1
ABSTRACT ALGEBRA
DAVID S. DUMMIT AND RICHARD M. FOOTE
Solutions provided by Scott Larson.
Contents
0. Preliminaries 3
0.1. Basics 3
0.2. Properties of the Integers 4
0.3. Z/nZ: The Integers Modulo n 6
Part I Group Theory 8
1. Introduction to Groups 8
1.1. Basic Axioms and Examples 8
1.2. Dihedral Groups 10
1.3. Symmetric Groups 11
1.4. Matrix Groups 11
1.5. Quaternion Groups 13
1.6. Homomorphisms and Isomorphisms 14
1.7. Group Actions 14
2. Subgroups 15
2.1. Denition and Examples 15
2.2. Centralizers and Normalizers, Stabilizers and Kernels 15
2.3. Cyclic Groups and Cyclic Subgroups 16
2.4. Subgroups Generated by Subsets of a Group 17
2.5. Denitions and Examples 19
2.6. The Lattice of Subgroups of a Group 20
3. Quotient Groups and Homomorphisms 20
3.1. Denitions and Examples 20
3.2. More on Cosets and Lagranges Theorem 21
3.3. The Isomorphism Theorems 23
3.4. Composition Series and the Holder Program 24
3.5. Transpositions and the Alternating Group 24
4. Group Actions 25
4.1. Group Actions and Permutation Representations 25
4.2. Groups Acting on Themselves by Left Multiplication Cayleys Theorem 25
4.3. Groups Acting on Themselves by Conjugation The Class Equation 25
1
2 DAVID S. DUMMIT AND RICHARD M. FOOTE
4.4. Automorphisms 26
4.5. The Sylow Theorems 26
4.6. The Simplicity of A
n
27
5. Direct and Semidirect Products and Abelian Groups 28
5.1. Direct Products 28
5.2. The Fundamental Theorem of Finitely Generated Abelian Groups 29
5.3. Table of Groups of Small Order 30
5.4. Recognizing Direct Products 30
5.5. Semidirect Products 30
6. Further Topics in Group Theory 31
6.1. pgroups, Nilpotent Groups, and Solvable Groups 31
Part II Ring Theory 33
7. Introduction to Rings 33
7.1. Basic Denitions and Examples 33
7.2. Examples: Polynomials Rings, Matrix Rings, and Group Rings 34
7.3. Ring Homomorphisms and Quotient Rings 36
7.4. Properties of Ideals 39
7.5. Rings of Fractions 42
8. Euclidean, Principal Ideal, and Unique Factorization Domains 45
8.1. Euclidean Domains 45
9. Polynomial Rings 46
9.1. Denitions and Basic Properties 46
9.2. Polynomial Rings Over Fields I 47
9.3. Polynomial Rings That are Unique Factorization Domains 48
9.4. Irreducibility Criteria 48
9.5. Polynomial Rings Over Fields II 50
9.6. Polynomials in Several Variables Over a Field and Grobner Bases 50
Part III Modules and Vector Spaces 53
10. Introduction to Module Theory 53
10.1. Basic Denitions and Examples 53
13. Field Theory 54
13.1. Basic Theory of Field Extensions 54
13.2. Basic Theory of Field Extensions 54
13.3. Classical Straightedge and Compass Constructions 56
13.4. Splitting Fields and Algebraic Closures 56
13.5. Separable and Inseparable Extensions 56
13.6. Cyclotomic Polynomials and Extensions 57
14. Galois Theory 57
14.1. Basic Denitions 57
14.2. The Fundamental Theorem of Galois Theory 57
15. Commutative Rings and Algebraic Geometry 58
15.1. Noetherian Rings and Ane Algebraic Sets 58
15.2. Radicals and Ane Varieties 58
15.3. Integral Extensions and Hilberts Nullstellensatz 58
15.4. Localization 58
3
0. Preliminaries
0.1. Basics.
Proposition 0.1. Let f : A B.
(1) The map f is injective if and only if f has a left inverse.
(2) The map f is surjective if and only if f has a right inverse.
(3) The map f is a bijection if and only if there exists g : B A such that f g is the identity map on B and
g f is the identity map on A.
(4) If A and B are nite sets with the same number of elements ([A[ = [B[), then f : A B is bijective if and
only if f is injective if and only if f is surjective.
Proof. (a) Suppose f is injective so that f
1
(b) contains a single element for b f(A). Thus g(b) = f
1
(b) is
welldened for b f(A). Hence g(f(a)) = a for all a A. Now suppose that f has a left inverse g so that
g(f(a)) = a for all a A. If f(a
1
) = f(a
2
) = b, then g(b) = a
1
and g(b) = a
2
. But g is a welldened map so
a
1
= a
2
and thus f is injective.
(b) Suppose f is surjective. Take some b B so we can choose a A such that f(a) = b. Thus we can dene
g(b) = a such that f(a) = b. Hence f(g(b)) = f(a) = b. Now suppose that f has a right inverse g so that f(g(b)) = b
for all b B. So f(g(B)) = B and thus f is surjective.
(c) Suppose that f is bijective so there is a left inverse, g
l
, and a right inverse, g
r
, of f. Let f(a) = b so that
a = g
l
(f(a)) = g
l
(b) and b = f(g
r
(b)). Since f is injective, g
r
(b) = a and thus g
l
= g
r
. Thus the right and left
inverses of f are the same map. Now suppose that there exists g : B A such that f g is the identity map on
B and g f is the identity map on A. Then f is surjective because f(g(B)) = B. Now let f(a
1
) = f(a
2
) so that
a
1
= g(f(a
1
)) = g(f(a
2
)) = a
2
. Thus f is injective.
(d) Suppose that f is injective. Then f takes elements of A to unique elements of B. Since [A[ = [B[, f is
surjective. Now suppose that f is surjective. Then since [A[ = [B[, f takes elements of A to unique elements of B
so f is injective. Therefore f is either bijective or neither injective or surjective.
Proposition 0.2. Let A be a nonepty set.
(1) If denes an equivalence relation on A then the set of equivalence classes of form a partition of A.
(2) If A
i
[ i I is a partition of A then there is an equivalence relation on A whose equivalence classes are
preciselly the sets A
i
, i I.
In Exercises 1 to 4 let / be the set of 2 2 matrices with real number entries. Recall that matrix multiplication
is dened by
a b
c d
p q
r s
ap +br aq +bs
cp +dr cq +ds
Let
M =
1 1
0 1
and let
B = X / [ MX = XM .
0.1.1. The following elements of / lie in B:
1 1
0 1
0 0
0 0
1 0
0 1
.
0.1.2. Prove that if P, Q B, then P +Q B (where + denotes the usual sum of two matrices).
Proof.
(P +Q)M = PM +QM = MP +MQ = M(P +Q).
0.1.3. Prove that if P, Q B, then P Q B (where denotes the usual product of two matrices).
Proof.
(P Q)M = P M Q = M P Q = M(P Q).
p q
r s
B.
r = 0, p = s.
0.1.5. Determine whether the following functions f are well dened:
(1) f : Q Z dened by f(a/b) = a.
(2) f : Q Q dened by f(a/b) = a
2
/b
2
.
Proof. (a) f is not a well dened map because 1/2 = 2/4 and 1 = 2.
(b) f is a well dened map because if a
1
/b
1
= a
2
/b
2
, then a
2
1
/b
2
1
= a
2
2
/b
2
2
.
0.1.6. Determine whether the function f : 1
+
Z dened by mapping a real number r to the rst digit to the
right of the decimal point in a decimal expansion of r is well dened.
f is well dened because the decimal expansion of a real number is unique. Thus there is one and only one
possible rst number to the right of the decimal for each real number.
0.1.7. Let f : A B be a surjective map of sets. Prove that the relation
a b if and only if f(a) = f(b)
is an equivalence relation whose equivalence classes are the bers of f.
Proof. f(a) = f(a) so is reexive. If f(a) = f(b), then f(b) = f(a) so is symmetric. If f(a) = f(b) and
f(b) = f(c), then f(a) = f(c) so is transitive.
If a
1
, a
2
f
1
(b), then f(a
1
) = f(a
2
) so a
1
a
2
. If a
1
a
2
then f(a
1
) = f(a
2
) so a
1
and a
2
are in the same
ber of f.
0.2. Properties of the Integers.
(1) (Well Ordering of Z) If A is any nonempty subset of Z
+
, there is some element m A such that m a,
for all a A.
(2) If a, b Z with a = 0, we say a divides b if there is an element c Z such that b = ac. In this case we write
a [ b; if a does not divide b we write a [ b.
(3) If a, b Z` 0, there is a unique positive integer d, called the greatest common divisor of a and b, satisfying:
(a) d [ a and d [ b,
(b) if e [ a and e [ b, then e [ d.
(4) If a, b Z` 0, there is a unique positive integer l, called the least common multiple of a and B, satisfying:
(a) a [ l and b [ l,
(b) if a [ m and b [ m, then l [ m. The connection between the greatest common divisor d and the least
common multiple l of the two integers a and b is given by dl = ab.
(5) The Division Algorithm: if a, b Z and b = 0, then there exist unique q, r Z such taht
a = qb +r and 0 r < [b[ ,
where q is the quotient and r the remainder.
(6) The Euclidean Algorithm is an important procedure which produces a greatest common divisor of two
integers a and b by iterating the Division Algorithm: if a, b Z` 0, then we obtain a sequence of quotients
and remainders
a = q
0
b +r
0
b = q
1
r
0
+r
1
r
0
= q
2
r
1
+r
2
r
1
= q
3
r
2
+r
3
.
.
.
r
n2
= q
n
r
n1
+r
n
r
n1
= q
n+1
r
n
where r
n
is the last nonzero remainder. Such an r
n
exists since [b[ > [r
0
[ > [r
1
[ > > [r
n
[ is a decreasing
sequence of strictly positive integers if the remainders are nonzero and such a sequence cannot continue
indenitely. Then r
n
= (a, b).
5
(7) One consequence of the Euclidean Algorithm which we chall use regularly is the following: if a, b Z` 0,
then there exist x, y Z such that
(a, b) = ax +by
that is, the gcd of a and b is a Zlinear combination of a and b. This follows by recursively writing the
element r
n
in the Euclidean Algorithm in terms of the previous remainders.
(8) An element p of Z
+
is calle a prime if p > 1 and the only posotive divisors of p are 1 and p. An integer
n > 1 which is not prime is called composite. An important property of primes is if p is a prime and p [ ab,
for some a, b Z then either p [ a or p [ b.
(9) The fundamental theorem of arithmetic says: if n Z, n > 1, then n can be factored uniquely into the
product of primes, i.e., there are distinct primes p
1
, p
2
, . . . , p
s
and positive integers
1
,
2
, . . . ,
s
such that
n = p
1
1
p
2
2
p
s
s
.
This factorization is unique in the sense that if q
1
, q
2
, . . . , q
t
are any distinct primes and
1
,
2
, . . . ,
t
, are
positive integers such that
n = q
1
1
q
2
2
q
t
t
,
then s = t and if we arrange the two sets of primes in increasing order, then q
i
= p
i
and
i
=
i
, 1 i s.
Suppose the positive integers a and b are expressed as products of prime powers:
a = p
1
1
p
2
2
p
s
s
, b = p
1
1
p
2
2
p
s
s
where p
1
, p
2
, . . . , p
s
are distinct and the exponents are 0. Then the gcd of a and b is
(a, b) = p
min(1,1)
1
p
min(2,2)
2
p
min(s,s)
s
and the lcm is obtained by taking the maximum of the
i
and
i
instead of the minimum.
(10) The Euler function is dened as follows: for n Z
+
let (n) be the number of positive integers a n
with a relatively prime to n, i.e., (a, n) = 1. For primes p, (p) = p 1, and, more generally, for all a 1
we have the formula
(p
a
) = p
a
p
a1
= p
a1
(p 1).
The function is multiplicative in the sense that
(ab) = (a)(b) if (a, b) = 1.
Together with the formula above this gives a general formula for the values of : if n = p
1
1
p
2
2
p
s
s
, then
(n) = (p
1
1
)(p
2
2
) (p
s
s
)
= p
11
1
(p
1
1)p
21
2
(p
2
1) p
s1
s
(p
s
1).
0.2.1. For each of the following paris of integers a and b, determine their gcd, their lcm, and write their gcd in the
form ax +by for some integers x and y.
(a) a = 20, b = 13.
(b) a = 69, b = 372.
(c) a = 792, b = 275.
(d) a = 11391, b = 5673.
(e) a = 1761, b = 1567.
(f) a = 507885, b = 60808.
Solution.
0.2.2. Prove that if the integer k divides the integers a and b then k divides as +bt for every pair of integers s and
t.
Proof.
0.2.3. Prove that if n is composite then there are integers a and b such that n divides ab but n does not divide
either a or b.
Proof.
6 DAVID S. DUMMIT AND RICHARD M. FOOTE
0.2.4. Let a, b and N be xed integers with a and b nonzero and let d = (a, b) be the gcd of a and b. Suppose x
0
and y
0
are particular solutions to ax
+
by = N. Prove for any integer t that the integers
x = x
0
+
b
d
t and y = y
0
a
d
t
are also solutions to ax +by = N (this is in fact the general solution).
Proof.
0.2.5. Determine the value (n) for each integer n 30 where denotes the Euler function.
Solution.
0.2.6. Prove the Well Ordering Property of Z by induction and prove the minimal element is unique.
Proof.
0.2.7. If p is a prime prove that there do not exist nonzero integers a and b such that a
2
= pb
2
.
Proof.
0.2.8. Let p be a prime, n Z
+
. Find a formula for the largest power of p which divides n! = n(n1)(n2) 2 1.
Solution.
0.2.9. Write a computer program to determine the greatest common divisor (a, b) of two integers a and b and to
express (a, b) in the form ax +by for some integers x and y.
0.2.10. Prove for any given positive integer N there exist only nitely many integers n with (n) = N where
denotes Eulers function. Conclude in particular that (n) tends to innity as n tends to innity.
Proof.
0.2.11. Prove that if d divides n then (d) divides (n) where denotes Eulers function.
Proof.
0.3. Z/nZ: The Integers Modulo n.
Theorem 0.3. The operations of addition and multiplication on Z/nZ are well dened, that is, they do not depend
on the choices of representatives for the classes involved. More precisely, if a
1
, a
2
Z and b
1
, b
2
Z with a
1
= b
1
and a
2
= b
2
, then a
1
+a
2
= b
1
+b
2
and a
1
a
2
= b
1
b
2
, i.e., if
a
1
b
1
mod n and a
2
b
2
mod n
then
a
1
+a
2
b
1
+b
2
mod n and a
1
a
2
b
1
b
2
mod n.
Proposition 0.4. Z/nZ)
= a Z/nZ [ (a, n) = 1.
0.3.1. Write down explicitly all the elements in the residue classes of Z/18Z.
Solution.
0.3.2. Prove that the distinct equivalence classes in Z/nZ are precisely 0, 1, 2, . . . , n 1.
Proof.
0.3.3. Prove that if a = a
n
10
n
+a
n1
n1
+ +a
1
10 +a
0
is any positive integer then a a
n
+a
n1
+ +a
1
+a
0
mod 9.
Proof.
0.3.4. Compute the remainder when 37
100
is divided by 29.
Solution.
0.3.5. Compute the last two digits of 9
1500
.
Solution.
0.3.6. Prove that the squares of the elements in Z/4Z are just 0 and 1.
7
Proof.
0.3.7. Prove for any integers a and b that a
2
+b
2
never leaves a remainder of 3 when divided by 4.
Proof.
0.3.8. Prove that for any integers a and b that a
2
+b
2
= 3c
2
has no solutions in nonzero integers a, b, and c.
Proof.
0.3.9. Prove that the square of any odd integer always leaves a remainder of 1 when divided by 8.
Proof.
0.3.10. Prove that the number of elements of (Z/nZ)
, then a b (Z/nZ)
.
Proof.
0.3.12. Let n Z, n > 1, and let z Z with 1 a n. Prove if a and n are not relatively prime then there exists
an itneger b with 1 b < n such that ab 0 mod n and deduce that there cannot be an itneger c such that ac 1
mod n.
Proof.
0.3.13. Let n Z, n > 1, and let a Z with 1 a n. Prove that if a and n are relatively prime then there is
an integer c such that ac 1 mod n.
Proof.
0.3.14. Conclude from the previous two exercises that (Z/nZ)
(a, b) (c, d)
(e, f) =
= (a, b)
(c, d) (e, f)
1.1.8. Let G = z C [ z
n
= 1 for some n Z
+
.
(a) Prove that G is a gorup under multiplication.
(b) Prove that G is not a group under addition.
Proof (a).
Proof (b).
1.1.9. Let G =
a +b
2 1 [ a, b Q
.
(a) Prove that G is a gorup under addition.
(b) Prove that the nonzero elements of G are a group under multiplication.
Proof (a).
Proof (b).
1.1.10. Prove that a nite gorup is abelian if and only if its group table is a symmetric matrix.
Proof.
1.1.14. Find the orders of the following elements of the multiplicative group (Z/36Z)
= 1,
= 2,
= 6,
13
= 3,
13
= 6,
17
= 2.
1.1.25. Prove that if x
2
= 1 for all x G then G is abelian.
10 DAVID S. DUMMIT AND RICHARD M. FOOTE
Proof. Since x
2
= 1 we know that x = x
1
for all x G. Thus
xy = (xy)
1
= y
1
x
1
= yx.
r, s [ r
n
= s
2
= 1, rs = sr
1
.
1.2.1. Compute the order of each of the elements in the following groups:
(a) D
6
The order of elements from (a) are: [1[ = 1, [r[ = 3,
r
2
= 3, [s[ = 2, [sr[ = 2,
sr
2
= 2.
1.2.2. Use the generators and relations above to show that if x is any element of D
2n
which is not a power of r,
then rx = xr
1
.
Proof. If x is not a power of r, then x = sr
k
for some 0 k < n. So
rsr
k
= sr
1
r
k
= sr
k1
= sr
k
r
1
.
1.2.3. Use the generators and relations above to show that every element of D
2n
which is not a power of r has
order 2. Deduce that D
2n
is generated by the two elements s and sr, both of which have order 2.
Proof. If x D
2n
is not a power of r, then x = sr
k
for some 0 k < n. Notice that sr
0
sr
0
= 1 and suppose that
sr
i
sr
i
= 1 for 0 i < k. Then
sr
k
sr
k
= sr
k1
sr
1
r
k
= sr
k1
sr
k1
= 1,
by induction.
1.2.7. Show that
a, b [ a
2
= b
2
= (ab)
n
= 1
In Exercise 1.2.9 you can nd the order of the group of rigid motions in 1
3
(also called the group of rotations)
of the given Platonic solid by following the proof for the order of D
2n
: nd the number of positions to which an
adjacent pair of vertices can be sent. Alternatively, you can nd the number of places to which a given face may
be sent and, once a face is xed, the number of positions to which a vertex on that face may be sent.
1.2.9. Let G be the group of rigid motions in 1
3
of a tetrahedron. Show that [G[ = 12.
Proof. We use the second method described above to show that [G[ = 12. Since there are 4 faces of a tetrahedron,
each face having 3 distinct rotations, there are 4 3 = 12 positions to which a vertex on each xed face may be
sent.
1.2.17. Let X
2n
=
x, y [ x
n
= y
2
= 1, xy = yx
2
.
(a) Show that if n = 3k, then X
2k
has order 6, and it has the same generators and relations as D
6
when x is
replaced by r and y by s.
(b) Show that if (3, n) = 1, then x satises the additional relation: x = 1. In this case deduce that X
2n
has
order 2. [Use the facts that x
n
= 1 and x
3
= 1.]
11
Proof (a). Suppose we have the relations given by D
6
. Then clearly x
3k
= (x
3
)
k
= y
2
= 1. We also have
xy = yx
1
= yx
31
= yx
2
.
Now suppose we have the relations given by X
2n
. Then clearly x
3
= y
2
= 1. The relations xy = yx
2
and x
3
= 1
show us
xy = yx
2
= yx
1
.
Therefore D
6
= X
6
and [X
6
[ = 6.
Proof (b). Since (3, n) = 1, there exist s, t Z such that 3s +nt = 1. So
x = x
3s+nt
= (x
3
)
s
(x
n
)
t
= 1.
Thus X
2n
has order 2, namely X
2n
= 1, y.
1.3. Symmetric Groups.
1.3.4. Compute the order of each of the elments in the following groups:
(a) S
3
(b) S
4
For elements in S
3
we calculate that [1[ = 1, [(12)[ = 2, [(13)[ = 2, [(23)[ = 2, [(123)[ = 3, [(132)[ = 3.
For elements in S
4
we calculate that [1[ = 1, [(12)[ = 2, [(13)[ = 2, [(14)[ = 2, [(23)[ = 2, [(24)[ = 2, [(34)[ = 2,
[(123)[ = 3, [(132)[ = 3, [(124)[ = 3, [(142)[ = 3, [(134)[ = 3, [(143)[ = 3, [(234)[ = 3, [(243)[ = 3, [(1234)[ = 4,
[(1243)[ = 4, [(1324)[ = 4, [(1342)[ = 4, [(1423)[ = 4, [(1432)[ = 4, [(12)(34)[ = 2, [(13)(24)[ = 2, [(14)(23)[ = 2.
1.3.7. Write out the cycle decomposition of each element of order 2 in S
4
.
(12), (13), (14), (23), (24), (34), (12)(34), (13)(24), (14)(23).
1.3.11. Let be the mcycle (1 2 . . . m). Show that
i
is also an mcycle if and only if i is relatively prime to m.
Proof. First suppose that
i
is an mcycle so that
i
= (a
1
a
2
. . . a
m
).
Then assume that [i and [m so there are s, t Z such that
s = i, t = m.
So
i
=
s
= [(a
1
a
+1
. . .
m+1
) (a
a
2
. . . a
m
)]
s
which can only be an mcycle if = 1. Thus i and m
are relatively prime.
Now suppose that i is relatively prime to m so that there are s, t Z such that
si +tm = 1.
But then =
si+tm
=
si
tm
= (
i
)
s
. Since a power of
i
is an mcycle, then
i
must be an mcycle as well.
1.3.20. Find a set of generators and relations for S
3
.
S
3
=
x
2
, x
3
[ x
2
2
= x
3
3
= 1, x
2
x
3
= x
1
3
x
2
a b
c d
in Figure 1.1 to be p
3
+p
2
p. Thus the order of GL
2
(F
p
) is p
4
p
3
p
2
+p.
1 1
0 1
0 1
1 0
1 1
1 0
0 1
1 0
1 1
0 1
0 1
1 1
so that GL
2
(F) is nonabelian for any F. Now notice that extending the matrices with zeros shows the result for
n > 2.
1.4.10. Let G =
a b
0 c
[ a, b, c 1, a = 0, c = 0
.
(a) Compute the product of
a
1
b
1
0 c
1
and
a
2
b
2
0 c
2
a b
0 c
a
1
b
1
0 c
1
a
2
b
2
0 c
2
a
1
a
2
a
1
b
2
+b
1
c
2
0 c
1
c
2
.
Since a
1
, a
2
= 0 and c
1
, c
2
= 0, this matrix is also in G.
(b). Notice that
a b
0 c
1/a b/ac
0 1/c
1 b/c +b/c
0 1
1 0
0 1
1/a b/ac
0 1/c
a b
0 c
1 b/a b/a
0 1
1 0
0 1
.
Since
1/a b/ac
0 1/c
1 0
0 1
a b
0 c
.
det M = ac = 0.
Thus G GL
2
(1). Note matrix multiplication is a binary operation on G, the identity is in G, and G is closed
under inverse. So G is a subgroup of GL
2
(1).
Proof (d). We rst show that matrix multiplication is a binary operation on G
= G [ a = c. So
a
1
b
1
0 a
1
a
2
b
2
0 a
2
a
1
a
2
a
1
b
2
+b
1
a
2
0 a
1
a
2
.
Since a
1
a
2
= a
1
a
2
, matrix multipliaction is indeed a binary operation G
.
The inverse is given by
a b
0 a
1
=
1/a b/a
2
0 1/a
G GL
2
(1), we must have
G
as a subgroup of GL
2
(1).
1.5. Quaternion Groups.
1.5.1. Compute the order of each of the elements in Q
8
.
Let Q
8
= 1, 1, i, i, j, j, k, k as given in the textbook. Then [1[ = 1, [1[ = 2, [i[ = 4, [i[ = 4, [j[ = 4,
[j[ = 4, [k[ = 4, [k[ = 4.
1.5.2. Write out the group tables for S
3
, D
8
, and Q
8
.
1.5.3. Find a set of generators and relations for Q
8
.
Q
8
=
1, i, j, k [ (1)
2
= 1, i
2
= j
2
= k
2
= ijk = 1
r
2
= 2. But Exercise 1.5.1 shows there are not two elements with order two in Q
8
. Thus
Exercise 1.5.2 shows that D
8
and Q
8
are not isomorphic.
1.6.13. Let G and H be groups and let : G H be a homomorphism. Prove that the image of , (G), is a
subgroup of H (cf. Exercise 26 of Section 1). Prove that if is injective then G
= (G).
Proof. Let h
1
, h
2
H so that there are g
1
, g
2
G such that (g
1
) = h
1
and (g
2
) = h
2
. Thus (g
1
g
2
) =
(g
1
)(g
2
) = h
1
h
2
so h
1
h
2
H. Hence
HH
: H H H. Notice that (1) = 1 so H is nonempty. Let h H
and let (g) = h. Then (g
1
) = (g)
1
= h
1
, so H is closed under inverse. Therefore H G is a subgroup.
1.6.14. Let Gand H be groups and let : G H be a homomorphism. Dene the kernel of to be g G [ (g) = 1
H
(so the kernel is the set of elements in G which map to the identity of H, i.e., is the ber over the identity of H).
Prove that the kernel of is a subgroup (cf. Exercise 26 of Section 1) of G. Prove that is injective if and only if
the kernel of is the identity subgroup of G.
Proof. First note that 1 ker so that ker = . Now let g
1
, g
2
ker so that (g
1
) = (g
2
) = 1. Then
(g
1
g
2
) = (g
1
)(g
2
) = 1 so that g
1
g
2
ker . Suppose that g ker so that (g) = 1. Then (g
1
) = (g)
1
= 1
so g
1
ker and ker is closed under inverse. Thus ker H is a subgroup.
Suppose that is injective. Thus there is only one element that maps to 1
H
. Since (1
G
) = 1
H
, ker = 1
G
.
Now suppose that ker = 1
G
and let (g
1
) = (g
2
). Then (g
1
)(g
2
)
1
= 1. So (g
1
g
1
2
) = 1 and since
ker = 1
G
, we must have g
1
g
1
2
= 1
G
and thus g
1
= g
2
. Therefore is injective.
1.6.18. Let G be any group. Prove that the map from G to itself dened by g g
2
is a homomorphism if and
only if G is abelian.
Proof. First suppose that g g
2
is a homomorphism. Let g
1
, g
2
G so that g
1
g
2
g
1
g
2
= (g
1
g
2
)
2
= g
2
1
g
2
2
. Thus
g
2
g
1
= g
1
g
2
showing that G is abelian.
Now suppose that G is abelian. Let g
1
, g
2
G so that (g
1
g
2
)
2
= g
1
g
2
g
1
g
2
= g
2
1
g
2
2
and thus g g
2
is a
homomorphism.
1.7. Group Actions.
1.7.4. Let G be a group acting on a set A and x some a A. Show that the following sets are subgroups of G
(cf. Exercise 26 of Section 1):
(a) the kernel of the action.
(b) g G [ ga = a this subgroup is called the stabilizer of a in G.
Proof (a). Notice that 1 a = a for all a A so the kernel of the action is nonempty. Let g
1
, g
2
be two elements
in the kernel of the action. Then (g
1
g
2
)(a) = g
1
(g
2
a) = g
1
a = a so the group operation on G is a binary operation
on the kernel of the action. Now let g be in the kernel of the action. Then g
1
a = g
1
(ga) = (g
1
g)a = a so the
kernel of the action is closed under inverse. Therefore the kernel of the action is a subgroup of G.
Proof (b). Notice that 1 a = a for all a A so the stabilizer of a in G is nonempty. Let g
1
, g
2
g G [ ga = a
so that (g
1
g
2
)a = g
1
(g
2
a) = g
1
a = a and hence the group operation on G is a binary operation on g G [ ga = a.
Now let g g G [ ga = a so that g
1
a = g
1
(ga) = (g
1
g)a = a and hence g G [ ga = a is closed under
inverse. Therefore g G [ ga = a is a subgroup of G.
1.7.5. Prove that the kernel of an action of the group G on the set A is the same as the kernel of the corresponding
permutation representation G S
A
(cf. Exercise 14 in Section 6).
15
Proof. First suppose that ga = a for all a A. Then
g
(a) = a for all a A hence
g
= 1
S
A
. Now suppose that
g
(a) = a for all a A. Then ga =
g
(a) = a. Therefore these kernels are equivalent.
1.7.10. With reference to the preceding two exercises determine:
(a) for which values of k the action of S
n
on kelement subsets is faithful.
Let S
n
act on subsets of order k for some 1 k n. Note if n = 1, then the group action is trivially faithful.
Now let S
n
be a nonidentity element for n > 1. Since is not the identity, let (a
i
) = a
j
for a
i
= a
j
. If k < n,
then we can choose a
1
, . . . , a
k
such that a
i
a
1
, . . . , a
k
and a
j
/ a
1
, . . . , a
k
. Thus
a
1
, . . . , a
k
= a
1
, . . . , a
k
.
Hence the homomorphism representing the group action is injective for k < n. If k = n, then is just a bijection
of nelement sets so the homomorphism representing the group action is trivial. Therefore the values of k giving
rise to a faithful group action are all values of k < n.
1.7.13. Find the kernel of the left regular action.
If g is in the kernel of the left regular action, then ga = a for all a G. Thus g = 1 so the kernel of the left
regular action is 1.
1.7.14. Let G be a group and let A = G. Show that if G is nonabelian then the maps dened by g a = ag for all
g, a G do not satisfy the axioms of a (left) group action of G on itself.
Proof. Since G is nonabelian we can nd g
1
, g
2
G such that g
1
g
2
= g
2
g
1
. We assume the axioms of a group
action hold. But g
1
g
2
= a
1
ag
1
g
2
= a
1
((g
1
g
2
) a) = a
1
(g
1
(g
2
a)) = a
1
(g
1
(ag
2
)) = a
1
(ag
2
g
1
) = g
2
g
1
, a
contradiction. Therefore this is not a group action.
2. Subgroups
2.1. Denition and Examples.
Proposition 2.1 (The Subgroup Criterion). A subset H of a group G is a subgroup if and only if
(1) H =
(2) for all x, y H, xy
1
H
Furthermore, if H is nite, then it suces to check that H is nonempty and closed under multiplication.
2.2. Centralizers and Normalizers, Stabilizers and Kernels.
2.2.2. Prove that C
G
(Z(G)) = G and deduce that N
G
(Z(G)) = G.
Proof. Notice C
G
(Z(G)) =
g G [ gag
1
= a, a Z(G)
1, r
2
, s, sr
2
, C
D8
(sr) =
1, r
2
, sr, sr
3
,
C
D8
(sr
2
) =
1, r
2
, s, sr
2
, C
D8
(sr
3
) =
1, r
2
, sr, sr
3
.
Q
8
. C
Q8
(1) = Q
8
, C
Q8
(1) = 1, 1, C
Q8
(i) = 1, i, C
Q8
(i) = 1, i, C
Q8
(j) = 1, j, C
Q8
(j) = 1, j,
C
Q8
(k) = 1, k, C
Q8
(k) = 1, k.
Note that Lagranges theorem lets us check our answers because the centralizer is a subgroup and thus must
divide the order of the group.
2.2.5. In each of parts (a) to (c) show that for the specied group G and subgroup A of G, C
G
(A) = A and
N
G
(A) = G.
(a) G = S
3
and A = 1, (1 2 3), (1 3 2).
(b) G = D
8
and A =
1, s, r
2
, sr
2
.
(c) G = D
10
and A =
1, r, r
2
, r
3
, r
4
.
16 DAVID S. DUMMIT AND RICHARD M. FOOTE
Proof (a). Note that C
G
(A) =
aA
C
G
(a) so by Exercise 2.2.4, C
S3
(A) = S
3
1, (1 2 3), (1 3 2)1, (1 2 3), (1 3 2) =
A.
Proof (b). Similarly to part (a) by Exercise 2.2.4, C
D8
(A) = D
8
1, r
2
, s, sr
2
D
8
1, r
2
, s, sr
2
= A.
Proof (c). Note that any power of r commutes with any power of r so A C
D10
(A). Since C
D10
(r) = 'r`, we also
must have C
D10
(A) A.
2.2.9. For any subgroup H of G and any nonempty subset A of G dene N
H
(A) to be the set
h H [ hAh
1
= A
.
Show that N
H
(A) = N
G
(A) H and deduce that N
H
(A) is a subgroup of H (note that A need not be a subset of
H).
Proof. Let x N
H
(A) so x H G and xAx
1
= A. Thus x N
G
(A) H. Now let x N
G
(A) H so that
x HandxAx
1
= A. Thus x G
H
(A) and hence N
H
(A) = N
G
(A) H.
Note that N
G
(A) and H are both subgroups of G and the intersection of two subgroups is again a subgroup.
Therefore N
H
(A) G.
2.2.10. Let H be a subgroup of order 2 in G. Show that N
G
(H) = C
G
(H). Deduce that if N
G
(H) = G then
H Z(G).
Proof. Let H = 1, h, x N
G
(H) so that x G and xHx
1
= H. Since x1x
1
= 1, we must have xhx
1
= h so
that x1, h x
1
= 1, h. Thus x C
G
(H). Now if x C
G
(H), then x1x
1
= 1 and xhx
1
= h so xHx
1
= H.
Thus x N
G
(H) and hence N
G
(H) = C
G
(H).
Note if N
G
(H) = G, then G = C
G
(H). Thus any element of G commutes with any element in H so H Z(G).
Since H and Z(G) are subgroups we get H Z(G).
2.3. Cyclic Groups and Cyclic Subgroups.
Proposition 2.2. If H = 'x`, then [H[ = [x[ (where if one side of this equality is innite, so is the other). More
specically
(1) if [H[ = n < , then x
n
= 1 and 1, x, x
2
, . . . , x
n1
are all the distinct elements of H
(2) if [H[ = , then x
n
= 1 for all n = 0 and x
a
= x
b
for all a = b in Z.
Proposition 2.3. Let G be an arbitrary group, x G and let m, n Z. If x
n
= 1 and x
m
= 1, then x
d
= 1, where
d = (m, n). In particular, if x
m
= 1 for some m Z, then [x[ divides m.
Theorem 2.4. Any two cyclic groups of the same order are isomorphic. More specically,
(1) if n Z
+
and 'x` and 'y` are both cyclic groups of order n, then the map
: 'x` 'y`
x
k
y
k
is well dened and is an isomorphism
(2) if 'x` is an innite cyclic group, the map
: Z 'x`
k x
k
is well dened and is an isomorphism.
Proposition 2.5. Let G be a group, let x G and let z Z` 0.
(1) If [x[ = , then [x
a
[ = .
(2) If [x[ = n < , then [x
a
[ =
n
(n,a)
.
(3) In particular, if [x[ = n < and a is a positive integer dividing n, then [x
a
[ =
n
a
.
Proposition 2.6. Let H = 'x`.
(1) Assume [x[ = . Then H = 'x
a
` if and only if a = 1.
(2) Assume [x[ = n < . Then H = 'x
a
` if and only if (a, n) = 1. In particular, then number of generators of
H is (n).
Theorem 2.7. Let H = 'x` be a cyclic group.
(1) Every subgroup of H is cyclic. More precisely, if K H, then either K = 1 or K =
x
d
, where d is the
smallest positive integer such that x
d
K.
17
(2) If [H[ = , then for any distinct nonnegative integers a and b, 'x
a
` =
x
b
x
m
, where [m[ denotes the absolute value of m, so that the nontrivial subgroups of H
correspond bijectively with the integers 1, 2, 3, . . ..
(3) If [H[ = n < , then for each positive integer a dividing n there is a unique subroups of H of order a.
This subgroup is the cyclic group
x
d
, where d =
n
a
. Furthermore, for every integer m, 'x
m
` =
x
(n,m)
,
so that the subgroups of H correspond bijectively with the positive divisors of n.
2.3.1. Find all subgroups of Z
45
= 'x`, giving a generator for each. Describe the containments between these
subgroups.
The subgroups of Z
45
are given by
x
0
, 'x` ,
x
3
x
5
x
9
x
15
x
3
x
5
x
9
x
15
x
0
.
2.3.5. Find the number of generators for Z/49000 Z.
Since Z/49000 Z is cyclic of order 49000, we count the number of relatively prime integers to 49000. This is given
by Eulers function, so
(49000) = (2
3
5
3
7
2
) = (2
3
)(5
3
)(7
2
) = 2
2
(1) 5
2
(4) 7(6) = 16800.
2.3.8. Let Z
48
= 'x`. For which integers a does the map
a
dened by
a
:
1 x
a
extend to an isomorphism from
Z/48Z onto Z
48
.
All integers a such that (a, 48) = 1. So a 1, 5, 7, 11, 13, 17, 19, 23, 25, 29, 31, 35, 37, 41, 43, 47.
2.3.12. Prove that the following groups are not cyclic:
(b) Z
2
Z
Proof. Suppose Z
2
Z is cyclic and let (x, y) be a generator. Note that x, y are not identity elements in their
respective groups. Then there exists k Z such that
(1, y) = (x, y)
k
= (x
k
, ky).
Note k = 1 because x = 1. But then y = ky shows that y = 0. This is a contradiction so this group is not cyclic.
2.3.13. Prove that the following pairs of groups are not isomorphic:
(b) QZ
2
and Q.
Proof. Suppose that : (Q Z
2
) Q is an isomorphism. Let (0, x) Q Z
2
be a nonidentity element. So
((0, x)) = y = 0 but
0 = ((0, 1)) = ((0, x)
2
) = 2y.
Since y = 0 this is a contradiction, thus no such isomorphism exists.
2.3.18. Show that if H is any group and h is an element of H with h
n
= 1, then there is a unique homomorphism
from Z
n
= 'x` to H such that x h.
Proof. Since h
n
= 1, there is an m N such that ['h`[ = m so m[n. Dene a map (x
k
) = h
k
. If x
k
= x
l
so
that k l (mod n), then k l = n for some Z. But then h
k
= h
n+l
= h
l
so is welldened. Now take
x
k
, x
l
'x` so that
(x
k
)(x
l
) = h
k
h
l
= h
k+l
= (x
k+l
)
shows that is such a homomorphism. Now suppose that and are two homomorphisms such that (x) = h =
(x). Take some x
k
'x` so that
(x
k
) = (x)
k
= h
k
= (x)
k
= (x
k
).
Therefore = shows there is a unique map with these properties.
2.4. Subgroups Generated by Subsets of a Group.
Proposition 2.8. If / is any nonempty collection of subgroups of G, then the intersection of all members of / is
also a subgroup of G.
Proposition 2.9. A = 'A`.
2.4.8. Prove that S
4
= '(1 2 3 4), (1 2 4 3)`.
18 DAVID S. DUMMIT AND RICHARD M. FOOTE
Proof. First note that [S
4
[ = 4! = 24. Since '(1 2 3 4), (1 2 4 3)` S
4
, Lagranges theorem states we need
to nd 13 elements generated by this set so that '(1 2 3 4), (1 2 4 3)` = S
4
. So we have (1 2 3 4), (1 2 3 4)
2
=
(1 3)(2 4), (1 2 3 4)
3
= (1 4 3 2), (1 2 4 3), (1 2 4 3)
2
= (1 4)(2 3), (1 2 4 3)
3
= (1 3 4 2), (1 2 3 4)(1 2 4 3) = (1 3 2),
(1 2 3 4)(1 4)(2 3) = (2 4), (1 2 3 4)(1 3 4 2) = (1 4 3), (1 3)(2 4)(1 2 4 3) = (1 4), (1 4 3 2)(1 2 4 3) = (2 3 4), (1 4)(2 4) =
(1 4 2), (1 3)(2 4)(2 4) = (1 3).
2.4.9. Prove that SL
2
(F
3
) is the subgroup of GL
2
(F
3
) generated by
1 1
0 1
and
1 0
1 1
.
Proof. First note that [SL
2
(F
3
)[ = 24 so we use a method similar to exercise 2.4.8 to show this subgroup is
generated by the two matrices. So we have
1 1
0 1
1 1
0 1
2
=
1 2
0 1
1 1
0 1
3
=
1 0
0 1
1 0
1 1
1 0
1 1
2
=
1 0
2 1
1 1
0 1
1 0
1 1
2 1
1 1
1 2
0 1
1 0
1 1
0 2
1 1
1 1
0 1
1 0
2 1
1 1
2 1
1 0
1 1
1 1
0 1
1 1
1 2
1 0
1 1
1 2
0 1
1 2
1 0
1 0
2 1
1 1
0 1
1 1
2 0
1 0
2 1
2 1
1 1
2 1
2 0
1 1
1 2
1 2
1 0
2 2
0 2
.
2.4.10. Prove that the subgroup of SL
2
(F
3
) generated by
0 1
1 0
and
1 1
1 1
i, j [ i
4
= 1 = j
4
, ij = ji
.
Notice that
0 1
1 0
4
= I =
1 1
1 1
4
, and
0 1
1 0
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
0 1
1 0
.
Thus the relations of the subgroup of SL
2
(F
3
) generated by the matrices are contained in the presentation of Q
8
.
So the order of the matrix subgroup is less than or equal to 8. Since we have found 5 matrices in this subgroup and
[SL
2
(F
3
)[ = 24, Lagranges theorem shows that the order of this subgroup is indeed 8. Therefore these groups are
isomorphic.
2.4.11. Prove that SL
2
(F
3
) and S
4
are two nonisomorphic groups of order 24.
Proof. First notice that there is more than one element of order two in S
4
, namely (1 2) and (1 3). Now let
a b
c d
SL
2
(F
3
) be an element of order 2. So
a b
c d
2
=
a
2
+bc ab +bd
ac +cd bc +d
2
1 0
0 1
1 0
0 1
A =
1 1 1
0 1 0
0 0 1
, B =
1 1 0
0 1 1
0 0 1
19
where AB = B
1
A and B
4
= I = A
2
. Calculating we nd
1 1 1
0 1 0
0 0 1
2
=
1 0 0
0 1 0
0 0 1
1 1 0
0 1 1
0 0 1
4
=
1 1 0
0 1 1
0 0 1
1 0 1
0 1 0
0 0 1
1 1 0
0 1 1
0 0 1
1 1 1
0 1 1
0 0 1
1 0 0
0 1 0
0 0 1
.
Thus we found ve matrices generated by A and B to show that it generates the whole group by Lagranges theorem.
Therefore these two groups are isomorphic.
2.5. Denitions and Examples. Let G and H be groups.
2.5.1. Let : G H be a homomorphism and let E be a subgroup of H. Prove that
1
(E) G (ie., the
preimage or pullback of a subgroup under a homormorphism is a subgroup). If E H prove that
1
(E) G.
Deduce that ker G.
Proof. Let g
1
, g
2
1
(E) so (g
1
) = h
1
and (g
2
) = h
2
where h
1
, h
2
E. But (1
G
) = (1
H
) and (g
1
g
1
2
) =
h
1
h
1
2
E so
1
(E) G.
If E H then hEh
1
E for all h H. So take
g
1
(E) g
1
(g) E (g)
1
E. Thus g
1
(E) g
1
1
(E) shows
1
(E) G. Since 1
H
is normal in H then ker G.
2.5.5. Use the preceding exercise to prove that the order of the element gN in G/N is n, where n is the smallest
positive integer such that g
n
N (and gN has innite order if no such positive integer exists). Give an example to
show that the order of gN in G/N may be strictly smaller than the order of g in G.
Proof. First note if 1 < m < n then (gN)
m
= g
m
N _ N for g / N so the order of gN n. But (gN)
n
= g
n
N = N
so the order of gN is n.
An example with the desired property would be D
8
/
r
2
. Note that
r
2
= 2 in D
8
but
r
2
r
2
= 1 in
D
8
/
r
2
.
2.5.6. Dene : 1
1 by letting (x) be x divided by the absolute value of x. Describe the bers of and
prove that is a homomorphism.
Proof. The bers of are given by
1
(1) = 1
[ > 0 and
1
(1) = 1
[ < 0. To show is a
homomorphism we calculate
() =
1 > 0,
1 < 0,
()() =
1 > 0,
1 < 0.
2.5.9. Dene : C
by (a + bi) = a
2
+ b
2
. Prove that is a homomorphism and nd the image of .
Describe the kernel and the bers of geometrically (as subsets of the plane).
Proof. We calculate,
((a +bi)(c +di)) = (ac +adi +bci bc)
= (ac bd)
2
+ (ad +bd)
2
= a
2
c
2
2abcd +b
2
d
2
+a
2
d
2
+ 2abcd +b
2
d
2
(a +bi)(c +di) = (a
2
+b
2
)(c
2
+d
2
)
= a
2
c
2
+a
2
d
2
+b
2
c
2
+b
2
d
2
20 DAVID S. DUMMIT AND RICHARD M. FOOTE
so is a homomorphism.
2.5.11. Let F be a eld and let G =
a b
0 c
[ a, b, c F, ac = 0
GL
2
(F).
(a) Prove that the map :
a b
0 c
(recall that F
is the
multiplicative group of nonzero elements in F). Describe the bers and kernel of .
(b) Prove that the map :
a b
0 c
. Describe
the bers and kernel of .
(c) Let H =
1 b
0 1
[ b F
a
1
b
1
0 c
1
a
2
b
2
0 c
2
a
1
a
2
a
1
b
2
+b
1
c
2
0 c
1
c
2
.
The bers of are given by
1
(a) =
a b
0 c
[ b F, c = 0
a b
0 c
[ b F
1 b
1
0 1
1 b
2
0 1
1 b
1
+b
2
0 1
1
(i) = (1/4)(1
+
/Z), (e
4i/3
) = (2/3)(1
+
/Z).
2.6. The Lattice of Subgroups of a Group.
3. Quotient Groups and Homomorphisms
3.1. Denitions and Examples.
Proposition 3.1. Let G and H be groups and let : G H be a homomorphism.
(1) (1
G
) = 1
H
(2) (g
1
) = (g)
1
for all g G.
(3) (g
n
) = (g)
n
for all n Z.
(4) ker is a subgroup of G.
(5) im() is a subgroup of H.
Proposition 3.2. Let : G H be a homomorphism of groups with kernel K. Let X G/K be the ber above
a, i.e., X =
1
(a). Then
(1) For any u X, X = uk [ k K
(2) For any u X, X = ku [ k K.
Theorem 3.3. Let G be a group and let K be the kernel of some homomorphism from G to another group. Then
the set whose eelements are the left cosets of K in G with operation dened by
uK vK = (uv)K
forms a group, G/K. In particular, this operation is well dened in the sense that if u
1
is any element in uK and
v
1
is any element in vK, then u
1
v
1
uvK, e.e., u
1
v
1
K = uvK so that the multiplication does not depend on the
choice of representatives for the cosets. The same statement is true with right coset in place of left coset.
21
Proposition 3.4. Let N be any subgroup of the group G. The set of left cosets of N in G form a partition of G.
Furthermore, for all u, v G, uN = vN if and only if v
1
u N and in particular, uN = vN if and only if u and
v are representatives of the same coset.
Proposition 3.5. Let G be a group and let N be a subgroup of G.
(1) The operation on the set of left cosets of N in G described by
uN vN = (uv)N
is well dened if and only if gng
1
N for all g G and all n N.
(2) If the above operation is well dened, then it makes the set of left cosets of N in G into a group. In particular
the identity of this group is the coset 1N and the inverse of gN is the coset g
1
N i.e., (gN)
1
= g
1
N.
Theorem 3.6. Let N be a subgroup of the group G. The following are equivalent:
(1) N G
(2) N
G
(N) = G
(3) gN = Ng for all g G
(4) the operation on left cosets of N in G described in proposition 3.5 makes the set of left cosets into a group
(5) gNg
1
N for all g G.
Proposition 3.7. A subgroup N of the group G is normal if and only if it is the kernel of some homomorphism.
3.1.36. Prove that if G/Z(G) is cyclic then G is abelian.
Proof. If G/Z(G) is cyclic with generator xZ(G), then if gZ(G) G/Z(G) we have gZ(G) = (xZ(G))
n
= x
n
Z(G)
for some n N. Thus there is some z Z(G) such that g = x
n
z. Now take g
1
, g
2
G such that g
1
= x
m
z
1
and
g
2
= x
n
z
2
. Then
g
1
g
2
= x
m
z
1
x
n
z
2
= x
m
x
n
z
2
z
1
= x
n
x
m
z
2
z
1
= x
n
z
2
x
m
z
1
= g
2
g
1
.
= Z
p
.
Theorem 3.11 (Cauchys Theorem). If G is a nite group and p is a prime dividing [G[, then G has an element
of order p.
Theorem 3.12 (Sylow). If G is a nite group of order p
.
Proposition 3.13. If H and K are nite subgroups of a group then
[HK[ =
[H[ [K[
[H K[
.
Proposition 3.14. If H and K are subgroups of a group, HK is a subgroup if and only if HK = KH.
Corollary 3.15. If H and K are subgroups of G and H and H N
G
(K), then HK is a subgroup of G. In
particular, if K G then HK G for any H G.
3.2.8. Prove that if H and K are nite subgroups of G whose orders are relatively prime then H K = 1.
Proof. Let [H[ = m and [K[ = n with (m, n) = 1. Now take x H K so [x[ [m and [x[ [n so x = 1.
22 DAVID S. DUMMIT AND RICHARD M. FOOTE
3.2.9. This exercise outlines a proof of Cauchys Theorem due to James McKay. Let G be a nite group and let p
be a prime dividing [G[. Let o denote the set of ptuples of elements of G the product of whose coordinates is 1:
o = (x
1
, x
2
, . . . , x
p
) [ x
i
G and x
1
x
2
x
p
= 1 .
Dene the relation on o by letting if is a cyclic permutation of .
(a) Show that o has [G[
p1
elements, hence has order divisible by p.
(b) Show that a cyclic permutation of an element of o is again an element of S.
(c) Prove that is an equivalence relation on o.
(d) Prove that an equivalence class contains a single element if and only if it is of the form (x, x, . . . , x) with
x
p
= 1.
(e) Prove that every equivalence class has order 1 or p (this uses the fact that p is a prime). Deduce that
[G[
p1
= k +pd, where k is the number of classes of size 1 and d is the number of classes of size p.
(f) Since (1, 1, . . . , 1) is an equivalence class of size 1, conclude from (e) that there must be a nonidentity
element x in G with x
p
= 1, i.e., G contains an element of order p. [Show p[k and so k > 1.]
Proof (a). The rst p 1 coordinates have [G[ choices and the pth coordinate must be the inverse of the product
of the rst p 1 coordinates. Thus [o[ = [G[
p1
.
Proof (b). Let (x
1
, x
2
, . . . , x
p
) o and notice x
p
x
1
x
2
x
p1
= 1 just by multiplication on the right by (x
p
)
1
,
then the left by x
p
. Thus (x
p
, x
1
, . . . , x
p1
) o and any cyclic permutation can be realized by repeating this
process.
Proof (c). Reexive is obvious. If is a cyclic permutation of , then is a cyclic permutation of so is
symmetric. If is a cyclic permutation of and is a cyclic permutation of , then is a cyclic permutation of
so is transitive. Thus is an equivalence relation.
Proof (d). If some x is changed to a y = x, then a cyclic permutation would yield a dierent element of the
equivalence class. If all the x are the same, then any cyclic permutation is indeed the same element of o.
Proof (e). It is clear that every equivalence class has order less than or equal to p. Suppose the equivalence class
is not of the form from part (d) and let S
p
be given by (i) i + j mod p for 0 j < p and suppose that
(x
1
, . . . , x
p
) (x
(1)
, . . . , x
(p)
). Then x
i
= x
(i)
= x
2
(i)
= = x
p1
(i)
for all i. Since not every coordinate is
allowed to be equal, j = 0 because [[ = p (for p prime, = 1).
Thus the number of equivalence classes is the sum of size 1 and size p classes. Since equivalence classes are
disjoint, [G[
p1
= k +pd.
Proof (f ). Since p divides [G[, then p divides [G[
p1
and specically p divides k. Thus k > 1 and hence there is an
element x in G with order p.
3.2.10. Suppose H and K are subgroups of nite index in the (possibly innity) group G with [G : H[ = m and
[G : K[ = n. Prove that lcm(m, n) [G : H K[ mn. Deduce that if m and n are relatively prime then
[G : H K[ = [G : H[ [G : K[.
Proof by Dr. Schulze. By Problem 11, you have [G : H K] = [G : K][K : H K]. So n divides [G : H K], and
also m divides by the same argument. Therefore lcm(m, n) [G : H K].
Now note that the rst isomorphism theorem works also for nonnormal subgroups, but you get a statement on
maps of sets (not groups). Apply this to the map G GG G/H G/K sending g (g, g) (gH, gK).
Its kernel is H K, so the (set version of the) rst isomorphism theorem gives an injective map G/(H K)
G/H G/K. Thus, [G : H K] = [G/(H K)[ [G/H G/K[ = [G/H[[G/K[ = n m.
3.2.22. Use Lagranges Theorem in the multiplicative group (Z/nZ)
[
= 1
(Z\nZ)

1
mod n.
23
3.3. The Isomorphism Theorems.
Theorem 3.16 (The First Isomorphism Theorem). If : G H is a homomorphism of groups, then ker G
and G/ ker
= (G).
Corollary 3.17. Let : G H be a homomorphism of groups.
(1) is injective if and only if ker = 1.
(2) [G: ker [ = [(G)[.
Theorem 3.18 (The Second Isomorphism Theorem). Let G be a group, let A and B be subgroups of G and assume
A N
G
(B). Then AB is a subgroup of G, B AB, A B A and AB/B
= A/A B.
Theorem 3.19 (The Third Isomorphism Theorem). Let G be a group and let H and K be normal subgroups of G
with H K. Then K/H G/H and
(G/H)/(K/H)
= G/K.
If we denote the quotient by H with a bar, this can be written
G/K
= G/K.
Theorem 3.20 (The Fourth Isomorphism Theorem). Let G be a group and let N be a normal subgroup of G.
Then there is a bijection from the set of subgroups A of G which contain N onto the set of subgroups A = A/N
of G/N. In particular, every subgroup of G is of the form A/N for some subgroup A of G containing N (namely,
its preimage in G under the natural projection homomorphism from G to G/N). This bijection has the following
properties: for all A, B G with N A and N B,
(1) A B if and only if A B,
(2) if A B, then [B: A[ =
B: A
,
(3) 'A, B` =
A, B
,
(4) A B = A B, and
(5) A G if and only if A G.
3.3.3. Prove that if H is a normal subgroup of G of prime index p then for all K G either
(i) K H or
(ii) G = HK and [K: K H[ = p.
Proof. Suppose that K < K so since K N
G
(H), HK = KH G. Since [K[ = 1, [HK[ > H so HK = G because
[G : H[ = p and Lagranges theorem. The fact that [K : H H[ = p is immediate by the diamond isomorphism
theorem.
3.3.4. Let C be a normal subgroup of the group A and let D be a normal subgroup of the group B. Prove that
(C D) (AB) and (AB)/(C D)
= (A/C) (B/D).
Proof. Let
1
: A A/C and
2
: B B/D which exist by the normality of C and D. Now let : (A B)
(A/C B/D) be dened by
1
and
2
. Thus ker = C D and thus (C D) (A B). Since is surjective,
the rst isomorphism theorem shows that (AB)/(C D)
= (A/C) (B/D).
3.3.7. Let M and N be normal subgroups of G such that G = MN. Prove that G/(M N)
= (G/M) (G/N).
[Draw the lattice.]
Proof. Dene a homomorphism : G (G/M) (G/N) by (g) = (g, g). Take some (m, n) (G/M) (G/N).
Then nm
n
1
= m so take nm
= (G/M) (G/N).
3.3.9. Let p be a prime and let G be a group of order p
a
m, where p does not divide m. Assume P is a subgroup of
G of order p
a
and N is a normal subgroup of G of order p
b
n, where p does not divide n. Prove that [P N[ = p
b
and [PN/N[ = p
ab
. (The subgroup P of G is called a Sylow psubgroup of G. This exercise shows that the
intersection of any Sylow psubgroup of G with a normal subgroup N is a Sylow psubgroup of N.)
Proof. First note that [PN[ = ([P[ [N[)/ [P N[ = (p
a+b
n)/ [P N[. But this divides p
a
m so [P N[ = p
b
.
But then p
b
divides p
b
n so divides n shows does not divide p. But now p
b
divides p
a
so = 1 and hence
[P N[ = p
b
. The diamond isomorphism theorem shows that [PN/N[ = p
ab
.
24 DAVID S. DUMMIT AND RICHARD M. FOOTE
3.4. Composition Series and the Holder Program.
Proposition 3.21. If G is a nite abelian group and P is a prime dividing [G[, then G contains an element of
order p.
Theorem 3.22 (JordanHolder). Let G be a nite group with G = 1. Then
(1) G has a composition series and
(2) The composition factors in a composition sereis are unique, namely, if 1 = N
0
N
1
N
r
= G and
1 = M
0
M
1
M
s
= G are two composition series for G, then r = s and there is some permutation,
, of 1, 2, . . . , r such taht
M
(i)
/M
(i)1
= N
i
/N
i1
, 1 i r.
3.5. Transpositions and the Alternating Group.
Proposition 3.23. The map : S
n
1 is a homomorphism (where 1 is a multiplicative version of the
cyclic group of order 2).
Proposition 3.24. Transpositions are all odd permutations and is a surjective homomorphism.
Proposition 3.25. The permutation is odd if and only if the number of cycles of even length in its cycle
decomposition is odd.
3.5.3. Prove that S
n
= '(i i + 1) [ 1 i n 1`.
Proof. Let S
n
and decompose into a product of transpositions. But if (i j) is any transposition, then
(i j) = (j 1 j) (i i + 1) (j 1 j).
Therefore this set generates S
n
.
3.5.5. Show that if p is prime, S
p
= ', ` where is any transposition and is any pcycle.
Proof. Without loss of generality, suppose = (1 2) and = (1 p). If 1 i n 1, then
(i i + 1) =
i1
i+1
.
Exercise 3.5.3 shows this generates S
p
.
3.5.6. Show that '(1 3), (1 2 3 4)` is a proper subgroup of S
4
. What is the isomorphism type of this subgroup?
Proof. Notice that (1 3)(1 2 3 4)
2
(1 3) = (1 2 3 4)
2
shows that (1 2 3 4)
2
Z('(1 3), (1 2 3 4)`). But
(1 2 3 4)
2
(1 4) = (1 2 4 3) = (1 3 4 2) = (1 4)(1 2 3 4)
2
.
Therefore '(1 3), (1 2 3 4)` < S
4
.
The isomorphism type of this subgroup is D
8
because we can construct 8 elements and we have
(1 3)
2
= (1 2 3 4)
4
= (1 3)(1 2 3 4)(1 3)(1 2 3 4) = 1.
3.5.7. Prove that the group of rigid motions of a tetrahedron is isomorphic to A
4
.
Proof. First notice that both groups are order 12. Now we represent the group of rigid motions of a tetrahedron as
a subgroup of S
n
. The elements in this subgroup are
1, (1 2 3), (1 3 2), (1 3 4), (1 4 3), (1 2 4), (1 4 2), (2 3 4), (2 4 3), (1 3)(2 4), (1 2)(3 4), (1 4)(2 3).
Since all 12 of these elements are even, this is A
4
.
3.5.10. Find a composition series for A
4
. Deduce that A
4
is solvable.
Solution. Let N
1
= 1, (1 3)(2 4) and N
2
= 1, (1 3)(2 4), (1 2)(3 4), (1 4)(2 3). Simple calculations show that
1 = N
0
N
1
N
2
N
3
= A
4
, while [N
i+1
/N
i
[ = p shows that each composition factor is simple. Since the order
of each composition factor is prime, they are also abelian and hence A
4
is solvable.
3.5.17. If x and y are 3cycles in S
n
, prove that 'x, y` is isomorphic to Z
3
, A
4
, A
5
, or Z
3
Z
3
.
25
Proof. If x = y, then 'x, y` = 'x`
= Z
3
by order argument. Now let x = (i j k) and y = (j k l) and notice these
are both even permutations hence generate even permutations. Since 1, (i j k), (i k j), (j k l), (j l k), (i k l) and
(i j)(k l) are all generated by x and y, then 'x, y`
= A
4
by Lagranges theorem. Now let x = (i j k), y = (k l m)
and we will show that 'x, y` A
5
, so they are equal because A
5
is simple. If we can show conjugating (i j k) by any
transposition remains in 'x, y`, then we are done because any permutation can be represented by a transposition, and
the argument for (k l m) will be similar. But (i j k)
(i j)
= (i j k)
(j k)
= (i j k)
(i k)
= (i j k)
1
, (i j k)
(l m)
= (i j k),
(i j k)
(i l)
= (j k l) = [(i j k)
(k l m)
]
(i j k)
, (i j k)
(j l)
= [(j i k)
(k l m)
]
(j i k)
, (i j k)
(i m)
= [(i j k)
(k m l)
]
(i j k)
,
(i j k)
(j m)
= [(j i k)
(k m l)
]
(j i k)
, (i j k)
(k l)
= (i j k)
(k l m)
, and (i j k)
(k m)
= (i j k)
(k m l)
. If x and y are
disjoint 3cycles, then it is clear that 'x, y`
= Z
3
Z
3
.
4. Group Actions
4.1. Group Actions and Permutation Representations.
4.2. Groups Acting on Themselves by Left Multiplication Cayleys Theorem.
4.2.2. List the elements of S
3
as 1, (1 2), (2 3), (1 3), (1 2 3), (1 3 2) and label these with the integers 1,2,3,4,5,6
respectively. Exhibit the image of each element of S
3
under the left regular representation of S
3
into S
6
.
Solution.
4.2.4. Use the left regular representation of Q
8
to produce two elements of S
8
which generate a subgroup of S
8
isomorphic to the quaternion group Q
8
.
Solution.
4.2.8. Prove that if H has nite index n then there is a normal subgroup K of G with K H and [G : K[ n!.
Proof.
4.2.10. Prove that every nonabelian group of order 6 has a nonnormal subgroup of order 2. Use this to classify
groups of order 6. [Produce an injective homomorphism into S
3
.]
Proof.
4.3. Groups Acting on Themselves by Conjugation The Class Equation.
4.3.3. Find all conjugacy classes and their sizes in the following groups:
(c) A
4
.
Solution (c).
4.3.6. Assume G is a nonabelian group of order 15. Prove that Z(G) = 1. Use the fact that 'g` C
G
(g) for all
g G to show that there is at most one possible class equation for G. [Use exercise 36, section 3.1.]
Proof.
4.3.13. Find all nite groups which have exactly two conjugacy classes.
Solution.
4.3.27. Let g
1
, g
2
, . . . , g
r
be representatives of the conjugacy classes of the nite group G and assume these elements
pairwise commute. Prove that G is abelian.
Proof.
4.3.29. Let p be a prime and let G be a group of order p
, for every
with 0 . [Use theorem 8 and induction on .]
Proof.
26 DAVID S. DUMMIT AND RICHARD M. FOOTE
4.4. Automorphisms. Let G be a group.
4.4.1. If Aut(G) and
g
is conjugation by g prove
g
1
=
(g)
. Deduce that Inn(G) Aut(G).
Proof. Let x G, so
1
(x) = (g
1
(x)g
1
)
= (g)x(g)
1
=
(g)
(x)
shows the desired equation. Since conjugating an element of Inn(G) with an element of Aut(G) is contained in
Inn(G), normality must hold.
4.4.2. Prove that if G is an abelian group of order pq, where p and q are distinct primes, then G is cyclic.
Proof. We use Cauchys theorem to pick x, y G with [x[ = p and [y[ = q. But if 1 = (xy)
k
= x
k
y
k
for k 1, then
k [ p and k [ q. Since x = y
1
, then k pq. Thus 'xy` = G and therefore G is cyclic.
4.4.3. Prove that under any automorphism of D
8
, r has at most 2 possible images and s has at most 4 possible
images (r and s are the usual generators). Deduce that [Aut(D
8
)[ 8.
Proof. Since there is only one subgroup of order 4, 'r` is characteristic so if Aut(D
8
), then ('r`) = 'r`. Since
[r[ = 4, then (r)
r, r
3
s, sr, sr
2
, sr
3
.
If x D
8
, then x = sr
k
for some 0 k 3. So (x) = (sr
k
) = (s)(r)
k
and thus the automorphism is
uniquely determined by the value on s and r. So [Aut(D
8
)[ 2 4 = 8.
4.4.11. If p is a prime and P is a subgroup of S
p
of order p, prove N
Sp
(P)/C
Sp
(P)
= Aut(P).
Proof. We know from a previous exercise that
N
Sp
(P)
=
H Aut(P). Since C
Sp
(P) = P and [Aut(P)[ = p 1, we have the desired isomorphism.
4.4.18. This exercise shows that for n = 6 every automorhpism of S
n
is inner. Fix an integer n 2 with n = 6.
(a) Prove that the automorphism group of a group G permutes the conjugacy classes of G, i.e., for each
Aut(G) and each conjugacy class / of G the set (/) is also a conjugacy class of G.
(b) Let / be the conjugacy class of transpositions in S
n
and let /
[H[
[G[, so P Syl
p
(H).
Example. Let 1 = H _ G and p
p
k
and let [x[ = p
l
. Then
x
l
= p.
4.5.9. Exhibit all Sylow 3subgroups of SL
2
(F
3
).
27
Solution. Note [SL
2
(F
3
)[ = 24 so if P Syl
3
(SL
2
(F
3
)) then [P[ = 3. Since n
3
1 mod 3, n
3
1, 4. But
1 1
0 1
1 0
1 1
0 1
2 2
0 2
1 2
,
all generate dierent subgroups of order 3.
4.5.13. Prove that a group, G, of order 56 has a normal Sylow psubgroup for some prime p dividing its order.
Proof. Note 56 = 2
3
7. Suppose n
2
, n
7
> 1. Then (
Syl
2
(G))`1 14 and (
Syl
7
(G))`1 48. These are clearly
all distinct elements, so n
2
or n
7
is equal to 1.
4.5.18. Prove that a group of order 200 has a normal Sylow 5subgroup.
Proof. Note 200 = 2
3
5
2
. But n
5
1, 6 and since 6 [ 8 we must have n
5
= 1.
4.5.19. Prove that if [G[ = 6545 then G is not simple.
Proof. Note 6545 = 5 7 11 17. But since n
11
1 mod 11 and n
11
[ 595, we must have n
11
= 1.
4.5.30. How many elements of order 7 must there be in a simple group of order 168?
Solution. Note 168 = 2
3
3 7. Since the group is simple, n
7
> 1. Since n
7
1 mod 7 and n
7
[ 24, we must have
n
7
= 8. Therefore there are 6 8 = 48 elements of order 7.
4.5.32. Let P be a Sylow psubgroup of H and let H be a subgroup of K. If P H and H K, prove that P is
normal in K. Deduce that if P Syl
p
(G) and H = N
G
(P), then N
G
(H) = H.
Proof. If P H, then P is characteristic in H. Since P is characteristic in H and H K, then P K.
Now if P Syl
p
(G) and H = N
G
(P), then P Syl
p
(H) and P is characteristic in H. Thus if H
g
= H, then
P
g
= P so g N
G
(P) = H. Therefore N
G
(H) = H.
4.5.40. Prove that the number of Sylow psubgroups of G = GL
2
(F
p
) is p + 1.
Proof. Note [GL
2
(F
p
)[ = (p 1)
2
p(p + 1) and n
p
1 mod p. Since
P
1
=
1
0 1
, P
2
=
1 0
1
Syl
p
(G),
for F
p
` 0, 1, we must have n
p
p + 1.
Now we show that the upper triangular matrices are a subgroup of N
G
(P
1
) with order (p 1)
2
p. Simple
calculations show that for a, b, c F
p
, we have
a b
0 c
1
=
a
1
a
1
bc
1
0 c
1
.
Now if F
p
` 0, 1 we get
a b
0 c
1
0 1
a b
0 c
1 a
1
b a
1
bc
1
b
0 1
.
Since the upper diagonal matrices have nonzero diagonal elements and any upper right entry, the order must be
(p 1)
2
p. Thus (p 1)
2
p [ N
G
(P
1
).
Sylows theorem states that n
p
= [G: N
G
(P
1
)], so n
p
[ p + 1. Therefore n
p
= p + 1.
4.6. The Simplicity of A
n
.
28 DAVID S. DUMMIT AND RICHARD M. FOOTE
5. Direct and Semidirect Products and Abelian Groups
5.1. Direct Products.
5.1.10. Let p be a prime. Let A and B be two cyclic groups of order p with generators x and y, respectively. Set
E = AB so that E is the elementary abelian groups of order p
2
: E
p
2. Prove that the distinct subgroups of E of
order p are
'x` , 'xy` ,
xy
2
, . . . ,
xy
p1
, 'y`
(note that there are p + 1 of them).
Proof. Example 3 in this section shows that there are p + 1 subgroups of order p. Note that these p + 1 subgroups
are all distinct and nontrivial, so if the pth power of each generator is the identity element, then we are done. But
x
p
= 1, (xy)
p
= x
p
y
p
= 1, (xy
k
)
p
= x
p
(y
k
)
p
= 1, y
p
= 1.
5.1.11. Let p be a prime and let n Z
+
. Find a formula for the number of subgroups of order p in the elementary
abelian group E
p
n.
Solution. Every nonidentity element has order p, so there are p
n
1 many elements of order p. By Lagranges
theorem, each distinct subgroup intersects trivially. So there are (p
n
1)/(p 1) = p
n1
+ p
n2
+ + 1 many
distinct subgroups of order p.
5.1.12. Let A and B be groups. Assume Z(A) contains a subgroup Z
1
and Z(B) contains a subgroup Z
2
with
Z
1
= Z
2
. Let this isomorphism be given by the map x
i
y
i
for all x
i
Z
1
. A central product of A and B is a
quotient
(AB)/Z where Z =
(x
i
, y
1
i
) [ x
i
Z
1
x
2
, Z
2
=
r
2
and the
isomorphism is x
2
r
2
) and let Z
4
Q
8
be the central product of Z
4
and Q
8
which identies x
2
and 1.
Prove that Z
4
D
8
= Z
4
Q
8
.
Proof (a).
Proof (b).
5.1.18. In each of (a) to (e) give an example of a group with the specied properties:
(a) an innite group in which every element has order 1 or 2
(b) an innite group in which every element has nite order but for each positive integer n there is an element
of order n
(c) a group with an element of innite order and an element of order 2
(d) a group G such that every nite group is isomorphic to some subgroup of G
(e) a nontrivial group G such that G
= GG.
Solution (a). The direct sum Z
2
Z
2
.
Solution (b). The direct sum Z
2
Z
3
.
Solution (c). Z Z
2
.
Solution (d). S
2
S
3
.
Solution (e). G = Z
2
Z
2
.
29
5.2. The Fundamental Theorem of Finitely Generated Abelian Groups.
5.2.2. In each of parts (a) to (e) give the lists of invariant factors for all abelian groups of the specied order:
(a) order 270
Solution (a). Since 270 = 2 3
3
5, the list of invariant factors is given by (30, 3, 3), (90, 3).
5.2.3. In each of parts (a) to (e) give the lists of elementary divisors for all abelian groups of the specied order
and then match each list with the corresponding list of invariant factors found in the preceding exercise:
(a) order 270
Solution (a). The elementary divisors of an abelian group of order 270 are either (2, 3, 3, 5) or (2, 9, 5) with lists
corresponding to respective orderings.
5.2.7. Let p be a prime and let A = 'x
1
` 'x
2
` 'x
n
` be an abelian pgroup, where [x
i
[ = p
i
> 1 for all i.
Dene the pthpower map
: A A by : x x
p
.
(a) Prove that is a homomorphism.
(b) Describe the image and kernel of in terms of the given generators.
(c) Prove both ker and A/im have rank n (i.e., have the same rank as A) and prove these groups are both
isomorphic to the elementary abelian group, E
p
n, of order p
n
.
Proof (a). This map is clearly welldened, so we calculate
(xy) = (xy)
p
= x
p
y
p
= (x)(y).
i
)
p
, Z .
The kernel of is given by (k
1
, . . . , k
n
) where
k
i
x
i
[ (x
i
)
p
= 1, Z .
i
)
p
= 1, then [x
i
[
p so
x
p
i
1
i
=
n
i=1
x
p
i
1
so it is isomorphic to E
p
n. Since : A A and
ker [
xi
= p, then
im[
xi
=
p1
i
. Thus 'x
i
` /im[
xi
= Z
p
.
Therefore A/im
= E
pn
.
5.2.5. Prove that A
n
is the commutator subgroup of S
n
for all n 5.
Proof. Since A
n
S
n
and S
n
/A
n
= Z
2
is abelian, proposition 7 part (4) states S
n
A
n
. Also by proposition 7
part (3), S
n
is characteristic in S
n
. Since A
n
is simple, S
n
A
n
, and S
n
is non abelian, we must have A
n
= S
n
.
5.2.11. Prove that if G = HK where H and K are characteristic subgroups of G with H K = 1 then Aut(G)
=
Aut(H) Aut(K). Deduce that if G is an abelian group of nite order then Aut(G) is isomorphic to the direct
product of the automorphism groups of its Sylow subgroups.
30 DAVID S. DUMMIT AND RICHARD M. FOOTE
Proof. Let , Aut(G) and notice that [
H
Aut(H) and [
K
Aut(K) since H, K are characteristic in G.
Now let f : Aut(G) Aut(H) Aut(K) be given by
f() = ([
H
, [
K
).
If = , then ([
H
, [
K
) = ([
H
, [
K
) so f is welldened. Now we calculate
f( ) = (( )[
H
, ( )[
K
)
= ([
H
[
H
, [
K
[
K
)
= ([
H
, [
K
)([
H
, [
K
)
= f()f(),
so f is a homomorphism. Now if
H
Aut(H) and
K
Aut(K), then let Aut(G) be given by [
H
=
H
and [
K
=
K
. So f() = ([
H
, [
K
) = (
H
,
K
) shows f is surjective. Now if ([
H
, [
K
) = ([
H
, [
K
), then for
g = hk G,
(hk) = (h)(k)
= [
H
(h)[
K
(k)
= [
H
(h)[
K
(k)
= (h)(k)
= (hk),
so = . Therefore f shows Aut(G)
= Aut(H) Aut(K).
If G is nite abelian, then it is the direct product of its (characteristic) Sylow subgroups. Therefore Aut(G) is
isomorphic to the direct product of the automorphism groups of the Sylow subgroups.
5.2.14. Let G = (a
ij
) GL
n
(F) [ a
ij
= 0 if i > j, and a
11
= a
22
= = a
nn
, where F is a eld, be the group
of upper triangular matrices all of whose diagonal entries are equal. Prove that G
a a
12
a
1n
a a
2n
.
.
.
.
.
.
a
, B =
b b
12
b
1n
b b
2n
.
.
.
.
.
.
b
= D U.
5.3. Table of Groups of Small Order.
5.4. Recognizing Direct Products.
5.5. Semidirect Products.
5.5.7. This exercise describes thirteen isomorphism types of groups of order 56.
(a) Prove that there are three abelian groups of order 56.
(b) Prove that every group of order 56 has either a normal Sylow 2subgroup or a normal Sylow 7subgroup.
(c) Construct the following nonabelian groups of order 56 which have a normal Sylow 7subgroup and whose
Sylow 2subgroup S is as specied:
one group when S
= Z
2
Z
2
Z
2
two nonisomorphic groups when S
= Z
4
Z
2
one group when S
= Z
8
two nonisomorphic groups when S
= Q
8
three nonisomorphic groups when S
= D
8
.
31
(d) Let G be a group of order 56 with a nonnormal Sylow 7subgroup. Prove that if S is the Sylow 2subgroup
of G then S
= Z
2
Z
2
Z
2
.
(e) Prove that there is a unique group of order 56 with a nonnormal Sylow 7subgroup.
Proof (a). Suppose that G is an abelian group of order 56 = 2
3
7. The invariant factors are given by (2
3
7), (2
2
7, 2), (2 7, 2, 2). Thus there are three abelian groups of order 56 given by Z
56
, Z
28
Z
2
, Z
14
Z
2
Z
2
.
Proof (b). Suppose that n
2
, n
7
> 1. Since n
p
1 mod p, we have n
2
3 and n
7
8. Counting [
Syl
2
(G)`1[ +
[
Syl
7
(G)`1[ = 3 7 + 8 6 > 56 reaches a contradiction. Therefore every group of order 56 has either a normal
Sylow 2subgroup or a normal Sylow 7subgroup.
Solution (c). Note for all of the Sylow 2subgroup isomorphism types, if P Syl
7
(G), we have P S = 1 and
P G so G is a semidirect product.
If S
= Z
2
Z
2
Z
2
, then we consider the nontrivial homomorphisms : S Aut(P), where Aut(P)
= Z
6
.
Thus [ker [ = 4 and [(S)[ = 2 for any such . Therefore P
Z
4
, Z
2
2
r
2
, s
= Z
2
Z
2
. Since these kernels are not isomorphic, the comment in part (c) states there
are three isomorphism types.
Proof (d). If n
7
> 1, then the map : P Aut(S) is nontrivial so that P is not normal. Thus ker = 1 and
(P)
= Z
7
. Let (P) = 'x` so if s S`1, then s
x
k
shows all elements in S`1 have the same order. Since every
nonidentity element in S must have order 2, it is given by the elementary abelian group, S
= Z
3
2
.
Proof (e). If S
= Z
3
2
, then Aut(S)
= GL
3
(F
2
) with [Aut(S)[ = 7 6 4 = 168. Let P
1
, P
2
Syl
7
(G) so that for any
nontrivial : P
i
Aut(S), ker = 1. Thus (P
i
) Syl
7
(Aut(S)), so P
1
= P
2
for some Aut(S). Therefore
by exercise 6, S
P
1
= S
P
2
.
5.5.20. Let p be an odd prime. Prove that if P is a noncyclic pgroup then P contains a normal subgroup U with
U
= Z
p
Z
p
. Deduce that for odd primes p a pgroup that contains a unique subgroup of order p is cyclic.
Proof. Let [P[ = p
n
with P nonabelian. If n = 2, then P
= Z
p
Z
p
since P is not cyclic. Now suppose there exists
U P with U
= Z
p
Z
p
for all k < n. Since p [ [Z(P)[, let Z Z(P) with [Z[ = p. Thus the noncyclic quotient
group, P/Z, has a subgroup U P/Z with U
= Z
p
Z
p
. Now we have the noncyclic U P with [U[ = p
3
, so
there is a U
P with U
= Z
p
Z
p
.
6. Further Topics in Group Theory
6.1. pgroups, Nilpotent Groups, and Solvable Groups.
6.1.3. If G is nite prove that G is nilpotent if and only if it has a normal subgroup of each order dividing [G[, and
is cyclic if and only if it has a unique subgroup of each order dividing [G[.
Proof. Suppose G is nilpotent so theorem 3 states G
= P
1
P
s
for P
i
Syl
pi
. But if n [ [G[, then n = p
k1
1
p
ks
s
.
Theorem 1 states each P
i
has a normal subgroup, P
i
, with order p
ki
i
. Thus [P
1
P
s
[ = n with the desired
normality property.
Now suppose G has a normal subgroup of each order dividing [G[. Then each Sylow subgroup is normal in G,
so theorem 3 states G is nilpotent.
If G is cyclic, then theorem 2.7 states it has a unique subgroup of each order dividing [G[.
If G has a unique subgroup of each order dividing [G[, then proposition 5 shows G is cyclic.
6.1.4. Prove that a maximal subgroup of a nite nilpotent group has prime index.
Proof. Let M < G be a maximal subgroup of G. Since G is nilpotent, MG. Since G/M is nilpotent, exercise 6.1.3
shows it has a normal subgroup of each order dividing [G/M[. If P G/M with
= Z
p
Z
p
.
32 DAVID S. DUMMIT AND RICHARD M. FOOTE
Proof. Since [P[ = p
3
we know P is nilpotent. Since P is a pgroup Z(P) = 1, since P is nonabelian, Z(P) = P.
Thus [Z(P)[
p, p
2
. If [Z(P)[ = p
2
, then P/Z(P) is cyclic contradicting P being nonabelian. Thus [Z(P)[ = p.
Since P/Z(P) is not cyclic, we must have P/Z(P)
= Z
p
Z
p
.
6.1.12. Find the upper and lower central series for A
4
and S
4
.
Solution. The upper central series of A
4
is Z
i
(A
4
) = 1 and of S
4
is Z
i
(S
4
) = 1 for all i since the center of S
4
is
trivial.
Since s
1
a
1
sa is an even permutation, S
1
4
A
4
. But if A
4
, we can choose s S
4
such that s
1
a
1
s = a
so that S
1
4
. The same argument shows S
2
4
= A
4
. Since S
1
4
= A
4
and S
2
4
= A
4
, we have found the lower central
series for S
4
.
Now N = 1, (12)(34), (13)(24), (14)(23) is the only proper normal subgroup of A
4
. Thus A
i
4
= N for all i so
this is the lower central series for A
4
.
33
Part II Ring Theory
7. Introduction to Rings
7.1. Basic Denitions and Examples.
Proposition 7.1. Let R be a ring. Then
(1) 0a = a0 = 0 for all a R.
(2) (a)b = a(b) = (ab) for all a, b R.
(3) (a)(b) = ab for all a, b R.
(4) if R has an identity 1, then the identity is unique and a = (1)a.
Proposition 7.2. Assume a, b and c are element of any ring with A not a zero divisor. If ab = ac, then either
a = 0 or b = c. In particular, if a, b, c are any elements in an integral domain and ab = ac, then either a = 0 or
b = c.
Corollary 7.3. Any nite integral domain is a eld.
7.1.3. Let R be a ring with identity and let S be subring of R containing the identity. Prove that if u is a unit in
S then u is a unit in R. Show by example that the converse if false.
Proof. If u is a unit in S, then there is a v S such that uv = 1 = vu. But then v R and hence u is a unit in R.
Now consider 2 Q and notice that it is a unit because 1/2 Q. But Z Q is a subring of Q and 2 is not a
unit. Thus the converse is false.
7.1.5. Decide which of the following (a)(f) are subrings of Q:
(a) the set of all rational numbers with odd denominators (when written in lowest terms)
(b) the set of all rational numbers with even denominators (when written in lowest terms)
(c) the set of nonnegative rational numbers
(d) the set of squares of rational numbers
(e) the set of all rational numbers with odd numerators (when written in lowest terms)
(f) the set of all rational numbers with even numerators (when written in lowest terms).
Solution. (a), (f).
7.1.6. Decide which of the following are subrings of the ring of all functions from the closed interval [0, 1] to 1.
(a) the set of all functions f(x) such that f(q) = 0 for all q Q [0, 1]
(b) the set of all polynomial functions
(c) the set of all functions which have only a nite number of zeros, together with the zero function
(d) the set of all functions which have an innite number of zeros
(e) the set of all functions f such that lim
x1
f(x) = 0
(f) the set of all rational linear combinations of the functions sin nx and cos mx, where m, n Z
+
.
Solution. (a), (b), (c), (e).
7.1.7. The center of a ring R is z R [ zr = rz, r R (i.e., is the set of all elements which commute with every
element of R). Prove that the center of a division ring is a eld.
Proof.
7.1.12. Prove that any subring of a eld which contains the identity is an integral domain.
Proof. Let F be a eld and let R F be a subring of F with 1 R. Then since R F, every element of R is a
unit. Therefore there are no zero divisors in R.
7.1.13. An element x in R is called nilpotent if x
m
= 0 for some m Z
+
.
(a) Show that if n = a
k
b for some integers a and b then ab is a nilpotent element of Z/nZ.
(b) If a Z is an integer, show that the element a Z/nZ is nilpotent if and only if every prime divisor of n
is also a divisor of a. In particular, determine the nilpotent elements of Z/72Z explicitly.
(c) Let R be the ring of functions from a nonempty set X to a eld F. Prove that R contains no nonzero
nilpotent elements.
7.1.14. Let x be a nilpotent element of the commutative ring R (cf. Exercise 7.1.13).
(a) Prove that x is either zero or a zero divisor.
34 DAVID S. DUMMIT AND RICHARD M. FOOTE
(b) Prove that rx is nilpotent for all r R.
(c) Prove that 1 +x is a unit in R.
(d) Deduce that the sum of a nilpotent element and a unit is a unit.
Proof (a). Suppose x = 0 and since x
m
= 0, let n be the smallest positive integer such that x
n
= 0. Then xx
n1
= 0
where x
n1
= 0 so x is a zero divisor.
Proof (b). Since R is commutative, we have (rx)
m
= r
m
x
m
= 0. Thus rx is nilpotent for all r R.
Proof (c). Observing the calculation
(1 +x)(1 x +x
2
x
3
+ x
m1
) = 1
shows that 1 +x is a unit.
Solution (d). Let u be a unit so vu = 1 and let x be our nilpotent element. Then part (b) shows vx is nilpotent
and part (c) shows 1 +vx is a unit. Thus v(u +x) is a unit and hence u +x is a unit.
7.1.23. Let D be a squarefree integer, and let O be the ring of integers in the quadratic eld Q(
D).
Proof. Let a + bf O
f
. Since bf Z, then a + bf O so O
f
O. Now (a + bf)(c + df) = ac + (ad +
bd)f +bdf
2
2
O
f
and (a +bf) (c +df) = (a c) + (b d)f O
f
. Thus O
f
is a subring of O.
Let a +b +O
f
= a
+b
+O
f
. Then (a +b) (a
+b
) O
f
and (a a
) +(b b
) O
f
. So b b
= b
f
thus b b
such that a + b + O
f
= a
+ b
+ O
f
.
But this is a contradiction because f would have to divide b b
. Thus a +b +O
f
and a
+b
+O
f
are the same
coset if and only if b b
Z satisfying
(i) (ab) = (a) +(b)
(ii) is surjective, and
(iii) (x +y) min (x), (y) for all x, y K
with x +y = 0.
The set R = x K
Solution (b).
p(x) +q(x) = x
3
x
2
+x 1
p(x)q(x) = x
5
x
Solution (c).
p(x) +q(x) = x
p(x)q(x) = 2x
6
+x
4
+x
3
+ 2x
Solution (c).
= 108(1) + 81(1 2) 21(1 3) 30(2 3) + 90(1 2 3)
Solution (d).
= 108(1) + 18(1 2) + 63(1 3) 51(2 3) + 84(1 2 3) + 6(1 3 2)
Solution (e).
2
= 34(1) 70(1 2) 28(1 3) + 42(2 3) 15(1 2 3) + 196(1 3 2)
7.2.12. Let G = g
1
, . . . , g
n
be a nite group. Prove that the element N = g
1
+ g
2
+ + g
n
is in the center of
the group ring RG (cf. Exercise 7.1.7).
Proof. Let X = x
1
g
1
+ +x
n
g
n
for x
i
R. Then
NX =
n
k=1
gigj=g
k
x
j
g
k
=
n
k=1
j=1
x
j
g
k
=
n
k=1
gigj=g
k
x
i
g
k
= XN.
= R/J.
(3) (The Fourth Isomorphism Theorem for Rings) Let I be an ideal of R. The correspondence A A/I is an
inclusion preserving bijection between the wset of subrings A of R that contain I and the set of subrings of
R/I. Furthermore, A (a subring containing I) is an ideal of R if and only if A/I is an ideal of R/I.
7.3.12. Let D be an integer that is not a perfect square in Z and let S =
a b
Db a
[ a, b Z
.
(a) Prove that S is a subring of M
2
(Z).
(b) If D is not a perfect square in Z prove that the map : Z[
D] S dened by (a +b
D) =
a b
Db a
is
a ring isomorphism.
(c) If D 1 mod 4 is squarefree, prove that the set
a b
(D 1)b/4 a +b
[ a, b Z
is a subring of M
2
(Z)
and is isomorphic to the quadratic integer ring O.
Proof (a). S is clearly a group under addition, so we show that S is closed under multiplication. But
a
1
b
1
Db
1
a
1
a
2
b
2
Db
2
a
2
a
1
a
2
+Db
1
b
2
a
1
b
2
+a
2
b
1
D(a
2
b
1
+a
1
b
2
) Db
1
b
2
+a
1
a
2
so S is a subring of M
2
(Z).
Proof (b). We calculate
((a
1
+b
1
D) + (a
2
+b
2
D)) = ((a
1
+a
2
) + (b
1
+b
2
)
D)
=
a
1
+a
2
b
1
+b
2
D(b
1
+b
2
) a
1
+a
2
a
1
b
1
Db
1
a
1
a
2
b
2
Db
2
a
2
((a
1
+b
1
D)(a
2
+b
2
D)) = ((a
1
a
2
+b
1
b
2
D) + (a
1
b
2
+a
2
b
1
)
D)
=
a
1
a
2
+b
1
b
2
D a
1
b
2
+a
2
b
1
D(a
1
b
2
+a
2
b
1
) a
1
a
2
+b
1
b
2
D
a
1
b
1
Db
1
a
1
a
2
b
2
Db
2
a
2
a
1
b
1
Db
1
a
1
a
2
b
2
Db
2
a
2
, so a
1
= a
1
, b
1
= b
2
and thus
a
1
+b
1
D = a
2
+b
2
D. Now if
a b
Db a
S, then maps a +b
a
1
b
1
(D 1)b
1
/4 a
1
+b
1
a
2
b
2
(D 1)b
2
/4 a
2
+b
2
a
1
a
2
+ (D 1)a
1
b
2
/4 a
1
b
2
+a
2
b
1
+b
1
b
2
(D 1)(a
1
b
2
+b
1
b
2
)/4 (D 1)b
1
b
2
/4 +a
1
a
2
+a
1
b
2
+a
2
b
1
+b
1
b
2
so this is a subring of M
2
(Z). Similarly to the proof in part (b), this is isomorphic to the quadratic integer
ring O.
7.3.15. Let X be a nonempty set and let {(X) be the Boolean ring of all subsets of X dened in exercise 21 of
section 1. Let R be the ring of all functions from X into Z/2Z. For each A {(X) dene the function
A
: X Z/2Z by
A
(x) =
1, x A,
0, x / A,
A
is called the characteristic function of A with values in Z/2Z. Prove that the map {(X) R dened by
A
A
is a ring isomorphism.
38 DAVID S. DUMMIT AND RICHARD M. FOOTE
Proof. Let A = A
so that (A) =
A
=
A
= (A
(A\B)(B\A)
=
1, x A`B,
1, x B`A,
0 x A B,
0, x X`(A B)
But simple calculations show this is (A) +(B). Similarly,
(AB) = (A B)
=
AB
=
0, x A`B,
0, x B`A,
1, x A B,
0, x X`(A B),
= (A)(B).
Now let f R so let A X be the set such that f(A) = 1 and f(X`A) = 0. So (A) =
A
= f and is
surjective. Now let f = g R so there exist A, B such that (A) = f = g = (B). So
A
=
B
, thus A = B and
is injective. Therefore is a ring isomorphism.
7.3.17. Let R and S be nonzero rings with identity and denote their respective identities by 1
R
and 1
S
. Let
: R S be a nonzero homomorphism of rings.
(a) Prove that if (1
R
) = 1
S
then (1
R
) is a zero divisor in S. Deduce that if S is an integral domain then
every ring homomorphism from R to S sends the identity of R to the identity of S.
(b) Prove that if (1
R
) = 1
S
then (u) is a unit in S and that (u
1
) = (u)
1
for reach unit u of R.
Proof (a). If (1
R
) = 1
S
, then
((1
R
) 1
S
)(1
R
) = (1
R
)(1
R
) 1
S
(1
R
)
= (1
R
) (1
R
)
= 0.
So (1
R
) is a zero divisor in S. If S is an integral domain, then there are no zero divisors. Since an integral domain
has a 1 = 0 and (1
R
) = 0 if (R) = 0, then (1
R
) 1
S
= 0. Therefore (1
R
) = 1
S
.
Proof (b). If u R is a unit, there exists v R such that uv = 1
R
. So (u)(v) = (uv) = (1
R
) = 1
S
. Thus
(u) is a unit, with (u)
1
= (v) = (u
1
).
7.3.27. Prove that a nonzero Boolean ring has characteristic 2 (cf. exercise 15, section 1).
Proof. Let a = 1 so that
0 = (1)
2
+ 1 = 1 + 1.
Since 1 = 0, then R has characteristic 2.
7.3.29. Let R be a commutative ring. Recall (cf. exercise 7.1.13) that an element x R is nilpotent if x
n
= 0 for
some n Z
+
. Prove that the set of nilpotent element form an ideal called the nilradical of R and denoted by
A(R).
Proof.
7.3.30. Prove that if R is a commutative ring and A(R) is its nilradical (cf. exercise 7.3.29) then zero is the only
nilpotent element of R/A(R) i.e., prove that A(R/A(R)) = 0.
Proof. Let x R/A(R) with x
n
= 0 for some n N. Then there exists r R such that x = r +A(R). But then
0 = x
n
= (r +A(R))
n
= r
n
+A(R). This implies r
n
A(R) so r A(R) and thus x = 0.
7.3.33. Assume R is commutative. Let p(x) = a
n
x
n
+a
n1
x
n1
+ +a
1
x +a
0
be an element of the polynomial
ring R[x].
(a) Prove tha tp(x) is a unit in R[x] if and only if a
0
is a unit and a
1
, a
2
, . . . , a
n
are nilpotent in R. [See
exercise 7.1.14.]
39
(b) Prove that p(x) is nilpotent in R[x] if and only if a
0
, a
1
, . . . , a
n
are nilpotenet elements of R.
Proof (a).
Proof (b).
7.4. Properties of Ideals.
Proposition 7.9. Let I be an ideal of R.
(1) I = R if and only if I contains a unit.
(2) Assume R is commutative. Then R is a eld if and only if its only ideals are 0 and R.
Corollary 7.10. If R is a eld then any nonzero ring homomorphism from R into another ring is an injection.
Proposition 7.11. In a ring with identity every proper ideal is contained in a maximal ideal.
Proposition 7.12. Assume R is commutative. The ideal M is a maximal ideal if and only if the quotient ring
R/M is a eld.
Proposition 7.13. Assume R is commutative. Then the ideal P is a prime ideal in R if and only if the quotient
ring R/P is an integral domain.
Corollary 7.14. Assume R is commutative. Every maximal ideal of R is a prime ideal.
Let R be a ring with identity 1 = 0.
7.4.2. Assume R is commutative. Prove that the augmentation ideal in the group ring RG is generated by
g 1 [ g G. Prove that if G = '` is cyclic then the augmentation ideal is generated by 1.
7.4.3.
(a) Let p be a prime and let G be an abelian group of order p
n
. Prove that the nilradical of the group ring
F
p
G is the augmentation ideal (cf. exercise 7.3.29).
(b) Let G = g
1
, . . . , g
n
be a nite group and assume R is commutative. Prove that if r is any element of the
augmentation ideal of RG then r(g
1
+ +g
n
) = 0.
Proof (a). Suppose there exists m such that (
p
n
i=1
a
i
g
i
)
m
= 0. If is the augmentation map, then
0 = (
p
n
i=1
a
i
g
i
)
m
= (
p
n
i=1
a
i
)
m
. Since F
p
` 0 is a multiplicative group, we must have
p
n
i=1
a
i
= 0. Thus
p
n
i=1
a
i
g
i
ker .
Now let g 1 ker . Then (g 1)
p
n
= g
p
n
+p(. . .) 1 = 0 in F
p
G. Since exercise 7.4.2 states ker is generated
by g 1 [ g G, we must have ker A(F
p
G). Therefore ker = A(F
p
G).
Proof (b). By exercise 7.4.2 we may assume r = g
i
1. But then
(g
i
1)(g
1
+ +g
n
) = g
i
g
1
+ g
i
g
n
g
1
g
n
= g
1
+ +g
n
g
1
g
n
= 0.
7.4.7. Let R be a commutative ring with 1. Prove that the principal ideal generated by x in the polynomial ring
R[x] is a prime ideal if and only if R is an integral domain. Prove that (x) is a maximal ideal if and only if R is a
eld.
Proof. Let : R[x] R be given by (p(x)) = p(0) and notice ker = (x). Thus R[x]/(x)
= R so by proposi
tion 7.13, (x) is prime if and only if R is an integral domain.
Similarly, proposition 7.12 shows (x) is maximal if and only if R is a eld.
7.4.10. Assume R is commutative. Prove that if P is a prime ideal of R and P contains no zero divisors then R is
an integral domain.
Proof. Since P is prime, we know that R = R/P is an integral domain. So if r, a R`P, then ra = r a = 0. Thus
ra = 0 and R is an integral domain.
7.4.13. Let : R S be a homomorphism of commutative rings.
40 DAVID S. DUMMIT AND RICHARD M. FOOTE
(a) Prove that if P is a prime ideal of S then either
1
(P) = R or
1
(P) is a prime ideal of R. Apply this
to the special case when R is a subring of S and is the inclusion homomorphism to deduce that if P is a
prime ideal of S then P R is either R or a prime ideal of R.
(b) Prove that if M is a maximal ideal of S and is surjective then
1
(M) is a maximal ideal of R. Give an
example to show that this need not be the case if is surjective.
Proof (a). Let r
1
(P), so (rs) = (r)(s) P because (r) P and P is an ideal. Since R is commutative,
1
(P) is an ideal. Now let rs
1
(P), so (r)(s) = (rs) P so (r) or (s) is in P because P is prime. So
r or s is in
1
(P) so
1
(P) is prime or
1
(P) = R.
Specically if R S and is the inclusion homomorphism, then
1
(P) is either R or a prime ideal of R. But
1
(P) = P R.
Proof (b). Let
1
: S S/M and
2
: R/ ker(
1
) be the canonical projection maps. Since
1
is surjective,
R/ ker(
1
)
n
i=1
a
i
x
i
be an element of F
2
[x]. If n = 1 then
1
i=0
= a
0
+ a
1
x. Suppose there exists c
0
, c
1
such
that
n1
i=0
a
i
x
i
= c
0
+c
1
x. Then if n is even,
n
i=0
a
i
x
i
=
n1
i=0
a
i
x
i
+a
n
x
n
=
n1
i=0
a
i
x
i
+a
n
(x + 1)
n/2
=
n1
i=0
c
i
x
i
= c
0
+c
1
x.
Now if n > 2 is odd,
n
i=0
a
i
x
i
=
n1
i=0
a
i
x
i
+a
n
x
n
=
n1
i=0
a
i
x
i
+a
n
x(x + 1)
(n1)/2
=
n1
i=0
b
i
x
i
= c
0
+c
1
x.
So there exist c
0
, c
1
such that
n
i=0
a
i
x
i
= c
0
+c
1
x for all n. Since c
0
, c
1
0, 1, F
2
[x]/(x
2
+x+1)
0, 1, x, x + 1
= V
4
.
Solution (c). The multiplication table in table 7.2 shows E
= Z
3
. Therefore E is a eld.
7.4.30. Let I be an ideal of the commutative ring R and dene
rad I =
r R [ r
n
I for some n Z
+
called the radical of I. Prove that rad I is an ideal containing I and that (rad I)/I = A(R/I) (cf. exercise 7.3.29).
41
Addition Table
0 1 x 1 +x
0 0 1 x 1 +x
1 1 0 1 +x x
x x 1 +x 0 1
1 +x 1 +x x 1 0
Table 7.1. F
2
[x]/(x
2
+x + 1).
Multiplication Table
0 1 x 1 +x
0 0 0 0 0
1 0 1 x 1 +x
x 0 x 1 +x 1
1 +x 0 1 +x 1 x
Table 7.2. F
2
[x]/(x
2
+x + 1).
7.4.31. An ideal I of the commutative ring R is called a radical ideal if rad I = I.
(a) Prove that every prime ideal of R is a radical ideal.
(b) Let n > 1 be an integer. Prove that 0 is a radical ideal in Z/nZ if and only if n is a product of distinct
primes to the rst power (i.e., n is square free). Deduce that (n) is a radical of Z if and only if n is a
product of distinct primes in Z.
Proof (a). Note P rad P. Let r R so if r
n
P, either r or r
n1
is in P. Induction shows r P, so
rad P = P.
Proof (b). Suppose rad 0 = 0 and n = mp
2
. Then mp = 0 mod n and (mp)
2
= 0 mod n which is a contradiction.
Thus n is a product of distinct primes to the rst power.
Now suppose n is a product of distinct primes to the rst power. If
m
= 0 mod n, then
m
contains all of the
prime factors of n. Thus must contain all prime factors of n so = 0 mod n. Therefore rad 0 = 0.
7.4.39. Following the notation of exercise 7.1.26, let K be a eld, let be a discrete valuation on K and let R be
the valuation ring of . For each integer k 0 dene A
k
= r R [ (r) k 0.
(a) Prove that A
k
is a principal ideal and that A
0
A
1
A
2
.
(b) Prove that if I is any nonzero ideal of R, then I = A
k
for some k 0. Deduce that R is a local ring with
unique maximal ideal A
1
.
Proof (a). Since is surjective, A
k
is nonempty. Since 1 R
is isomorphic to Q.
Let R be a commutative ring with identity 1 = 0.
7.5.1. Prove theorem 7.15.
Proof. Let T = (r, d) [ r R, d D and dene the relation on T by
(r, d) (s, e) if and only if re = sd.
It is immediate that this relation is reexive and symmetric. Suppose (r, d) (s, e) and (s, e) (t, f). Then
re sd = 0 and sf te = 0. Multiplying the rst of these equations by f and the second by d and adding them
gives rf td)e = 0. Since e D is neither zero nor a zero divisor we must have rf td = 0, i.e., (r, d) (t, f).
This proves is transitive, hence an equivalence relation. Done the equivalence class of (r, d) by
r
d
:
r
d
= (a, b) [ a R, b D and rb = ad .
Let Q be the set of equivalence classes under . Note that
r
d
=
re
de
in Q for all e D, since D is closed under
multiplication.
We now dene an additive and multiplicative structure on Q:
a
b
+
c
d
=
ad +bc
bd
and
a
b
c
d
=
ac
bd
.
In order to prove that Q is a commutative ring with identity there are a number of things to check:
(1) these operations are well dened
(2) Q is an abelian group under addition, where the additive identity is
0
d
for any d D and the additive
inverse of
a
d
is
a
d
,
(3) multiplication is associative, distributive and commutative, and
(4) Q has an identity.
43
To check that addition is well dened assume
a
b
=
a
(i.e., ab
= a
b) and
c
d
=
c
+b
, i.e.,
(ad +bc)(b
) = (a
+b
)(bd).
The left hand side of this equation is ab
dd
+ cd
bb
substituting a
b for ab
and c
d for cd
gives a
bdd
+ c
dbb
,
which is the right hand side. Hence addition of fractions is well dened.
To check that multiplication is well dened assume
a
b
=
a
and
c
d
=
c
. We must show
(ac)(b
) = (a
)(bd).
The left hand side of this equation is (ab
)(cd
) substituting a
b for ab
and c
d for cd
gives (a
b)(c
d) which is the
right hand side. Hence multiplication of fractions is well dened.
To show Q is an abelian group under addition note that
0
d
Q with
0
d1
+
r
d2
=
d1r
d1d2
=
r
d2
so there exists an
additive identity in Q. Now if
r1
d1
,
r2
d2
,
r3
d3
Q, then
(
r
1
d
1
+
r
2
d
2
) +
r
3
d
3
=
r
1
d
2
+r
2
d
1
d
1
d
2
+
r
3
d
3
=
r
1
d
2
d
3
+r
2
d
1
d
3
+r
3
d
1
d
2
d
1
d
2
d
3
=
r
1
d
1
+
r
2
d
3
+d
1
d
2
d
2
d
3
=
r
1
d
1
+ (
r
2
d
2
+
r
3
d
3
).
Thus addition is associative. Now if
r
d
Q, then
r
d
Q with
r
d
+
r
d
=
rd rd
d
2
=
0
d
.
So Q is closed under inverse. Since addition is clearly commutative in Q, we see Q is an abelian group under
addition.
To show multiplication is associative, let
r1
d1
,
r2
d2
,
r3
d3
Q so
(
r
1
d
1
r
2
d
2
)
r
3
d
3
=
r
1
r
2
d
1
d
2
r
3
d
3
=
r
1
r
2
r
3
d
1
d
2
d
3
=
r
1
d
1
r
2
r
3
d
2
d
3
=
r
1
d
1
(
r
2
d
2
r
3
d
3
).
To show multiplication is distributive, let
r1
d1
,
r2
d2
,
r3
d3
Q so
(
r
1
d
1
+
r
2
d
2
)
r
3
d
3
=
r
1
d
2
+r
2
d
1
d
1
d
2
r
3
d
3
=
r
1
r
3
d
2
+r
2
r
3
d
1
d
1
d
2
d
3
=
r
1
d
1
r
3
d
2
d
2
d
3
+
r
2
d
2
r
3
d
1
d
1
d
3
=
r
1
d
1
r
3
d
3
+
r
2
d
2
r
3
d
3
r
1
d
1
(
r
2
d
2
+
r
3
d
3
) =
r
1
d
1
r
2
d
3
+r
3
d
2
d
2
d
3
=
r
1
r
2
d
3
+r
1
r
3
d
2
d
1
d
2
d
3
=
r
1
d
1
r
2
d
3
d
2
d
3
+
r
1
d
1
r
3
d
2
d
2
d
3
=
r
1
d
1
r
2
d
2
+
r
1
d
1
r
3
d
3
.
44 DAVID S. DUMMIT AND RICHARD M. FOOTE
To show multiplication is commutative, let
r1
d1
,
r2
d2
Q so
r
1
d
1
r
2
d
2
=
r
1
r
2
d
1
d
2
=
r
2
r
1
d
2
d
1
=
r
2
d
2
r
1
d
1
.
To show Q has an identity, notice that
d
d
Q for any d D. Now let
r
d1
Q so
r
d
1
d
d
=
rd
d
1
d
=
r
d
1
.
Next we embed R into Q by dening
: R Q by : r
rd
d
where d is any element of D.
Since
rd
d
=
re
e
for all d, e D, (r) does not depend on the choice of d D. Now let r
1
, r
2
R so
(r
1
+r
2
) =
(r
1
+r
2
)d
d
=
r
1
d
d
+
r
2
d
d
= (r
1
) +(r
2
)
(r
1
r
2
) =
(r
1
r
2
)d
d
=
r
1
d
d
r
2
d
d
= (r
1
)(r
2
),
so is a ring homomorphism. Furthermore, is injective because
(r) = 0
rd
d
=
0
d
rd
2
= 0 r = 0
because d is neither zero nor a zero divisor. The subring (R) of Q is therefore isomorphic to R. We henceforth
identify each r R with (r) and so consider R as a subring of Q.
Next note that each d D has a multiplicative inverse in Q: namely, if d is represented by the fraction
de
e
then
its multiplicative inverse is
e
de
. One then sees that every element of Q may be written as r d
1
for some r R and
some d D. In particular, if D = R` 0, every nonzero element of Q has a multiplicative inverse and Q is a eld.
It remains to establish the uniqueness property of Q. Assume : R S is an injective ring homomorphism such
that (d) is a unit in S for all d D. Extend to a map : Q S by dening (rd
1
) = (r)(d)
1
for all
r R, d D. This map is well dened, since rd
1
= se
1
implies re = sd, so (r)(e) = (s)(d), and then
(rd
1
) = (r)(d)
1
= (s)(e)
1
= (se
1
).
Now let r
1
d
1
1
, r
2
d
1
2
Q. Then
(r
1
d
1
1
+r
2
d
1
2
) = ((r
1
d
2
+r
2
d
1
)(d
1
d
2
)
1
)
= (r
1
d
2
+r
2
d
1
)(d
1
d
2
)
1
= (r
1
)(d
2
)(d
1
)
1
(d
2
)
1
+(r
2
)(d
1
)(d
1
)
1
(d
2
)
1
= (r
1
)(d
1
)
1
+(r
2
)(d
2
)
1
= (r
1
d
1
1
) + (r
2
d
1
2
)
(r
1
d
1
1
r
2
d
1
2
) = ((r
1
r
2
)(d
1
d
2
)
1
)
= (r
1
r
2
)(d
1
d
2
)
1
= (r
1
)(d
1
)
1
(r
2
)(d
2
)
1
= (r
1
d
1
1
)(r
2
d
1
2
).
So is a ring homomorphism. Finally, is injective because rd
1
ker implies r ker R = ker ; since is
injective this forces r and hence also rd
1
to be zero.
45
7.5.2. Let R be an integral domain and let D be a nonempty subset of R that is closed under multiplication.
Prove that the ring of fractions D
1
R is isomorphic to a subring of the quotient eld of R (hence is also an integral
domain).
Proof. Let Q be the quotient eld of R. Since R is a domain, D` 0 is also a multiplicative set of R. By theorem 7.15
there exists an injective ring homomorhpism : R Q with (R` 0) Q
), then d
= ud for some u R
. In particular, if d and d
D) be a quadratic eld with associated quadratic integer ring O and eld norm N as in section
7.1.
(a) Suppose D is 1, 2, 3, 7 or 11. Prove that O is a Euclidean Domain with respect to N.
(b) Suppose that D = 43, 67 or 163. Prove that O is not a Euclidean Domain with respect to any norm.
8.1.10. Prove that the quotient ring Z[i]/I is nite for any nonzero ideal I of Z[i].
Proof. By exercise 8.1.8(a), Z[i] is a Euclidean Domain. So I = () for some Z[i]. Let x + I be a coset of I.
Then there exists q, r such that
x + = q +r.
Since x / (), then r = 0 so N(r) < N(). But now x+I = r +I, so any coset of I has a representative with norm
less than N().
Let F = f R [ N(f) < N(). Then if f = f
1
+f
2
i and = a
1
+a
2
i we have f
2
1
+f
2
2
< a
2
1
+a
2
2
. Since there
are only a nite number of (f
1
, f
2
) Z Z with this property, Z[i]/I is nite for any nonzero ideal I of Z[i].
46 DAVID S. DUMMIT AND RICHARD M. FOOTE
8.1.11. Let R be a commutative ring with 1 and let a and b be nonzero elements of R. A least common multiple
of a and b is an element e of R such that
(i) a [ e and b [ e, and
(ii) if a [ e
and b [ e
then e [ e
.
(a) Prove that a least common multiple of a and b (if such exists) is a generator for the unique largest principal
ideal contained in (a) (b).
(b) Deduce that any two nonzero elements in a Euclidean Domain have a least common multiple which is unique
up to multiplication by a unit.
(c) Prove that in a Euclidean Domain the least common multiple of a and b is
ab
(a,b)
, where (a, b) is the greatest
common divisor of a and b.
Proof (a). First suppose there exists a least common multiple e of a and b. Since a [ e and b [ e we must have
(e) (a) and (e) (b). If (e
[ a and e
[ b. Thus e [ e
and
(e
and b [ e
,
then (e
) (a) (b). Since (e) is the largest such principal ideal, then (e
.
Proof (b). If a and b are nonzero elements of R then (a) (b) is the largest ideal contained in (a) (b). Since R is
a Euclidean Domain, there exists e R with (e) = (a) (b). So part (a) shows e is a least common multiple of a
and b.
Now let e, e
and e
[ e. So e = se
and et = e
. Since a
and b are nonzero then e is also nonzero so e = ste and e(1 st) = 0 implies 1 = st because R is a domain. Thus
least common multiples are unique up to multiplication by a unit.
Proof (c). First note that (
ab
(a,b)
) (a) (b) so (
ab
(a,b)
) (e). Now let ax = e = by. Then abx = eb and aby = ea
so ab [ eb and ab [ ea. Thus ab [ (ea, eb) and
ab
(a,b)
[ e. So we have (e) (
ab
(a,b)
). Therefore
ab
(a,b)
is a least common
multiple of a and b.
8.1.12 (A Public Key Code). Let N be a positive integer. Let M be an integer relatively prime to N and let d
be an integer relatively prime to (N), where denotes Eulers function. Prove that if M
1
M
d
mod N then
M M
d
1
mod N where d
1 mod (N).
Proof. Since (M, N) = 1 we know that M
(N)
1 mod N. Thus
M
d
1
M
dd
mod N
M
(N)+1
mod N
M mod N.
9. Polynomial Rings
9.1. Denitions and Basic Properties.
Proposition 9.1. Let R be an integral domain. Then
(1) degree p(x)q(x) = degree p(x) + degree q(x) if p(x), q(x) are nonzero
(2) the units of R[x] are just the units of R
(3) R[x] is an integral domain.
Proposition 9.2. Let I be an ideal of the ring R and let (I) = I[x] denote the ideal of R[x] generated by I (the
set of polynomials with coecients in I). Then
R[x]/(I)
= (R/I)[x].
In particular, if I is a prime ideal of R then (I) is a prime ideal of R[x].
9.1.13. Prove that the rings F[x, y]/(y
2
x) and F[x, y]/(y
2
x
2
) are not isomorphic for any eld F.
Proof. Since y
2
x
2
= (y x)(y +x) then F[x, y]/(y
2
x
2
) is not a domain. Not let : R[x, y] R[y] be given by
(p(x, y)) = p(y
2
, y).
Let p(x, y)(y
2
x) (y
2
x) so that
(p(x, y)(y
2
x)) = p(y
2
, y)(y
2
y
2
) = 0.
47
Thus (y
2
x) ker . Now let p(x, y) ker so that 0 = (p(x, y)) = p(y
2
, y). Note that we can write p(x, y) =
r(x)+ys(x)+q(x, y)(y
2
x) by considering cosets of F[x, y]/(y
2
x). But then 0 = (r(x)+ys(x)) = r(y
2
)+ys(y
2
).
So r = 0 and s = 0. Therefore p(x, y) = q(x, y)(y
2
x) so ker (y
2
x).
9.1.14. Let R be an integral domain and let i, j be relatively prime integers. Prove that the ideal (x
i
y
j
) is a
prime ideal in R[x, y].
Proof.
9.1.15. Let p(x
1
, x
2
, . . . , x
n
) be a homogeneous polynomial of degree k in R[x
1
, . . . , x
n
]. Prove that for all R
we have
p(x
1
, x
2
, . . . , x
n
) =
k
p(x
1
, x
2
, . . . , x
n
).
Proof. First we write
p(x
1
, x
2
, . . . , x
n
) =
m
i=0
a
i
n
j=1
x
kj
j
,
where
n
j=1
k
j
= k. Then
p(x
1
, x
2
, . . . , x
n
) =
m
i=0
a
i
n
j=1
(x
j
)
kj
=
k
m
i=0
a
i
n
j=1
x
kj
j
=
k
p(x
1
, x
2
, . . . , x
n
).
n
i=1
p
i
(x) and
notice 1 +p(x) F[x]. Since F[x] is a unique factorization domain we can write
1 +p(x) =
m
i=1
p
i
(x)
where each p
i
(x) is irreducible. But now
1 =
m
i=1
p
i
(x) p(x)
= p
1
(x)
i=2
p
i
(x)
p(x)
p
1
(x)
.
This is a contradiction because a prime is not a unit.
9.2.5. Exhibit all the ideals in the ring F[x]/(p(x)), where F is a eld and p(x) is a polynomial in F[x].
48 DAVID S. DUMMIT AND RICHARD M. FOOTE
Proof. Since F is a eld, F[x] is a Euclidean domain and in particular a unique factorization domain. So let
p(x) =
n
i=1
p
i
(x),
where each p
i
(x) is irreducible. Let I/(p(x)) F[x]/(p(x)) be an ideal. By the lattice isomorphism theorem,
(p(x)) I, and I is an ideal of F[x]. Since F[x] is a Euclidean domain, it is also a principal ideal domain so let I =
(q(x)). Since (p(x)) (q(x)), p(x) = r(x)q(x) for some r(x) F[x]. Let r(x) =
s
i=1
r
i
(x) and q(x) =
t
i=1
q
i
(x)
for r
i
(x), q
i
(x) irreducible. Then
s
i=1
r
i
(x)
t
i=1
q
i
(x) =
n
i=1
p
i
(x), for some F. So there exists S
n
such that
i
q
i
(x) = p
(i)
(x) for all 1 i t,
i
F. Therefore (q(x)) = (
t
i=1
q
i
(x)) = (
t
i=1
p
(i)
(x)).
9.2.6. Describe the ring structure of the following rings:
(d) Z[x, y]/(x
2
, y
2
, 2).
Show that
2
= 0 or 1 for every in the last ring and determine those elements with
2
= 0. Determine the
characteristics of each of these rings.
Proof (d). The highest power of x and y is 1 along with the largest coecient. Let = (x, y) + (x
2
, y
2
, 2) so
2
= ((x, y))
2
+ (x
2
, y
2
, 2)
0, 1
because if there is no constant term then every term has even degree and if
there is a constant term, let (x, y) = (x, y) +1 where (x, y) has no constant term. So (x, y)
2
= ((x, y) +1)
2
=
(x, y)
2
+ 2(x, y) + 1 so this is in 1. The elements with = 0 are the elements with no constant term.
9.2.10. Determine the greatest common divisor of a(x) = x
3
+ 4x
2
+ x 6 and b(x) = x
5
6x + 5 in Q[x] and
write it as a linear combination in Q[x] of a(x) and b(x).
Solution. Using division of polynomials we calculate
x
5
6x + 5 = (x
3
+ 4x
2
x 6)(x
2
4x + 15) (50x
2
+ 45x 95)
x
3
+ 4x
2
+x 6 = (50x
2
+ 45x 95)(x/50 + 31/500) + (31/10 x
2
+ 1395/500 x 2945/500)
50x
2
+ 45x 95 = (31/10 x
2
+ 1395/500 x 2945/500)(500/31).
So we write the greatest common divisor as
31/10 x
2
+ 1395/500 x 2945/500 =
= (x
3
+ 4x +x 6)[(x
2
4x + 15)(x/50 + 31/500) 1] (x
5
6x + 5)[x/50 + 31/500].
= F[x]/(f
1
(x)
n1
) F[x]/(f
2
(x)
n2
) F[x]/(f
k
(x)
n
k
).
Proposition 9.17. If the polynomial f(x) has roots
1
,
2
, . . . ,
k
in F (not necessarily distinct), then f(x) has
(x
1
) (x
k
) as a factor. In particular, a polynomial of degree n in one variable over a eld F has at most
n roots in F, even counted with multiplicity.
Proposition 9.18. A nite subgroup of the multiplicative group of a eld is cyclic. In particular, if F is a nite
eld, then the multiplicative group F
(Z/p
r
r
Z)
(2) (Z/2
Z)
is the direct product of a cyclic group of order 2 and a cyclic group of order 2
2
, for all 2
(3) (Z/p
Z)
i=2
b
i
S(f
i1
, f
i
), for some constants b
i
F.
51
Proposition 9.26 (Buchbergers Criterion). Let R = F[x
1
, . . . , x
n
] and x a monomial ordering on R. If I =
(g
1
, . . . , g
m
) is a nonzero ideal in R, then G = g
1
, . . . , g
m
is a Gr obner basis for I if and only if S(g
i
, g
j
) 0
mod G for 1 i < j m.
Theorem 9.27. Fix a monomial ordering on R = F[x
1
, . . . , x
n
]. Then there is a unique reduced Grobner basis for
every nonzero ideal I in R.
Corollary 9.28. Let I and J be two ideal in F[x
1
, . . . , x
n
]. Then I = J if and only if I and J have the same
reduced Grobner basis with respect to any xed monomial ordering on F[x
1
, . . . , x
n
].
Proposition 9.29 (Elimination). Suppose G = g
1
, . . . , g
m
is a Grobner basis for the nonzero ideal I in F[x
1
, . . . , x
n
]
with respect to the lexicographic monomial ordering x
1
> > x
n
. Then G F[x
i+1
, . . . , x
n
] is a Grobner basis
of the i
th
elimination ideal I
i
= I F[x
i+1
, . . . , x
n
] of I. In particular, I F[x
i+1
, . . . x
n
] = 0 if and only if
G F[x
i+1
, . . . , x
n
] = .
Proposition 9.30. If I and J are any two ideals in F[x
1
, . . . , x
n
] then tI + (1 t)J is an ideal in F[t, x
1
, . . . , x
n
]
and I J = (tI + (1 t)J) F[x
1
, . . . , x
n
]. In particular, I J is the rst elimination ideal of tI + (1 t)J with
respect to the ordering t > x
1
> > x
n
.
52 DAVID S. DUMMIT AND RICHARD M. FOOTE
9.6.24. Use reduced Grobner bases to show that the ideal I = (x
3
yz, yz +y) and the ideal J = (x
3
z +x
3
, x
3
+y)
in F[x, y] are equal.
Proof. By corollary 9.28 two ideals in F[x, y, z] are equal if and only if I, J have the same reduced Grobner basis
for any monomial ordering. So x the lexicographical ordering x > y > z for F[x, y, z]. To nd a Grobner basis for
I,
S(f
1
, f
2
) = yzf
1
x
3
f
2
= y
2
z
2
x
3
y 0 mod f
1
, f
2
.
Since x
3
yz = 1(yz +y) + (x
3
+y), then our reduced Grobner basis for I is G =
x
3
+y, yz +y
.
Now to nd the reduced Grobner basis for J,
S(g
1
, g
2
) = g
1
zg
2
= x
3
yz yz y mod g
1
, g
2
S(g
1
, g
3
) = yg
1
x
3
g
3
= x
3
y +x
3
y 0 mod g
1
, g
2
, g
3
S(g
2
, g
3
) = x
3
g
2
+yzg
3
= x
3
y
2
z 0 mod g
1
, g
2
, g
3
.
Thus g
1
, g
2
, g
3
is a Grobner basis for J. Since LT(g
2
) mod LT(g
1
), G =
x
3
+y, yz +y
xy, x yz, y
2
z
is a Grobner basis so
x yz, y
2
z
f
1
f
2
0 0
but
1 1
1 1
f
1
f
2
0 0
f
1
f
2
f
1
f
2
a
0
+a
1
+a
2
2
+ +a
n1
n1
[ a
0
, a
1
, . . . , a
n1
F
= F[x]/(p(x)).
Corollary 13.7. Suppose in theorem 13.6 that p(x) is of degree n. Then
F() =
a
0
+a
1
+a
2
2
+ +a
n1
n1
[ a
0
, a
1
, . . . , a
n1
F
K.
Theorem 13.8. Let : F F
(x) F
(x) be the irreducible polynomial obtained by applying the map to the coecients of p(x). Let be
a root of p(x) (in some extension of F) and let be a root of p
). Then there is an
isomorphism
: F() F
()
mapping to and extending , i.e., such that restricted to F is the isomorphism .
13.2. Basic Theory of Field Extensions.
Proposition 13.9. Let be algebraic over F. Then there is a unique monic irreducible polynomial m
,F
(x) F[x]
which has as a root. A polynomial f(x) F[x] has as a root if and only if m
,F
(x) divides f(x) in F[x].
Corollary 13.10. If L F is an extension of elds and is algebraic over both F and L, then m
,L
(x) divides
m
,F
(x) in L[x].
Proposition 13.11. Let be algebraic over the eld F and let F() be the eld generated by over F. Then
F()
= F[x]/(m
(x))
so that in particular
[F(): F] = deg m
(x) = deg ,
i.e., the degree of over F is the degree of the extension it generates over F.
Proposition 13.12. The element is algebraic over F if and only if the simple extension F F() is nite.
More precisely, if is an element of an extension of degree n over F then satises a polynomial of degree at most
n over F and if satises a polynomial of degree n over F then the degree of F() over F is at most n.
Corollary 13.13. If the extension F K is nite, then it is algebraic.
55
Theorem 13.14. Let F K L be elds. Then
[L: F] = [L: K][K: F],
i.e., extension degrees are multiplicative, where if one side of the equation is innite, the other side is also innite.
Corollary 13.15. Suppose F L is a nite extension and let K be any subeld of L containing F, F K L.
Then [K: F] divides [L: F].
Lemma 13.16. F(, ) = (F())(), i.e., the eld generated over F by and is the eld generated by over
the eld F() generated by .
Theorem 13.17. The extension F K is nite if and only if K is generated by a nite number of algebraic
elements over F. More precisely, a eld generated over F by a nite4 number of algebraic elements of degrees
n
1
, n
2
, . . . , n
k
is algebraic of degree n
1
n
2
n
k
.
Corollary 13.18. Suppose and are algebraic over F. Then , , / (for = 0), (in particular
1
for = 0) are all algebraic.
Corollary 13.19. Let F L be an arbitrary extension. Then the collection of elements of L that are algebraic
over F form a subeld K of L.
Theorem 13.20. If K is algebraic over F and L is algebraic over K, then L is algebraic over F.
Proposition 13.21. Let K
1
and K
2
be two nite extensions of a eld F contained in K. Then
[K
1
K
2
: F] [K
1
: F][K
2
: F]
with equality if and only if an Fbasis for one of the elds remains linearly independent over the other eld.
If
1
,
2
, . . . ,
n
and
1
,
2
, . . . ,
m
are bases for K
1
and K
2
over F, respectively, then the elements
i
j
for
i = 1, . . . , n and j = 1, . . . , m span K
1
K
2
over F.
Corollary 13.22. Suppose that [K
1
: F] = n, [K
2
: F] = m in proposition 13.21, where n and m are relatively
prime: (n, m) = 1. Then [K
1
K
2
: F] = [K
1
: F][K
2
: F] = nm.
13.2.1. Let F be a nite eld of characteristic p. Prove that [F[ = p
n
for some positive integer n.
Proof. Let : Z F be the ring map given by (n) = n 1
F
. Since pZ ker then : F
p
F is well dened. But
is a ring map of elds so it is trivial or injective. Notice (1
Fp
) = 1
F
so is injective. Thus we can consider F as
a vector space over F
p
. Since F is nite then
[F[ = [F
p
[
[F: Fp]
= p
n
for some n Z
+
.
13.2.7. Prove that Q(
2+
3) = Q(
2,
2+
2 +
3.
Proof. Since Q,
2,
3 are contained in Q(
2,
3), then Q,
2 +
3 are contained in Q(
2,
3). Hence by
denition Q(
2 +
3) Q(
2,
3).
To prove the other inclusion notice that
2 +
3)
3
9(
2 +
3)
/2 =
2. Thus
2 Q(
2 +
3) thus
3 Q(
2 +
2,
3) Q(
2 +
2,
3) = Q(
2 +
3) then [Q(
2 +
3): Q] = [Q(
2,
2 +
3) = 0
where
p(x) = x
4
10x
2
+ 1.
If p were to be reducible, then there would be an irreducible polynomial with degree less than 4. This contradicts
the minimality of the degree of m
2+
3
= 4. Therefore p is irreducible in Q.
13.2.10. Determine the degree of the extension Q(
3 + 2
2) over Q.
Solution. First note that [Q(
3 + 2
2): Q] = [Q(
3 + 2
2: Q(
2)][Q(
2 = (x 1 +
2
2)(x + 1 + 2
2) so [Q(
3 + 2
2): Q(
3 + 2
2): Q] = 2.
13.2.13. Suppose F = Q(
1
, . . . ,
n
) where
2
i
Q for i = 1, . . . , n. Prove that
3
2 / F.
Proof. Since [Q(
1
, . . . ,
k
): Q(
1
, . . . ,
k1
)] 1, 2, then [Q(
1
, . . . ,
n
): Q] = 2
l
for some l Z
+
. Now if
3
2 F then Q Q(
3
2) F then [Q(
3
2 / F.
56 DAVID S. DUMMIT AND RICHARD M. FOOTE
13.2.16. Let F K be an algebraic extension and let R be a ring contained in K and containing F. Show that
R is a subeld of K containing F.
Proof. Let R` 0 so since F then there is some irreducible p R[x] such that p() = 0. Let p(x) be given
by
p(x) = a
n
x
n
+ +a
0
,
and notice since p is irreducible, a
0
= 0. Since p() = 0 we have
1
= a
1
0
(a
n
n1
+ +a
1
).
Since a
i
F R and R then
1
R and R is a eld.
13.3. Classical Straightedge and Compass Constructions.
Proposition 13.23. If the element 1 is obtained from a eld F 1 by a series of compass and straightedge
constructions then [F(): F] = 2
k
for some integer k 0.
Theorem 13.24. None of the classical Greek problems: (I) Doubling the Cube, (II) Trisecting an Angle, and (III)
Squaring the Circle, is possible.
13.4. Splitting Fields and Algebraic Closures.
Theorem 13.25. For any eld F, if f(x) F[x] then there exists an extension K of F which is a splitting eld
for f(x).
Proposition 13.26. A splitting eld of a polynomial of degree n over F is of degree at most n! over F.
Theorem 13.27. Let : F F
(x) F
[x]
be the polynomial obtained by applying to the coecients of f(x). Let E be a splitting eld for f(x) over F and
let E
(x) over F
, i.e.,
restricted to F is the isomorphism .
Corollary 13.28 (Uniqueness of Splitting Fields). Any two splitting elds for a polynomial f(x) F[x] over a
eld F are isomorphic.
Proposition 13.29. Let F be an algebraic closure of F. Then F is algebraically closed.
Proposition 13.30. For any eld F there exists an algebraically closed eld K containing F.
Proposition 13.31. Let K be an algebraically closed eld and let F be a subeld of K. Then the collection of
elements F of K that are algebraic over F is an algebraic closure of F. An algebraic closure of F is unique up to
isomorphism.
Theorem (Fundamental Theorem of Algebra). The eld C is algebraically closed.
Corollary 13.32. The eld C contains an algebraic closure for any of its subelds. In particular, Q, the collection
of complex numbers algebraic over Q, is an algebraic closure of Q.
13.5. Separable and Inseparable Extensions.
Proposition 13.33. A polynomial f(x) has a multpile root if and only if is also a root of D
x
f(x), i.e., f(x)
and D
x
f(x) are both divisible by the minimal polynomial for . In particular, f(x) is separable if and only if it is
relatively prime to its derivative: (f(x), D
x
f(x)) = 1.
Corollary 13.34. Every irreducible polynomial over a eld of characteristic 0 is separable. A polynomial over such
a eld is separable if and only if it is the product of distinct irreducible polynomials.
Proposition 13.35. Let F be a eld of characteristic p. Then for any a, b F,
(a +p)
p
= a
p
+b
p
, and (ab)
p
= a
p
b
p
.
Put another way, the p
th
power map dened by (a) = a
p
is an injective eld homomorphism from F to F.
Corollary 13.36. Suppose that F is a nite eld of characteristic p. Then every element of F is a p
th
power in F,
i.e., F = F
p
.
Proposition 13.37. Every irreducible polynomial over a nite eld F is separable. A polynomial in F[x] is separable
if and only if it is the product of distinct irreducible polynomials in F[x].
57
Proposition 13.38. Let p(x) be an irreducible polynomial over a eld F of characteristic p. Then there is a unique
integer k 0 and a unique irreducible separable polynomial p
sep
(x) F[x] such that
p(x) = p
sep
(x
p
k
).
Corollary 13.39. Every nite extension of a perfect eld is separable. In particular, every nite extension of either
Q or a nite eld is separable.
13.6. Cyclotomic Polynomials and Extensions.
Lemma 13.40. The cyclotomic polynomial
n
(x) is a monic polynomial in Z[x] of degree (n).
Theorem 13.41. The cyclotomic polynomial
n
(x) is an irreducible monic polynomial in Z[x] of degree (n).
Corollary 13.42. The degree over Q of the cyclotomic eld of n
th
roots of unity is (n):
[Q(
n
): Q] = (n).
14. Galois Theory
14.1. Basic Denitions.
Proposition 14.1. Aut(K) is a group under composition and Aut(K/F) is a subgroup.
Proposition 14.2. Let K/F be a eld extension and let K be algebraic over F. Then for any Aut(K/F),
is a root of the minimal polynomial for over F i.e., Aut(K/F) permutes the roots of irreducible polynomials.
Equivalently, any polynomial with coecients in F having as a root also has as a root.
Proposition 14.3. Let H Aut(K) be a subgroup of the group of automorphisms of K. Then the collection F of
elements of K xed by all the elements of H is a subeld of K.
Proposition 14.4. The association of groups to elds and elds to groups dened above is inclusion reversing,
namely
(1) if F
1
F
2
K are two subelds of K then Aut(K/F
2
) Aut(K/F
1
), and
(2) if H
1
H
2
Aut(K) are two subgroups of automorphisms with associated xed elds F
1
and F
2
, respec
tively, then F
2
F
1
.
Proposition 14.5. Let E be the splitting eld over F of the polynomial f(x) F[x]. Then
[Aut(E/F)[ [E: F]
with equality if f(x) is separable over F.
Corollary 14.6. If K is the splitting eld over F of a separable polynomial f(x) then K/F is Galois.
14.2. The Fundamental Theorem of Galois Theory.
Theorem 14.7 (Linear Independence of Characters). If
1
, . . . ,
n
are distinct characters of G with values in L
then they are linearly independent over L.
Corollary 14.8. If
1
, . . . ,
n
are distinct embeddings of a eld K into a eld L, then they are linearly independent
as functions on K. In particular distinct automorphisms of a eld K are linearly independent as functions on K.
Theorem 14.9. Let G =
1
= 1,
2
, . . . ,
n
be a subgroup of automorphisms of a eld K and let F be the xed
eld. Then
[K: F] = n = [G[ .
Corollary 14.10. Let K/F be any nite extension. Then
[Aut(K/F)[ [K: F]
with equality if and only if F is the xed eld of Aut(K/F). Put another way, K/F is Galois if and only if F is
the xed eld of Aut(K/F).
Corollary 14.11. Let G be a nite subgroup of automorphisms of a eld K and let F be the xed eld. Then every
automorphism of K xing F is contained in G, i.e., Aut(K/F) = G, so that K/F is Galois, with Galois group G.
Corollary 14.12. If G
1
= G
2
are distinct nite subgroups of automorphisms of a eld K then their xed elds are
also distinct.
Theorem 14.13. The extension K/F is Galois if and only if K is the splitting eld of some separable polynomial
over F. Furthermore, if this is the case then every irreducible polynomial with coeecients in F which has a root
in K is separable and has all its roots in K (so in particular K/F is a separable extension).
Theorem 14.14 (Fundamental Theorem of Galois Theory).
58 DAVID S. DUMMIT AND RICHARD M. FOOTE
15. Commutative Rings and Algebraic Geometry
15.1. Noetherian Rings and Ane Algebraic Sets.
15.2. Radicals and Ane Varieties.
15.3. Integral Extensions and Hilberts Nullstellensatz.
15.4. Localization.
Theorem 15.36 (cf. theorem 7.15). let R be a commutative ring with 1 = 0 and let D be a multiplicatively closed
subset of R containing 1. Then there is a commutative ring D
1
R and a ring homomorphism : R D
1
R
satisfying the following universal property: for any homomorphism : R S of commutative rings that sends 1
to 1 such that (d) is a unit in S for every d D, there is a unique homomorphism : D
1
R S such that
= .
Corollary 15.37. In the notation of theorem 15.36,
(1) ker = r R [ xr = 0 for some x D; in particular, : R D
1
R is an injetion if and only if D
contains neither zero nor any zero divisors of R, and
(2) D
1
R = 0 if and only if 0 D, hence if and only if D contains nilpotent elements.
15.4.18. Let R be any commutative ring with 1 and let f be any element of R. Let D be the multiplicative set
f
n
[ n 0 of nonnegative powers of f in R. Dene R
f
= D
1
R. Note that R
f
= 0 if and only if f is nilpotent.
If f is not nilpotent, then f becomes a unit in R
f
. Prove that R
f
= R[x]/(fx 1) where R[x] is the polynomial
ring in the variable x, if f is not nilpotent in R.
Proof.
15.4.21. Suppose : R S is a ring homomorphism with (1
R
) = 1
S
and D
: D
1
R D
1
S given
by
) because D
is a multiplicative set.
Let
r1
d1
=
r2
d2
. Then r
1
d
2
= r
2
d
1
so
(
r
1
d
1
) =
(
r
1
r
2
d
2
r
2
d
1
d
2
)
=
(r
1
r
2
d
2
)
(r
2
d
1
d
2
)
=
(r
2
d
1
)(r
2
)
(r
2
d
1
)(d
2
)
=
(
r
2
d
2
)
59
So
(
r
1
d
1
+
r
2
d
2
) =
(
r
1
d
2
+r
2
d
1
d
1
d
2
)
=
(r
1
d
2
+r
2
d
1
)
(d
1
d
2
)
=
(r
1
)
(d
1
)
+
(r
2
)
(d
2
)
=
(
r
1
d
1
) +
(
r
2
d
2
)
(
r
1
d
1
r
2
d
2
) =
(
r
1
r
2
d
1
d
2
)
=
(r
1
r
2
)
(d
1
d
2
)
=
(r
1
)
(d
1
)
(r
2
)
(d
2
)
=
(
r
1
d
1
)
(
r
2
d
2
).