= anb−1
= n1 ab−1
= n1 m ∈ N.
Therefore, N (ac) = N (bd) by Proposition 2.2. □
The following result lists equivalent properties of a normal subgroup. There are scenarios where
one of the properties will be easier to check than the others.
Proposition 2.5. Let N be a subgroup of G with aN a−1 = {ana−1 : n ∈ N } for a ∈ G. The
following are equivalent.
(1) N is a normal subgroup of G.
(2) aN a−1 ⊂ N for all a ∈ G.
(3) aN a−1 = N for all a ∈ G.
Proof. (1) ⇒ (2): If N is normal, then aN = N a for all a ∈ G. Fix a ∈ G. Then for each n ∈ N ,
an = n1 a for some n1 ∈ N . Thus ana−1 = n1 ∈ N by multiplying on the right by a−1 . We have
aN a−1 ⊂ N for all a ∈ G.
(2) ⇒ (3): Assume that aN a−1 ⊂ N for each a ∈ G. Let n ∈ N . Then a−1 na = n1 so
n = an1 a−1 for some n1 ∈ N . Thus N ⊂ aN a−1 and aN a−1 = N for each a ∈ G.
(3) ⇒ (1): If aN a−1 = N for each a ∈ G, then each n ∈ N satisfies ana−1 = n1 for some
n1 ∈ N . Thus an = n1 a via right multiplication by a. We conclude that aN ⊂ N a. Each n ∈ N
also satisfies an2 a−1 = n for some n2 ∈ N . Then an2 = na via right multiplication by a. Thus
aN ⊃ N a and aN = N a. □
Example 2.9. Let G = S3 and N = {e, (123), (132)}. For n ∈ N , we always have nN n−1 . We
only need to check conjugation by a ∈ G for transpositions a. Then
(12)(123)(12) = (132)
(12)(132)(12) = (123).
We can do the same computation for (13) and (23). We don’t need to check conjugation by elements
that are in the subgroup since any product of elements in N will remain in N We conclude that
aN a−1 = N for each a ∈ G so N is normal.
End of lecture 8
Remark 2.2. It is sufficient to check normality on a set of generators of G. Let G = ⟨S⟩ and
g ∈ G. Then g = s1 · · · sk for some si ∈ S. For each n ∈ N , we have
gng −1 = (s1 · · · sk )n(s1 · · · sk )−1
= (s1 · · · (sk−1 (sk ns−1 −1
k )sk−1 ) · · · s1 ).
We can inductively prove that gng −1 ∈ N if si ns−1i ∈ N for each si ∈ S.

Further, we can check normality on a set of generators for N . Let N = ⟨T ⟩. For each n ∈ N ,
we can write n = t1 · · · tk for some ti ∈ S. Then
gng −1 = g(t1 · · · tk )g −1
= gs1 (g −1 g)t1 (g −1 g) · · · (g −1 g)tk g −1
= (gt1 g −1 ) · · · (gtk g −1 ).
If gti g −1 ∈ N for each ti ∈ S, then gng −1 ∈ N for all n ∈ N . Therefore, N is normal.
Lemma 2.1. Let H be a subgroup of G with [G : H] = 2. Then H is normal.
Proof. Let {H, Ha} be the set of right cosets of H. Then a ̸∈ H and the left cosets aH and H are
not equal. Since [G : H] = 2, the set of left cosets is {H, aH}. We conclude that aH = Ha. □
22 MATTHEW GHERMAN
2.3. Quotient Groups.
Definition 2.6. Let N be a normal subgroup of G. We will denote the sets of right cosets of N
in G by G/N . We will call G/N the quotient of G by N .
The following result partially motivates the fraction notation.
Proposition 2.6. Let N be a normal subgroup of G with a, b ∈ G.

(1) G/N is a group via the operation (N a)(N b) = N (ab).
(2) If G is finite, then |G/N | = |G|/|N |.
(3) If G is abelian, then G/N is abelian.
Proof. (1) The operation is well-defined by Proposition 2.4. Let a, b, c ∈ G. The coset N = N e
is the identity element since
(N e)(N a) = N (ea) = N a
(N a)(N e) = N (ae) = N a.
The inverse of N a is N a−1 since
(N a)(N a−1 ) = N (aa−1 ) = N e = N

(N a−1 )(N a) = N (a−1 a) = N e = N.
The operation is associative since
((N a)(N b))(N c) = N (ab)N c

= N ((ab)c)
= N (a(bc))
= N a(N (bc))
= N a((N b)(N c)).
(2) The order |G/N | is [G : H] which, by Lagrange’s Theorem, is |G|/|N |.

(3) Let a, b ∈ G. Then (N a)(N b) = N (ab) = N (ba) = (N b)(N a).
□
Example 2.10. Let G = D4 and N = ⟨r⟩. Then G/N is a group of order |G|/|N | = 2 by
Proposition 2.6(2). The only group of order 2 by Example 2.4 is Z/2Z. Thus G/N ≃ Z/2Z.
Example 2.11. Let G = D4 and K = Z(G). By Example 1.13, Z(G) = ⟨r2 ⟩, and Z(G) is normal
by Example 2.8. Proposition 2.6 implies that |G/K| = 4. The cosets are K = {e, r2 }, Kr = {r, r3 }
Ks = {s, r2 s = sr2 }, and K(sr) = {sr, r2 sr = sr3 }. We have the following multiplication table.
· K Kr Ks K(sr)
K K Kr Ks K(sr)
Kr Kr K K(sr) Ks
Ks Ks K(sr) K Kr
K(sr) K(sr) Ks Kr K
By Example 2.5, G/K is isomorphic to Z/4Z or Z/2Z × Z/2Z. The multiplication table is that of
Z/2Z × Z/2Z since every non-identity element has order 2. Therefore, G/K ≃ Z/2Z × Z/2Z.
Example 2.12. Let G = Z/2Z × Z/4Z and N = ⟨(1, 2)⟩. Since G is abelian, N is normal. The
order of N is 2 so Proposition 2.6 implies |G/N | = |G|/|N | = 4. The cosets are
N = {(0, 0), (1, 2)}
N + (1, 0) = {(1, 0), (0, 2)}
N + (0, 1) = {(0, 1), (1, 3)}
N + (1, 1) = {(1, 1), (0, 3)}.
The element N + (0, 1) generates G/N so G/N ≃ Z/4Z.
Example 2.13. The additive group Q is abelian so the subgroup Z is normal. The quotient group
Q/Z is an infinite group by Example 2.3. An element rs has order dividing s. Thus every element
of Q/Z has finite order. Interestingly, the group has at least one element of each finite order.
2.3.1. The Structure of Groups. The general method for studying a group G is first find normal
subgroups N of G. If we know information about N and the quotient G/N , we can often learn
information about the group G.
Proposition 2.7. Let N be a normal subgroup of G. Then G/N is abelian if and only if
aba−1 b−1 ∈ N
for all a, b ∈ G.
Proof. (⇒) Assume that G/N is abelian. Let a, b ∈ G. Then (N a)(N b) = N (ab) is equal to
(N b)(N a) = N (ba) so ab(ba)−1 ∈ N . Thus aba−1 b−1 ∈ N for every a, b ∈ G.
(⇐) Assume that aba−1 b−1 ∈ N for every a, b ∈ G. We can rewrite aba−1 b−1 = (ab)(ba)−1 .
Then (N a)(N b) = N (ab) = N (ba) = (N b)(N a) for every a, b ∈ G. We conclude that G/N is
abelian. □
Definition 2.7. The commutator subgroup of a group G is [G, G] = ⟨aba−1 b−1 : a, b ∈ G⟩.
Proposition 2.7 proves that every normal subgroup for which the quotient group is abelian must
contain the commutator subgroup. Thus the commutator subgroup will play a significant role in
studying almost every non-abelian group.
End of lecture 9
Proposition 2.8. The subgroup [G, G] of G is normal.
Proof. Let g ∈ G. We have
g(aba−1 b−1 )g −1 = gaba−1 (g −1 g)b−1 g −1
= (ga)b(a−1 g −1 )gb−1 g −1
= (ga)b(ga)−1 (b−1 b)gb−1 g −1
= ((ga)b(ga)−1 b−1 )(bgb−1 g −1 ) ∈ [G, G].
The conjugation of every generator of [G, G] is an element of [G, G]. Remark 2.2 implies that
[G, G] is normal. □
Example 2.14. Let G be an abelian group. Then [G, G] = {e}. In fact, G is abelian if and only
if [G, G] = {e}.
The following result is another useful tool in classifying groups.
Proposition 2.9. If G is a group for which G/Z(G) is cyclic, then G is abelian.
24 MATTHEW GHERMAN
Proof. Note that Z(G) is a normal subgroup of G by Example 2.8. Let Z(G)a ∈ G/Z(G) be a
generator where a ∈ Z(G)a is an element of G. Then we can write every element of G as nai for
some n ∈ Z(G). For n1 , n2 ∈ Z(G),
(n1 ai )(n2 aj ) = ai n1 aj n2
= ai aj n1 n2
= ai+j n1 n2
= aj (ai n1 )n2
= (aj n2 )(ai n1 ).
Every element of G commutes with every other element of G so G is abelian. □
2.4. Isomorphism Theorems. There is a close connection between normal subgroups, quotient
groups, and surjective group homomorphisms. Recall the definition of the kernel of a group
homomorphism f : G → H from Definition 1.14. In Proposition 1.17, we prove that the kernel is
a subgroup of G.
Proposition 2.10. Let f : G → H be a group homomorphism. Then ker(f ) is a normal subgroup
of G.
Proof. Let g ∈ G and k ∈ ker(f ). Then
f (gkg −1 ) = f (g)f (k)f (g −1 )
= f (g)eG f (g)−1
= f (g)f (g)−1
= eG .
Thus gkg −1 ∈ ker(f ) and ker(f ) is a normal subgroup of G. □
As the next result shows, the kernel measures how far a group homomorphism is from being
injective.
Proposition 2.11. Let f : G → H be a group homomorphism with K = ker(f ). Then K = {eG }
if and only if f is injective.
Proof. (⇒) Assume K = {eG }. Then f (a) = f (b) implies
f (a)f (b)−1 = f (a)f (b−1 ) = f (ab−1 ) = eH .
We have ab−1 = eG so a = b and f is injective.
(⇐) Assume f is injective. Since f (eG ) = eH , we conclude that f (a) = eH implies a = eG . Thus
K = {eG }. □
The following result develops a one-to-one correspondence between normal subgroups of G and
kernels of homomorphisms out of G. We have already seen in Proposition 2.10 that any kernel is
a normal subgroup. Every normal subgroup is also the kernel of some group homomorphism.
Proposition 2.12. If N is a normal subgroup of G, the function π : G → G/N given by π(a) = N a
is a surjective group homomorphism with kernel N .
Proof. Let a, b ∈ G. Then π(ab) = N (ab) = (N a)(N b) = π(a)π(b) so π is a group homomorphism.
Given some right coset N a, the element a ∈ G satisfies π(a) = N a. Thus π is surjective.
Let a ∈ ker(f ). Then f (a) = N a = N , which is equivalent to a ∈ N . Therefore, ker(f ) = N . □
Definition 2.8. We call π : G → G/N a quotient group homomorphism.
The following are some of the most useful results in our study of groups. The Correspondence
Theorem will relate the subgroups of G to that of a quotient G/N . In most situations, we will
have information about the structure of either G or G/N . The Correspondence Theorem often
allows us to translate that information to the other setting. The Isomorphism Theorems provide
tools for describing groups as the quotient of a possibly more familiar group.
Theorem 2.2 (Correspondence Theorem). Let N be a normal subgroup of G. Then there is a
bijection between the set of all subgroups of G that contain N and the set of subgroups of G/N .
Proof. Assume that K is a subgroup of G that contains N . Since N is a normal subgroup of G,
N is also a normal subgroup of K. Then K/N = {N k : k ∈ K} is a non-empty subset of G/N
that is closed under the operation and inverses. Thus K/N is a subgroup of G/N .
26 MATTHEW GHERMAN
Assume that K is a subgroup of G/N . Define K = {k ∈ G : π(k) ∈ K}, and we want to show
that K is a subgroup of G that contains N . Clearly, K is not empty since π(eG ) ∈ K. Given
k1 , k2 ∈ K, we have π(k1 k2 ) = π(k1 )π(k2 ) ∈ K since K is closed under the group operation. Thus
k1 k2 ∈ K. We have π(k1−1 ) = π(k1 )−1 ∈ K since K is closed under inverses. Thus k1−1 ∈ K. We
conclude that K is a subgroup of G. For any n ∈ N , π(n) is the identity of G/N so n ∈ K.
Therefore, K is a subgroup of G that contains N .
The two operations described are inverse to one another, producing the desired bijection. □
End of lecture 10
Corollary 2.3 (Correspondence Theorem for Normal Subgroups). Let N be a normal subgroup
of G. Then there is a bijection between the set of all normal subgroups of G that contain N and
the set of normal subgroups of G/N .
Proof. The Correspondence Theorem proves the bijection for subgroups. We need only show that
normal subgroups map to normal subgroups under the correspondence.
Let K be a normal subgroup of G that contains N . Then for N k ∈ K/N and N g ∈ G/N , we
have (N g)(N k)(N g)−1 = N (gkg −1 ) ∈ K/N . Thus K/N is a normal subgroup of G/N .
Let K/N be a normal subgroup of G/N . For k ∈ K and g ∈ G, we have (N g)(N k)(N g)−1 ∈
K/N so gkg −1 ∈ K. Thus K is a normal subgroup of G. □
Example 2.15. Let G = Z and N = nZ. Since G is abelian, N is automatically normal in G.
Any subgroup of Z containing nZ will be of the form mZ for m|n. The corresponding subgroup
of Z/nZ will be mZ/nZ = {nZ + h : h ∈ mZ}.
Take, for example, n = 6 and m = 2. Then the subgroup 2Z of Z contains 6Z and corresponds
to 2Z/6Z = {Z + 0, Z + 2, Z + 4} ≃ Z/3Z. Visually, we reduce the fraction 2/6 to 1/3. This
intuition is a bit dangerous, however.
Lemma 2.2. Let f : G → H be a group homomorphism with kernel K. Let a, b ∈ G. Then
f (a) = f (b) if and only if Ka = Kb.
Proof. (⇒) Assume f (a) = f (b). Then
f (a)f (b)−1 = f (a)f (b−1 ) = f (ab−1 ) = eH .
so ab−1 ∈ K. We conclude Ka = Kb.
(⇐) Assume Ka = Kb. Then ab−1 ∈ K so f (ab−1 ) = eH . We have f (a) = f (b) by a similar
argument as above. □
Theorem 2.3 (First Isomorphism Theorem). Let f : G → H be a group homomorphism with
kernel K. Then the quotient group G/K is isomorphic to im(f ).
Proof. We want to define φ : G/K → im(f ) by φ(Ka) = f (a), but cosets can have different labels.
We need to show that each label results in the same image under φ. Suppose Ka = Kb for a, b ∈ G.
Then f (a) = f (b) by Lemma 2.2 so φ(Ka) = φ(Kb). Therefore, φ as written is well-defined.
The function φ is a group homomorphism since
φ((Ka)(Kb)) = φ(K(ab)) = f (ab) = f (a)f (b) = φ(Ka)φ(Kb).
Suppose h ∈ im(f ). Then there is some g ∈ G for which f (g) = h by f surjective. Thus
φ(Kg) = f (g) = h
and φ is surjective. Suppose φ(Ka) = φ(Kb) for a, b ∈ G. Then f (a) = f (b) so Ka = Kb by
Lemma 2.2. We conclude that φ is injective and, as a result, an isomorphism. □
The First Isomorphism Theorem will allow us to describe a group H by finding a group G and
a surjective homomorphism f : G → H. Then G/ ker(f ) ≃ H.
Example 2.16. Define projection onto the first coordinate f : G × H → G as f (g, h) = g. The
group homomorphism is surjective by Example 1.28. With H = ker(f ) = {(eG , h) : h ∈ H},
the First Isomorphism Theorem implies (G × H)/H ≃ G. We can show that H ≃ H via the
identification (eG , h) 7→ h.
Example 2.17. Let f : C× → R× be f (a + bi) = a2 + b2 . We have
f ((a + bi)(c + di)) = f ((ac − bd) + (ad + bc)i)
= (ac − bd)2 + (ad + bc)2
= (a2 c2 − 2abcd + b2 d2 ) + (a2 d2 + 2abcd + b2 c2 )
= a2 c2 + b2 d2 + a2 d2 + b2 c2
= (a2 + b2 )(c2 + d2 )
= f (a + bi)f (c + di)
so f is a group homomorphism. Let K = im(f ) be the subgroup of all positive real numbers. Since
1 is the identity of R× , N = ker(f ) = {a + bi : a2 + b2 = 1} is the circle of radius 1 in the complex
plane. The First Isomorphism Theorem implies C× /N ≃ K.
Example 2.18. The function f : R× → R× where f (r) = r2 is a group homomorphism by
Example 1.26. We have ker(f ) = {1, −1} since these are the real numbers that square to 1. Let
K = im(f ) be the subgroup of all positive real numbers. The First Isomorphism Theorem implies
R× /{1, −1} ≃ K.
Lemma 2.3. Let K and N be subgroups of G with N normal in G. Then
N K = {nk : n ∈ N, k ∈ K}
is a subgroup of G that contains N and K.
Proof. The identity element of N K is eG eG . Let n1 k1 , n2 k2 ∈ N K. Since N is normal, k1 n2 = n3 k1
for some n3 ∈ N . Thus
(n1 k1 )(n2 k2 ) = n1 (k1 n2 )k2
= n1 (n3 k1 )k2
= (n1 n3 )(k1 k2 ) ∈ N K.
Further, N normal implies that (n1 k1 )−1 = k1−1 n−1 −1
1 = n4 k1 ∈ N K for some n4 ∈ N . For any
n ∈ N and k ∈ K, we have n = neG ∈ N K and k = eG k ∈ N K. Therefore, N K is a subgroup of
G that contains N and K. □
Of particular interest is the case when G = N K. Then we can build the group G out of possibly
more recognizable subgroups.
Theorem 2.4 (Second Isomorphism Theorem). Let K and N be subgroups of G with N normal
in G. Then N ∩ K is a normal subgroup of K and N K/N ≃ K/(N ∩ K).
Proof. Let a ∈ N ∩ K and k ∈ K. Then kak −1 ∈ K since k, a, k −1 ∈ K. Since N is normal in G,
ka = a1 k for some a1 ∈ N . Thus kak −1 = a1 kk −1 = a1 ∈ N . We conclude that kak −1 ∈ N ∩ K so
N ∩ K is a normal subgroup of K.
Define the function f : N K → K/(N ∩ K) as f (nk) = (N ∩ K)k. There could be multiple ways
of writing elements of N K so we need to check if f is well-defined. Let n1 k1 = n2 k2 for ni ∈ N and
28 MATTHEW GHERMAN
kj ∈ K. Then f (n1 k1 ) = (N ∩ K)k1 while f (n2 k2 ) = (N ∩ K)k2 . However, k2 k1−1 = n−1

2 n1 ∈ N ∩ K
so (N ∩ K)k1 = (N ∩ K)k2 . Thus f is well-defined. Let n1 k1 , n2 k2 ∈ N K. Since N is normal,
(n1 k1 )(n2 k2 ) = nk1 k2 for some n ∈ N . Then
f ((n1 k1 )(n2 k2 )) = f (nk1 k2 ) = (N ∩ K)k1 k2 = f (n1 k1 )f (n2 k2 )
and f is a group homomorphism. Given some (N ∩ K)k, we have f (eG k) = (N ∩ K)k so f is
surjective. For n ∈ N , we have f (neG ) = (N ∩ K) so N ⊂ ker(f ). If f (nk) = (N ∩ K), then
k ∈ N ∩ K which implies nk ∈ N . Thus ker(f ) = N . The First Isomorphism Theorem implies
that N K/ ker(f ) = N K/N ≃ K/(N ∩ K). □
End of lecture 11
Often, we will use a special case of the Second Isomorphism Theorem where N ∩ K = {eG }.
Corollary 2.4. Let K and N be subgroups of G with N ◁ G. If N ∩ K = {eG }, then N K/N ≃ K.
The Second Isomorphism Theorem produces a useful order argument for finite groups G.
Corollary 2.5. Let K and N be subgroups of G with N ◁ G. If G is a finite group, then
|N K||N ∩ K| = |N ||K|.
Proof. By the Second Isomorphism Theorem, N K/N ≃ K/(N ∩K). Thus |N K/N | = |K/(N ∩K)|.
We have |N K/N | = |N K|/|N | and |K/(N ∩ K) = |K|/|N ∩ K|. By clearing denominators,
|N K||N ∩ K| = |N ||K|. □
Example 2.19. Let G = Dn , K = ⟨s⟩, and N = ⟨r⟩. By Lemma 2.1, the index 2 subgroup N is
normal. Lemma 2.3 proves that N K is a subgroup of G. In this case, every element of G can be
written as ri sj for 0 ≤ i ≤ n − 1 and 0 ≤ j ≤ 1 so G = N K. Further, N ∩ K = {eG }. The Second
Isomorphism Theorem implies G/N = N K/N ≃ K. In other words, the cosets {N eG , N s} under
multiplication behave the same way as the elements {eG , s} of G.
The Third Isomorphism Theorem is a particularly useful case of the Correspondence Theorem
when K is also a normal subgroup of G.
Theorem 2.5 (Third Isomorphism Theorem). Let K and N be normal subgroups of G with
N ⊂ K ⊂ G. Then K/N is a normal subgroup of G/N , and the quotient group (G/N )/(K/N ) is
isomorphic to G/K.
Note that the cancellation of N in (G/N )/(K/N ) motivates the fraction notation for quotients.
However, we have to be careful because this only works when K is also a normal subgroup of G.
Proof. The Correspondence Theorem for Normal Subgroups proves K/N ◁ G/N .
Define the function f : G/N → G/K as f (N g) = Kg. We need to check whether f is well-
defined. Let N g1 = N g2 for gi ∈ G so g2 g1−1 ∈ N . Then f (N g1 ) = Kg1 and f (N g2 ) = Kg2 . Since
N ⊂ K, we have g2 g1−1 ∈ K and f is well-defined. For N g1 , N g2 ∈ G/N , we have
f ((N g1 )(N g2 )) = f (N (g1 g2 )) = K(g1 g2 ) = (Kg1 )(Kg2 ) = f (N g1 )f (N g2 )
so f is a group homomorphism. Given Kg ∈ G/K, we have f (N g) = Kg so f is surjective.
If f (N g) = Kg = K, then g ∈ K and N g ∈ K/N . Thus ker(f ) ⊂ K/N . For N k ∈ K/N ,
f (N k) = Kk = K so K/N ⊂ ker(f ) and ker(f ) = K/N . The First Isomorphism Theorem implies
(G/N )/(K/N ) ≃ G/K. □
Example 2.20. Let m and n be integers such that n|m. Since Z is abelian, we have a chain of
normal subgroups mZ ⊂ nZ ⊂ Z. The Third Isomorphism Theorem gives an isomorphism
f : (Z/mZ)/(nZ/mZ) → Z/nZ
where f (nZ/mZ + (mZ + k)) = nZ + k. In other words, if you reduce modulo m and then modulo
n, it is the same thing as reducing modulo n to begin with.
2.4.1. Simple Groups. The techniques developed in this section often require normal subgroups of
a group G. A particularly interesting and tricky example of groups are those without any normal
subgroups.
Definition 2.9. Let G be a group. Then G is simple if its only normal subgroups are {e} and G.
Example 2.21. Let p be prime. By Example 2.4, a group G of order p is isomorphic to Z/pZ.
Any subgroup of G will have order 1 or p by Lagrange’s Theorem. In other words, the element is
either the identity or a generator. Thus Z/pZ is simple.
Proposition 2.13. A group G is simple abelian if and only if G is isomorphic to Z/pZ for some
prime p.
Proof. (⇒) Suppose G is a simple abelian group. Since G is abelian, every subgroup of G is
normal. Thus G has no non-trivial subgroups so ⟨a⟩ = G for a non-identity element a ∈ G. If ⟨a⟩
is not finite, then ⟨a⟩ ≃ Z by Proposition 1.16(1). The group Z is not simple so ⟨a⟩ is finite. Let
|a| = n and n = td. Then ⟨at ⟩ is a subgroup of G of order d by Proposition 1.4(3). Therefore, G
is a cyclic group of order p or G ≃ Z/pZ.
(⇐) See Example 2.21. □
Example 2.22. Let G be a simple group. Let f : G → H be a group homomorphism. We
know that ker(f ) is normal in G by Proposition 2.10. By simplicity of G, ker(f ) = {eG } or
ker(f ) = G. Thus f is injective by Proposition 2.11 or f (a) = eH for all a ∈ G. A kernel of a
group homomorphism out of a simple group is all or nothing.
30 MATTHEW GHERMAN
2.5. The Symmetric and Alternating Groups. Group theory began with the study of per-
mutations and groups of permutations. The abstract definition of a group came later and may
appear to be far more general than the concept of a group of permutations. The next theorem
shows that this is not the case, however.
Let A(G) be the set of bijective functions from G to G. These functions need not be homo-
morphisms. On the homework, we proved that A(G) is group under composition. The proof of
Cayley’s Theorem is based on an injection from G to A(G).
Theorem 2.6 (Cayley’s Theorem). Every group G is isomorphic to a group of permutations
Proof. Let a ∈ G. Define the set map φa : G → G as φa (x) = ax. Since a has an inverse,
φa−1 ◦ φa = φa ◦ φa−1 is the identity set map. Thus φa is a bijection of sets and, thus, an element
of A(G).
Define f : G → A(G) as f (a) = φa . For a, b ∈ G, we have f (ab) = φab . We have
φab (x) = (ab)x = a(bx) = φa (φb (x))
so φab = φa ◦ φb . Then f (ab) = f (a)f (b) and f is a group homomorphism. Suppose f (a) = f (b)
for a, b ∈ G so φa (x) = φb (x) for all x ∈ G. Thus ax = bx. Multiply on the right by x−1 to obtain
a = b. Therefore, f is an injective group homomorphism. By the First Isomorphism Theorem,
G ≃ im(f ) ⊂ A(G). □
Corollary 2.6. Every finite group G of order n is isomorphic to a subgroup of Sn .
Proof. By Cayley’s Theorem, G is isomorphic to a subgroup of A(G). Since G is a set of n elements,
A(G) ≃ Sn . □
Definition 2.10. We refer to an element (a1 a2 . . . ak ) ∈ Sn as a k-cycle. Two cycles σ1 and σ2 are
disjoint if they have no elements in common.
Example 2.23. The cycles (12) and (34) are disjoint in S4 .
Example 2.24. The inverse of a k-cycle (a1 a2 a3 . . . ak−1 ak ) is (a1 ak ak−1 . . . a3 a2 ).
Proposition 2.14. Let σ = (a1 . . . ak ) and τ = (b1 . . . bℓ ) be disjoint cycles of Sn . Then στ = τ σ.
Proof. Let c ∈ Xn . Then σ(c) = c or τ (c) = c. Without loss of generality, assume τ (c) = c.
Then (στ )(c) = σ(c). Since σ and τ are disjoint, τ (σ(c)) = σ(c). Thus (στ )(c) = (τ σ)(c) for all
c ∈ Xn . □
Proposition 2.15. Every permutation of Sn is the product of disjoint cycles.
Proof. Let σ ∈ Sn . For each element c ∈ Xn , trace where c goes via powers of σ to produce the
k-cycle τ = (c σ(c) σ 2 (c) . . . σ k−1 (c)). For each element of Xn that is not in τ , perform the same
process. Eventually each element of Xn is in one and only one of the cycles. Then σ is the product
of these disjoint cycles. □
End of lecture 12
Proposition 2.16. The order of the permutation τ in Sn is the least common multiple of the
lengths of the disjoint cycles whose product is τ .
Proof. Let τ = ℓi=1 σi for disjoint ki -cycles σi and m = lcm(k1 , . . . , kℓ ). We have σim = e for all
Q
1 ≤ i ≤ ℓ. Since the σi are disjoint, they will commute by Proposition 2.14. Thus
ℓ
!m ℓ
Y Y
m
τ = σi = σim = e.
i=1 i=1
Qℓ
Assume that τ M = e. Then τ M = i=1 σiM . The σi are disjoint so σiM must be the identity for each
1 ≤ i ≤ ℓ. By Proposition 1.4(1), ki divides M for each 1 ≤ i ≤ ℓ. Thus m|M and |τ | = m. □
2.5.1. Alternating Groups.

Definition 2.11. A 2-cycle of Sn is called a transposition.
Proposition 2.17. Every element of Sn is a product of (not necessarily disjoint) transpositions.
Proof. By Proposition 2.15, every element of Sn is a product of disjoint cycles. Thus we only need
to show that a k-cycle is a product of transpositions:
(a1 a2 . . . ak ) = (a1 a2 )(a2 a3 ) · · · (ak−1 ak ).
□
Definition 2.12. A permutation is even if it can be written as a product of an even number of
transpositions. Likewise, a permutation is odd if it can be written as a product of an odd number
of transpositions.
Strictly from the definitions, a permutation could be both even and odd at this point. We will
prove that this cannot occur in the following results.
Lemma 2.4. The identity element is even but not odd.
Proof. We can write the identity permutation as (12)(12) so it is even. Assume e = τk · · · τ2 τ1
for transpositions τi and k odd. Let c ∈ Xn and τr be the lowest index transposition for which c
appears. Write τr = (cd) for d ∈ Xn . If r = k, then e does not fix c, a contradiction. Thus r < k.
Let c, d, x, y be distinct elements of Xn . Then τr+1 has one of the following forms.
(1) (xy)
(2) (xd)
(3) (cy)
(4) (cd)
In Case (1), Proposition 2.14 implies that (xy)(cd) = (cd)(xy) so we can move τr to the left
one position. In Case (2), (xd)(cd) = (xc)(xd) so the first appearance of c can be moved to the
left one position. In Case (3), (cy)(cd) = (cd)(dy) so the first appearance of c can be moved
to the left one position. By repeated use of Cases (1)-(3), we will eventually be in Case (4).
Then τr+1 τr = (cd)(cd) is the identity. We can write the identity with two fewer transpositions.
Repeat this procedure for any element of Xn to eventually reduce to a single transposition. This
is a contradiction since the identity fixes each element of Xn . Therefore, the identity is an even
permutation. □
Proposition 2.18. Each permutation in Sn is exclusively even or odd.
Proof. Assume α ∈ Sn can be written as a product of transpositions σ1 · · · σk and transpositions
τ1 · · · τℓ . Then
e = αα−1 = (σ1 · · · σk )(τ1 · · · τℓ )−1 = σ1 · · · σk τℓ−1 · · · τ1−1
is a description of the identity with k + ℓ transpositions. Lemma 2.4 implies that k + ℓ is even so
k and ℓ are either both even or both odd. □
Definition 2.13. The alternating group An is the subset of all even permutations of Sn .
n!
Proposition 2.19. An is a group of order 2
(for n ≥ 2).
Proof. By Lemma 2.4, e ∈ An so An is non-empty. Let α, β ∈ An . Then α = σ1 · · · σ2k can be
written as a product of 2k transpositions and β = τ1 · · · τ2ℓ can be written as a product of 2ℓ
transpositions by Proposition 2.17. We conclude that
αβ = (σ1 · · · σ2k )(τ1 · · · τ2ℓ )
= σ1 · · · σ2k τ1 · · · τ2ℓ
32 MATTHEW GHERMAN
is a product of 2(k + ℓ) transpositions. Thus An is closed under the group operation. The inverse
−1
of α can be written α−1 = σ2k · · · σ1−1 so An is closed under inverses.
In the case that n = 1, A1 = S1 is trivial. Let n ≥ 2. Define f : An → Bn by f (α) = (12)α where
Bn is the set of odd permutations of Sn . Assume f (α) = f (β) so (12)α = (12)β. Multiply on the
left by (12) to obtain α = β. Thus f is injective. For γ ∈ Bn , the permutation is (12)γ ∈ An .
Then f ((12)γ) = (12)(12)γ = γ so f is surjective. Therefore, |An | = |Bn |. Each permutation is
either even or odd so |Sn | = |An | + |Bn | and |An | = |S2n | = n!2 . □
End of lecture 13
2.5.2. The Simplicity of the Alternating Group.
Example 2.25. The group A4 is not simple. We will prove on the homework that the subgroup
{e, (12)(34), (13)(24), (14)(23)} is normal in A4 .
Theorem 2.7. For n ̸= 4, the group An is simple.
Corollary 2.7. For n ≥ 5, the only normal subgroups of Sn are {e}, An , and Sn .
Section 8.5 of Hungerford has the whole proof of the theorem. In the interest of time, we will
only prove two helpful lemmas.
Lemma 2.5. Every element of An for n ≥ 3 is a product of 3-cycles.
Proof. Let α ∈ An . Then α can be written as a product of transpositions by Proposition 2.17. Let
a, b, c, d ∈ Xn . A pair of transpositions has the form (ab)(ac) or (ab)(cd).
(ab)(ac) = (acb)
(ab)(cd) = (adb)(adc)
Since there are an even number of transpositions in the description of α, we pair them up and
follow the above formulas to write α as a product of 3-cycles. □
Lemma 2.6. Let n ≥ 3. If N is a normal subgroup of An and N contains a 3-cycle, then N = An .
Proof. Without loss of generality, assume (123) ∈ N . Then (123)2 = (132) ∈ N as well. Let
x = (12)(3k) for k ∈ Xn and k > 3. Since N is normal, x(123)x−1 ∈ N . We have
(12)(3k)(123)(3k)(12) = (1k2) ∈ N.
The inverse of (1k2) = (12k) ∈ N as well. Thus N contains all 3-cycles of the form (1k2) and
(12k). For a, b, c ∈ Xn with a, b, c ≥ 3, every 3-cycle is either of the form or inverse to one of the
form (1a2), (1ab), (2ab), or (abc).
(1ab) = (12b)(1a2) ∈ N
(2ab) = (1b2)(12a) ∈ N
(abc) = (1a2)(12c)(1b2)(12a) ∈ N
Thus N contains all 3-cycles of Sn , and Lemma 2.5 proves that N = An . □
End of midterm material
3. Topics in Group Theory

3.1. Direct Products. If G and H are groups, then their Cartesian product G × H is also a
group, with the operation defined coordinate-wise. In this section we extend this notion to more
than two groups. Then we examine the conditions under which a group is isomorphic to a direct
product of certain of its subgroups. We will then define a more general product structure called
the semi-direct product. The semi-direct product can simplify the classification of finite groups of
small order.
Definition 3.1. If G1 , . . . , Gn are groups, then the Cartesian product G1 × · · · × Gn is a group
under coordinate-wise multiplication. For ai , bi ∈ Gi ,
(a1 , . . . , an ) · (b1 , . . . , bn ) = (a1 b1 , . . . , an bn ).
The identity element is (eG1 , . . . , eGn ) and the inverse of (a1 , . . . , an ) is (a−1 −1
1 , . . . , an ). We call the
group G1 × · · · × Gn the direct product.
Example 3.1. Note that Gi is not directly a subgroup of the direct product G = G1 × · · · × Gn .
However, there is a subgroup of the direct product that is isomorphic to Gi . Define the inclusion
ιi : Gi → G as ιi (ai ) = (0, . . . , ai , . . . , 0) where ai ∈ Gi is placed in the ith component of the
direct product. The inclusion is an injective group homomorphism so the First Isomorphism
Theorem implies that Gi ≃ im(ιi ), which is a subgroup of the direct product. In particular,
im(ιi ) = (0, . . . , Gi , . . . , 0).
Example 3.2. If G1 , . . . , Gn are finite groups, then |G1 × · · · × Gn | = ni=1 |Gi | and the direct
Q
product is finite.
Example 3.3. The direct product G1 × · · · × Gn is abelian if and only if Gi is abelian for each
1 ≤ i ≤ n.
Example 3.4. Let G = Z/6Z. Then M = ⟨3⟩ and N = ⟨2⟩ are normal subgroups of G since G is
abelian. The direct product M × N is cyclic generated by (3, 2) so M × N ≃ Z/6Z by Proposition
1.16. Note also that every element in Z/6Z can be written uniquely as a sum of an element in M
and an element in N . Consequently, we can describe Z/6Z internally as a product of subgroups
without making use of the externally defined direct product.
The following lemma will help to prove in which scenarios a group is isomorphic to a direct
product of subgroups.
Lemma 3.1. Let M and N be normal subgroups of G with M ∩ N = {e}. If a ∈ M and b ∈ N ,
then ab = ba.
Proof. Consider aba−1 b−1 . Since M is normal, bab−1 ∈ M so aba−1 b−1 ∈ M . Similarly, aba−1 ∈ N
so aba−1 b−1 ∈ N . Since M ∩ N is trivial, we conclude that aba−1 b−1 = e or ab = ba. □
Recall that a product of subgroups is given by M N = {mn ∈ G : m ∈ M, n ∈ N }.
Proposition 3.1. Let N1 , . . . , Nk be normal subgroups of a group G such that G = N1 · · · Nk
and Ni ∩ (N1 · · · Ni−1 Ni+1 · · · Nk ) = {eG } for each 1 ≤ i ≤ k. Then G is isomorphic to the direct
product N1 × · · · × Nk .
Proof. Define f : N1 × · · · × Nk → G as f (a1 , . . . , ak ) = a1 · · · ak . Let ai , bi ∈ Ni . By Lemma 3.1,
f ((a1 , . . . , ak )(b1 , . . . , bk )) = f (a1 b1 , . . . , ak bk )
= (a1 b1 ) · · · (ak bk )
= (a1 · · · ak )(b1 · · · bk )
= f (a1 , . . . , ak )f (b1 , . . . , bk ).
34 MATTHEW GHERMAN
Thus f is a homomorphism. Assume f (a1 , . . . , ak ) = eG . Then a1 · · · ak = eG or

−1 −1 −1
ai = a−1
1 · · · ai−1 ak · · · ai+1 .
Since ai ∈ Ni and a−1 −1 −1 −1 −1 −1 −1 −1

1 · · · ai−1 ak · · · ai+1 = a1 · · · ai−1 ai+1 · · · ak ∈ (N1 · · · Ni−1 Ni+1 · · · Nk ) by
Lemma 3.1, we conclude ai = eG for each 1 ≤ i ≤ k. Thus f is injective. Since G = N1 · · · Nk , f
is surjective. Therefore, f is an isomorphism. □
End of lecture 14
Corollary 3.1. Let G be a group with normal subgroups M and N such that M ∩ N = {eG } and
M N = G. Then G ≃ M × N .
Example 3.5. The multiplicative group G = (Z/15Z)× = {1, 2, 4, 7, 8, 11, 13, 14} can be written
as a direct product of two cyclic groups. Let M = ⟨11⟩ = {1, 11} and N = ⟨2⟩ = {1, 2, 4, 8}. Since
G is abelian, M and N are normal. Further, M ∩ N = {1}. Corollary 3.1 implies
(Z/15Z)× ≃ M × N ≃ Z/2Z × Z/4Z.
3.1.1. Semi-direct Product. Many group examples cannot be written as a direct product of sub-
groups. The semi-direct product will allow us to describe a larger number of groups in terms of
subgroups.
Example 3.6. Every proper subgroup of D4 has order 1, 2, or 4 by Lagrange’s Theorem. Thus
every subgroup is abelian by Examples 2.4 and 2.5. Since D4 is not abelian, D4 cannot be a direct
product of subgroups.
Recall that in homework problem Section 7.4 # 36, we prove the following result.
Proposition 3.2. For a group G, an automorphism of G is a group isomorphism f : G → G. The

set of all automorphisms of G, denoted Aut(G), is a group under composition.
Definition 3.2. Let H and N be groups. Let φ : H → Aut(N ) be a group homomorphism. Then
the semi-direct product N ⋊φ H is the set N × H with the following operation
(n1 , h1 ) · (n2 , h2 ) = (n1 φ(h1 )(n2 ), h1 h2 ).
Proposition 3.3. The semi-direct product N ⋊φ H is a group with identity (eN , eH ) and
(n, h)−1 = (φ(h)−1 (n−1 ), h−1 ) = (φ(h−1 )(n−1 ), h−1 ).
Proof. The multiplication is associative. The identity element satisfies
(n, h)(eN , eH ) = (nφ(h)(eN ), h)

= (neN , h)
= (n, h)
(eN , eH )(n, h) = (eN φ(eH )(n), h)
= (eN idN (n), h)
= (n, h).
The inverse element satisfies

(n, h)(φ(h)−1 (n−1 ), h−1 ) = (nφ(h)(φ(h)−1 (n−1 )), eH )
= (nn−1 , eH )
= (eN , eH )
(φ(h)−1 (n−1 ), h−1 )(n, h) = (φ(h)−1 (n−1 )φ(h−1 )(n), eH )
= (φ(h−1 )(n−1 )φ(h−1 )(n), eH )
= (φ(h−1 )(n−1 n), eH )
= (φ(h−1 )(eN ), eH )
= (eN , eH ).
□
Example 3.7. We will write D4 as a semi-direct product of subgroups. Let N = ⟨r⟩ and H = ⟨s⟩.
The group homomorphism h : N → N defined by h(r) = r3 is an automorphism of N . Define
φ : H → Aut(N ) as φ(s) = h. The multiplication in N ⋊φ H works as follows.
(r, eH )(eN , s) = (rφ(eH )(eN ), s)
= (ridN (e), s)
= (r, s)
(eN , s)(r, eH ) = (eN φ(s)(r), s)
= (eN h(r), s)
= (r3 , s).
By identifying (a, b) ∈ N ⋊φ H with ab ∈ D4 , we can prove that N ⋊φ H ≃ D4 .
The symbol ⋊ is a combination of the direct product symbol and the normal subgroup symbol.
The next result motivates the notation since N is isomorphic to a normal subgroup of the semi-
direct product.
Proposition 3.4. The subgroup {(n, eH ) : n ∈ N } is a normal subgroup of N ⋊φ H.
Proof. Let k, n ∈ N and h ∈ H. Then
(k, h)(n, eH )(k, h)−1 = (kφ(h)(n), h)(φ(h)−1 (k −1 ), h−1 )
= (kφ(h)(n)φ(h)(φ(h)−1 (k −1 )), eH )
= (kφ(h)(n)k −1 , eH ).
Since k and φ(h)(n) are elements of N , we conclude that the desired subgroup is normal in
N ⋊φ H. □
Remark 3.1. There is some intuition for building a group out of the semi-direct product of
subgroups. Take a normal subgroup N and a subgroup H of G. As motivation, observe
(eN , h)(n, eH )(eN , h)−1 = (φ(h)(n), h)(φ(h)−1 (e−1 −1
N ), h )
= (φ(h)(n)φ(h)(φ(h)−1 (e−1
N )), eH )
= (φ(h)(n), eH ).
The element φ(h) ∈ Aut(N ) represents how the conjugation by h behaves in G.
Proposition 3.5. Let H be a subgroup of G and N a normal subgroup of G. If G = N H and
N ∩ H = {eG }, then G ≃ N ⋊φ H for some φ : H → Aut(N ).
36 MATTHEW GHERMAN
Proof. Let φ : H → Aut(N ) be the group homomorphism φ(h) = (n 7→ hnh−1 ). Define

f : N ⋊φ H → G
as f (n, h) = nh. We have
f ((n1 , h1 ) · (n2 , h2 )) = f (n1 φ(h1 )(n2 ), h1 h2 )
= f (n1 h1 n2 h−1
1 , h1 h2 )
= (n1 h1 n2 h−1
1 )(h1 h2 )
= n1 h1 n2 h2
= f (n1 , h1 )f (n2 , h2 )
so f is a group homomorphism. Let f (n, h) = eG so nh = eG . Then h = n−1 ∈ N ∩H. We conclude
that h = n = eG and f is injective. Since G = N H, f is surjective. Therefore, G ≃ N ⋊φ H. □
The following shows that the semi-direct product is a generalization of the usual direct product.
Example 3.8. Let φ : H → Aut(N ) be the group homomorphism φ(h) = idN . The multiplication
in the semi-direct product is
(n1 , h1 ) · (n2 , h2 ) = (n1 φ(h1 )(n2 ), h1 h2 )
= (n1 idN (n2 ), h1 h2 )
= (n1 n2 , h1 h2 ).
Thus N ⋊φ H ≃ N × H and the semi-direct product is the usual direct product.
Example 3.9. Let N = Z/5Z and H = Z/6Z. We want to find all possibilities for N ⋊φ H.
As we know from the homework, Aut(Z/5Z) ≃ (Z/5Z)× = {1, 2, 3, 4}. The automorphism that
i ∈ (Z/5Z)× represents is the unique group isomorphism Z/5Z → Z/5Z defined by sending 1 to
i. A generator of H has order 6 so φ(1) must have order 1, 2, 3, or 6. The elements 2 and 3 are
order 4 in (Z/5Z)× so φ(1) = 1 or φ(1) = 4. In the first case, φ(1) = 1 represents the identity
automorphism. We obtain N ⋊φ H = N × H by Example 3.8. In the second case, φ(1) = 4
represents the automorphism of N where 1 maps to 4. We have a multiplication example
(0, 1) · (2, 5) = (0 + φ(1)(2), 0)
= (3, 1)
(2, 5) · (0, 1) = (2 + φ(5)(0), 0)
= (2, 0)
so N ⋊φ H is a non-abelian group of order 30.
End of lecture 15
3.2. Finite Abelian Groups. We have the tools to classify all finite abelian groups. We will
prove that every finite abelian group is isomorphic to a direct product of cyclic groups.
Definition 3.3. If G is an abelian group and p is a prime integer, let G(p) be the set of all elements
whose order is some power of p. In other words,
G(p) = {a ∈ G : |a| = pk for some k ≥ 0}.
We call G(p) the p-primary subgroup of G.
Proposition 3.6. The subset G(p) of an abelian group G is a subgroup.
Proof. The identity has order 1 = p0 so G(p) is non-empty. Let a, b ∈ G(p) with |a| = pk and
k+ℓ k+ℓ k+ℓ
|b| = pℓ . Then (ab)p = ap bp = eG eG = eG since G is abelian. We conclude |ab| divides pk+ℓ
by Proposition 1.4(1) so |ab| is a power of p since p is prime. Thus ab ∈ G(p), and G(p) is closed
k k
under the group operation. The inverse a−1 to a ∈ G(p) satisfies (a−1 )p = (ap )−1 = e−1
G = eG so
|a−1 | divides pk and a−1 ∈ G(p). Therefore, G(p) is a subgroup of G. □
Example 3.10. If G = Z/12Z, then G(2) = {0, 3, 6, 9} and G(3) = {0, 4, 8}.
If G = Z/3Z × Z/3Z, then G(3) = G since every element has order a power of 3.
Lemma 3.2. Let G be an abelian group with a ∈ G an element of finite order. Let {pi }ti=1 be the
distinct primes that divide |a|. Then a = a1 · · · at for some ai ∈ G(pi ).
Proof. We will proceed by induction on the number t of distinct primes that divide |a|. If t = 1,
then a ∈ G(p1 ) = G and the result holds. Assume the result holds for t ≤ k − 1, and we will prove
it for t = k. We have |a| = pr11 · · · prkk for distinct primes pi and ri > 0. Let m = pr22 · · · prkk and
n = pr11 . We have gcd(m, n) = 1 so 1 = mu + nv for some u, v ∈ Z. Then
a = amu+nv = amu anv .
Since |a| = mn, amu ∈ G(p1 ). Define a1 = amu . Similarly, |anv | is divisible by only the distinct
primes p2 , . . . , pk . By the inductive hypothesis, anv = a2 · · · ak for ai ∈ G(pi ). Therefore, a =
a1 a2 · · · ak as desired. □
Proposition 3.7. If G is a finite abelian group, then
G ≃ G(p1 ) × · · · × G(pt )
where p1 , . . . , pt are the distinct primes that divide |G|.
Proof. By Lemma 3.2, G = G(p1 ) · · · G(pt ). We will now show that
G(pi ) ∩ (G(p1 ) · · · G(pi−1 )G(pi+1 ) · · · G(pk )) = {eG }
for each 1 ≤ i ≤ k. Let H = G(p1 ) · · · G(pi−1 )G(pi+1 ) · · · G(pt ). Let a ∈ G(pi ) ∩ H so |a| = pri i
for some ri ≥ 0. Further, a = a1 · · · ai−1 ai+1 · · · at for aj ∈ G(pj ). By Lemma 1.2, |a| divides
lcm(|a1 |, . . . , |ai−1 |, |ai+1 |, . . . , |at |), which is a product of powers of p1 , . . . , pi−1 , pi+1 , . . . , pt . The
only way this is possible is if |a| = 1 and a = eG . Apply Proposition 3.1 to complete the proof. □
Definition 3.4. A group G of order pk is a finite p-group.
An element of maximal order a in a finite p-group G is one for which |b| ≤ |a| for all b ∈ G.
Lemma 3.3. Let G be a finite abelian p-group and a an element of maximal order in G. Then
there is a subgroup K of G such that G ≃ ⟨a⟩ × K.
The proof of Lemma 3.3 is technical and too long for lecture. We will outline the proof, and the
rest can be found as Hungerford Lemma 9.6.
38 MATTHEW GHERMAN
Proof. Consider the set S of all subgroups H of G for which ⟨a⟩ ∩ H = {eG }. S is non-empty since
⟨eG ⟩ ∈ S. With G finite, there is a maximal element K ∈ S. If we show ⟨a⟩K = G, then Corollary
3.1 would imply that G ≃ ⟨a⟩ × K.
Assume ⟨a⟩K ̸= G. Then there is some non-identity b ∈ G for which b ̸∈ ⟨a⟩K. Since G is a
j
p-group, there is some pj for which bp = eG ∈ ⟨a⟩K. Let k be the smallest positive integer for
k k−1
which bp ∈ ⟨a⟩K. Then c = bp ̸∈ ⟨a⟩K but cp ∈ ⟨a⟩K. With some work and the maximality of
a, we can show that c ∈ ⟨a⟩K, contradicting the choice of c. Therefore, ⟨a⟩K = G. □
Theorem 3.1 (Fundamental Theorem of Finite Abelian Groups). Every finite abelian group G is
the direct product of cyclic groups, each of prime power order.
Proof. By Proposition 3.7, G is isomorphic to the direct product of its subgroups G(pi ), one for
each pi dividing |G|. To complete the proof, we need to show that each finite abelian p-group is a
direct product of finite cyclic p-groups.
We proceed by induction on |H|. If |H| = p, then H ≃ Z/pZ by Example 2.4. Assume the
result holds for every finite abelian p-group of order less than that of H. Let a ∈ H be an
element of maximal order pk in H. Then Lemma 3.3 implies there is a subgroup K of H for which
H ≃ ⟨a⟩ × K. By the inductive hypothesis, K is isomorphic to a direct product of finite cyclic
p-groups. Therefore, H is isomorphic to a direct product of finite cyclic p-groups. □
End of lecture 16
Definition 3.5. The order of the cyclic factors in the Fundamental Theorem of Finite Abelian
Groups are the elementary divisors of a finite abelian group G.
A corollary to Lagrange’s Theorem states that the order of each element divides the order
of a finite group. When k divides the order of a finite group, do we have an element of that
order? Cauchy’s Theorem for Finite Abelian Groups provides a partial converse to the corollary
to Lagrange’s Theorem when a group is finite abelian. We will eventually extend this result to all
finite groups.
Corollary 3.2 (Cauchy’s Theorem for Finite Abelian Groups). If G is a finite abelian group G
and p is a prime dividing the order of G, then G contains an element of order p.
Proof. By the Fundamental Theorem of Finite Abelian Groups, we can write G as a product of
cyclic groups
G ≃ G1 × · · · × Gm .
There is some cyclic component Z/pk Z, say G1 without loss of generality, that contains an element
a of order p. Then (a, eG2 , . . . , eGm ) is an element of order p in G. □
The Fundamental Theorem of Finite Abelian Groups reduces the description of finite abelian
groups to a question of elementary number theory.
Example 3.11. Let G be an abelian group of order 36. We can factor 36 = 22 · 32 . By the
Fundamental Theorem of Finite Abelian Groups, G is isomorphic to one of the following.
Z/2Z × Z/2Z × Z/3Z × Z/3Z
Z/2Z × Z/2Z × Z/9Z
Z/4Z × Z/3Z × Z/3Z
Z/4Z × Z/9Z
We can verify that no two of the groups are isomorphic by counting elements of certain orders.
Note that one of the groups in the list of Example 3.11 must be cyclic of order 36. The next
result will help us determine which groups in such a list are cyclic.
Lemma 3.4. If gcd(m, n) = 1, then Z/mZ × Z/nZ ≃ Z/(mn)Z.
Proof. The order of (1, 1) in Z/mZ × Z/nZ ≃ Z/(mn)Z is the smallest t ∈ Z such that
t(1, 1) = (t, t) = (0, 0).
Then t ≡ 0 (mod m) and t ≡ 0 (mod n). We have m|t and n|t. Since gcd(m, n) = 1, (mn)|t. The
order of Z/mZ × Z/nZ is mn so (1, 1) generates the group and Z/mZ × Z/nZ is cyclic. Therefore,
Z/mZ × Z/nZ ≃ Z/(mn)Z. □
Proposition 3.8. Let n = pn1 1 · · · pnk k with the pi distinct primes. Then
Z/nZ ≃ Z/pn1 1 × · · · × Z/pnk k Z.
Proof. We will proceed by induction on n. By Example 2.4, the result holds for n = 2. Assume that
the result holds for all groups of order less than n. Apply Lemma 3.4 to pn1 1 and m = pn2 2 · · · pnk k .
Then Z/nZ ≃ Z/pn1 1 Z × Z/mZ. The inductive hypothesis provides the result. □
Example 3.12. Consider the group G = Z/2Z × Z/4Z × Z/4Z × Z/3Z × Z/3Z × Z/5Z. Arrange
the prime powers of the cyclic factors by size where each prime has a row as follows.
2 4 4
3 3
5
Reorganize by multiplying the elements in each column.
2 4 4
3 3
5
2 12 60
Via this procedure, each number will divide the next as we move left to right. By Proposition 3.8,
we have
G ≃ Z/2Z × Z/12Z × Z/60Z.
Theorem 3.2 (Invariant Factors). Every finite abelian group is the direct product of cyclic groups
of orders m1 , . . . , mt where mi−1 divides mi for each 2 ≤ i ≤ t.
Definition 3.6. The m1 , . . . , mt from a finite abelian group G are the invariant factors of G.
In order to prove the uniqueness of our elementary divisor and invariant factor descriptions of
finite abelian groups, we prove the following lemma.
Lemma 3.5. Let G and H be abelian groups.
(1) The subset pG = {xp : x ∈ G} is a subgroup of G.
(2) If G and H are isomorphic, pG ≃ pH.
(3) If G is a finite abelian p-group, then pG ̸= G.
Proof. (1) The element epG = eG so Gp is non-empty. Let a, b ∈ pG. Then a = xp and b = y p
and ab = xp y p = (xy)p . Thus pG is closed under the group operation. For a = xp , we have
a−1 = (xp )−1 = (x−1 )p so pG is closed under inverses. Therefore, pG is a subgroup of G.
(2) Let f : G → H be a group isomorphism. Then f (xp ) = f (x)p so f (pG) ⊂ pH. Since f is
surjective, every element h ∈ pH satisfies h = y p = f (x)p = f (xp ). Therefore, f (pG) = pH,
and f injective implies pG ≃ pH.
40 MATTHEW GHERMAN
(3) Assume, first, that G = ⟨a⟩ is cyclic. Then pG = ⟨ap ⟩ and pG ̸= G. By the Fundamental
Theorem of Finite Abelian Groups, a general finite abelian group G is isomorphic to a
direct product of cyclic groups Ci . For each cyclic component, pCi ̸= Ci so pG ̸= G by (2).
□
End of lecture 17
Theorem 3.3 (Uniqueness of Elementary Divisors). Let G and H be finite abelian groups. Then
G is isomorphic to H if and only if G and H have the same elementary divisors up to reordering.
Proof. (⇒) Assume f : G → H is a group isomorphism. Then a and f (a) have the same order for
each a ∈ G so f (G(p)) = H(p). Thus G(p) ≃ H(p) via f , and we need only prove the result when
G and H are p-groups.
Assume G and H are isomorphic p-groups. We will proceed by induction on the order of G to
prove that G and H have the same elementary divisors. When |G| = 2, all groups of order 2 will
have the same elementary divisors. Assume the result holds for all groups of order less than |G|.
Suppose the elementary divisors of G and H are respectively
pn1 , pn2 , . . . , pnt , p, . . . , p
pm1 , pm2 , . . . , pmk , p, . . . , p
with each mi , nj > 1, r copies of p at the end of the list for G, and s copies of p at the end of the
list for H. By Lemma 3.5, the elementary divisors of pG and pH are respectively
pn1 −1 , pn2 −1 , . . . , pnt −1
pm1 −1 , pm2 −1 , . . . , pmk −1 .
By Lemma 3.5(3), |pG| < |G| and |pH| < |H|. Apply the inductive hypothesis so t = k and
ni − 1 = mi − 1 or ni = mi . Since |G| and |H| are the products of the respective elementary
divisors, r = s, and G and H have the same elementary divisors.
(⇐) If G and H have the same elementary divisors, then G ≃ H up to possibly rearranging
cyclic factors. □
Theorem 3.4 (Uniqueness of Invariant Factors). Let G and H be finite abelian groups. Then G
is isomorphic to H if and only if G and H have the same invariant factors.
Example 3.13. We will classify up to isomorphism all the finite abelian groups of order 72 with
elementary divisors. The prime factorization is 72 = 23 · 32 . The possibilities for the 2-primary
part are Z/8Z, Z/2Z × Z/4Z, and Z/2Z × Z/2Z × Z/2Z. The possibilities for the 3-primary part
are Z/9Z and Z/3Z × Z/3Z. Thus a finite abelian group of order 72 is isomorphic to one of the
following.
Z/8Z × Z/9Z
Z/8Z × Z/3Z × Z/3Z
Z/2Z × Z/4Z × Z/9Z
Z/2Z × Z/4Z × Z/3Z × Z/3Z
Z/2Z × Z/2Z × Z/2Z × Z/9Z
Z/2Z × Z/2Z × Z/2Z × Z/3Z × Z/3Z
The following table shows how to obtain invariant factors for the last element on the list.
2 2 2
3 3
2 6 6
The classification of finite abelian groups of order 72 via invariant factors is as follows.
Z/72Z
Z/3Z × Z/24Z
Z/2Z × Z/36Z
Z/6Z × Z/12Z
Z/2Z × Z/2Z × Z/18Z
Z/2Z × Z/6Z × Z/6Z
42 MATTHEW GHERMAN
3.3. The Sylow Theorems. Finite non-abelian groups are much more complicated than finite
abelian groups. The Sylow Theorems provide the most useful tools for studying non-abelian finite
groups. We will not prove the results in this section but, instead, focus on applications. The proofs
can be found in the next section.
Let G be a group with subgroup H. Lagrange’s Theorem shows that |H| divides |G|. Sylow’s
First Theorem provides a partial converse.
Theorem 3.5 (Sylow’s First Theorem). Let G be a finite group and p a prime. If pk divides |G|,
then G has a subgroup of order pk .
Example 3.14. The symmetric group S6 has order 6! = 720 = 24 · 32 · 5. Thus S6 has subgroups of
order 2, 4, 8, 16, 3, 9, and 5. The other two Sylow theorems will help us prove how many distinct
subgroups of these orders exist in a group.
Corollary 3.3 (Cauchy’s Theorem). If G is a finite group whose order is divisible by p, then G
contains an element of order p.
Proof. Let p divide |G|. Then Sylow’s First Theorem implies that G has a subgroup H of order p.
Since |H| = p, H is cyclic generated by some order p element of G. □
Definition 3.7. Let G be a finite group and p a prime. If pn is the largest power of p that
divides |G|, then a subgroup of G of order pn is called a Sylow p-subgroup. The existence of such
a subgroup is guaranteed by Sylow’s First Theorem.
Example 3.15. Since S4 has order 4! = 24 = 23 · 3, a subgroup of order 8 will be a Sylow
2-subgroup. The Sylow 2-subgroups of S4 are
{e, (13), (24), (12)(34), (13)(24), (14)(23), (1234), (1432)}
{e, (12), (34), (12)(34), (13)(24), (14)(23), (1324), (1423)}
{e, (14), (23), (12)(34), (13)(24), (14)(23), (1243), (1342)}.
The Sylow 3-subgroups of S4 are order 3 and, thus, cyclic. The order 3 elements of S4 are 3-cycles.
Example 3.16. Let p be a prime and G a finite abelian group of order pn m where p does not
divide m. Then G(p) is a Sylow p-subgroup of G. In fact, we will show that G(p) is the unique
Sylow p-subgroup of G.
End of lecture 18
In Example 1.32, we showed that the homomorphism f : G → G defined by f (a) = xax−1 is an
isomorphism. We referred to f as conjugation by x. Let K be a subgroup of G. Then the image
of K under f , denoted xKx−1 , is isomorphic to K. In particular, |xKx−1 | = |K| for all x ∈ G. If
K is a Sylow p-subgroup of G, then so is xKx−1 .
Sylow’s Second Theorem provides a relationship among the Sylow p-subgroups of a group.
Theorem 3.6 (Sylow’s Second Theorem). If P and K are Sylow p-subgroups of a group G, then
there is some x ∈ G for which P = xKx−1 .
Any two Sylow p-subgroups of a group are isomorphic by Sylow’s Second Theorem.
Corollary 3.4. Let G be a finite group and K a Sylow p-subgroup of G for some prime p. Then
K is normal if and only if K is the unique Sylow p-subgroup of G.
Proof. (⇒) Assume that K is normal so xKx−1 = K for all x ∈ G. By Sylow’s Second Theorem,
any two Sylow p-subgroups are conjugate. Thus K is the unique Sylow p-subgroup of G.
(⇐) For each x ∈ G, xKx−1 is a Sylow p-subgroup of G. Since K is the unique Sylow p-subgroup,
we conclude that xKx−1 = K for all x ∈ G. Thus K is normal. □
Sylow’s Third Theorem helps count the number of Sylow p-subgroups in a group G.
Theorem 3.7 (Sylow’s Third Theorem). The number mp of Sylow p-subgroups of a finite group
G divides |G| and mp ≡ 1 (mod p).
3.3.1. Applications of Sylow Theorems. Simple groups are the basic building blocks of all groups.
We will use Sylow theorems to determine when groups of a certain order are not simple.
Example 3.17. Let G be a group of order 63 = 32 · 7. By Sylow’s Third Theorem, the number
of Sylow 7-subgroups m7 divides 63 and mp ≡ 1 (mod 7). The only number that satisfies both
properties is m7 = 1. By Corollary 3.4, we conclude that the unique Sylow 7-subgroup of G is
normal. Therefore, there are no simple groups of order 63.
Example 3.18. We will show that there are no simple groups of order 56. Let G be a group of
order 56 = 23 · 7. The number of Sylow 7-subgroups m7 divides 56 and m7 ≡ 1 (mod 7). We
conclude m7 = 1 or m7 = 8. If m7 = 1, then the unique Sylow p-subgroup is normal by Corollary
3.4 and G is not simple.
Assume m7 = 8. Each Sylow 7-subgroup is a cyclic group of order 7. Let P1 and P2 be two
distinct Sylow 7-subgroups. Then |P1 ∩ P2 | divides |P1 | = 7 so |P1 ∩ P2 | = 1 or |P1 ∩ P2 | = 7.
Since P1 and P2 are distinct, |P1 ∩ P2 | = 1. Therefore, each Sylow 7-subgroup provides 6 elements
of order 7 so G has 8 · 6 = 48 elements of order 7. By Lagrange’s Theorem, the intersection of
a Sylow 2-subgroup and a Sylow 7-subgroup is trivial. Since the order of a Sylow 2-subgroup is
8, we have m2 = 1 so G has a normal subgroup by Corollary 3.4. We conclude that there are no
simple groups of order 56.
We can also use the Sylow theorems to classify finite groups.
Proposition 3.9. Let G be a group of order pq for primes p > q. If q does not divide p − 1, then
G ≃ Z/(pq)Z.
Proof. By Sylow’s Third Theorem, the number mp of Sylow p-subgroups must divide |G| = pq.
Thus mp is 1, q, p, or pq. Further, mp ≡ 1 (mod p) by Sylow’s Third Theorem. Since p > q, q ̸≡ 1
(mod p) and mp ̸= q. Clearly, p ≡ 0 (mod p) and pq ≡ 0 (mod p) so mp = 1. Therefore, there is
exactly one Sylow p-subgroup P , which is normal by Corollary 3.4.
By Sylow’s Third Theorem, the number mq of Sylow q-subgroups must divide |G| = pq. Thus
mq is 1, q, p, or pq. Further, mq ≡ 1 (mod q) by Sylow’s Third Theorem. By assumption, p ̸≡ 1
(mod q). Clearly, q ≡ 0 (mod q) and pq ≡ 0 (mod q) so mq = 1. Therefore, there is exactly one
Sylow q-subgroup K, which is normal by Corollary 3.4.
Since P ∩ K is a subgroup of P and K, its order must divide |P | = p and |K| = q by Lagrange’s
|P ||K|
Theorem. Thus P ∩ K = {eG }. The Second Isomorphism Theorem implies |P K| = |P ∩K|
= pq so
G = P K. Corollary 3.1 and Proposition 3.8 imply
G ≃ P × K ≃ Z/pZ × Z/qZ ≃ Z/(pq)Z.
□
Example 3.19. To obtain the result of Proposition 3.9, we need the assumption that q does not
divide p − 1. However, we can obtain a result in the general case too.
Let G be a group of order pq for distinct primes p > q. Following the proof of Proposition
3.9, if mq = 1, G ≃ Z/(pq)Z. If mq = p, let K be a Sylow q-subgroup. We can still argue that
P ∩ K = {eG } and G = P K. Proposition 3.5 implies that G ≃ P ⋊φ K for some φ : K → Aut(P ).
End of lecture 19
44 MATTHEW GHERMAN
3.4. Conjugacy and the Proof of Sylow’s Theorems.

Definition 3.8. Let G be a group with a, b ∈ G. We say that a is conjugate to b if there exists
x ∈ G for which b = xax−1 .
Proposition 3.10. Conjugacy is an equivalence relation on G.
Proof. The relation is reflexive. Choose x = eG . We have eG ae−1
G = a and a is conjugate to itself.
The relation is symmetric. If a is conjugate to b, then b = xax−1 or x−1 bx = a. Thus b is
conjugate to a.
The relation is transitive. Assume a is conjugate to b and b is conjugate to c. Then b = xax−1
and c = yby −1 for x, y ∈ G. We have c = (yx)a(yx)−1 and a is conjugate to c. □
Definition 3.9. The conjugacy class of a ∈ G is the set of all elements conjugate to a.
By Proposition 3.10, G is partitioned into conjugacy classes. In other words, G is the disjoint
union of conjugacy classes of G where each element is in one and only one conjugacy class. Since
{eG } is a conjugacy class, each conjugacy class of a non-trivial group will satisfy |Ci | < |G|.
Example 3.20. We will determine the conjugacy classes of S3 . The identity element is always in
its own conjugacy class. We have
(13)(12)(13)−1 = (123)(13) = (23)
(23)(12)(23)−1 = (132)(23) = (13).
Conjugation by an element of S3 will not change whether an element is even or odd. The remaining
elements (123) and (132) are even while any conjugate of (12) is odd. Thus {(12), (13), (23)} is
one conjugacy class. Finally,
(12)(123)(12)−1 = (23)(12) = (132).
The final conjugacy class {(123), (132)}.
We can write each element of Sn uniquely up to rearrangement as a product of disjoint cycles.
The cycle type is the list of cycle lengths for a given element of Sn . In general, each conjugacy
class of Sn contains only the elements of a certain cycle type and contains all elements of a certain
cycle type.
Definition 3.10. Let G be a group and a ∈ G. The centralizer of a is the set of all elements of
g ∈ G that commute with a,
CG (a) = {g ∈ G : ga = ag}.
Proposition 3.11. For a group G and a ∈ G, the centralizer CG (a) is a subgroup of G.
Proof. The identity eG ∈ CG (a) since eG a = aeG . Let g, h ∈ CG (a). Then
(gh)a = g(ha) = g(ah) = (ga)h = (ag)h = a(gh)
so gh ∈ CG (a). For g ∈ CG (a), we also have a−1 g = ga−1 . Thus
g −1 a = (a−1 g)−1 = (ga−1 )−1 = ag −1
so g −1 ∈ CG (a). □
Example 3.21. The element (12) ∈ S4 commutes with itself, disjoint elements, and products
thereof. Thus CG ((12)) = {eG , (12), (34), (12)(34)}.
Proposition 3.12. Let G be a finite group with a ∈ G. The number of elements in the conjugacy
class of a is the index [G : CG (a)]. The number of elements in the conjugacy class of a divides |G|.
Proof. Let S be the set of distinct right cosets of C = CG (a) in G. Let T be the conjugacy class
of a in G. Define the function f : S → T as f (Cx) = xax−1 . We will show that f is a bijection
of sets. First, we need to show that f is well-defined. Assume Cx = Cy for x, y ∈ G. Then
x−1 y ∈ CG (a) and (x−1 y)a = a(x−1 y) for all a ∈ G. We rearrange to obtain yay −1 = xax−1
or f (Cx) = f (Cy). To show f is injective, assume f (Cx) = f (Cy) for x, y ∈ G. The previous
argument backward proves that Cx = Cy. For each element b ∈ T , we have b = xax−1 for some
x ∈ G. Thus f (Cx) = b and f is surjective.
The number of elements in S is [G : CG (a)] and the number of elements of T is the number of
distinct conjugates of a. By Lagrange’s Theorem, [G : CG (a)] divides |G|. □
Proposition 3.13 (Class Equation). Let G be a group with distinct conjugacy classes C1 , . . . , Ct
with representatives ai ∈ Ci . The class equation states
t
X
|G| = [G : CG (ai )].
i=1
Further,
r
X
|G| = |Z(G)| + |Cj |
j=1
for each Cj containing more than one element.
Proof. We have |G| = ti=1 |Ci |. Apply Proposition 3.12.
P
If Ci has only one element a ∈ C, then ag = ga for all g ∈ G and a ∈ Z(G). Thus the center
Z(G) is the union of all the one-element conjugacy classes of G. □
End of lecture 20
Example 3.22. The class equation applied to G = S3 shows that
|G| = |Z(G)| + |(12)| + |(123)|
6 = 1 + 3 + 2.
The class equation applied to G = S4 shows that
|G| = |Z(G)| + |(12)| + |(12)(34)| + |(123)| + |(1234)|
24 = 1 + 6 + 3 + 8 + 6.
Note that the size of each conjugacy class divides the order of the group.
Sylow’s First Theorem (Theorem 3.5) takes a finite group G and a prime p. If pk divides |G|,
then G has a subgroup of order pk .
Proof of Sylow’s First Theorem. We will proceed via induction on the order of G. If |G| = 1, then
the G has a subgroup of order p0 . Assume that the statement is true for all groups of order less
than |G|. By the class equation,
Xr
|G| = |Z(G)| + [G : CG (aj )]
j=1
for some elements aj ∈ G. We have [G : CG (aj )] > 1 and |Z(G)| ≥ 1 since eG ∈ Z(G). Suppose
there is some j for which p does not divide [G : CG (aj )]. By Lagrange’s Theorem,
|G| = |CG (aj )|[G : CG (aj )].
Thus pk divides |CG (aj )|. The inductive hypothesis implies that CG (aj ) and, thus, G has a subgroup
of order pk , completing the proof.
46 MATTHEW GHERMAN
If p divides [G : CG (aj )] for each 1 ≤ j ≤ r, then p divides |G| − rj=1 |CG (aj )| = |Z(G)|. Since
P
Z(G) is abelian, Cauchy’s Theorem for Abelian Groups implies there is an element c of order p
in Z(G). Let N = ⟨c⟩. Since c ∈ Z(G), N is a normal subgroup of G of order p. Then pk−1 is
the largest power of p that divides G/N . By the inductive hypothesis, G/N has a subgroup T of
order pk−1 . The Correspondence Theorem implies that there is a subgroup H of G containing N
such that H/N ≃ T . By Lagrange’s Theorem, |H| = |N ||T | = pk . □
In order to prove the remaining two Sylow theorems, we will introduce a subgroup notion of
conjugacy.
Definition 3.11. Let A, B, H be subgroups of G. Then A is H-conjugate to B if there exists an
x ∈ H such that B = xAx−1 . In the special case that H = G, we say that A and B are conjugate.
Proposition 3.14. Let H be a subgroup of G. Then H-conjugacy is an equivalence relation on
the set of all subgroups of G.
Proof. Refer to the proof of Proposition 3.10 using subgroups A, B, C in place of a, b, c. □
Definition 3.12. Let A be a subgroup of G. The normalizer of A in G is
NG (A) = {g ∈ G : gAg −1 = A}.
Proposition 3.15. For a subgroup A of G, the normalizer NG (A) is a subgroup of G. Further, A
is a normal subgroup of NG (A).
Proof. Since eG Ae−1
G = A, eG ∈ NG (A) and NG (A) is non-empty. Let x, y ∈ NG (A). Then
(xy)A(xy)−1 = x(yAy −1 )x−1 = xAx−1 = A

and xy ∈ NG (A). For each a ∈ A, xbx−1 = a for some b ∈ A. Thus x−1 ax ∈ A for all a ∈ A.
Conjugation by x−1 is an isomorphism so x−1 Ax = A and x−1 ∈ NG (A). By definition, xAx−1 = A
so A is normal in NG (A). □
Example 3.23. Let A = ⟨(12)⟩ in S4 . The elements that fix A under conjugation would be
{e, (12), (34), (12)(34)}. Looking back at Example 3.21, we notice that CG (a) = NG (A) when
A is cyclic generated by a. In general, though, centralizers fix elements via conjugation while
normalizers fix subgroups via conjugation.
If we instead take A = {e, (12)(34), (13)(24), (14)(23)}, we find NG (A) = G since A ◁ S4 .
Proposition 3.16. Let A and H be subgroups of a finite group G. The number of elements in
the equivalence class of A under H-conjugacy is [H : H ∩ NG (A)] and, therefore, divides |H|.
Proof. Refer to the proof of Proposition 3.12 where G is replaced by H, a is replaced by A, and
CG (a) is replaced by H ∩ NG (A). □
Lemma 3.6. Let Q be a Sylow p-subgroup of a finite group G. If x ∈ G has order a power of p
and xQx−1 = Q, then x ∈ Q.
Proof. Proposition 3.15 implies that Q is normal in NG (Q). By assumption, x ∈ NG (Q). Since |x|
is a power of p, the order of Qx in NG (Q)/Q is a power of p. Denote by T the cyclic subgroup
of NG (Q)/Q generated by Qx. By the Correspondence Theorem, there is some subgroup H of G
containing Q for which T = H/Q. Lagrange’s Theorem implies that |H| = |T ||Q| is a power of p.
However, Q ⊂ H with |Q| the largest power of p dividing the order of G. Thus Q = H and T is
the trivial subgroup of NG (Q)/Q. Therefore, Qx = Q or x ∈ Q. □
End of lecture 21
Sylow’s Second Theorem (Theorem 3.6) takes a finite group with Sylow p-subgroups P and K.
There is some x ∈ G for which P = xKx−1 .
Proof of Sylow’s Second Theorem. Let |G| = pn m where p does not divide m. Then |K| = pn . Let
K = K1 , . . . , Kt be the distinct conjugates of K in G. By Proposition 3.16, t = [G : NG (K)].
Since K is a subgroup of NG (K), we have pn divides |NG (K)|. By Lagrange’s Theorem,
|G| = |NG (K)|[G : NG (K)] = |NG (K)|t
so p does not divide t. We want to show that P is one of the Ki .
The set S = {K1 , . . . , Kt } is a union of distinct equivalence classes under P -conjugacy. By
Proposition 3.16, the number of elements P -conjugate to some Ki is [P : P ∩ NG (Ki )], which is a
divisor of |P | = pn by Lagrange’s Theorem. The number of subgroups in each equivalence class is
a power of p so t is a sum of powers of p. Since p does not divide t, at least one of the powers of p
in the sum is p0 = 1. Therefore, there is some i for which xKi x−1 = Ki for all x ∈ P . By Lemma
3.6, x ∈ Ki for each x ∈ P . Thus P ⊂ Ki and, by order considerations, P = Ki . □
Sylow’s Third Theorem (Theorem 3.7) states that the number of Sylow p-subgroups t divides
|G| and is congruent to 1 modulo p.
Proof of Sylow’s Third Theorem. Let S = {K1 , . . . , Kt } be the set of all Sylow p-subgroups of G.
By Sylow’s Second Theorem, these are all conjugates of K1 . The proof of Sylow’s Second Theorem
shows t = [G : NG (K1 )], and t divides the order of G by Lagrange’s Theorem.
Let P = Kj for some 1 ≤ j ≤ t. The only P -conjugate of P is P . Lemma 3.6 shows that the
only equivalence class consisting of a single element is the class consisting of P . Further, S is the
disjoint union of distinct equivalence classes, and each equivalence class has order a power of p.
Only the class containing P has order p0 = 1. Hence, t ≡ 1 (mod p). □
48 MATTHEW GHERMAN
3.5. The Structure of Finite Groups. We will apply the results of this chapter to the classi-
fication of finite groups. In particular, we will classify all groups of order up to 15. We begin the
section with helpful results about p-groups.
Proposition 3.17. If G is a group of order pn with p prime and n ≥ 1, then the center Z(G)
contains more than one element. In particular, |Z(G)| = pk for 1 ≤ k ≤ n.
Proof. By Lagrange’s Theorem, |Z(G)| = pk for 0 ≤ k ≤ n. The class equation implies
r
X
|Z(G)| = |G| − |Cj |
j=1
where each |Cj | is a number larger than 1 that divides |G|. Thus each |Cj | is divisible by p. Since
|G| is divisible by p, the righthand side of the equation is divisible by p. Therefore, p divides
|Z(G)| and 1 ≤ k ≤ n. □
Corollary 3.5. If p is prime and n > 1, then there is no simple group of order pn .
Proof. By Proposition 3.17, a group G of order pn has non-trivial center. The center Z(G) is
normal. If Z(G) ̸= G, then G is not simple. If Z(G) = G, then G is abelian and G is not simple
by Proposition 2.13. □
End of lecture 22
Corollary 3.6. If G is a group of order p2 for p prime, then G is abelian. Hence, G is isomorphic
to Z/p2 Z or Z/pZ × Z/pZ.
Proof. Note that Z(G) is normal so G/Z(G) is a group. By Proposition 3.17, the order of Z(G)
is either p or p2 . If Z(G) has order p2 , then G is abelian. If Z(G) has order p, then |G/Z(G)| = p
and G/Z(G) ≃ Z/pZ by Example 2.4. Since G/Z(G) is cyclic, G is abelian by Proposition 2.9.
We draw a contradiction because Z(G) = G has order p2 . Thus G is abelian of order p2 , and the
Fundamental Theorem of Finite Abelian Groups completes the proof. □
Proposition 3.18. Let p and q be distinct primes such that q ̸≡ 1 (mod p) and p2 ̸≡ 1 (mod q).
If G is a group of order p2 q, then G is isomorphic to Z/(p2 qZ) or Z/pZ × Z/pZ × Z/qZ.
Proof. By Sylow’s Third Theorem, the number mp of Sylow p-subgroups must divide p2 q so the
only options are 1, p, p2 , q, pq, and p2 q. Further, mp ≡ 1 (mod p) so mp = 1 or mp = q. Since
q ̸≡ 1 (mod p) by assumption, mp = 1 and the Sylow p-subgroup P is normal by Proposition 3.4.
By Sylow’s Third Theorem, mq can only be 1, p, or p2 . Since p2 ̸≡ 1 (mod q) by assumption,
neither p nor p2 is an option so mq = 1. The Sylow q-subgroup Q is normal.
The order of P ∩ Q must divide |P | = p2 and |Q| = q so P ∩ Q = {eG }. Thus
|P ||Q|
|P Q| = = p2 q = |G|
|P ∩ Q|
by the Second Isomorphism Theorem. Corollary 3.1 implies G ≃ P × Q. Finally, Corollary 3.6
provides the possibilities for P , and Example 2.4 provides the possibilities for Q. □
Example 3.24. The group A4 has order 12 = 22 · 3. However, A4 cannot be written as a direct
product of abelian groups because A4 is not abelian.
Proposition 3.19. If p and q are distinct primes, then there is no simple group of order p2 q.
Proof. Suppose |G| = p2 q. If either q ̸≡ 1 (mod p) or p2 ̸≡ 1 (mod q), then the proof of Proposition
3.18 proves that G has a non-trivial normal subgroup.
Assume q ≡ 1 (mod p) and p2 ≡ 1 (mod q). Then p|(q − 1) and p ≤ q − 1. Since q is prime and
p2 − 1 = (p − 1)(p + 1), q divides p − 1 or p + 1. However, q cannot divide p − 1 since p ≤ q − 1.
Thus q = p + 1. Since p and q are prime, the only possibility is p = 2 and q = 3.
We will now prove that no group of order 12 is simple. Sylow’s Third Theorem implies that
m3 = 1 or m3 = 4. If m3 = 1, then the Sylow 3-subgroup is normal by Corollary 3.4. Assume
m3 = 4. Each Sylow 3-subgroup is cyclic of order 3. There are 8 elements of order 3 in G. Every
non-trivial element of a Sylow 2-subgroup must have even order so there can only be one Sylow
2-subgroup. By Corollary 3.4, G has a normal Sylow 2-subgroup. □
3.5.1. Dihedral Groups.

Definition 3.13. Let the dihedral group Dn of order 2n be the group generated by {r, s} where
|r| = n, |s| = 2, and sr = r−1 s.
Note that one of the first dihedral group defined, D4 , satisfies the above criteria. As mentioned,
the dihedral group represents the symmetries of a regular n-gon. We can write the elements as
Dn = {e, r, r2 , . . . , rn−1 , s, sr, sr2 , . . . , srn−1 }
where sri = rn−i s for 0 ≤ i < n.
Proposition 3.20. If |G| = 2p for p ̸= 2 a prime, then G is isomorphic to Z/(2p)Z or Dp .
Proof. By Cauchy’s Theorem, G contains an element a of order p and an element b of order 2.
Then b2 = eG implies b−1 = b. The index 2 subgroup H = ⟨a⟩ is normal in G by Lemma 2.1. Thus
bab−1 = bab ∈ H so bab = at for some 0 ≤ t ≤ p − 1. We have
2
at = (at )t = (bab)t = (bab) · · · (bab) = bat b = b(bab)b = a
so t2 ≡ 1 (mod p) by Proposition 1.4(2). Since p divides t2 − 1, p divides t − 1 or t + 1. Rewriting
this, t ≡ 1 (mod p) or t ≡ −1 (mod p).
Assume t ≡ 1 (mod p). Then bab = a or ba = ab. By Proposition 1.1, |ab| = |a||b| = 2p and G
is cyclic generated by ab. Therefore, G ≃ Z/(2p)Z by Proposition 1.16.
End of lecture 23
Assume t ≡ −1 (mod p). Then bab = a−1 . Define f : Dp → G by f (ri sj ) = ai bj . Note that
|a| = |r|, |b| = |s|, and bab = a−1 while srs = r−1 . We can thus show that f is well-defined.
Further, f is a group homomorphism via
f ((ri sj )(rk sℓ )) = f (ri−k sj+ℓ )
= ai−k bj+ℓ
= (ai bj )(ak bℓ )
= f (ri sj )f (rk sℓ ).
Let K = ⟨b⟩. By Lagrange’s Theorem, H∩K = {eG }. Then |HK| = 2p by the Second Isomorphism
Theorem so G = HK. Every element can be written as ai bj , which implies that f is surjective.
Since |Dp | = |G|, f is an isomorphism. □
Example 3.25. The symmetric group S3 is not abelian of order 6 = 2 · 3. Proposition 3.20 implies
that S3 ≃ D3 . Is this true for other symmetric groups Sn ?
3.5.2. Groups of Small Order. The table below contains the finite groups of order at most 15 that
we have classified thus far. We will fill in rows 8 and 12 in this section.
50 MATTHEW GHERMAN
Order Groups Proof

1 {e}
2 Z/2Z Example 2.4
3 Z/3Z Example 2.4
4 Z/4Z, Z/2Z × Z/2Z Example 2.5
5 Z/5Z Example 2.4
6 Z/6Z, D3 Proposition 3.20
7 Z/7Z Example 2.4
8 ?
9 Z/9Z, Z/3Z × Z/3Z Corollary 3.6
11 Z/11Z Example 2.4
12 ?
15 Z/15Z Proposition 3.9
Example 3.26. Let G be a non-abelian group of order 8. In particular, G is not cyclic. Thus
every non-identity element must have order 2 or 4 by Lagrange’s Theorem. If every non-identity
element of G has order 2, then every element of G is its own inverse so
ab = (ab)−1 = b−1 a−1 = ba
for all a, b ∈ G and G is abelian. We conclude there is some element a ∈ G of order 4. Let b ∈ G
be such that b ̸∈ ⟨a⟩. Any two elements in {eG , a, a2 , a3 , b, ab, a2 b, a3 b} are distinct since ai = aj b
implies b ∈ ⟨a⟩. Thus G = {eG , a, a2 , a3 , b, ab, a2 b, a3 b}.
By Lemma 2.1, the index 2 subgroup ⟨a⟩ is normal. The conjugation group homomorphism
f : G → G defined as f (a) = bab−1 is an isomorphism. Thus element bab−1 has order |a| = 4
and bab−1 ∈ ⟨a⟩ by normality. Then bab−1 = a or bab−1 = a3 by order considerations. However,
bab−1 = a would imply G is abelian so bab−1 = a3 or ba = a3 b. We can produce the following part
of the multiplication table of G.
· eG a a2 a3 b ab a2 b a3 b
eG eG a a2 a3 b ab a2 b a3 b
a a a2 a3 eG ab a2 b a3 b b
2
a a2 a3 eG a a2 b a3 b b ab
a3 a3 eG a a2 a3 b b ab a2 b
b b a3 b a2 b ab
ab ab b a3 b a2 b
a2 b a2 b ab b a3 b
a3 b a3 b a2 b ab b
End of lecture 24
In order to complete the table, we want to find b2 . Since b2 = ai b implies b ∈ ⟨a⟩, we conclude
that b2 ∈ ⟨a⟩. If b2 = a, then
ab = b2 b = bb2 = ba
and G is abelian. Similarly, b2 = a3 implies
ab = (a3 )−1 b = b−2 b = bb−2 = b(a3 )−1 = ba
and G is abelian. Therefore, b2 = eG or b2 = a2 . In the first case, the multiplication table will be
that of D4 . In the second case, the multiplication table is that of Q8 .
Definition 3.14. The quaternion group is the set Q8 = {1, −1, i, −i, j, −j, k, −k} with relations
(−1)2 = 1
i2 = j 2 = k 2 = −1
ijk = −1.
Proposition 3.21. If G is a non-abelian group of order 8, then G ≃ D4 or G ≃ Q8 .
Proof. It remains to show that D4 and Q8 are not isomorphic. We see that D4 has two elements
of order 4, {r, r3 } while Q8 has six elements of order 4, {i, −i, j, −j, k, −k}. Thus D4 and Q8 are
not isomorphic. □
Example 3.27. Let G be a non-abelian group of order 12. As a result, G cannot be cyclic and
every non-identity element of G has order 2, 3, 4, or 6 by Lagrange’s Theorem. By Proposition
3.19, either a Sylow 2-subgroup P is normal or a Sylow 3-subgroup Q is normal. Example 2.4
implies Q ≃ Z/3Z. Example 2.5 shows P ≃ Z/4Z or P ≃ Z/2Z × Z/2Z. Lagrange’s Theorem
implies P ∩ Q = {eG } so |P Q| = 12 by Second Isomorphism Theorem and G = P Q.
Case 1: Assume that m2 = 1 so P is normal. Assume that P is not cyclic. Then P = ⟨a1 , a2 ⟩ for
|ai | = 2 and a1 a2 = a2 a1 . Let Q = ⟨b⟩. Since P is normal, order considerations imply ba1 b−1 = a1 ,
ba1 b−1 = a2 , or ba1 b−1 = a1 a2 . The same can be said for conjugating a2 by b. If bai b−1 = ai ,
then baj b−1 = aj for j ̸= i and G is abelian. Assume ba1 b−1 = a2 . Then ba2 b−1 = a1 would
imply b2 a2 b−2 = a2 and b3 a2 b−3 = a1 . We need a2 = b3 a2 b−3 , which is a contradiction. Thus
ba2 b−1 = a1 a2 . If ba1 b−1 = a1 a2 , a similar argument shows that ba2 b−1 = a1 . In either case, a
relabeling results in the same multiplication table. We can show that G is isomorphic to A4 .
Assume that P is cyclic so P = ⟨a⟩. G is generated by {a, b}. Since P is normal, order
considerations imply bab−1 = a or bab−1 = a3 . If bab−1 = a, then G is abelian. If bab−1 = a3 , then
b2 ab−2 = a and, since b3 = eG ,
a = b3 ab−3 = bab−1 = a3 ,
which contradicts |a| = 4. Therefore, G does not have a normal cyclic subgroup of order 4.
Case 2: Assume that m3 = 1 so Q is normal. Let Q = ⟨b⟩. At first, assume P is not cyclic so
P = ⟨a1 , a2 ⟩ for |ai | = 2 and a1 a2 = a2 a1 . Then a1 ba−1 −1 2
1 = b or a1 ba1 = b since order is preserved
and Q is normal. The same can be said of conjugation by a2 . If a1 ba−1 −1
1 = b, then a2 ba2 = b
2
−1 2 −1 2 −1
since G is not abelian. If a1 ba1 = b and a2 ba2 = b , then (a1 a2 )b(a1 a2 ) = b. By possibly
relabeling, we can assume a1 ba−1 −1 2
1 = b and a2 ba2 = b . Then a1 and b commute so Lemma 1.1
implies |a1 b| = lcm(|a1 |, |b|) = 6. We can write
G = {(a1 b)i , a2 (a1 b)j }
for 0 ≤ i, j ≤ 6. Further,
a2 (a1 b)a−1 −1 2 −1
2 = a1 (a2 ba2 ) = a1 b = (a1 b) .
We can show that G ≃ D6 .

Assume that P = ⟨a⟩ is cyclic. Once again, take Q = ⟨b⟩ to be normal in G. G is generated by
{a, b}. If aba−1 = b, then a and b commute and G is abelian. Thus aba−1 = b2 since b2 is the only
other order 3 element in Q. We can write out a unique multiplication table. Denote the group T .
Proposition 3.22. If G is a non-abelian group of order 12, then G is isomorphic to A4 , D6 , or T .
Proof. It remains to check that no two of the groups are not isomorphic. A4 does not have an
element of order 6 while D6 has two elements of order 6. Additionally, the group T has an element
of order 6, namely a2 b in the notation of Example 3.27. The groups D6 and T are not isomorphic
since T has an element of order 4. □
52 MATTHEW GHERMAN
We can now complete the table.

Order Groups Proof
1 {e}
2 Z/2Z Example 2.4
3 Z/3Z Example 2.4
4 Z/4Z, (Z/2Z)2 Example 2.5
5 Z/5Z Example 2.4
7 Z/7Z Example 2.4
8 Z/8Z, Z/4Z × Z/2Z, (Z/2Z)3 , D4 , Q8 Proposition 3.21
9 Z/9Z, Z/3Z × Z/3Z Corollary 3.6
12 Z/12Z, Z/3Z × (Z/2Z)2 , A4 , D6 , T Proposition 3.22
15 Z/15Z Proposition 3.9
End of material covered on the final
End of lecture 25
4. Applications of Group Theory

4.1. Group Actions. Loosely speaking, a group action will be a permutation of a set by a
group. In other areas of mathematics, groups arise mostly as the objects that act on a set. In
combinatorics, groups act on finite sets of objects. In topology, groups like the fundamental group
act on topological spaces. In Galois theory, groups arise as automorphisms of fields.
Definition 4.1. Let G be a group and X a set. G acts on X if there is an operation g · x such
that the following properties hold.
(1) For every g, h ∈ G and x ∈ X, (gh) · x = g · (h · x).
(2) For every x ∈ X, eG · x = x.
We have already seen numerous examples of group actions.
Example 4.1. The group Sn acts on Xn = {1, 2, . . . , n}.
Example 4.2. The group Dn acts on the set of vertices of the regular n-gon.
Example 4.3. Let G be the group of the Rubik’s cube, which is all sequences of motions on the
cube (keeping center colors in fixed locations). The group acts on two different sets: the 12 edge
cubelets and the 8 corner cubelets. We could further let G act on the set of all 20 non-center face
cubelets together.
Example 4.4. The proof of Cayley’s Theorem, Theorem 2.6, relied on a group action. We let the
group G act on the set G by left multiplication. In other words g · x = gx for g ∈ G and x ∈ X.
We have a generalization of Cayley’s Theorem for every group action. Cayley’s Theorem pro-
duces an injective group homomorphism, but that need not be the case in general.
Proposition 4.1. The group G acts on a set X if and only if there exists a group homomorphism
φ : G → A(X) where A(X) is the group of set bijections of X.
Proof. (⇒) Assume that G acts on X. Let g ∈ G. Define φ(g) as the bijection φ(g)(x) = g · x for
x ∈ X. Then φ(gh)(x) = (gh)·x = g·(h·x) = φ(g)(φ(h)(x)). We conclude that φ(gh) = φ(g)◦φ(h)
so φ is a group homomorphism.
(⇐) Assume that φ : G → A(X) is a group homomorphism. Define the group action g ·
x = φ(g)(x). Then eG · x = φ(eG )(x) = x since φ(eG ) is the identity set bijection. Further,
(gh) · x = φ(gh)(x) = φ(g)(φ(h)(x)) = g · (h · x). □
We can use this equivalent condition of a group action to prove the following generalization of
index 2 subgroups being normal. The result can be very useful for future group classification.
Corollary 4.1. Let p be the smallest prime divisor of |G| for a finite group G. A subgroup H of
G with [G : H] = p is normal in G.
Proof. Let G act on the left cosets G/H by left multiplication. There is some group homomorphism
φ : G → A(G/H) by Proposition 4.1. The homomorphism is non-trivial since H is not all of G.
With |G/H| = [G : H] = p, we can identify A(G/H) ≃ Sp . Note that im(φ) is a subgroup of Sp
so |im(φ)| divides |Sp | = p! by Lagrange’s Theorem. In order for φ(g) = idX , we need g · H = H
or g ∈ H. Thus H ⊂ ker(φ).
The First Isomorphism Theorem implies that G/ ker(φ) ≃ im(φ). In particular, |G/ ker(φ)| =
|G|/| ker(φ)| divides p!. The only factor that |G| and p! share is p since p is the smallest prime di-
viding |G|. We conclude that |G/ ker(φ)| is 1 or p. With φ non-trivial, |G/ ker(φ)| = p. Therefore,
| ker(φ)| = |G|
p
= |H| and H = ker(φ). By Proposition 2.10, H is a normal subgroup of G. □
54 MATTHEW GHERMAN
Example 4.5. In the proof of Sylow’s Theorems, we made use of a different group action of the
group G on itself. A group G can act on the set G via conjugation. In other words, g · x = gxg −1 .
End of lecture 26
Definition 4.2. Let G be a group that acts on a set X. For x ∈ X, the stabilizer of x in G,
written StabG (x), is the set of elements such that g · x = x.
StabG (x) = {g ∈ G : g · x = x}
Note that StabG (x) is a subset of G.
Definition 4.3. Let G be a group that acts on a set X. For x ∈ X, the orbit of x under G,
written OrbG (x) is the set of all elements in X of the form g · x for g ∈ G.
OrbG (x) = {y ∈ Y : g · x = y}
Note that OrbG (x) is a subset of X.
Proposition 4.2. Let G be a group that acts on the set X. For x, y ∈ X, the relation x ∼ y if
and only if g · x = y is an equivalence relation. We can partition X into a disjoint union of orbits.
Proof. Reflexivity: x ∼ x since eG · x = x.
Symmetry: Let x ∼ y. Then g · x = y and x = g −1 · y so y ∼ x.
Transitivity: Let x ∼ y, y ∼ z. Then g · x = y and h · y = z. We have (hg) · x = z. □
Example 4.6. Let 1 ∈ X3 and G = S3 . Then StabG (1) = {e, (23)} and OrbG (1) = {1, 2, 3} = X3 .
Definition 4.4. A group action is transitive if for any pair x, y ∈ X, there is some g ∈ G for
which y = g · x. In other words, the group action only has one orbit.
Example 4.7. Let G be a group that acts on the set G by conjugation. The orbit of some a ∈ G
is the conjugacy class of a. The stabilizer of a ∈ G is the centralizer CG (a).
Proposition 4.3 (Cauchy’s Theorem). Let G be a finite group with prime p dividing |G|. Then
G has an element of order p.
Proof. Let X = {(x1 , . . . , xp ) ∈ Gp : x1 · · · xp = eG }. Each element in the p-tuple is an independent
choice except for the last element xp = x−1 −1
p−1 · · · x1 . Thus |X| = |G|
p−1
and p divides |X|. Let
Z/pZ act on X by cyclically permuting the elements of each tuple. In particular,
1 · (x1 , . . . , xp−1 , xp ) = (xp , x1 , . . . , xp−1 ).
The action sends an element of X to another element of X since
eG = x1 · · · xp−1 xp = (x1 · · · xp−1 )(x1 · · · xp−1 )−1 = (x1 · · · xp−1 )−1 (x1 · · · xp−1 ) = xp (x1 · · · xp−1 )
in G. The orbit of each point is either size 1 or p. The orbit of an element x ∈ X is size 1 if and
only if x = (a, . . . , a) or ap = eG . Partition X into a disjoint union of orbits. Thus
(# orbits of size 1) = |X| − p · (# orbits of size p) ≡ 0 (mod p).
Since (eG , . . . , eG ) ∈ X is an element with orbit size 1, there must be at least p − 1 other elements
with an orbit of size 1. □
Proposition 4.4. Let the group G act on X. For each x ∈ X, StabG (x) is a subgroup of G.
Proof. We have eG ∈ StabG (x) by property (2) of the group action definition. Let g, h ∈ StabG (x).
Then (gh) · x = g · (h · x) = g · x = x by property (1) of the group action definition. Further,
x = eG · x = (g −1 g) · x = g −1 · (g · x) = g −1 · x. □
We now find a generalization of Proposition 3.12. There is an inverse relationship between

elements that fix x ∈ X, the stabilizer, and all the elements one can obtain as g · x, the orbit.
With the orbit and stabilizer, we often get a complete picture of the group action.
Theorem 4.1 (Orbit-Stabilizer). Suppose G is a finite group that acts on the set X. For any
x ∈ X, then |G| = |StabG (x)||OrbG (x)|.
Proof. Let S be the set of right cosets of StabG (x) in G. Let T = Orbx (G). Define the set map
f : T → S such that f (g·x) = StabG (x)g. We need to check that f is well-defined. Given g·x = h·x,
then h−1 g ∈ StabG (x). Thus StabG (x)g = StabG (x)h. We want to show that f is injective. Assume
f (g · x) = f (h · x). Then StabG (x)g = StabG (x)h which implies h−1 g ∈ StabG (x). As a result,
g · x = h · x. To show f is surjective, let StabG (x)g be a right coset. Then f (g · x) = StabG (x)g.
We conclude that f is a set bijection. □
Example 4.8. Let the group G = S3 act on X3 = {1, 2, 3}. Pick 1 ∈ X3 so OrbG (1) = {1, 2, 3}
and StabG (1) = {e, (23)}. We confirm Orbit-Stabilizer below.
|G| = |StabG (x)||OrbG (x)|

6=2·3
Example 4.9. Let G = D4 act on the vertices V = {1, 2, 3, 4} of a square.
Define r as counterclockwise rotation of the square by π2 and s the leftmost reflection. Pick 1 ∈ V
so OrbG (1) = {1, 2, 3, 4} and StabG (1) = {e, rs}. We once again confirm Orbit-Stabilizer.
Example 4.10. Let G = GLn (R) act on the set Rn where A · v is the matrix multiplication
Av. The stabilizer StabG (v) is the non-trivial subgroup of matrices that have invariant subspace
Span(v). The orbit of some v is the entirety of Rn so the group action is transitive. Note that
Orbit-Stabilizer does not apply since G is not a finite group.
4.1.1. Burnside’s Lemma. We can rearrange the result of Orbit-Stabilizer to obtain a sometimes
more useful result. Burnside’s Lemma usually is applied to problems in combinatorics.
Definition 4.5. Let the group G act on a set X. A fixed point of g ∈ G is an element x ∈ X for
which g · x = x. Denote by Fix(g) the subset of fixed points of g in X.
Definition 4.6. Let the group G act on a set X. Denote by X/G the set of orbits of X with
respect to the group action.
Lemma 4.1 (Burnside’s Lemma). Let the group G act on the set X. Then
1 X
|X/G| = |Fix(g)|.
|G| g∈G
56 MATTHEW GHERMAN
Proof. We can write the sum in two different ways

X X
|Fix(g)| = |{x ∈ X : g · x = x}|
g∈G g∈G
= |{(g, x) ∈ G × X : g · x = x}|
X
= |{g ∈ G : g · x = x}|
x∈X
X
= StabG (x).
x∈X
In |X/G|, each orbit provides 1 to the sum. The size of each orbit is P
|OrbG (x)| so we can view
each x ∈ X as contributing |OrbG (x)| to the sum |X/G|. Thus |X/G| = x∈X |Orb1G (x)| .
1
By Orbit-Stabilizer and the above results, we have

X 1
|X/G| =
x∈X
|OrbG (x)|
X |StabG (x)|
=
x∈X
|G|
1 X
= |StabG (x)|
|G| x∈X
1 X
= |Fix(g)|.
|G| g∈G
□
Example 4.11. We can use Burnside’s Lemma to count the number of rotationally different
3-colorings of the cube.
Let X be the set of colorings of a cube with fixed orientation so |X| = 36 . The group G is the
rotational symmetry group of a cube. Each rotation of a cube has an axis of rotation through the
center of two opposite faces, edges, or vertices. We organize the 24 distinct rotations below.
• The identity element of G
• The six 90-degree face rotations
• The three 180-degree face rotations
• The eight 120-degree vertex rotations
• The six 180-degree edge rotations
Two colorings belong to the same orbit when one coloring can be obtained from another by rotating
the cube. We want to count the number of orbits, |X/G|.
• The identity element of G leaves all 36 elements of X fixed.
• The six 90-degree face rotations each leave 33 of the elements of X fixed.
• The three 180-degree face rotations each leave 34 of the elements of X fixed.
• The eight 120-degree vertex rotations each leave 32 of the elements of X fixed.
• The six 180-degree edge rotations each leave 33 of the elements of X fixed.
Burnside’s Lemma implies the number of rotationally distinct 3-colorings of the cube is
1
|X/G| = (36 + 6 · 33 + 3 · 34 + 8 · 32 + 6 · 33 ) = 57.
24
58 MATTHEW GHERMAN
4.2. Solvable and Nilpotent Groups. The Classification of Finite Abelian groups allows us to
easily list all finite abelian groups of any order up to isomorphism. In this section, we will define
more general classes of groups called nilpotent and solvable groups that can be approximated well
by abelian groups. We hope that the power of the Classification of Finite Abelian Groups can be
extended to such a class of groups.
4.2.1. Nilpotent groups.
Definition 4.7. Let G be a group. A central series is a sequence of subgroups
G = H0 ▷ H1 ▷ · · · ▷ Hn−1 ▷ Hn = {eG }
such that, for each 0 ≤ i ≤ n − 1, Hi+1 is normal in G and Hi /Hi+1 is a subgroup of Z(G/Hi+1 ).
The condition implies Hi /Hi+1 is abelian for each 0 ≤ i ≤ n − 1.
A group does not necessarily have a central series. However, there is a minimal choice at each
step in the construction of such a sequence. In order for Hi /Hi+1 to be in the center of G/Hi+1 , we
need elements of Hi to commute with elements of G modulo Hi+1 . Defining Hi+1 = [G, Hi ] where
[G, Hi ] is the subgroup of G generated by {ghg −1 h−1 : g ∈ G, h ∈ Hi } could produce a central
series.
Definition 4.8. Let G be a group. A lower central series is a sequence of subgroups
G = H0 ▷ H1 ▷ · · · ▷ Hn ▷ . . .
such that Hi+1 = [G, Hi ] for 0 ≤ i ≤ n − 1.
Definition 4.9. A nilpotent group is one that has a central series. Equivalently, a nilpotent group
is one that has a lower central series that terminates to the trivial subgroup in finitely many steps.
We write
G = H0 ▷ H1 ▷ · · · ▷ Hn−1 ▷ Hn = {eG }
such that Hi+1 = [G, Hi ] for 0 ≤ i ≤ n − 1.
We can think of central series as approximating a nilpotent group G by abelian groups. We
hope that many of the techniques used to study abelian groups can be extended to the study of
nilpotent groups. The following examples will illustrate that nilpotent groups really do extend the
notion of an abelian group.
Example 4.12. Abelian groups are nilpotent. Let G be an abelian group. To construct a lower
central series, H0 = G and H1 = [G, G] = {eG }.
Example 4.13. The quaternion group Q8 is nilpotent. The commutator subgroup
[Q8 , Q8 ] = {±1}
−1 −1
by computing all products similar to iji j = kk = −1. Thus we have the following finite lower
central series.
G = H0 ▷ {±1} ▷ {1}
Example 4.14. The direct product of two nilpotent groups is nilpotent. Let G1 and G2 be
nilpotent groups. We have the following two lower central series where, without loss of generality,
m ≤ n.
G1 = H10 ▷ H11 ▷ · · · ▷ H1n = {eG1 }
G2 = H20 ▷ H21 ▷ · · · ▷ H2m = {eG2 }
Then the center Z(G1 × G2 ) = Z(G1 ) × Z(G2 ). One can check that the following sequence is a
lower central series for G1 × G2 .
G1 × G2 = H10 × H20 ▷ H11 × H21 ▷ · · · ▷ H1m × H2m ▷ · · · ▷ H1n × {eG2 } = {eG1 } × {eG2 }
Proposition 4.5. All finite p-groups are nilpotent.

Proof. We will proceed via induction on the order of G. As a base case, the trivial group is
nilpotent. If Z(G) = G, then G is abelian and nilpotent by Example 4.12. Assume without loss
of generality that G is not abelian. Then Proposition 3.17 implies that Z(G) is non-trivial. The
group G/Z(G) is a p-group of order less than |G| so the inductive hypothesis implies that G/Z(G)
is nilpotent. We can build the following central series for G/Z(G) where Hi+1 ◁ G/Z(G) and
Hi /Hi+1 is a subgroup of Z((G/Z(G))/Hi+1 ) for each 0 ≤ i ≤ n − 1.
G/Z(G) = H0 ▷ H1 ▷ · · · ▷ Hn−1 ▷ Hn = {eG/Z(G) }
By the Correspondence Theorem for Normal Subgroups, we can lift to the following sequence
where Hfi ◁ G and, by the Third Isomorphism Theorem, H fi /H
]i+1 is a subgroup of Z(G/Hi+1 ) for
]
each 0 ≤ i ≤ n − 1.
G=H f1 ▷ · · · ▷ H
f0 ▷ H ]n−1 ▷ Hn = Z(G)
f
Since Z(G) is abelian, we can construct the central series.
G=H f1 ▷ · · · ▷ H
f0 ▷ H ]n−1 ▷ Hn ▷ Hn+1 = {eG }
f ]
Therefore, G is nilpotent. □
Corollary 4.2. Any finite direct product of finite p-groups is nilpotent.
Proof. Proceed by induction on the number of components in the direct product. Proposition 4.5
is the base case. Example 4.14 completes the inductive step. □
Theorem 4.2. A finite group G is nilpotent if and only if G is a direct product of p-groups.
Equivalently, a finite group G is nilpotent if and only if each of the Sylow p-subgroups of G is
normal.
Although the class of nilpotent groups generalizes the class of abelian groups, Theorem 4.2
proves that it is not a novel enough definition. In the next section we introduce solvable groups, a
sufficiently distinct generalization of abelian groups to encompass what we need in future studies
like Galois theory.
4.2.2. Solvable groups.

Definition 4.10. A subnormal series of a group G is a sequence of subgroups
G = H0 ▷ H1 ▷ H2 ▷ · · · ▷ Hn−1 ▷ Hn = {eG }
such that Hi+1 ◁ Hi and Hi /Hi+1 is an abelian group for each 0 ≤ i ≤ n − 1. Note that Hn need
not be normal in G.
Definition 4.11. Let G be a group and define G(0) = G and G(i+1) = [G(i) , G(i) ]. Then the
sequence of subgroups
G = G(0) ▷ G(1) ▷ G(2) ▷ . . .
is a derived series.
Definition 4.12. A group is solvable if it has a subnormal series. Equivalently, a group is solvable
if it has a derived series that terminates to {eG } in finitely many steps.
The next example shows that solvable groups extend the idea of nilpotent, and thus abelian,
groups.
60 MATTHEW GHERMAN
Example 4.15. A central series is an example of a subnormal series. The condition that Hi+1
is normal in G and Hi /Hi+1 is contained in Z(G/Hi+1 ) implies that Hi+1 is normal in Hi and
Hi /Hi+1 is abelian. Thus nilpotent groups are solvable.
cyclic groups ⊂ abelian groups ⊂ nilpotent groups ⊂ solvable groups
Example 4.16. In order to prove that not all solvable groups are nilpotent, take G = S3 . We can
show that [S3 , S3 ] = A3 . Since A3 ≃ Z/3Z is abelian, [A3 , A3 ] = {e}. The derived series of S3 is
S3 ▷ A3 ▷ {e}.
Thus S3 is solvable. However, S3 does not have a normal Sylow 2-subgroup. Theorem 4.2 implies
S3 is not nilpotent.
Example 4.17. S5 is not solvable. We can show that the commutator subgroup [S5 , S5 ] = A5 .
Since A5 is simple and non-abelian, [A5 , A5 ] = A5 . The derived series of S5 does not terminate to
the trivial subgroup.
In general, Sn is not solvable for n ≥ 5. This is a key step in the proof of Abel–Ruffini Theorem
that there are polynomials of each degree n ≥ 5 that are not solvable by radicals.
Theorem 4.3 (Feit-Thompson). Every finite group of odd order is solvable.
Corollary 4.3. If a finite group is simple, it is either cyclic of prime order or of even order.
Proof. If G is simple abelian, then G ≃ Z/pZ for prime p by Proposition 2.13. Assume G is
non-abelian. If G is solvable, then
{eG } ⊂ [G, G] ⊂ G.
Feit-Thompson implies that a non-abelian simple group G is even order. □

Math 110B - Group Theory Lecture Notes

Uploaded by

Copyright:

Available Formats

Math 110B - Group Theory Lecture Notes

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Math 110B - Group Theory Lecture Notes

Uploaded by

Copyright:

Available Formats

MATH 110B - GROUP THEORY

These notes are based on Hungerford, Abstract Algebra 3rd edition.

Likewise, we can rotate by an angle of 2π3

1.2. Basic properties of groups.

1.3.1. Center of a group.

Definition 1.7. The center of a group G is

Z(G) = {h ∈ G : gh = hg for all g ∈ G}.

Proposition 1.6. The center of G is a subgroup of G.

Proof. The identity of G is always in the center so Z(G) is non-empty.

Let r be a rotation by π2 counterclockwise and s be a reflection. Then the elements of D4 are

Example 1.24. Let E = 2Z be the additive group of even integers. Define f : Z → E as

Proof. Let n be the order of a and k the order of f (a). Then

f (k + ℓ) = f (k + ℓ) = 2(k + ℓ) = 2k + 2ℓ = f (k) + f (ℓ)

2. Normal Subgroups and Quotient Groups

For a subgroup K of G and an element a ∈ G, define the congruence class of a modulo K to

(2) If |G| = k, then ak = e for every a ∈ G.

We can inductively prove that gng −1 ∈ N if si ns−1i ∈ N for each si ∈ S.

2.3. Quotient Groups.

The following result partially motivates the fraction notation.

Proposition 2.6. Let N be a normal subgroup of G with a, b ∈ G.

The inverse of N a is N a−1 since

(N a)(N a−1 ) = N (aa−1 ) = N e = N

The operation is associative since

((N a)(N b))(N c) = N (ab)N c

(2) The order |G/N | is [G : H] which, by Lagrange’s Theorem, is |G|/|N |.

kj ∈ K. Then f (n1 k1 ) = (N ∩ K)k1 while f (n2 k2 ) = (N ∩ K)k2 . However, k2 k1−1 = n−1

2.5.1. Alternating Groups.

3. Topics in Group Theory

Thus f is a homomorphism. Assume f (a1 , . . . , ak ) = eG . Then a1 · · · ak = eG or

Since ai ∈ Ni and a−1 −1 −1 −1 −1 −1 −1 −1

(Z/15Z)× ≃ M × N ≃ Z/2Z × Z/4Z.

Proposition 3.2. For a group G, an automorphism of G is a group isomorphism f : G → G. The

(n1 , h1 ) · (n2 , h2 ) = (n1 φ(h1 )(n2 ), h1 h2 ).

(n, h)−1 = (φ(h)−1 (n−1 ), h−1 ) = (φ(h−1 )(n−1 ), h−1 ).

Proof. The multiplication is associative. The identity element satisfies

(n, h)(eN , eH ) = (nφ(h)(eN ), h)

The inverse element satisfies

Proof. Let φ : H → Aut(N ) be the group homomorphism φ(h) = (n 7→ hnh−1 ). Define

3.4. Conjugacy and the Proof of Sylow’s Theorems.

(xy)A(xy)−1 = x(yAy −1 )x−1 = xAx−1 = A

3.5.1. Dihedral Groups.

Order Groups Proof

We can show that G ≃ D6 .

We can now complete the table.

4. Applications of Group Theory

We now find a generalization of Proposition 3.12. There is an inverse relationship between

|G| = |StabG (x)||OrbG (x)|

Example 4.9. Let G = D4 act on the vertices V = {1, 2, 3, 4} of a square.

Proof. We can write the sum in two different ways

By Orbit-Stabilizer and the above results, we have

Proposition 4.5. All finite p-groups are nilpotent.

4.2.2. Solvable groups.

You might also like