Вы находитесь на странице: 1из 141

Solutions

to

A First Course in
Abstract Algebra
John B. Fraleigh
sixth edition
ISBN 0-201-33596-4
Addison Wesley Longman
by
Ben Hekster
PO Box 391852
Mountain View, CA 94039-1852
heksterb@acm.org
http://www.hekster.org/Academic/Mathematics/
These are completely unofficial and unverified worked solutions by me. Corrections welcome.
Typeset in Galliard and Gill Sans using AppleWorks 5 on Apple Macintosh.
Copyright 1999-2005 All Rights Reserved by: Ben Hekster.

Abstract Algebras
set
+ binary operation

binary algebraic structure


+ associative binary operation

semigroup
+ identity

monoid
+ inverse

group
+ finite order

+ commutative operation

finite group

commutative group

+ prime order

symmetric group

+ generator

cyclic group
+ prime order

alternating group

prime order group

Glossary
:
+ i , i

reads as so that
summation, multiplication over i

(i {i
, , <, >
, , ,

ordered, unordered set over i


scalar operators
set operators

=n
<

congruent modulo n
is normal to, is ideal to

fx

function application f x

commutative group
maximal p-group

abelian group
Sylow p-group

()

0.1 Preliminaries
1.
2.
3.
4.
5.

6.
7.

8.
9.
10.
11.

12.
13.
14.

15.
16.
17.
18.

proving theorems
set
precision?
definition
A triangle with vertices P, Q, R is the collection of points X such that
X is in the line segment PQ, or
X is in the line segment QR, or
X is in the line segment RP.
An equilateral triangle is a triangle with vertices P, Q, R such that the length of the line segment PQ equals both
the length of the line segment QR and the length of the line segment RP.
A right triangle is a triangle with vertices P, Q, R in which the two line segments through one of its vertices (say PQ
and PR) are such, that for any point X on PQ there is no point Y on PR such that the length of the line segment XY
is less than the length of the line segment XP.
The interior of a triangle is the collection of points X such that the line segments XP, XQ, XR from X to its vertices
P, Q, R have only the vertices in common with the triangle.
A circle with center C and radius r is the collection of points X such that the length of the line segment XC equals
r.
A disk with center C and radius r is the collection of points X such that the length of the line segment XC is less
than or equal to r.
Define the relationship between PQ and PR in 7. to be a right angle. Then, a rectangle with vertices P, Q, R, S is
the collection of points formed by the four line segments PQ, QR, RS, SP, where PQ is at a right angle to QR, QR to
RS, RS to SP, and SP to PQ.
Let n and m be even integers. Then by (2), there are integers p, q such that n = 2p, m = 2q. Then n + m = 2p + 2q
= 2(p + q), so n + m is even.
Let n, m, p, q as in 12. Then nm = 2p 2q = 4pq. Since pq is an integer, 4pq is an integral multiple of 4.
Define an odd integer m to be an integer such that there exists another integer n such that m = 2n + 1.
Let r be an even integer and s an odd integer. Then there are integers p, q such that r = 2p, s = 2q + 1. So r + s =
2p + 2q + 1 = 2(p + q) + 1, so r + s is odd.
counterexample
A B F G M, C D J, E H K N, I, L, O.
1, 2, 4, 8, 16, 31 (the conjecture is false).
Suppose that i is the square of an odd integer k. Then
i  : k = 2l + 1 i = k 2 = (2l + 1) 2 = 4l 2 + 4l + 1
Since i is also even,
j  :

i = 2j

4l 2 + 4l + 1 = 2 j

2l 2 + 2l +

1
2

= j 

which is a contradiction, so k cannot be odd. Since k must be even,


l  : k = 2l i = k 2 = (2l) 2 = 4l 2
so i is indeed an integral multiple of 4.
19.

Let n = 0, then (n + 3) 2 = 3 2 = 9 >/ 9 .

20.

Let n 2 + 2 = 3

21.

Let n = 2

n 2 + 4 = 22 + 4 = 8 .

22.

Let n = 3

n 2 + 5 = 3 2 + 5 = 14 .

23.

Let n = 3

24.

Let n = 0:

25.

Let n ,n < 0

26.

Let x =

1
2

n2 = 1

n 2 + 5 = (3) 2 + 5 = 14 . With 22., n is not unique.

n2 > n

n = 1 n = +1, so n is not unique.

0 2 > 0 , which is a contradiction.


n2 > 0

x2 < x

()
1
2

n2 > 0 >n .
<

1
2

1
4

< 1 , which is a contradiction.


2

n2 > n

22 > 2

27.

Let n = 2:

28.

n = 0: n 2 = n 0 2 = 0
Let
2
2
n = 1: n = n 1 = 1

29.

4 > 2.
0=0
1=1

, so x is not unique.

Let j be an odd integer, so


k  : j = 2k + 1
j

= (2k + 1) 2

= 4k 2 + 4k + 1
= k 2 + k 
m  :

31.

Let n = 2:

32.

Let n = 2, m = 1:

33.
34.

n = 3m + 1 , so n 2 = (3m + 1) 2 = 9m 2 + 6m + 1 = 3 3m 2 + 2m + 1, and 3m 2 + 2m is integral.

30.

n3 < n

(2)3 < 2

( ) =( )
n
m

2
1

8 < 2 .

= (2) 2 = 4 </ 1 .

( ) < (m = 0) n < nm (n < 0) n m n </ m .

mn (n 0) n m n < m
. So let m = 1 and n = 2:
( ) ( ) mm 00:: nn mn
(n 0) n m n </ m
n
m
n
m

n
m

n
m

( ) ( )
n
m

n
m

( ) ( )
2
1

2
1

8 4 , which is a contradiction.

0.2 Sets and Relations


17. An equivalence relation extracts a property from the whole identity of its arguments and asserts the equality of just
this property: equivalence is property equality. For example, congruence modulo asserts equality of the
remainder under division.
1.

{x  | x 2 = 3} = { 3 , + 3}

2.

{m  | m 2 = 3} =

3.
4.

{m  | mn = 60 for some

{m  | m

m < 115 . Solve the inequality:

m 2 m = 115
m=

n } = {1, 2, 3, 4, 5, 6, 10, 12, 15, 20, 30}

m 2 m 115 = 0

+1 (1) 2 4 1 115

=
2 1
= 1 (1 461) 10.2, 11.2
2

5.
6.
7.
8.
9.
10.
11.
12.

1 1 + 460
2
1
(1
2

461)

1
(1 +
2

461)

so m {10, 9, , 10, 11}.


not a set




m
2

| m 

{(a,1), (a,2), (a,c), (b,1), (b,2), (b,c), (c,1), (c,2), (c,c)}


function
one-to-one
onto
a.
yes
no
no
b.
yes
no
no
c.
no
3

13.

d.
yes
e.
yes
f.
no
Map x to y(x).
P

yes
no

yes
no

B
x

D
y(x)

14

a. f : [0, 1] [0, 2]: x a 2x


b. f : [1, 3] [5, 25]: x a (x 1) 20 + 5
2

d c
c. f : [a , b] [c , d]: x a (x a)
+c
b a
15.
16.

f : S : x a tan x 12
a. P () = ,

P () = 1

b. P ({a}) = ,{a} ,

P ({a}) = 2

c. P ({a , b}) = , {a}, {b}, {a , b} ,

P ({a , b}) = 4

d. P ({a , b , c}) = , {a}, {b}, {a , b}, {c}, {a , c}, {b , c}, {a , b , c} ,


17.

P ({a , b , c}) = 8

Conjecture P (A) = 2 .
Let An be a series of sets such that |An| = n, and An An +1 .

P (A0 ) = P () = 1.

Let P (An ) = 2

An

There is sn + 1 An such that An + 1 = An sn + 1 . Consider the set


An =

P An

P P {sn + 1}

Since every element of An is a subset of An +1, An P (An + 1) .


Every subset P of An +1 either does or does not contain sn +1 :
sn + 1 P
sn + 1 P

P An P An
P \ {sn + 1} An P An

so P (An + 1) An .
So P (An + 1) = An , and P (An + 1) = 2 P (An ) = 2 2

An

=2

An + 1

=2

An +1

18.

Let f : A B BA .

For each subset P A , there is a corresponding function


a P: 0
f P : A B: a a
a P : 1
Let there be two such subsets P , P A such that f P = f P . Then a A:
a P f P (a) = f P (a) = 1
so P = P .

a P ; a P a P
4

Conversely, for each function f BA there is a corresponding subset Pf A :

Pf = a A | f (a) = 1 .
Let there be two functions f , f BA such that Pf = Pf . Then a A:
f (a) = 0

f (a) = 1

a Pf

a Pf

f (a) = 0

f (a) = 1

so f = f .
So, Pf : BA P (A) is a bijection, and B A = P (A) .
19.

For every element of A there is a distinct singleton subset containing just that element, which is an element of P (A)
. is not such a singleton set, yet is an element of P (A) . So P (A) > A .

Let A be such that A = . Then the power set of A has P (A) >, and P P (A) > P (A) , ad infinitum.
20.

a. It is possible to define addition in  in terms of the union of disjoint sets, so


2+3 = 5

A = 2, B = 3, A B = 5 .
(*)

i. 3 + 0 = {0}  + =  + = 0 , where (*) :

:  + {0}  + : m a m 1 .

m odd :
ii. 0 + 0 =   =  = 0 , where (*) :    : m a
m even :
b. It is possible to define multiplication in  in terms of a Cartesian product:

1
2

(*)

1
2

1
m
2
1
m
2

1) +

1
2

2 3 = {1, 2} {1, 2, 3} = 6 , so
0 0 =  +  +
21.

fig 14

 + = 0 .

102 digits, 105 digits. By extrapolation, 100 would equal the number of digits of the form 0.###, where # is
repeated 0 times name this set R. Since any number in R = {x  | 0 x < 1} can be expressed arbitrarily
precise by an element of R, R R. Since R R , R = R . By Exercise 15, R = , so R = and 100 = .

22.

Similar arguments can be made in terms of duodecimal and binary expansions of numbers of R , so 120 = 20 = .
Since
(17)

P() = 2

(18)

= 20 = ;

(19)

P() = {0, 1}

The next higher cardinals after 0 are


= {0, 1} = {0, 1} exp ,


{0, 1}{0 , 1}
28.

= {0, 1} exp {0, 1} exp 

et cetera.
x R y i: x , y Pi y , x Pi y R x (symmetric)
x R x i: x Pi x S (reflexive)
x R y y R z i: x , y Pi j : y , z Pj

29.
30.

(Pi disjoint i = j ) i: x , y , z Pi (transitive).


y Rz
not reflexive because 0 0 >/ 0 0 R/ 0
not symmetric because 2 1, 1 / 2 2 R 1, 1 R/ 2 .
5

31.

R is a relation, because
y

x = x

x = y y = x

x = y y = x x = z
32.
33.

x R x

x R y y R x
x R y y R z x R z

0 3 = 3 3,

3 6 = 3 3,

0 R 3,

0 R/ 6

3 R 6,

0 6 = 6 / 3

R
x
.

so R is not transitive.

+
The number of digits of n  is base 10 notation is 1 +

transitive.

10

log n . Obviously R is reflexive and symmetric, and

R
0
34.
35.

10

100

R is congruence modulo 10 on  *.
a. {1, 3, 5, }, {2, 4, 6, }
b. {1, 4, 7, }, {2, 5, 8, }, {3, 6, 9, }
c. {1, 6, 11, }, {2, 7, 12, }, {3, 8, 13, }, {4, 9, 14, }, {5, 10, 15, }

36a.
r  : r r = 0 = 0 n r ~ r
r, s ,r ~ s : q  : r s = qn s r = (qn ) = ( q )n s ~ r
r, s ,t,r ~ s , s ~ t :

b.

p ,q  :

r s = pn , s t = qn
r s + s t = pn + qn
r t = (p + q)n r ~ t
r s r s
r, s  + ,r ~ s : q  : r s = qn (n )
= =q
n
n n
rn , sn , rn, s n  : r = r n n + rn, s = sn n + s n , 0 r n, s n < n
r s = qn
r n n + rn s n n s n = qn

(r n sn )n + (rn s n ) = qn
(r n sn )

r n sn +

r s
Since rn , sn  , q , n n = 0
n

=q

r n sn
<1
n

rn s n .

c.
{, 2, 1, 3, }, {, 2, 0, 2, }
{, 2, 1, 4, }, {, 1, 2, 5, }, {, 3, 0, 3, }
{, 4, 1, 6, }, {, 3, 2, 7, }, {, 2, 3, 8, }, {, 1, 4, 9, }, {, 5, 0, 5, }

0.3 Mathematical Induction


1.

Prove that

+ i2 =

i = 1n

n = 1: 12 =

( )(

)(

).

n n + 1 2n + 1
6

) = 23 = 1

1 1 + 1 2 1 + 1
6

n + 1:

+ i = + i + n +1

i n + 1

i n

) +n

)(

n n + 1 2n + 1

+ 2n + 1 =

)(

)(

n n + 1 2n + 1 + 6 n 2 + 2n + 1

)(

= n 2n + 3n + 1 + 6n + 12n + 6 = = n + 1 n + 2 2n + 3
n 2(n + 1 )
Prove that + i =
, n  + .
i =1n
4
2

2.

n = 1: 1 =
n + 1:

( )

12 1 + 1

4
3

1 22
=1
4

+ i = + i + n +1

1n + 1

i n
2

) + (n + 1)(n + 1)

n2 n + 1

) ( )(

)=n

n n + 2n + 1 + 4 n + 1 n 2 + 2n + 1

)(
2

+ 2n3 + n 2 + 4n3 + 8n 2 + 4n + 4n 2 + 8n + 4
4

n +1 n + 2
n + 6n + 13n + 12n + 4
=
==
4
4
4

3.

(2i 1) = n .
2

Prove that

i = 1n

n = 1: 1 = 12
n + 1:
4.

(2i 1) =

i = 1n + 1

i = i n

(2i 1) + 2(n + 1) 1 = n

n + 1:

( )

1 1+1

1
2

1
=
1+1

n
1
+
n +1 n +1 n + 2

( ) (n + 1)(n + 2)
( )( )
n(n + 2) + 1
(n + 1) = n + 1
n + 2n + 1
=
=
=
(n + 1)(n + 2) (n + 1)(n + 2) (n + 1)(n + 2) n + 2
a (1 r
).
Prove that a,r , r 1, n  : + ar =
i = 1n + 1 i

(i + 1)

1
2

i = 1n

i i +1

n + 1:

a + ar =
n +1

1 r

1 r

) = a(1 r )(1 + r ) = a(1 r )


1 r

i =0

=a

7.

+ ar = + ar + ar

i =0

i =0

a 1 r

n +1

n = 1:

6.

+ 2n + 1 = n + 1

1
n
Prove that +
=
,n  + .
i =1n i (i + 1 )
n +1
n = 1:

5.

1 r

n +1

n +1

a 1 r n +1

+ 1 r r
1 r

1 r

n +1

=a

) + ar

n +1

) ( )

a 1 r n + 1 + 1 r ar n + 1
1 r

n+2
1 r n +1 + r n +1 r n + 2 a 1 r
=
1 r
1 r

max is only defined on  , so max(i 1, j 1) is undefined.


the concept interesting property is not well defined

0.4 Complex and Matrix Algebra


1.

(2 + 3i ) + (4 + 5i ) = 6 + 2i .

2.

i + 5 3i = 5 2i .
7

3.
4.

(5 + 7i ) (3 2i ) = 2 + 5i .
(1 3i ) (4 + 2i ) = 5 5i .

5.

i 3 = ii 2 = i .

6.

i 4 = i 2 i 2 = 1 1 = +1.

7.

i 23 = i 20i 3 = (i 4 )5 i 3 = 15 i 3 = i .

8.
9.
10.
11.
12.

(i ) = i = (i i ) = (1 i ) = i .
(4 i )(5 + 3i ) = 20 + 12i 5i 3i = 23 + 7i .
(8 + 2i )(3 i ) = 24 + 6i 8i 2i = 26 2i .
(2 3i )(4 + i ) + (6 5i ) = 8 + 2i 12i 3i + 6 5i = 17 15i .
(1 + i ) = (1 + i )(1 + i ) = (1 + i )(1 + 2i + i ) = (1 + i )2i = 2i + 2i
7 5i (7 5i )(1 6i ) 7 42i 5i + 30i
=
= ( 23 47i ) .
=
1 + 6i (1 + 6i )(1 6i )
1 36
i (1 i )
i i
i +1
1
=
=
.
=
1 + i (1 + i )(1 i ) 1 i
2
1 i (1 i )i
= i i = 1 i .
=
35

35

32 3

= 2 + 2i .

14.

1
35

15.

16.
17.

i2

) = 1 i (3 + i )(1 + 2i ) = (3i 1)(1 + 2i ) = 1 + 3i 2i + 6i = (7 + i ) .


(
)
2 4i
2 (1 2i )(1 + 2i )
1 4i
(3 + 7i )(1 i )(2 + 3i ) = (3 3i + 7i 7i )(2 + 3i )
3 + 7i
=
(1 + i )(2 3i ) (1 + i )(1 i )(2 3i )(2 + 3i )
(1 i )(4 9i )
.
(10 + 4i )(2 + 3i ) = (5 + 2i )(2 + 3i ) = 10 + 15i + 4i + 6i = 4 + 19i
=
i 3+i

1
2

1
10

1
10

18.

2 13

13

13

19.

(1 i )(2 + i ) = (1 i ) (2 + i )(1 + 2i ) = (1 2i + i )(2 + 4i + i + 2i


(1 2i )(1 + i ) (1 2i )(1 + 2i )(1 + i )(1 i )
(1 4i )(1 i )

20.

3 4i = 5 .

21.

6 + 4i = 23 + 2i = 2 9 + 4 = 2 13 .

22.

3 4i = 5 3 4i = 5

23.

1 + i = 2

24.

12 + 5i = 144 + 25 = 169 = 13

25.

3 + 5i = 9 + 25 = 36 = 6

26.

3
5

13

) = 2i 5i = i
5 2

= 1.

4i .
5

1
1
1
1

1 + i = 2
+i
= 2 2 2 + 2 i 2 .

2
2

12 + 5i = 13

12
13

5
i
13

).

3 + 5i = 6 1 + 5 i .
2

r1 i ( 1 2 )
e
. So z 1 z 2 is the point in the complex
r2
point at the end of a line from the origin with length r1 r2 and angle 1 2 from the positive x-axis.
z 1 = r1e i 1 , z 2 = r2e i 2

z 1 z 2 = z 1z 21 = r1e i 1 r2e i 2

27.
28.

( ) = 1e
(re ) = 1e
4

z 4 = 1 re i

z 4 = 1

r = 1, 4 = 2 0

0i

1
2

1.1 Binary Operations


1.
2.

3.
4.

(( ) )

2 +i

2,

1
2

r = 1, = 1
2

1
2

2 1 i 2,
2

z 1, i , 1, i .

r = 4 1, 4 = 2 i

2 + 1 i 2,

1
2

r = 1, = 1 0

1
2

2 1i 2
2

( )

b d = e, c c = b , a c e a = c e a = a a = a .

(a b ) c = b c = a , so * could be, but is not necessarily, associative.


a (b c ) = a a = a
(b d ) c = e c = a , so * is not associative.
b (d c ) = b b = c

no, because e b b e .

5.

*
a
b
c
d

a
a
b
c
d

b
b
d
a
c

c
c
a
d
b

d
d
c
b
a

6.

*
a
b
c
d

a
a
b
c
d

b
b
a
d
c

c
c
c
c
c

d
d
d
d
d

( )
( )
d b = (c b ) b = c (b b ) = c a = c ,
d c = (c b ) c = c (b c ) = c c = c ,
d d = (c b ) d = c (b d ) = c d = d .
1 0 = 1 0 = 1, 0 1 = 0 1 = 1 , (a b ) c = (a b ) c = a b c , a (b c ) = a (b c ) = a b + c , so * is neither
d a = c b a = c b a = c b = d ,

7.

commutative nor associative.


8.

Let a,b  :

is commutative, but not associative.


9.

a,b  : a b = 12 ab = 12 ba = b a , a,b ,c  :
commutative and associative.

10.

Let a,b  + :
1 2 = 12 = 1;

( )

12.

1;

2 2 = 2 4 = 16;

13.

1;

22

( )

33 = 39 = 19683;

( ) = 21 = 2;

2 2 1

32

32

= 20 = 1 and 0 0 1 = 2 2
3

00

= 21 = 2 , so * is

= 26 , so * is neither commutative nor

nn ; the table defining * has n 2 entries, each having n possible values.

( ) = 33 = 27;

3 3 1

01

) ( ) = 2 ; (2 3) 2 = (2 )

2 1 = 21 = 2 and 2 3 2 = 2

associative.
2

(a b ) c = 12 ( 12 ab )c = 12 a ( 12 bc ) = a (b c) , so * is

a b = 2 ab = 2 ba = b a , then 0 0 1 = 20 2

commutative, but not associative.


11.

a b = ab + 1 = ba + 1 = b a , 0 0 1 = 0 0 + 1 1 + 1 = 2 , 0 0 1 = 0 0 1 + 1 + 1 = 1, so *

n2

( ) ; the table defining commutative * has

n n 1

(n 1) entries, each

1
n
2

having n possible values.


9

14.
15.
16.

A binary operation on a set S is commutative if and only if for all a , b S:


well defined
Correct the last part to read a , b H .

17.
18.
19.
20.
21.

C1 good; 1 2 = 1 2 = 1 Z + , so C2 is ill defined.


C1, C2 good.
C1, C2 good.
C1, C2 good.
C1 is not well defined, C2 is good.

22.

C1 good; 1 1 = 0  + , so C2 is ill defined.

23.

a
M =
b
a
a.
b

24.
25.

26.
27.
28.

c
b
, N =
a
d

b c
+
a d

a b = b a .

d
H :
c

d a + c
=
c b d

b d
H
a +c

a b c d a c b d
a d b c ac + bd ad bc
b.

=
=
H
b a d c b c + a d b d + a c ad bc ac + bd
a. false; b. true; c. false; d. false; e. false; f. true; g. true; h. true; i. true; j. false.
Let * be addition and *' subtraction on the set of colors {K, R, G, B, C, M, Y, W} (black, red, green, blue, cyan,
magenta, yellow, and white).
+ K R G B C MY W - K R G B C MY W
K K
K K K K K K K K K
R R R
R R K R R R K K K
G G Y G
G G G K G K G K K
B B MC B
B B G B K K K B K
C C WC C C
C C C B G K G B K
MMMWMWM
MMB MR R K B K
Y Y Y Y WWWY
Y Y Y R Y R G K K
WWWWWWWWW WWWWWWWWW

(a b ) (c d ) = (c d ) (a b ) = (d c ) (a b ) = ((d c ) a ) b
Let S be a set with single element s.
A binary operation * on S always maps its operands to s, so * must be associative and commutative.
Let * be the binary operation defined by the table. Then

(b a ) a = a a = b and b (a a ) = b b = a , so * is not associative.


* a b
a b a
b a a

29.

Let f , g , h F . Then x :

30.

Let f :  : x a 0,

31.

so is not commutative on F.
Let f :  : x a 1, then

(( f + g ) + h )(x ) = ( f + g )(x ) + h (x ) = f (x ) + g (x ) + h (x ) , so + is associative on F.


= f ( x ) + ( g + h )( x ) = ( f + ( g + h ))( x )
g :   : x a 1 , then

( f g )(0) = f (0) g (0) = 0 1 = 1


( g f )(0) = g (0) f (0) = 1 0 = 1

10

(( f f ) f )(0) = ( f f )(0) f (0) = ( f (0) f (0)) f (0) = (1 1) 1 = 1


( f ( f f ))(0) = f (0) ( f f )(0) = f (0) ( f (0) f (0)) = 1 (1 1) = 1
so is not associative on F.

( fg )(x ) = f (x ) g (x ) = g (x ) f (x ) = ( gf )(x ) , so multiplication is commutative on F.

32.

f , g F , x  :

33.

For f , g F , x ,

(( fg )h )(x ) = ( fg )(x ) h (x ) = ( f (x ) g (x )) h (x ) = f (x ) ( g (x ) h (x )) = f (x ) ( gh )(x ) = ( f ( gh ))(x ) ,

so multiplication is associative on F.
g :   : x a x 2 , then

34.

Let f :  : x a x + 1,

35.

so concatenation is not commutative on F.


Let = +, = , S = . Then

( f o g )(1) = f ( g (1)) = f (1) = 2


( g o f )(1) = g ( f (1)) = g (2) = 4
(

) ( )
(1 0) (1 3) = (1 + 0) (1 + 3) = 1 4 = 4

1 0 3 = 1 + 0 3 = 1 + 0 = 1

36.

so the property does not hold.


For h , h H :

(h h ) x

associative

h hx

h commutative

(h x ) h

associative

( x h ) h

associative

h h x

37.

For a , b H :

38.
39.
40.

so a b H .
(deposit deposit) talk (deposit press press) = (deposit deposit deposit press) talk (press)
( doesn't affect whatever symbol is next on input.
a,c
s1

(a b ) (a b ) = (a b ) (b a ) = ((a b ) b ) a = (a (b b )) a
= (a b ) a = (b a ) a = b (a a ) = b a = a b

a,c

b
s0

a,c

s2

41.
a,b
s0

a,b
s1

a,b,c
s2

42.
a,b
s0

a,b
s1

a,b
s2

a,b
s3

a,b,c
s4

11

43.
0

s1
0

1
s0
1

s2

44.

45.

s0
s1
s2

0
s0
s0
s0

1
s1
s2
s2

s0
s1
s2

a
s0
s1
s2

b
s0
s1
s2

c
s1
s2
s2

1.2 Isomorphic Binary Structures


1.

is a surjection and injection (bijection)

( )

s , t S: s t = s t .
2.

m  :

n  :

n1 , n2  : n1 = n2

n = n = ( m ) = m (surjection)

n = m

n1 = n2

) (

n1 = n2 (injection)

) ( ) ( )

n1 , n2 : n1 + n2 = n1 + n2 = n1 + n2 = n1 + n2
3.
4.

1 Z, / n  : n = 2n = 1 , so is not surjective.
For n1 ,n 2  ,

) (
)
+ n = (n + 1) + (n

n1 + n2 = n1 + n2 + 1 = n1 + n2 + 1
n1

so is not an isomorphism.
5.

y  :

x  :

+ 1 = n1 + n2 + 2

x 1 , x 2  :

x 1 = x 2

x 1 , x 2  :

(x 1 + x 2 ) =

6.

1 , / x  :

7.

y :

x  :

x =

x = 2y

1
x
2 1
1
2

1
x
2

1
x
2 2

2y = y (surjection)
x 1 = x 2 (injection)

( x 1 + x 2 ) = 12 x 1 + 12 x 2 = x 1 + x 2 .

x 2 = 1, so is not surjective.
x = 3y

x = x 3 =
3

( y)
3

=y

x 1 , x 2 :

x 1 = x 2

x 1 , x 2 :

( x 1 x 2 ) = ( x 1x 2 ) = x 1 x 2 = x 1 x 2

8.

0 0
= 0,
0 0

9.

y :

x1 = x2
3

x1 = x2
3

0 0
= 0 , so is not injective.
1 0
X M 1:

X = [ y ] (surjective)
12

X 1 = X2

X1 ,X 2 M 1:

(X 1 X 2 ) = X 1X 2 = [ x 1] [ x 2] = [x 1x 2] = x 1x 2 = x 1 x 2 = X1 X 2 .

y  + :

10.

X1 = X2

x =2ln y

x  :

x1 =x2

x = 0.5 x = 0.5

x1

x2

2 ln y

=2

ln y

x 1 , x 2 :

x 1 = x 2

x 1 , x 2 :

( x 1 + x 2 ) = 0.5 x 1 + x 2 = 0.5 x 1 0.5 x 2 = x 1 x 2 .

0.5

= 0.5

g F :

X 1 = X 2 (injective)

= y (surjective)

x 1 = x 2 (injective)

11.

[ x 1] = [x 2]

X1 ,X 2 M 1:

( ) g (t ) dt (f )(x ) = ( f )(x ) = g (t ) dt (x ) = g (x )

f F :

f x =

f 1 = f 2

f 1 , f 2 F :

) (

f 1 , f 2 F : f 1 + f 2 = f 1 + f 2
12.

f 1 F : x a x 2 ,

13.

g F :

f 2 F : x a x 3 :

f F :

f = g

f 1 = f 2

f 1 , f 2 F :

f 1 ( x ) = f 2 ( x ) f 1 ( 0) = f 2( 0 ) = 0

x  :

)( ) ( f
(f )(x ) = ddx f (t ) dt = f (x )

f 1 = f2

) = f + f = f + f
f (0) = ( 2x )(0) = 0, f (0) = (3x )(0) = 0 , so is not injective.
2

x  : (f

f 1 , f 2 F : f 1 + f 2 x =

f 1 ( x ) = f 2( x )

) (x ) = (f 2 ) ( x )

)( )

+ f 2 t dt =

x :

f (t ) dt +
x

f 1 (t ) dt =

f (t ) dt
0

f1 = f2

()

f 2 t dt = f 1 + f 2

14.
15.
16.

x  : ( f 1 f 2 ) ( x ) = x ( f 1 f 2 )( x ) = x f 1 ( x ) f 2 (x )

f 1 , f 2 F :
a. ni  :

, so is not an isomorphism.

(f 1 f 2) ( x ) = (f 1 )( x ) (f 2 )( x ) = x f 1 ( x ) x f 2 (x )

mi  :

m i = ni

m i = ni 1

n1 n 2 = m 1 m2 = (m 1m2) = m 1m2 + 1 = (n 1 1 ) (n 2 1 ) + 1

(m 1 m 2) = m 1 + m 2

b. m1, 2  :
17.

a. ni  :

mi  :

m i = ni

(m 1 m2 ) + 1 = m 1 + 1 + m 2 + 1

m i + 1 = ni

m 1 m 2 = m1 + m 2 + 1

m i = ni 1

n1 n 2 = m 1 m2 = (m 1 m2 ) = m 1m2 + 1 = (n 1 1 ) (n 2 1 ) + 1

b. m1, 2  : (m1 m 2 ) = m 1 m2
18.

a. y i  :

x i  :

x i = y i

(m1 m 2 ) + 1 = (m1 + 1) (m2 + 1)

3x i 1 = y i

m1 m 2 = (m 1 + 1) (m 2 + 1) 1 .

3x i = y i + 1

y 1 y 2 = x 1 x 2 = ( x 1 + x 2 ) = 3 ( x 1 + x 2 ) 1 = ( y 1 + 1 ) + ( y 2 + 1 ) 1 = y 1 + y 2 + 1

b.
19.

x 1 , x 2  :

( x 1 x 2 ) = x 1 + x 2

3 ( x 1 x 2 ) 1 = (3x 1 1 ) + (3x 2 1 )

3 ( x 1 x 2 ) = 3x 1 + 3x 2 1

a. y i  :

x i  :

x i = y i

x1 x 2 = x 1 + x2 +

3x i 1 = y i

1
3

3x i = y i + 1

y 1 y 2 = x 1 x 2 = ( x 1x 2 ) = 3x 1x 2 1 = ( y 1 + 1 )( y 2 + 1 ) 1

b. x 1 , x 2  :

( x 1 x 2 ) = x 1 x 2

3 ( x 1 x 2 ) 1 = (3x 1 1 ) (3x 2 1 )

3 ( x 1 x 2 ) = (3x 1 1 ) (3x 2 1 ) + 1

x 1 x 2 = x1

1
3

)(x

1
3

)+

1
3

20.

The result of the operands after * then must be equal to that after then *.

21.

A function :S S is an isomorphism between binary structures S, and S , if and only if


a , b S:

( )

a b = a b .

22.

Exchange the phrases is an identity for * and for all s S .

23.

An element eL , eR is a left, right identity of a binary structure S, if and only if s S:

( )

eL s = s ,

s eR = s . Let
13

* be defined by the table. Then a and b are both such that s S: a s = s , b s = s , so left, right identites are not
unique. The proof of uniqueness of identity breaks down when applied to left, right identities at the point of the
role reversal of the two identities.
* a b
a a b
b a b
24.

( )

Let eL , eR be a left, right identity of a binary structure S, . Then


s S:
s S:

25.

eL eR = eR

eL eR = eL

eL s = s
s eR = s

s1 , s 2 S :

eL = eR

isomorphism
invertible

s1 s 2
s1 s 2
=
=
s1 s 2


, so is an isomorphism.

invertible

=
s 1 s 2
s1 s 2 = s1 s 2

isomorph

isomorph

26.
27.

( o )(s s ) = ((s s )) = (s s ) = (s ) (s ) = ( o )s ( o )s
reflexive: (S , ) (S , ) by I : S S: s a s .
symmetric: (S , ) (S , ) by :S S . Then, by Exercise 25 (S , ) (S , ) by .
s1 , s 2 S:

inv

( )(

) (S , ) (S , ) , then by Exercise 26, (S, ) (S , ) .

transitive: If S , S , ,

28.

si S :

si S : si = si : s1 s 2

isomorphism

( ) (
(s s )
=

Let be associative on S, and : S , S , , and si S :

s1 s 2 s3 = s1

((

isomorph

) )

isomorph

= s1 s 2 s3
30.

commutative

commutative on S .
29.

( )( )
(s s )
=
(s

Let be commutative on S, and let be an isomorphism S , S , . So,

c S :

c S:

c = c , x S:

Let b S:

32?

Let :  H : a + bi , a,b R :

b b S . Then b = b S :

isomorph

isomorphism

( (

s1 s 2 s3

s3 = s1 s 2 s3

b = b = b b = b b = b b .

a b
a
, and let v,w ;
b a

v = v + iv , w = w + iw , v , v , w , w .

(
) ((
) (
)) ((
) (
))
(
) (
)
b. (v w ) = ((v + iv ) (w + iw )) = (v w v w + i (v w + v w ))
v w v w (v w + v w ) v w v w v w v w v
=
=
=
v w v w v w + v w v w + v w
v w + v w
= v + iv + w + iw = v w

(s s )
c = c = ( x x ) = x x , so x x = c has a solution
=

) (

s 2 s1 , is

))

v + w v + w
a. v + w = v + iv + w + iw = v + w + i v + w =
v + w
v + w
= v + iv + w + iw = v + w

si = si . Then

isomorph

s1 s 2 s3

x x =c

x = x S .
31.

s1 s 2 s3

si S:

2 s1

) = v

v w w
+

v v w w

v w w

v v w w

14

( ) ( )

The two isomorphisms possible are the identity and : a , b a b , a , so the equivalence classes have either one or

33.

two elements. Calculate the number of equivalence classes with one element these are the ones where coincides
with the identity:
C = F , F = C
C = F

, which corresponds to the four tables where C , D a , a , a , b , b , a , b , b . So

E = D , D = E
D = E

16 4
= 4 + 6 = 10 equivalence classes.
2
b a
b a
a b
b C' D' a E' F' a F' E'
a E' F' b C' D' b D' C'

there are 4 +

a
b

) {( ) ( ) ( ) ( )}

a b
C D
E F

1.3 Groups
1.

2.

 is closed under .
G1. a,b ,c  : (a b ) c = ( ab ) c = (ab )c = a (bc ) = a (bc ) = a (b c ) .
G2. a  :

1 a = 1 a = a,

G3. a  :

/ a  :

a 1 = a1 = a.

a a = aa = 1

a =

1
.
a
a b = a + b = 2m + 2n = 2 (m + n ) , m + n  , so 2 is closed

a,b 2 : m,n  : a = 2m, b = 2n


under .
G1. a,b ,c 2 : ( a b ) c = ( a + b ) + c = a + (b + c ) = a (b c ) .
G2. a 2 :

a + 0 = 0 + a = a.

G3. a 2 :

n  :

a,b  + :

3.

(a b ) c = (

a,b  :

( )

a a = a + a = 0 .

ab c = c ab = abc 2 .

a (b c ) = a
4.

( )

ab  + , so + is closed under .

ab =

G1. a,b ,c  + :

( ) ( )

a = 2n . Let a = a = 2n = 2 n 2Z, so a a = a + a = 0,

( bc ) =

a bc = a 2bc

a b = ab  , so  is closed under .

(a b ) c = ( ab ) c = (ab )c = a (bc ) = a (bc ) = a (b c ) .

G1. a,b ,c  :
G2. a  :
G3. a  :

1 a = 1 a = a, a 1 = a 1 = a .
/ a  : 0 a = 1
+
a,b  : a b = a b  + , so + is closed under .

5.

G1. a,b ,c  + :
a,b  :

6.

( a b ) c = ( a b ) c = a (bc ) .
a (b c ) = a (b c ) = ac b

a b = ab , so  is closed under .

G1. a,b ,c  :

a (b c ) = a bc = a bc = abc .

(a b ) c =
/ e  :

G2. a :

i e = ie = i .

a , b 0, , 999 :

7.
G1.

ab c = ab c = abc

a b = a + b mod 1000 0, , 999 , so the set is closed under .

a , b , c 0, , 999 :

(a b ) c = ((a + b ) mod1000 + c ) mod1000 = (a + b + c ) mod1000 .


a (b c ) = (a + (b + c ) mod 1000) mod 1000 = (a + b + c ) mod 1000
15

8.

( )
.
a 0 = (a + 0) mod 1000 = a mod 1000 = a
a a = a + ( a ) mod 1000 = 0 mod 1000 = 0
(
)

G3. a {0, , 999}: a = ( a ) mod 1000:


.
a a = (( a ) + a ) mod 1000 = 0 mod 1000 = 0

U = {e }
G2.

a 0, , 999 :

( a b ) c = (e

G2. e = 1 = e 0i U :

x U :

)e

x y = e i e

=e

,z = e

=e

i ( +

) = a (b c )

U , so U is closed under multiplication.

e x = x, x e = x .

x = e i

 :

x U :

x = e i , y = e

, , :

G1. x , y , z U :

G3.

x = e i , y = e

, :

x , y U :

9.

0 a = 0 + a mod 1000 = a mod 1000 = a

x = e i U : x x = e i e i = e 0 = 1 = e
x x = e i e i = e 0 = 1 = e

x U :
yy = e

x = e i

 :

( + 2 )i

y , y U :

y =e

( )

1
i
2

,y =e

+ i

, y y where y y = e i = x and

= e i = x . So U , has two distinct halves of each of its elements this is an algebraic property of

the group. Now


x : y  :

y +y = x

y = 12 x

x , x < 0 : y y = x y = x y 
so (,+ ) has just exactly one half for each element, and (, ) has elements with none. So neither of the three
groups are isomorphic.

10.

a. a,b (n, + ) :

l,m  + :

G1. + is associative.
G2. 0 (n, + ) : a (n, + ) :

a = ln, b = mn

a + b = ( l + m )n (n, + ) , so the operation is closed.

0 + a = a, a + 0 = a .

G3. a (n, + ) : m  : a = mn a = ( m )n (n, +)


b. Define isomorphisms by n  :
m  :

: (n, +) ( , + ) : nm a m . Then

(nm ) = m (surjective)

nm n

m,p  : m n ,p n  : n = m nn , p = p nn : m = p

12.

mn = p n

m = p (injective)

The operation is closed, associative, with identity 0, and inverse A.

[ ] , then the operation is closed by A B = [a b ] . Also


G1. (A B ) C = [a b ] [c ] = [a b c ] = [a ] [b c ] = A (B C ) .
G2. A + 0 = [a ] + [0] = [a ] = A, 0 + A = = A .
G3. A = [a ] : A = A = [ a ] : A + A = [a ] + [ a ] = [0] = 0, A + A = = 0 .
Write these matrices as ai
i i

13.
14.

(m nn ) = (p nn )

a + a = ( m )n + mn = 0 .

(nm + np ) = (n(m + p ) ) = m + p = (nm ) + (np )

m,p  :
11.

a + a = mn + ( m )n = 0,

n 1

i i

i =0

i i i

n =1

i i

i =0

By the calculations in Exercise 12, the operation is closed, associative, with identity 0, and inverse A.
As Exercise 13.
n 1

15.

n 1

In our notation, A B = + aik bkj


. The elements of A and B under the diagonal ai > j , j = bi > j , j = 0 are zero,
k = 0

i , j =0
so the elements of AB under the diagonal are:
16

n 1
i
n 1
j
= + aik bkj = + +
+ a b = + 0 bkj
k = 0 k = j + 1 k = i + 1 ik kj k j < i
i > j, j
k =0

so the operation is closed.

[ ]
AB

n 1

n 1
cij
G1. A B C = + aik bkj

k = 0
i , j =0

[ ]

n 1
i , j =0

+ aik 0 + aik 0 = 0

j <k i

j <i <k

n 1

n 1 n 1

n 1

= + + aik bkl clj = + ail blkckj
and
k , l =0

i =0 k =0

i , j

i , j =0

n 1

n 1

n 1 n 1

n 1
n 1
+ bikckj
= + ail + blkckj = + ail blkckj
A B C = aij
.
i , j = 0 k = 0
l =0 k =0

k , l = 0

i , j =0
i , j =0
G2. A I = I A = A .

) [ ]

n 1

n 1

G3. A = + aii , so A is not invertible if A = 0.


i =0

16.

The operation is closed, associative, with identity 0, and inverse A.

17.

The operation is closed, associative, and identity by Exercise 15G2. Since A = 1, an inverse exists:
A = A 1

A 1A = AA 1 = I . Is the inverse in the group? Suppose that A 1 is not upper-triangular, then by the

calculation in Exercise 15, neither is A 1A = I , which is a contradiction.


18.
19.

A, B:

AB = A B , so the operation is closed. It is associative, with identity I, and the regular matrix inverse.

a. a,b  \ {1} :

a b = a + b + ab ;

so a b  \ {1} .

b.G1. a,b ,c  \ {1} :


a,b ,c  \ {1} :

a + b + ab = 1

a a = 0

Conversely, a a = a
c. 2 x 3 = 7

20.

a
b
c
d

a
e
a
b
c

b
a
e
c
b

c
b
c
e
a

d
c
b
a
e

a + a + a a = 0

ae = e

e = 0 a = 1

a
e
a
b
c

b +1
= 1,
b +1

b
a
e
c
b

e = 0 . Conversely,

(1 + a ) a = a (a 1 1 + a 0)

a =

a
.
a +1

a a +1 a a2
a
a2

= 0 , so a is the inverse.
a +1 a +1
a +1
a +1

(2 + x + 2x ) 3 = 7 (2 + x + 2x ) + 3 + (2 + x + 2x )3 = 7
a
b
c
d

b = 1 a =

( a b ) c = (a + b + ab ) c = ( a + b + ab ) + c + ( a + b + ab)c = a + b + c + ab + ac + db + abc ,
a (b c ) = a (b + c + bc ) = a + (b + c + bc ) + a (b + c + bc ) = a + b + c + ab + ac + bc + abc .

G2. a  \ {1} : a e = a a + e + ae = a
0 a = 0 + a + 0a = a , so 0 is the identity.
G3. a  \ {1} :

(b + 1) a = (b + 1)

c
b
c
a
e

d
c
b
e
a

a
b
c
d

a
e
a
b
c

b
a
b
c
e

c
b
c
e
a

d
c
e
a
b

1 i -1 -i
1 1
i
-1
-1
1
-i
-1

12x = 4

) (

x = 1.
3

The groups represented by the second and third tables are isomorphic by : e , a , b , c a e , b , a , c .
a. commutative
b. See fourth table it is isomorphic to the group represented by the second and third tables.
c. Since the group has four elements, n must equal two. The four elements are thus represented by
1 1
1 1

,
,
,
. Each of these squared equals the identity matrix, so this group must be
1
1
1
1
isomorphic to that represented by the first table.
21.
A two-element group must be isomorphic to the one represented by Table 1.3.18. A three-element group must be
isomorphic to the one represented by Table 1.3.19.
22.
The definition of an inverse depends on that of identity, so G2 must precede G3. So the logically possible orders
are G1-G2-G3, G2-G1-G3, and G2-G3-G1.
17

23.

a. associativity might be defined; the statement x = identity is false; the operation is not defined
b. a group is a set with an operation; associativity might be defined; identity should be defined; inverse should
be defined
c. the statement the binary operation is defined is redundant; associativity axiom is omitted; identity should be
defined; inverse should be defined, after identity
d. a set is called a group is incorrect, rather a set with an operation; associativity might be defined; the statement
an operation is associative under addition is meaningless, an operation is either associative or not; define what a
is; {e} is a set, many groups do not have a set as an identity element; define a' as the inverse; define a and a' as
elements of the group
24.
Name this group S.
S e a b
e e a b
a a e e
b b e e
x S: e S: x e = e x = x (G2)
x S: x x = e (G3)

(e a ) e = a e = a , so G1 is not satisfied.
e (a a ) = e e = e

25.

a. false; b. true; c. true; d. false; e. false; f. true, assuming the text is correct; g. by Table 18 and 19, true; h. true
(see calculation); i. false, no identity element; j.true.
a a x b = a c a a x b = a c e x b = a c x b = a c

(x b ) = a c (x b ) b = (a c ) b

26.

x = a c b
a A: a A:

x e = a c b

a a = e

( )

e = a a = a a
27.

x b b = a c b

(a ) e = (a ) a (a ) (a ) = (a )

By contradiction. Since G is finite, there are an odd number of elements in G besides e. Reduce by pairs until there
is just one element left.
Take any a G , a e . If a a = e , we stop; otherwise, a a = b G , b e

() (

b = b

= a a

(17)

= a a . If

b = e , then a a = e and we can stop; otherwise, b e . If a = a then e = a a = a a , which is a


contradiction, so a a , that is, a and a' are distinct elements that do not square to identity.
Continue this process until an appropriate element is found, or there is just one element left; call this element c.
Suppose c c = d e c c = d . If d = d , then e = d d and we can stop. If d = e , then c c = e , and we
can stop. Otherwise c c = b for some b we considered in the reduction process, so c c = d = b , which is
impossible because we already removed b'.
So we must have stopped at some point previous and found an appropriate element.
28.

a. For a,b ,c  :

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

b. 1  :

a  : 1 a = 1 a = a
1
1
1
a  :
 : a
= a
=1
a
a
a

c. 1  :

/ a  :

a 1 = a 1 = a = 1 , so it is not a group.

d. The group axioms with left identity and inverse, or with right identity and inverse, both define groups; the group
axioms with left identity and right inverse do not.
29.
x x = x x x x = x x x = e , and the identity is unique.
30.
For a , b G ,
18

(a b ) (a b ) = ((a b ) a ) b = e ((a b ) a ) b b = e b ((a b ) a ) e = b (a b ) a = b


(a b ) a a = b a (a b ) e = b a a b = b a
so G is commutative.
31.

n  + , let U n = z i C

n-1

be the roots of z n = 1. Then z i , j U n :

i=0

(z

z j

= z i z j = 1 1 = 1, so

z i z j U n and the set is closed under multiplication.


G1. multiplication is associative
G2. 1 U n : z i U n : 1 z i = z i 1 = z i .
G3. z i U n :
32.

(a b )

C:

zi zi

0
n
1
n
= zi = zi
= z i = 1, and z i

( )( )
(

= a b

) (a b ) = (a b )
n

( ) (b ) a b

a b = a

n +1

n +1

Let m = G , and consider the m + 1 elements a 0 , , a m . Since G has only m elements, i , j : a i = a j . Assume

(a b ) = a b (a b ) (a b ) = a a b b
(a b ) = a b a b (a b ) = a b (a b )
2

(a ) a = (a ) a
i

a j i

e = a j i .

a a b a b a = a a a b b b
e = a b a b

b = a b a

38.

39.

Let :

40.

g G :

i g : G G : x a gxg . Then

x G :

g xg G :

37.

U n .

(a ) a (b ) b = (a ) (b )

n times abelian

b c = a b c a = e .
Suppose x x e , then x x x e x x e x x x e x x
Suppose x e x , then x x e x x e e e e e .
Define e by e a = a for some a G . Then
b G : y G : a y = b e a = a e a y = a y e b = b ,
so e is a left identity. Also,
a G : a G : a a = e ,
so a' is a left inverse for a. By Exercise 37, G is a group.

36.

without loss of generality that i j , so a i = a j = a i a j i

35.

= 11 = 1, so z i

= an ab

n +1

34.

(a b ) = (a b ) = (a ) (b ) = (a ) (b )

a , b G :

(a b )
33.

zi

b a = a b.

b a = a b .

a b c = e

e e e

e e.

(G , ) (G , ): a a a . Then

a (G , ): a (G , ): (a ) = (a ) = a because (a ) a = e (a ) a a = a (a ) = a , so is surjective.
a , b (G , ): a = b a = b a a = b a e = b a b e = b b a b = a , so is injective.

a , b (G , ): (a b ) = (a b ) = (b a ) = a b = a b , so (G , ) (G , ) .
(

i g g xg = gg xg g = x (surjective)

x , y G :

ig x = ig y

x , y G :

i g xy = gxyg = gxeyg = gxg gyg = i g x i g y ,

( )

gxg = gyg

g gxg g = g gyg g

x = y (injective)

so G i g G .
41.

a. monoid
b. semigroup ( is the identity element)

1.4 Subgroups
19

1.

x , y  :
+

x + y  (closed), e R = eC = 0 (identity), x :

2.

1 :

1 

3.

x , y 7 :
x 7 :

4.

7.

x 7 , y 7  :

x , y i :

x i :

x , y  :

{ }

2 n

x , y  :

n 

1. 0 

x = x , y = y

{ }

2 + 2 n

x + y = i ( x i + y i ) i (closed), e i = e  = 0 (identity),

x = i ( x i ) i (inverse).
x = ( x

x = x

x  :

x 7 (inverse).

x = ix i , y = iy i

x = ix i

x  (inverse).

x + y = 7( x 7 + y 7) 7 (closed), e 7 = e  = 0 (identity),

x = 7x 7 , y = 7y 7

x = ( 7)x 7

x i , y i  :

x  :
6.

(inverse), so not a subgroup.

x 7  :

x i:
5.

x + y = (x + y

)  (inverse).

(not closed), so not a subgroup.

n 

x y  + (closed), e + = e  = 1 (identity), x  + :

x 1  + (inverse)

3. e  = 1 7 (identity not in subset), so not a subgroup.


4. i i, i i = 1 i (not closed), so not a subgroup.
5. e  = 1 / q  : q = 1 (identity), so not a subgroup.

{ }

6. x , y i

i 

10.

11.
12.

x , y  :

{ }
Let A, B GL(n, R ):
(identity), x

9.

= e  = 0 (identity),


/ 

2. x , y  + :

8.

)  (closed), e 

i 

x = x ,y =

x  :

A = B = 2

x =

xy =
=

{ }

x y

{ }

i 

i 

{ }

(closed), e  = 1 = 0 i

i 

(inverse).

A B = 4, so multiplication is not closed on that set.

[ ]

By 1.3.12, the set is closed. The identity is in the set. For all A = aii

ii

1
A 1 = aii is in the set also.

ii

1 1 1
1 1
By 1.3.15, the set is closed. The identity is in the set. By

==
we see that the inverse of at
1 1 1
1
least one element of the set is itself not in the set.
By the argument of Exercise 8, multiplication is not closed on that set.
By the argument of Exercise 8, multiplication is closed on that set. The identity has a determinant of one, and so is
in the set. Since A 1 = 1 A , the inverse is also in the set.

13.

( ) (AB) = B

Let A, B be orthogonal matrices, then AB

A TAB = B TB = I , so AB is orthogonal. Also, I TI = I so the

identity is orthogonal also. Since the transpose of an orthogonal matrix is its inverse, the inverse is also orthogonal.
14. a. +1 F :  : x a 1, 1 F :   : x a 1 ( +1) + ( 1) = 0 F , so the set is not closed under
addition.
b. The question of whether F is a subgroup of itself is answered by whether F is a group.
15. a. f , g G : f + g :  : ( f + g)1 = f 1 + g1 = 0 f + g G (closed)
0  :
f G :
b. f G :
16.

f G :

f + 0  : x a ( f + 0 ) x = fx + 0 (x ) = fx

f R R: x a fx

()

f 1 =0

x R:

f + 0 = f (identity)

( f + f )x = fx + f x = fx fx = 0

f + f = 0 G (inverse)

f F , so the set is not a subset of F .

x = 1 : 1
a. Let f  : x a
f G , then 1 + f 0 = 1 0 + f 0 = 1 + 1 = 0
x 1 : 1
not closed under addition.
( fg )x = fx gx 0 fg F
b. f , g G : x :
fg G (closed)
( fg )1 = f 1 g1 = 1 1 = 1
1 G : f G : x : (1 f ) x = 1 ( x ) fx = 1 fx = fx 1 f = f (identity)

()

( )

1 + f F , so the set is

20

f   : x a ( fx )

f G :

f x = ( fx )

x  :

17.

()

a. 1 F
b.

x  :

= 1 1 = 1

( ff ) x

= fx f x = fx ( fx )

= 1

(inverse).

f G

1 + 1 F , so the set is not closed.

x  :

ff = 1

(inverse)

(1)0 = 1 (1 + 1)0 = (1)0 + (1)0 = 1 + 1 = 2 , so the set is not closed under addition.
( fg )0 = f 0 g 0 = 1 1 = 1, so the set is not closed under multiplication.

a. 1 F
b. f , g F:

a. Let a :
a,b  :
a :

ff = 1

f F

( fg) x = fx gx 0 fg F
(closed)
( fg ) 0 = f ( 0) g ( 0) = 1 1 = 1
1 ( 0) = 1 : f F : x  : ( f 1 ) x = fx 1 ( x ) = fx f 1 = f (identity)
1
1
f   : x a ( fx ) : x  : ( ff ) x = fx f x = fx ( fx ) = 1
1
f 0 = 1 f 0 = ( f 0 ) = 1

f , g F :

f F :

19.

(1 + 1)0 = 1(0) + 1(0) = 1 + 1 = 2

1 0 =1

1 F

18.

f 1 = ( f 1)

f a  {a} . Then

x :
x  :

a : f a = f a
b. f F .

( f a + f b ) x = f ax + f b x = a + b f a + f b = f a + b (closed)
( f a + f 0 )x = f ax + f 0 x = a + 0 = a f a + f 0 = f a (identity)
x  : ( f a + f a ) x = ( f a + f a ) x = f ax + f ax = a + ( a) = 0

f a + f a = 0 (inverse)

20.
21.

See table.
a. 50, 25, 0, 25, 50
b. 4, 2, 1, 1 ,

1
4

c.

1
, 1, , 2 .

22.

0 1
0 1
0 0 1 1 0 1 1 0 1 0
1 0 + 0 1 1 1 + 0 0 0 1

=
=
=
=
and
, so
1 0
1 0
1 0 + 0 1 1 1 + 0 0 0 1
0 0 + 1 1 0 1 + 1 0 1 0
1 0 0 1
0 1

=
,
.
0 1 1 0
1 0
0

23.

1 1
1 0 1 n

=
=
,
0 1
0 1 0 1n = 0
1 1

0 1
1 1

0 1

n +1

=
n 1

1 1 1 1 1 n 1 1 1 1 + n 0 1 1 + n 1 1 n + 1
=
,

=

=
=
1
0 1 0 1 0 1 0 1 0 1 1 0 0 1 + 1 1 0
1 1 1 1 n
,

1 0 1 0 1n = 1
n

1 1 1 1
=

0 1 0 1
1 1
so by induction
=
0 1
1 1

0 1

24.

3 0
1 0 3n

=
=
0 2
0 1 0

1 n 1 1 1 1 + n 0 1 1 + n 1 1 n 1
=
,
=

=
1
0 1 1 0 1 + 1 0 0 1 + 1 1 0
1 n

0 1 n 

0
,
2n n = 0
21

3 0

0 2
3 0

0 2

n +1

3 0 3 0 3n
=

=
0 2 0 2 0

1 2 0 3 1

=
6 0 3 0

=
n 1

3 0 3
=

0 2 0
3 0
so by induction

0 2
3 0

0 2

25.

0 2
0 1

= 2

2 0
1 0

26.

G 1 = ,+ = 1 ,

0 3 0 3n 3 + 0 0 3n 0 + 0 2 3n + 1

=
=
2n 0 2 0 3 + 2n 0 0 0 + 2n 2 0

0 3n
=
2 1 0
1

0
,

2n n = 1
0 3n 3 1 + 0 0 3n 0 + 0 2 1 3n 1
=
=
2 1 3n 0 + 2n 0 0 0 + 2n 2 1 0

3n
=
0

3 n
=
0

0
,
2
n +1

0 3 1

2n 0
0
.

2 n
n 

0 2

2 0

Exercise 22

0
,
2n 1

1 0 0 1
2
,
.
0 1 1 0

G 2 = ,+ is not cyclic,


G 3 =  + , is not cyclic, because q 1 :

p  + ,p > q ,p prime :

p q , and the same argument can be

made for numbers of the form 1 q when q 1.


G 4 = 6, + = 6 ,

{ }

G 5 = 6n

n 

= 6 ,

G6 = a + b 2

a, b 

27.

3 0 = 0, 3 1 = 3,

28.

c 0 = e,

29.

U 6 = e 2ji

30.

3
i
2

, so

3 2 = 2,

3 3 = 1,

c2 = e

i =0

3 = 4 = 4 .

c = 2.

34 = 0

32 i
e ,

1 ji
2 i
= e 3 , so e 3 = 1,

i = 0

4
i
32 i
e = e 3 ,

6
i
32 i
2i
= e0 = 1
e = e 3 = e

4 i
, so e 5 = 1,

j =0

54 i
e ,

8
i
54 i
e = e 5 ,

12
2
i
i
54 i
e = e 5 = e 5 ,

16
6
i
i
54 i
e = e 5 = e 5

i =0

1 ij
3 i
= e 4 , so e 2 = e 0 ,

i = 0

3
i
32 i
e = e 2 ,

2
i
32 i
e = e 2 ,

1
i 3 i
32 i
0
e = e 2 , e 2 = e

= 4.
0

32.

n 

= U 5 = 5.

U 8 = e 2ij
e

= 3.

U 5 = e 2ij
e5

31.

c 1 = c,

a + b 2 = na + nb 2

na + (n + 1 )b 2 a + b 2 .

n  :

2
i
3

is not cyclic, because a,b  :

54 i
0
e = e ,

5
i
54 i
e = e 4 ,

3
i
54 i
e = e 4

e4

2
7
4
1
6
i 5 i
i 5 i
i 5 i
i 5 i
i
54 i
e = e 4 , e 4 = e 4 , e 4 = e 4 , e 4 = e 4 , e 4 = e 4 ,

= U8 = 8.

22

33.

,
1

=
1

34.

,
1

1
,

1
, so

[]

= 3.

a. See table.

1
,

, so

[]

= 2.

35.

36.

1
, so

[]

= 2.

{}

b. 0 = 0

1 = {0,1, 2,3, 4,5} =  6 ,


2 = {0,2, 4}  3 ,

3 = {0,3}  2 ,

4 = {0,4, 2}  3 ,

5 = {0,5, 4,3, 2,1} =  6 .


c. 1 and 5.
d.
<1>, <5>

<2>, <4>

37.
38.
39.

<3>

<0>
Replace is a subset H of G with is a group on the subset of elements H of G, with the induced operation from
G.
Ok.
a. true (G1); b. false; c. true; d. false (the group itself is the only improper subgroup of itself); e. false; f. false;
g. false; h. false; i. true (under addition); j. false.

40.

In + , , e = 1,

41.

H G (subset)
h 1 , h 2 H:
(1.2.14)

eG

eG

h H:

12 = 1,

(1)

= 1.

h 1 , h 2 H: h 1 = h 1 , h 2 = h 2
= eH H (identity)

h 1 H:

h 1 h = eH

If G is cyclic, then g0 G :
g G :

43.

g G :

{ }

Write HK = hk

G = g0

g = g :

h H , k K

h 1 h = e

(inverse)
42.

h 1 h 2 = h 1 h 2 = h 1 h 2 H (closed)

(H G)

m  :

g G :
g = g0

h 1 h = e

m  :

g = g0

( )

h 1 h = eH , so h

= h 1

g = g0 .
m

= (g 0 ) , so G = g 0 .
m

, then
23

(h k )(h k )

h 1k1 , h 2k 2 HK:

1 1

abelian

(h h )(k k )

2 2

1 2

h1 h2 H , k 1k 2 K

1 2

HK (closed)

H, K G

eG = eG eG

hk HK:

h 1 H , k 1 K:

(hk )

eH eK HK (identity)

G3

a H,

46.

a 1 G ,

a G,

G1

a G a 1 H ,

a G a 1 = eG

45.

) ( ) = (h h )(k k ) = e
abelian

h 1k 1 hk

eG H , , which proves G2?!


G3

( )

H G

( )

a , b H:

a =1 H ,

ab 1 H

1
a H: aa = eG H (G2)

1
1
b H: eb = b H (G3)

a , b H: c H: c = b 1

ab 1 H

Let G = g 0 , so g G :

G1

b 1 H

a , b H:

m  :

g = g0

1
g = g0

( )

ac 1 = a b 1

g 0 = e,

g0 = g0,

g0 = g0 g0 = g0 g0

(h h )

h 1 , h 2 H :

commutative

1 2

e2 = e

e H (identity)

h H :

(h )

(inverse).
48.

h 2

, and so a cyclic group must have at least g 0

Let Hn N + = x G

and

g0 2 .

(h h )
1

= e 2 = e;

(h ) = (h )
1

( )

e = h 1 h 2

(h h )

commut.

=e

(h )
1

. Then

x n =e
commut.
n
n n
h 1h 2
= h1 h 2

h 1 , h 2 Hn :
en = e

= e , so

h 1 h 2 = e e = e (closed)

commut.

= ab H (G1)

and its inverse as a generator. If a cyclic group has only one generator, then g 0 = g 0

47.

= eHK , so

H eK

= h 1k 1 (inverse).
H G

44.

h 1k 1 HK

= e e = e (closed)

e Hn (identity)

h Hn :

(h )
1

hn

(h h )

commut.

= e n = e;

(h ) = (h )
1

( )

e = h 1 h n

(h h )

commut.

49.

(inverse).
See Exercise 1.3.33.

50.

Since H , h H . Since H is closed under the operation of G, m  :

= en = e

(h )
1

Hn

h m H . Since H  ,

h m = h n . Suppose without loss of generality that m < n , so h m = h m h n m . Since this is an

m,n  :

expression in G also, and since the identity of G is unique, eG = h n m H . Also, h h n m 1 = h n m = e , so


h 1 = h n m 1 H . So H is a subgroup of G.
51.

x , y Ha :

xa = ax , ya = ay

ea = a = ae
x Ha :
52.

a. x , y H s :

(xy )a = xay = a(xy )

xy Ha (closed)

e Ha (identity)

x a = x 1ae = x 1axx 1 = x 1xax 1 = eax 1 = ax 1


s S:

xs = sx , ys = sy

(xy )s = xsy = s(xy )

x 1 Ha (inverse).

xy H s (closed)
24

s S:

es = s = se

x s = x 1se = x 1sxx 1 = x 1xsx 1 = esx 1 = sx 1

x H s : s S:
b. By definition.
53.

a G :

aa 1 = eG = eH H

a , b G :

54.

e H s (identity)
1

a~b

ab

x 1 H s (inverse).

a~a

(ab ) = ba
1

b~a

a , b , c G : a ~ b b ~ c ab 1 , bc 1 H ab 1bc 1 = ac 1 H a ~ c .
q , r H K: q H , q K , r H , r K qr H qr K qr H K (closed)
H G eH = eG H
eG H K (identity)

K G eK = eG K
q H K: q 1 H q 1 K q 1 H K (inverse).

55.

g 1 , g 2 g 0 :

56.

This is the case if G is commutative:


g n , h n G n :

m 1 , m 2 Z:

g nh n

g1 = g0

commutative

( gh )

m1

, g 2 = g0

m2

g1 g 2 = g0

m1 + m 2

= g0

m 2 + m1

= g0

m2

g0

m1

= g 2 g1 .

G n (closed)

e = e G n (identity)
g n G n :
57.

( )
g 1

gn

(g g)

commutative

= e n = e,

(g )
1

G n (inverse).

By contradiction: suppose G is not cyclic. If / g G , g e , then G = E = e , which is a contradiction. So


g G , g e and by (17), the the nontrivial cyclic group g G . But G is not cyclic, so g G :

g g , so

g is proper.

1.5 Cyclic Groups and Generators


1.
2.
3.
4.

42 = 4 9 + 6
42 = 5 9 + 3
50 = 7 8 + 6
50 = 6 8 + 2

5.

gcd 32, 24 = gcd 25 , 23 3 = 23 = 8 .

6.
7.

( ) (
)
gcd( 48, 88) = gcd( 2 3, 2 11) = 2 = 8 .
gcd(360, 420) = gcd( 2 3 5, 2 3 5 7 ) = 2
4

3 5 = 60 .

8.

13 + 17 8 = 21 mod 17 = 4 .

9.

21 + 30 19 = 40 mod 30 = 10 .

10.

26 + 42 16 = 42 mod 42 = 0 .

11.
12.
13.
14.
15.
16.

39 + 54 17 = 56 mod 54 = 2 .
1, 2, 3, 4: 4 (by relative primes).
1, 3, 5, 7: 4.
1, 5, 7, 11: 4.
1, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 57, 59: 16.
The image of a generator under an isomorphism must be another generator. By Exercise 52, an isomorphism is
defined completely by its action on a generator. Therefore, there is one automorphism for each generator that one
specific generator could be mapped onto that is, the number of automorphisms on a cyclic group is the number
of generators of that group.
1: 1.
1, 5: 2.
25

17.

18.
19.
20.

1, 3, 5, 7: 4.
1, 1: 2.
1, 5, 7, 11: 4.

21.

30 gcd 25, 30 = 30 5 = 6 .

22.

( )
42 gcd(30, 42) = 42 6 = 7 .

23.

{i, i

24.

(1 + i )

25.

1+ i = 2 e 4 ,

= 1, i 3 = i , i 4 = 1 : 4.
1

2=e4 ,
1

e4

e4

26.

= 8.
= 0 .
27.

23

22

28.

23

22

23

22 3
22 3

2 32

22 32
29.

6 = 2 3 , so the cyclic subgroups are the ones generated by 20 30 = 1,

30.

8 = 23

31.

12 = 2 2 3

20 30 = 1,

21 30 = 2,

20 31 = 3,

2 2 30 = 4,

21 31 = 6,

2 2 31 = 12 .

32.

20 = 2 2 5

20 50 = 1,

21 50 = 2,

2 2 50 = 4,

20 51 = 5,

21 51 = 10,

2 2 51 = 20 .

33.

17 = 17 1

34.
35.
36.
37.
38.
39.
40.
41.

20 = 2,

21 = 2,

17 0 = 1,

2 2 = 4,

21 30 = 2,

20 31 = 3,

21 31 = 6 .

23 = 8 .

17 1 = 17 .

{2, 3} = 1 .
{4, 6} = {2 , 2 3} = 2 .
{8, 10} = {2 , 2 5} = 2 .
{12, 30} = {2 3, 2 3 5} = 2 3 = 6 .
{12, 42} = {2 3, 2 3 7} = 2 3 = 6 .
{18, 24, 39} = {2 3 , 2 3, 3 13} = 3 .
2

Insert the phrase [if and only if] n is the smallest nonnegative integer such that [ a n = e ].
Ok.
26

42.

a. true; b. false; c. true; d. false; e. true; f. false (the group of order 3 with the operation that takes the right
element); g. true; h. false (G and G' don't even have to be defined on the same set); i. true; j. true.

44.

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


(,+ ) .

45.
46.
47.

E.
Every infinite cyclic group is isomorphic to , which has two generators.
5 has generators 1, 2, 3, 4.

48.

j 2i

U 4  4 which has generators 1, 3, so e 4

43.

1 i 3 i
= e 2 , e 2 .

j = 1, 3

49.

U 6 6 :

2i
6

50.

U 8 8 :

2i
8

51.

U 12  12 :

52.

x G :

53.

p ,q S : p n ,p m ,qn ,q m  : p = p nn + p mm, q = q nn + qmm


p + q = (p n + q n )n + (p m + qm )m, p n + q n , p m + q m  (closed).
0 = 0n + 0m S (identity)
p S : p n ,p m  : p = p nn + p mm p = (p nn + p mm) = ( p n )n + ( p m )m S (inverse).

54.

(ab )
(ba )

1 i

= e 3 ,e

j =1, 5

= e

j =1, 3, 5, 7

= e

j =1, 5, 7, 11
x = an

n2

,e 4

1
i
6

,e 4

,e 6

x = a n

( ) ab = a(ba )
(ab ) = e , so ba =

ab ab

=e

1
i
4

2i
12

n  :

=e

5
i
3

n 1

,e

,e 6

7
i
4

,e

11
i
12

isomorphism

(ba )

n 1

b =e

(a) n .

(ba )

n 1

b = a 1

ba = e

(ba )

= e . Similarly,

ab .

a. The least common multiple of r, s  + is the generator of the group  r  s (which exists by Theorem 24). This
agrees with the intuitive notion because elements of the intersection must be multiples of both r and s.
b. When  r  s =  rs .
c.
56.
Show that an infinite group has an infinite number of subgroups. If there is a generator a of the group, then it is
isomorphic to  and thus has an infinite number of subgroups. If it does not have a generator, then
55.

57.

The group 0, 1, i , 1 + i under modulo addition is not cyclic, but all its proper subgroups 0 , 1 , i , 1 + i are.

58.

 n is closed under + n . For r, s ,t  n ,

(r + s ) +
n

t=

(associative)
r  n : r + 0 = r (identity)
r  n :

r = 0 :
r =
r 0 :

n r

r = 0 :
r +n r =
r 0 :

(((r + s ) mod n) + t ) mod n = ((r + s ) + t ) mod n


( ( ))

( ( )

= r + s + t mod n = r + s + t mod n mod n = r + n s + n t

0 +n 0 = 0

r + n (n r ) = n modn = 0

a 2 = e = xx 1 = xa 2 x 1 = xaax 1 = xax 1xax 1 = xax 1

(inverse)

a = xax 1

59.

x G :

ax = xa .

60.

 pq is generated by all relative primes to pq, that is, to p and q, less than pq. There are p 1 divisors by q of pq, and
27

) (

) (

) (

q 1divisors by p of pq, so there are pq 1 q 1 p 1 generators when p q and pq 1 p 1


generators when p = q .
61.
62.

This again amounts to finding the relative primes to p r , of which there are p r 1 1.

63.

n gcd n, m

64.

( )
( )

All the proper subgroups of  p are 1s = s , where s = p gcd s , p < p

( )

gcd s , p > 1, and p has no

denominator common with s except 1 if it is prime.


65.
66.

Every permutation of edges leads to the same vertex.

67.
68.
69.
70.

Not commutative, because a 3b b 1a = e , a 3b ab 1 = a 2 e .


Not obvious: one would need to find a path which generates the group.
No, because it does not include the identity element.
71.
0
0
7
5

6
4

2
5

72.

4
a. A relation is represented by a closed path from the identity element to itself.
b. b 2 = e ,

73.

74.

75.

76.

( )

abab = e ,

a 4 = e,

( )(

a 2ba 2b = e .

( )

a. a 2b a 3 = a 3b ; b. ab a 3b = a 2 ; c. b a 2b = a 2 .
e a b c
a e c b
b c e a
c b a e
See table, where c = ab .
e a b c d f
a e c b f d
b d e f a c
c f a d e b
d b f e c a
f c d a b e
See table, where c = ab ,
e
a
b
c
d
f

a
c
d
e
f
b

b
f
e
d
c
a

c
e
f
a
b
d

d
d
a
f
e
c

d = ba ,

f = aba .

f
b
c
b
a
e
28

See table, where c = a 1 ,


77.

Z4

f = ba 1.

d = ba ,

78.

(nonabelian)
b

2.1 Groups of Permutations


1.
2.
3.
4.
5.
6.

7.

= 1 2 3 6 5 4

(
)
= (3 4 1 6 2 5)
= (51 6 2 4 3)
= ( 2 6 1 5 4 3)
= (1 2 3 4 5 6) , = (3 1 4 5 6 2) , = ( 4 3 5 6 2 1) , = (5 4 6 2 1 3) , = (6 5 2 1 3 4) , = ( 2 6 1 3 4 5) ,
= (1 2 3 4 5 6) = , so = 6 .
( ) = (1 2 3 4 5 6), ( ) = (4 3 2 1 5 6) , ( ) = (1 2 3 4 5 6) = ( ) , so = 2 .
2 = 2 4 1 5 6 3
2

2
1

( )

100 = 616 + 4 = 6

9.

0 = 1 2 3 4 5 6 , 1

3
Z2 , +

(
)
= (5 2 4 3 1 6) , = (1 2 3 4 5 6) , so

16

8.

S2

4 = 4 = 6 5 213 4 .

17

3Z, +

17Z, +

6!

100

( )

= e 50 = e .

R, +

R *,

= 250 = 2

50

S6

cyclic
Z6 , +

(3 5 4 1 2)

cyclic

acyclic

Z, +

Q, +

Q*,
C* ,

R +,

10.

{
}
= {1, 2, 4, 3} .
= {1, 5} .

11.

O1, = 1, 3, 4, 5, 6, 2 .

12.

O1,

13.

O1,

14.

, 1 = , 2 = 2 , 1 = , 2 = , 3 = 2 .
29

15.

(
)
(
) (
)
= = (3 4 1 2) = ( 4 1 2 3) , = , = ( 2 3 41) = (1 4 3 2) = , = (3 4 1 2) = ( 4 3 2 1) =
= ( 4 1 2 3) = (3 2 1 4) = .
= 1 = 2 1 4 3 , = 0 , 1 = 1, 2 = 2 = 2 3 4 1 = 3 4 1 2 ,

= 1 = 2 3 4 1 ,
3

2,

16.

S4

17.

S5

18.

3 = 3
2=5

S3

S4

S4

S5

= S3 = 3! = 6 .

3 = 3

= S 4 = 4! = 24 .

2=5

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

= ( 2 3 1), = (1 2 3) = {, , } ,
= , = , = ( 2 1 4 3) = (1 2 3 4) = = {, } .
b. = {, } , = {, } , ,
= D , = (1 3 2) = ( 2 1 3) = ,
= ( 2 1 3) = (3 2 1) = .
0

( )
= (1 2 3),

( )
= (3 1 2),

a. 1 = 1 2 3 ,

1 = 2 3 1 ,
1

0
1

1
1

3
1

1 = 3 1 2 ,
2

i = 1, 2; j = 1, 2, 3

1 1

D3

1 = 2

19.

1 = ,

1 = 1 ,

) (

) (

1 = 3 4 1 2 = 2 , 1 = 1 3 4 1 2 = 4 1 2 3 = 3 , 1 = 1 4 1 2 3 = 1 2 3 4 = ,

1 = , 1 , 2 , 3 ;
= {, } ;
(
) (
)
= , = , = ( 4 1 2 3) = (3 4 1 2) = , = (3 4 1 2) = ( 2 3 4 1) = ,
= ( 2 3 4 1) = (1 2 3 4) = = { , , , } ;
= , = , = ( 2 1 4 3) = (1 2 3 4) = = {, } ;
= = {, } ;
= , = , = (3 2 1 4) = (1 2 3 4) = = {, } ;
= = {, }
= (1 2 3 4 5), = ( 2 4 51 3), = ( 4 1 3 2 5), = (1 2 5 4 3) , = ( 2 4 3 1 5) , = ( 4 1 5 2 3) ,
= (1 2 3 4 5) = .
Since ( 2 1 3) = , (1 3 2) = , there are two distinct elements that square to the identity, while has only one
0

2 = ,

2 = 2 ,

2 = 2 3 4 1 2 = 1 2 3 4 =

0
1

1
1

0
1

1
1

20.

( 3 ), so / S3 .

30

0
1
2
3
4
5

21.

1
2
3
4
5
0

2
3
4
5
0
1

3
4
5
0
1
2

4
5
0
1
2
3

5
0
1
2
3
4 .
1 1
1

a. 1 2 = 2

1 3 3

1 1
1

1 2 = 3

1 3 2

1 1 3

~ 2 3 1 , 1
2 = 1
1 3 2

1 1 2

~ 3 2 1 , 1
2 = 1

1 3 3

Since A B x = A Bx , the matrices form a group isomorphic to a group of permutations.

( )

1 1 2

~ 123 ,
1 2 = 3
3 1
1

1 1 3

~ 1 3 2 , 1 2 = 2
3 1
1

[ ] (

[ ] (

[] ~ (3 1 2) ,

[ ] (

[ ] (

[] ~ (2 1 3) .

b. S3 .

23.

1
1

1
1
1
1
1

,
~
, 1 ~
, 2 ~
, ~
, 1 ~

1
3 1

1
1
1

1
1
1

1
1

1
1
1
, ~
2 ~
, 1 ~
.
1

1
2
1

1
1

S2 .

24.

S2 S2 .

25.

S4 .

26.

S .

22.

27.

) (

( ) (

) (

) (

0 = 0 + 4 = 0 1 2 3 , 1 = 1 + 4 = 1 2 3 0 , 2 = 2 + 4 = 2 3 0 1 , 3 = 3 + 4 = 3 0 1 2 , the left regular


representation is :  4 S4 : x a x .

With S3 = , 1 , 2 , 1 , 2 , 3 :

) (
) = (

1 2 1 2 3 = 1 2 1 2 3 = 1 2 1 2 3 ,

1 1 2 1 2

2 1 2 3 1 = 1 2 2 3 1 ,

28.
29.
30.
31.
32.
33.

et cetera, reading off the columns of Table 2.1.8. Then the right regular representation is : S3 S3 : a .
The book definition states onto, but this is the same as to when a set is mapped to itself.
Okay.
Permutation.
Not surjective for negative numbers.
Permutation.
Not surjective for nonpositive numbers.

34.

f 5x = x 3 x 2 2x f 5 x = 3x 2 2x 2 f 5 x = 6x 2 . f 5 x = 0

6x = 2

f 5 0 = 6 0 2 = 2 < 0 and f 5 1 = 6 1 2 = 4 > 0 , and thus f 5 is not injective around


35

x = 1 , so
3

1
.
3

a. true; b. false, must map on the same set; c. true; d. true? (book says false); e. true; f. true (by Theorem
31

1.4.17); g. false, S10 = 10!; h. false, by Theorem 1.5.1. every cyclic subgroup is commutative, but in Example
2.1.17 S3 is shown not to be commutative and is thus not cyclic; i. true, any Sn has S3 as a subgroup and can
therefore be neither commutative nor cyclic; j. true
36.
37.

38.

and

commutative.
Function composition is associative by 1.1.13. The set has the identity function as an identity element, however the
set does not have an inverse for each of its elements. For example, let a A , then f a : x a a has no f a such that
f a o f a = 1. This algebraic structure is a monoid.

Let H = SA | b = b , then , H: 1 SA , b = b

1 SA , 1b = b = b
39.

(1 2) , (1 3) , and (2 3) , each isomorphic to S

S3 is not commutative, with proper subgroups

1b = b , and

1 H , so H is a subgroup by Exercise 1.5.45.

Let H = SA | b B . If B A then a A \ B:

H: a = b

1b = a

1 H , so H is not a

subgroup by Theorem 1.4.14.


40.
41.
42.

43.

44.

Let H = SA | B B . , H: B B ( bijective)

B = B

1B = B , so 1 SA , and

1B = B = B and 1 H .
By 40., also a subgroup.
A copy of an n-gon is any permutation of the vertices of the original in which neighbors of vertices remain
neighbors. There are n permutations that leave the orientation unchanged, and another n that reverse it ( n 3).
The first set form a group in itself, because any product of permutations that leave the orientation unchanged itself
leaves the orientation unchanged.
How many different ways can a cube be rotated? One of its six faces can be rotated upwards, then one of four faces
can be rotated leftward, which fixes the rotation. So there are 6 4 = 24 possible rotations. Three subgroups of
order four are formed by rotating the cube around its three perpendicular axes, and four subgroups of order three
are formed by rotating it around its four diagonal axes.

( )( )

( )( )(

) ( )(

) (

) ( )( )(

) ( )(

) (

For Sn 3 : 1 2 , 1 3 Sn , and 1 2 1 3 1, 2, 3 = 1 2 3, 2, 1 = 2, 3, 1 , 1 3 1 2 1, 2, 3 = 1 3 2, 1, 3 = 3, 1, 2
so the group is not commutative.

45.

Let Sn :

( )

= 1 . Suppose

Sn : =

) ( )
1

= j i . So i = i =
46.

{ }

n Z

= n +n a a

n Z

(i ) = j , but j i , so it cannot be that .

= n n a a

n Z

{ }

= nc

n Z

= n nb b

n Z

= n + nb b

n Z

{ }

= nb

n Z

= Ob , .

Number the elements of A by a0n 1 . Generate n permutations i SA by i a j = a j + n i . + n induces n distinct


permutations on A. Also, ai , j A:

48.

i: i i . Since n 3, j i , i , so define

Suppose c Oa , , Ob , , then na , nb Z: n a a = c , nb b = c . So
Oa , = n a

47.

j n i < n

j n i ai = a j .

If Oa, = A then it is possible to number the elements of A by a0n 1 such that n a0 = an . Then
ai , j A: j i ai = j i i a0 = j a0 = a j , and j i , so is transitive on A.
Conversely, let be transitive on A. Then for any given a A:

ai A:

j : j a = ai , so

Oa, = A .
49.

a. They will read every product a b = c as b a = c , and every instance of the associative property of the group
a b c = a b c as an associative property c b a = c b a of a corresponding, but different, group. Since

( ) (

( )

a group can be defined solely in terms of such expressions, their reversal defines a group also.

) ( )

( )

b. a b c = b a c = c b a = c b a = a c b = a b c (associativity)
e x = x e = e (left identity)
a a = a a = e (left inverse)
50.

( )

Show that the right regular representation : G G : g a g is an isomorphism. Obviously


32

( g ) = ( g ) g = g because * is a group operation, so is an injection, and surjective on G, so a bijection.


g , h G : ( g h ) = ( ( g h )) = ( g ) h = (g ) h = h o g , with Exercise 49 shows that (G, o) does indeed
form a group.
51.

52.
0

s0

s1

b.

s1

s0

c.

11101

s0

s1

d.

010100

s0

s1

Tx s 0

Tx s 1

Tx s 2

a.

0110

s0

s0

s0

b.

0110111

s2

s2

s2

c.

1101

s1

s1

s1

s1

s2

s2

s0

s1

s2

54.

(n + 1)

55.

yx is such a string.

57.

n +1

T 1
1
1 1 is a group, because it is a monoid with an inverse T 1 = T .
x
x

(
T (s

) (
s ) = (s

) (

) (

) (

) (

T s0 s1 s 2 = s0 s1 s 2 , T0 s0 s1 s 2 = s0 s0 s0 , T1 s0 s1 s 2 = s1 s 2 s 2 , T01 s0 s1 s 2 = s1 s1 s1 , and
11

0
1
01
11
58.

Tx s 1

a.

53.

56.

Tx s 0

0 s1 2

0
1
01
11

0
0
0
0
0
0

s2 s2 .

1
1
01
11
11
11

01
01
01
01
01
01

11
11
11
11
11
11

is not a group, because there is no inverse for any Tx except T .


1
0

1
1

33

59.

e
a

c
c

b
e

a
e

60.

The state transition function for an input string g = g 0 g n 1 of the automaton of a finite group G is a function
Tg : G G : x a x ni =01 g i . Since ni =01 g i G , T is a permutation of G.

61.

isomorphic to G.

2.2 Orbits, Cycles, and the Alternating Groups


1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13

{1, 5, 2}, {3}, {4, 6}


{1, 5, 8, 7}, {2, 6, 3}, {4}
{1, 2, 3, 5, 4}, {6}, {7, 8}


{2i}i ,

{3i}i ,

{2i + 1}i 
{3i + 1}i , {3i + 2}i 

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

a. 4
b. The order of a cycle is equal to the number of elements in its orbit.

( ), = (4 5)(2 3 7), = (2 7 3), = (4 5), = (2 3 7), = (4 5)(2 7 3), = ( ) =


= ( ), = (1 4)(3 5 7 8), = (3 7 )(5 8), = (1 4)(3 8 7 5), = ( ) = = 4 .
d. (1 8)(3 6 4)(5 7 ), (3 4 6), (1 8)(5 7 ), (3 6 4), (1 8)(3 4 6)(5 7 ), ( )
= 6,
(1 3 4)(2 6)(5 8 7), (1 4 3)(5 7 8), (2 6), (1 3 4)(5 8 7), (1 4 3)(2 6)(5 7 8), ( ) = 6 ,
(1 3 4 7 8 6 5 2), (1 4 8 5)(3 7 6 2), (1 7 5 3 8 2 4 6), = 8 .
c. 0 =
0

= 6,

e. The order of a permutation is equal to the least common multiple of the numbers of elements of the orbits in a
decomposition into disjoint cycles.
14.

5 = 2 + 3,

( )

lcm 2, 3 = 6
34

15.

6.

16.

7 = 3 + 4,

17.

10 = 5 + 3 + 2,

18.

15 = 3 + 5 + 7,

19.
20.
21.
22.

( )
lcm(5, 3, 2) = 30
lcm(7, 5, 3) = 105

lcm 3, 4 = 12

okay
[A cycle is a permutation having] at most [one] nontrivial [orbit].
For all positive n, An [is the] sub[group of all even permutations] of Sn .
a. false, but every permutation is a product of disjoint cycles.
b. true
c. true, but it wouldn't have been obvious that a permutation couldn't have been both even and odd

(1 2 3 4) S contains the odd permutation (1 2 3 4) = (1 4)(1 3)(1 2) but none of


(1 2 3 4) , (1 2 3 4) , (1 2 3 4) = (1 3)(2 4), (1 2 3 4) = (1 4 3 2) is a transposition.

d. false,

e. false, A5 =

1
2

S5 = 1 5! = 60 .
2

{( )} ( )

f. false, S1 = 1 = 1 .

g. true, A3 = , , 1 where = 1 2 3 ,


1 = 1 3 2 and the group is commutative:

h. true
i. true
j. false, (1 2) and (3 4) are both odd permutations but (1 2)(3 4) is even.
23.

( ) , = (1 2 3) = (1 3)(1 2) , = (1 3 2) = (1 2)(1 3) are even, = (2 3),


1

1 2
1 2
2 1
24.

( )

2 = 1 3 ,

( )

3 = 1 2 are odd.

{( )} can be written as a product of zero transpositions. For

a. By induction. For n = 1, the only element of S1 = 1

)
So (n n)(n n) = is a product of at most n transpositions.

n > 1, for any Sn , the permutation n n does not move n so is a permutation of Sn 1 and can be written as a
product of at most n 1 transpositions.

b. If a permutation Sn is not a cycle it consists of at least two cycles. Since by (a) each cycle can be written as a
product of at most n 1 transpositions, can be written as a product of at most n 2 .
c.
25.

( )(

)( ) (

a. i j b j i = b i j

b
35

( )( )( ) ( )

b. i j j i = i j

26.

Let H Sn . Either H where is odd, or all the permutations in H are even. In the first case let He be the
set of even permutations of H, and let : He H: a . Since is even and is odd, must also be odd. If
, He : = = , so is a bijection.

2.3 Cosets and the Theorem of Lagrange


1.

4 + 0 = {, 8, 4, 0, 4, 8, } ,
4 + 1 = {, 7, 3, 1, 5, } ,

4 + 2 = {, 6, 2, 2, 6, } ,
4 + 3 = {, 5, 1, 3, 7, } .
2.

2 = {, 4, 2, 0, 2, 4, }

3.

w12 =

4 + 0 = {, 4, 0, 4, } , 4 + 2 = {, 2, 2, 6, } .
2

12

12

4.

5.

{ 18

6.

7.

8.

12

({0, , 11}, + 12) ,

 12 2

{
}
+ 1 = {1, 3, 5, 7, 9, 11} .
+ 0 = {0, 4, 8} , 4 + 1 = {1,

12

= {0, 2, 4, 6, 8, 10}

+ 0 = 0, 2, 4, 6, 8, 10 ,

36

+i

12

5, 9 , 4

12

+ 2 = 2, 6, 10 , 4

12

+ 3 = 3, 7, 11 .

i 0, , 17

{ } { }
{ , } = { , } ,
{ , } = { , } ,
{ , } = { , } .
{ , } = { , } ,
{ , } = { , },
{ , } = { , } ,
{ , } = { , } .

0 0 , 2 = 0 , 2 ,
1

The left and right cosets are not the same.


Neither the left nor the right cosets form a group.

36

9.

10.

11.

0 2 1 2 1 2 1 3

0 2 1 1 1 2 2 3

0 0 2 1 2 1 2 1 3

0 0 2 1 1 1 2 2 3

2 2 0 1 3 2 1 1 2

2 2 0 1 1 2 1 3 2

1 1 2 2 1 1 3 2 0

1 1 2 2 2 1 3 1 0

2 2 1 2 0 3 1 2 1

1 1 3 1 0 1 2 2 2

1 1 2 2 1 0 2 3 1

1 1 2 2 3 0 2 1 1

2 2 1 3 1 2 0 2 1

2 2 1 3 2 2 0 1 1

1 1 3 1 2 1 2 0 2

2 2 1 2 2 3 1 0 1

3 3 1 0 2 2 1 1 2

3 3 1 0 1 2 1 2 2

{ } { }
{ , } = { , } ,
{ , } = { , } ,
{ , } = { , } .
{ , } = { , } ,
{ , } = { , },
{ , } = { , } ,
{ , } = { , } .
0 0 , 2 = 0 , 2 ,
1

The left and right cosets of this subgroup are the same.
So, even a noncommutative group may (must?) have left and right coset partitions that equal, and thus a coset
group, if the subgroup is appropriately chosen.
This subgroup induces a coset group isomorphic to the Klein 4-group.

0 2 1 3 1 2 1 2
0 0 2 1 3 1 2 1 2
2 2 0 3 1 2 1 2 1
1 1 3 2 0 1 2 2 1

12.

3 3 1 0 2 2 1 1 2

1 1 2 2 1 0 2 3 1

2 2 1 1 2 2 0 1 3

1 1 2 1 2 1 3 0 2

2 2 1 2 1 3 1 2 0

24

 24 : 3
13.

= 0, 3, , 21 ,

24

=  24

24

= 24 8 = 3 .

1 = 0 , 1 ,
S3 : 1 = S3

1 = 3! 2 = 3 .

37

14.

D 4 : 1 = D 4

15.

= 1

)(

= 1

20.
21.
22.
23.
24.

= 5,

= 5! 5 = 24 .

)(

= 22 = 4 ,

5 3

S6 : = S6
17.
18.
19.

) (

6 = 1

5 3

S5 : = S5
16.

1 = 8 2 = 4 .

= 6! 4 = 180 .

Insert where a G .
Amend H G (H is a subgroup of G).
a. true
b. true
c. true (every subgroup of prime order is cyclic (2.3.11), thus isomorphic to  n , and thus commutative)
d. false (the trivial subgroup of any infinite group obviously has left cosets)
e. true ( H = H )
f. false
g. true (by Theorem 2.2.20)
h. true
i. false (not necessarily if the group is noncommutative)
j. true (because cyclic groups are commutative (1.5.1) and by the remark after 2.3.14)
Impossible, by the boxed remark after Example 2.3.3.
The improper subgroup of any group G.
The trivial subgroup of any group of order 6 such as  6 .
Impossible, since a partition of a set can never produce more cells than the order of the set.
Impossible, since by the boxed remark before 2.3.10 the order of each cell of the partition must be equal, and thus
equal 6 4 = 1 1 , and the order of a set must obviously be integral.
2

25.

The relation ~ R is
reflexive: g G :

g ~R g

gg 1 = e H ,

symmetric: g , g G : g ~ R g

gg 1 H , and because H is a group, gg 1

g ~R g .

) = (g )
1

g 1 = g g 1 H so

transitive: g , g , g G : g ~ R g , g ~ R g gg 1 H g g 1 H , and because H is a group,

( gg ) ( g g ) = gg
1

26.

H , so g ~ R g ,

so it is an equivalence relation.
Let : H Hg : h a hg . This function is
surjective:

h H : h = hg Hg H Hg

hg Hg h H : h = hg Hg H

injective: hg , h g Hg : hg = h g
so it is bijective.
27.

h = h,

For every left coset defined by some g G , gh gH

( g ) h ( g ) H
1

ghg 1 H

H = Hg ,

h H , and then because g 1 G ,

h H : ghg 1 = h

gh = h g

gh Hg , so gH Hg . Conversely,

Hg gH , so Hg = gH .
hg gH

h H : hg = gh

g 1hg = h

g 1hg H .

28.

h H , g G : hg Hg

27 and 28 together state that H G induces the same left and right coset partition iff h H , g G : g 1hg H .
We already know from Example 7 that this is equivalent to the existence of a coset group.
38

29.

Counterexample: choose a = e

(H = bH

H = Hb

bH = Hb , which obviously does not always hold, as

in Example 7, where 1H H1.


1

30.

a , b G : Ha = Hb

h a H : h b H : h a a = h b b

hb ha a = b

b Ha .

31.

a , b G : aH = bH

h a H : h b H : ah a = bh b

h a = a 1bh b

ha hb

h H : ha 1 = hh a h b b 1

ha 1 Hb 1

that Ha 1 = Hb 1 .

= a 1b

h a h b b 1 = a 1 , so

Ha 1 Hb 1. Transposition of a and b gives the converse, so

(H = bH

H = b 2H

bH = b 2H

32.

Counterexample: choose a = e

33.

The order of any proper subgroup H G must divide the order pq of G, so H 1, p , q is prime, so by (11) H is
cyclic.

34.

{ }

Let : H
surjective:

{ }

{ }

{ }
g G : g H {H }
( g H ) = H ( g ) = Hg {H } {H }

injective: g , g G : gH , g H {H }
: ( gH ) = ( g H ) Hg
G

: gH a Hg 1, which is:

{ }

( )
g G : Hg {H }

g G : gH H

H = bH which is false if b H .

gH = Hg 1 H

(h

h H : h H : h 1 g 1 = h g 1

g 1

) = (h g )
1

{ }

{ }

{ }

= H

= Hg 1 , so

gh = g h 1

gh g H

gH g H .

Transposition of symbols gives the converse, so gH g H .


So the function is bijective, which shows the existence of an isomorphism between the left and right coset partition,
and thus (for infinite sets, by definition) their equal size.
35.

Suppose there were two elements c, d of order 2, then c , d would generate a subgroup of order 4 (remebering
that the group is commutative):
e
c

cd

cd

cc = e

cd

ccd = d

dc = cd

dd = e

dcd = ddc = c

cd

cd

cdc = ccd = d

cdd = c

cdcd = cddc = cc = e

By Lagrange, c , d = 4 would have to divide 2n = 4


36.

g G :

n
n
, but
is not integral.
2
2

g G . Since G has no proper subgroups, g = G . If G is of infinite order, then g 2 g = G

which contradicts G not having a proper subgroup, so G must be of finite order. Similarly, if
n > 1, then
37.

g is divisible by

g n g , again contradicting. So G must be of prime order.

We need to show that each of the elements is in fact a left coset of K in G, that every such coset is an element, and
that the elements are distinct. So, let ai
be such that ai H is the set of distinct left cosets of H in G, and

{b }
i

0 i < H :K

ai , b j

{ }

{ }

0 i < G :H

{ }

such that bi K is the set of distinct left cosets of K in H.


ai b j G

ai b j K is a left coset of K in G;

g G : gK is a left coset of K in G, since

U a H =G
i

ai : g ai H

h H : g = ai h , and since
39

U b K = H b : h b K k K : h = b k , so gK = a b kK = a b K ;
a , a {a }; b , b {b } : abK = a b K (a H are distinct in G ) a = a bK = b K , so
j

i j

i j

(bi K are distinct in H) b = b .


So
38.

39.
40.

i, j

i j

a G , so a divides G , that is m N : m a = G , so a n = a
The left cosets of  in (,+ ) are { + }  . Then :
x x = ( x ) ( x )

x , x 

=G :H H :K.

=a

m a

= em = e .

x , x + 

[ [

}  .

x , x [0,1[ ,

sense to write the sine function as sin : + 2

x = x.

Then

n  : x = + n 2

 + and x x 0, 1 < 1 , so x x = 0

The left cosets of 2 in (,+ ) are + 2


: x + 2

42.

Obviously H is itself one of the left cosets of H in G. Since there is just one other left coset, and since the cosets
form a partition of G, the other is G \ H . The same argument holds for the right coset partition, so the left and
right coset partitions are equal.

41.

U {a b K} = G is a distinct left coset partition of G, so G : K = {a } {b }

sin x = sin ( + n 2 ) = sin , so it does indeed make

}  [ 1, +1] .

a. The relation ~ is an equivalence relation because it is:


reflexive: a ~ a h H , k K : a = hak a = a ; h , k = e ;

symmetric: a , b G : a ~ b h H , k K : a = hbk h 1ak 1 = b , h 1 H , k 1 K b ~ a ;


transitive: a , b , c : a ~ b , b ~ c h , h H ; k , k K : a = hbk , b = h ck a = hh ck k , hh H , k k K , so
a ~c.
b. h H , k K : a = hbk a HbK .
43. a. Prove it is a subgroup because it satisfies the requirements of Theorem 1.4.14:

( o )c = c = c , so the subset is closed under the operation of S


The identity permutation e of A certainly has e (c ) = c , so e S ;

, Sc , c :

A;

c,c

Sc , c : 1c = c , so 1 Sc , c ;
so Sc , c SA .
b. The identity permutation of SA is not closed in Sc , d , so again by Theorem 1.4.14, Sc , d / SA .

c. Sc , d is one of the left cosets o Sc , c


44.
45.

SA

of Sc , c .

n N : i : 0 i < n , i is a generator of exactly one subgroup of  n , and conversely, any subgroup of  n must
be generated by i : 0 i < n , so it suffices to enumerate the generators of the subgroups. By Exercise 44, the
subgroups of  n are { d }d| n , and by Corollary 1.5.18,  d has d generators, so n = + d : d |n d .

46.

2.4 Direct Products and Finitely Generated Abelian Groups


1.

2.

(0, 0) (0,1) (0, 2) (0,3)


1 4 2 4
2 4 =
. The orders are
. There is no element of order
(1,0) (1,1) (1, 2) (1,3)
2 4 2 4
 2  4 = 8 , so it is not cyclic.
(0, 0) (0,1) (0, 2) (0,3)

 3  4 = (1,0) (1,1) (1, 2) (1,3) . The orders are


(2, 0) (2,1) (2, 2) (2,3)

3
3

4
12
12

2
6
6

12 . There are elements of order


12
40

 3  4 = 12 , so the group is cyclic.


3.
4.
5.
6.
7.
8.
9.
10.

11.

12.

lcm(2, 2) = 2 (by Theorem 9).


lcm(3, 5) = 15.
lcm(3, 9) = 9.
lcm(4, 6, 5) = 60.
lcm(4, 2, 5, 3) = 60.
 3  8  6  8,  3  8 = 24 (excepting the nonproper subgroup).

{ (0,1) , (1,0) , (1,1) } 2 2 .


{ (0, 0,1) , (0,1, 0) , (0,1,1) , (1, 0,0) ,
{(0, 0),(0,1), (0, 2), (0,3)}
{(0, 0),(1,1), (0, 2), (1,3)} .
{(0, 0),(0, 2), (1, 0), (1,2)}

14.

}
(1, 0, 2)} .
(0,1, 2)}

(0,1, 0), (1, 0, 0), (1,1, 0)


(0, 0, 2), (1, 0, 0),
(0, 0, 2), (0,1, 2),

60 = 2 2 3 5, so by Corollary 6
 60  4  3  5
 12  5
.
 20  3
 15  4

a. 4 ( 0, 18, 12, 6 ).
b. 12 (by Corollary 6,  3  4  12).
c. lcm(3, 4) = 12.
d. V  2  2.
e. 2 1 4 = 8.

15.

{2,3} = {1} =  12 . The left cosets are

16.

{4, 6} = {2} . The left cosets are

17.

{8, 6,10} = {2} . The left cosets are

18.

{ 2 , 1} = { 0 , 2 , 1 , 2}. The left cosets are

19.

{ 1 ,2} = { 0 , 1,2 , 1, 3,1 , 2, 2} = D 4 . The left cosets are

20.

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

21.
22.
23.

} 2 2 2 .

  2  1 V
2 2 V 2
, so the subgroups are
 2  2  2 / V

{(0, 0, 0),
{(0, 0, 0),
{(0, 0, 0),
13.

(1,0,1) , (1,1, 0) , (1,1,1) ,

{
Ui {(0, 0 ), (1, 0)} {(2, 0), (0,1)}

i {0}

i {0, 1}

 12 + i .

{2} + i .

i {0, 1}

{2} + i .

i { 0 , 1}

= (2,3) = (0,0), (2,3), (4, 2),(0,1)

8 = 2 3, giving  2  2  2,
4

{ 2 , 1} i .

i { 0}

{ 1 ,2} i .

= (2, 0),(0,1) . The left cosets are

i .
4 2 ,

16 = 2 , giving  2  2  2  2,

8 .

4 2 2 ,

4 4 ,

8 2 ,

 16 .

32 = 2 , giving
 2  2  2  2  2,

4 2 2 2 ,

4 4 2 ,

8 2 2 ,

8 4 ,

 16  2 ,

 32 .
41

24.

720 = 2 43 25, giving  2  2  2  2  3  3  5,  4  2  2  3  3  5,  4  4  3  3  5,


 8  2  3  3  5,  16  3  3  5,  2  2  2  2  9  5,  4  2  2  9  5,
 4  4  9  5,  8  2  9  5,  16  9  5.

25.

1089 = 3 2112, giving  3  3  11  11,

 9  11  11 ,

26.

24 = 2 33, giving 2 2 2 3,
25 = 5 2, giving 5 5, 25 ;
so 24 25 has 2 3 = 6.

8 3 ;

27.

Each commutative group of order m is isomorphic (by Theorem 12) to


order n to


i

p i

ri

4 2 3,

for some p i , p i , r i , r i . Then


i

pi

ri

 3  3  121 ,


i

p i


i

ri

pi

 9  121 .

ri

, and each commutative group of

is a group of order mn. Since p i p j

there is no rearrangement of factors between the two halves that gives the same order, so this product is unique for
the given halves.
Conversely, any commutative group of order nm can be written (by Theorem 12, reordering factors as required) as
 p ri
 p ri .

Thus there are exactly


28.


i

pi

ri


i

p i

ri

= rs groups.

10 5 = (2 5 ) = 2 555 . By (23) there are 7 groups of order k 5, so there are 49.


5

29.

a. For each order, the possible group factorings are:


2:
2, 11 (2)
3:
3, 21, 111 (3)
4:
4, 31, 22, 211, 1111 (5)
5:
5, 41, 32, 311, 221, 2111, 1111 (7)
6:
6, 51, 42, 411, 33, 321, 3111, 222, 2211, 21111, 111111 (11)
7:
7, 61, 52, 511, 43, 421, 4111, 331, 322, 3211, 31111, 2221 22111, 211111, 1111111 (15)
8:
8, 71, 62, 611, 53, 521, 5111, 44, 431, 422, 4211, 41111, 332, 3311, 3221, 32111, 311111, 2222, 22211,
221111, 21111111, 11111111 (22)
b. 3 5 15 = 225; 15 15 = 225 ; 22 5 = 110 .
30. a. true
b. true
c. false
d. true
e. false (  2  4 is not cyclic whereas  8 is)
f. false ( S8 = 8! whereas  2  4 = 8)
g. false? (there is no element of S4 of order 8 that generates the subgroup isomorphic to  8)
h. false ( = 1 )
i. true
j. true
31.
32.

{ }

Z 2 = 0, 1 .
a. 1, because every proper subgroup has fewer elements than the group.
b. , because n  * : n .

33.
34.

S3 = 3!= 6 .
a. true (Corollary 3.11)
b. false (the Klein 4-group V is not cyclic, and V = 4 = 2 2 )
c. false ( 1

{4, 6}

= 2 )

d. true ( {4, 5, 6} = 1 =  8)
42

e. true
f. false (  2 /  3 both have Betti number 0)
g. true (by Theorem 16 G  5i K, and  5i is cyclic)
h. false (it could be that G  2  2; but there exist G for which it is true)
i. false (by Theorem 16, there is no isomorphic factorization containing 6 )
j. true
35.

It is equal. For each commutative group of order p r the factorization of Theorem 16 gives the structure of a
corresponding group of order q r .

36.

72 = 2 332, so G must be isomorphic to one of  2  2  2 ,  4  2 ,  8 .


a. In each of the three cases, G has one subgroup of order 8.
b. In the first case, G has three subgroups of order 4; in the second case, two (  4 and  2  2  4  2); in the
third case, one.

37.

By Theorem 12, G i 

pi

Zp

ri

m . Then H = i 

pi

ri

E m , since  has no other elements of finite order, and

are finite so each of its elements are.

ri
i

38.

The torsion subgroup of  4   3 is  4 E  3, which has 4 3 = 12 elements; that of  12 Z  12 is


 12 E  12, which has 12 12 = 144 elements.

39.

The only elements of finite order in * form its torsion subgroup

40.

The only elements of finite order in * form its torsion subgroup

41.

By Theorem 12, every finitely generated commutative group is isomorphic to H = i 

.
pi

obviously torsion-free, and i 

pi

42.

{1, +1} .
{1, j , 1, j }

c. Let G = i j < ni 

pi

qij

ri

ri

m . E n is

is its torsion subgroup.

, i , j : qij qi , j + 1 , then T = j 

i :j <ni p i

qij

. For each prime pi , qij are its powers in the

factorization. Note that I reverse the order of the torsion coefficients because it simplifies the expressions.
a. G =  2 2 3 2 , so i = 0, 1; n0 = 1, n1 = 1; p0 = 2, p1 = 3; q00 = 2, q10 = 2 . Then T =  2 2 3 2 = 36 .
b. G =  2131  2 231  2 251 , so T =  2 2 31 51  2 2 31  21 = 60 12  2 from
43.

i = 0, 1, 2; p0 = 2, p1 = 3, p 2 = 5; n0 = 3, n1 = 2, n2 = 1; q00 = 2, q01 = 2, q02 = 1, q10 = 1, q11 = 1, q 20 = 1 .


( g , h), ( g , h ) G H : ( g , h) ( g , h ) = (g g , h h ) = ( g g , h h) = ( g , h ) (g , h) , so G H is
commutative.

44.

H = h G | h = 2 E

e H (identity)

{ }

{ }

h H : h = e , h .

h H : h 1 = h = e , h

h 1 = 2

h 1 H (inverse)

h, h H \ E : hh = h 1 h 1 = h 1 h 1 = (hh ) 1
so H G .
45.

a. H = h G | h = 3 E
e H (identity)

hh = 2 (closure)

h H : h = e , h , h 2 .

} {

h H : h 1 = h 2 = e , h 2 , h 4 = e , h 2 , h

h, h H \ E :

hh = e , h h

h 1 = 3

h 1 H (inverse)

h h : hh = e , hh ,(hh ) 2

h = h :
hh = h 2 = h 1

2
h = h : hh = e

hh H E (closure).

43

so H G .

b. H = h G | h = 4 E

{ }

h H : h = e , h , h 2 , h 3 . Then hh = e , h 2

hh 4

H
/ G.

c. For any n, the identity and inverse exist in the subgroup. Suppose n is divisible by m, then
h H :

(h )
m

n/ m

= hn = e

hm =

n
m

h m H , so n must be prime.

<n

47.

a. By Definition 1.
hk = (h,e ) (e, k) = (he ,ek) = (h, k)
b.
hk = kh .
kh = (e , k) (h,e ) = (eh, ke ) = (h, k)
c. h H, k K , h = k : (h,e ) = (e, k) h = e k = e H K = E .
48.
h, h H , k, k K : hk = h k (h,e ) (e , k) = (h ,e ) (e, k ) (he ,ek) = (h e ,ek ) h = h k = k . Also,
H K H K.
49.
Consider the factorization of any finite commutative group by Theorem 12. If it contains a factor of the form
 p  p , the group is not cyclic because that subgroup has no generator. Since the group is finite, it contains no 
factors. Any factors  p  q where p q have (1,1) as generator, but factors  p  p m  p  p have no
generator.
50.

By Theorem 12, any such group is isomorphic to G =


of p, so the order of any element of  p

51.

ri
i


i

pi

ri

. For each factor, p i

ri

is divisible only by a power

is a power of p. So the order of any element of G is (Theorem 9) the least

common multiple of powers of p, which is itself a power of p.


Is there a counterexample for noncommutative groups?
From the isomorphism,
: G K H K : ( g , k),( g , k ) G K : (( g , k) ) (( g , k ) ) = (( g , k) ( g , k ) ) = (( gg , k k ) ) , so then

: G H : g a 1( g ,e ) is an isomorphism between G and H.


52.

n r

It is easily verified that r < n 1 : (1 2 3 Kn ) (1 2 )(1 2 3 Kn ) = (r + 1 r + 2 ) . Then a,b : a < b ,


(a b) = (a a + 1)(a + 1 a + 2)K(b 2 b 1)(b 1 b)(b 2 b 1) K(a + 1 a + 2)(a a + 1). By Corollary
2.12, every g Sn is a product of such transpositions, so the given set indeed generates Sn .

53.

. (0,1)
. (1,0)

54.

a. G will be commutative when the inner and outer n-gons have the same orientation.
b.  2 n .

c. If n is odd, n = 2m + 1 :
d. The dihedral group.
55.
fx = sin 2x .

 2  n =  2  2m + 1
/  2  2 it is (49) cyclic.

44

56.
57.
58.
59.

60.
61.
62.

fx = sin

x.
3
f (x , y ) = sin 2x sin 2 y .
2
2
f (x , y ) = sin
x sin
y.
3
5
2
3
2 x + cos
2y
2
2

2
2
1(x , y ) = x +
, y ; 2(x , y ) = x , y +
2
3

f (x , y ) = sin 2x + cos3y = sin

y
f (x , y ) = sin12 arctan
x

y +5
f (x , y ) = sin12 arctan
.

x 3
the rotation over zero degrees is the identity isometry (identity)
if rot H

rot

if rot , rot H

rot

= rot H (inverse)

rot o rot G

rot o rot = rot + H (closure)

so H G . An isometry is either orientation-preserving or not, and isometry preservation is isomorphic to 2 (e.g.,


the composition of a preserving with a non-preserving function is non-preserving, 0 + 1 = 1). The isometries in H
are all the orientation-preserving ones. If there is at least one orientation non-preserving isometry in G, then
G H  2 G = 2 H . Otherwise, G = H.
63.
64.
65.

rotation
N
N

h-reflection
N
N

v-reflection
N
Y

glide
N
N

66.

Z Z2

67.

68.
69.
70.

Y
N
Y

Y
N
N

Y
N
Y

N
Y
Y

D Z 2
Z
Z D ?

72.

{
a. {0,180 } ;

73.
74.

a. no; b. no; c. no.


a. no; b. yes; c. no.

75.

a. 0,180 ; b. yes; c. no.

71.

isomorphism
Z
D

a. 0, 90,180 , 270 ; b. yes; c. no. (see left figure)


b. yes; c. yes. (see center figure)

77.

{
}
a. {0,120 , 240 } ;
a. {0,120 , 240 } ;

78.

a. no; b. no; c. yes; d. (1,0) and (0,1).

79.

a. 0, 90,180 , 270 ; b. yes; c. no; d. (2,0) and (0,2) (why does the book say (1,1) and not just
(1,0)?).

80.

a. 0,120 , 240 ; b. no; c. yes; d. (1,0) and (0,2)

76.

81.
82.

{
}
a. {0,120 , 240 } ;

b. yes; c. no. (see right figure)


b. yes; c. yes (? book says no).

b. yes; c. no; d. (0,1) and (1, 3) .

Space rotation of a cube is a permutation of its four diagonal axes, so G S4 . How many ways are there of
permuting them? Fix one arbitrary axis there are 2 4 = 8 ways of doing this. Then there remain three 120
45

rotations along that axis, giving a total of 8 3 = 24 permutations. So G = S4 .

2.5 Binary Linear Codes

1
6

3
2
3

1.
2.
3.
4.
5.

6.

0B007F7F2500257F39 (hexadecimal).
GONE_HOME.
x 4 = x 1 + x 2; x 5 = x 1 + x 3; x 6 = x 2 + x 3 .
000000, 001011, 010101, 011110, 100110, 101101, 110011, 111000.
An error in one bit generates an error in two parity bits; an error in two bits also generates an error in two parity
bits; an error in three bits is never detected.
One- and two-bit errors both generate errors in two parity bits, so only one type can be reliably corrected.

7.

C + {000111} = 000111 , 001100 , 010010 , 011001 , 100001 , 101010 , 110100 , 111111 .

8.

a. 110;
1

H = 1
0

9.

10.

b. 001;
1 0
0 1
1 1

c. 110; d. 001, 100, 111; e. 101.


1 0 0

0 1 0 .
0 0 1

a. H [110111 ] = [100 ]

110 .

b. H [001011 ] = [ 000 ]

001 .

c. H [111011 ] = [ 011 ]

110 .

d. H [101010 ] = [111 ]

not decodable.

e. H [100101 ] = [ 011 ]

101 .

T
T
T

11.

12.

000000
000
000001
001
000010
010
000100
100
001000
011
010000
101
100000
110
Hw
corrected
code
100
110011
110
000
001011
001
011
110011
110
111
incorrigible
011
101101
101
a. wt(u) = 7 ; b. wt(v) = 6 ; c. u + v = 1010011001 ; d. 5.

13.
14.

v n : v 1 n : v + v 1 = 0 i : v 1i = v i v 1 = v , so u v = u + v 1 = u + v.
Because it has a 1 bit in each position where a transmission error occurred.
46

15.
16.

17.
18.

Because u v has a 1 bit in each position where u differs from v.


a. d(u , v) = 0 u and v agree in each bit position u = v .
b. In every bit position where u differs from v, v differs from u.
c. If u differs from w in some bit position, then u differs from v or w differs from v in that position, so
d(u , v) + d(w, v) d(u, w) .
d. If ui = v i for some bit position i, then also ui + 0 = v i + 0, ui + 1 = v i + 1 , so u + w = v + w .
 n = ( 2 ) .
n

0000000 = e  n ; 0000000 C e C (identity)


x C : x = (x 1x 2x 3x 4 x 5x 6 x 7 )

)(

(inverse)

)(

= (x 1x 2x 3x 4 ) x 1 + x 2 + x 3 x 1 + x 2 + x 3 x 1 + x 2 + x 3

= (x 1x 2x 3x 4 )(x 1 + x 2 + x 3)(x 1 + x 3 + x 4 )(x 2 + x 3 + x 4 ) C


x , y C : x + y =
(x 1x 2x 3x 4 x 5x 6 x 7 ) + (y 1y 2y 3y 4 y 5y 6 y 7) =

(x 1 + y 1
(x 1 + y 1

x2 + y 2

x 3 + y3

x4 +y4

x2 + y 2

x 3 + y3

x4 +y4

x 5 + y5

x6 +y6

x7 + y 7) =

(x 1 + x 2 + x 3) + (y 1 + y 2 + y 3)
(x 1 + x 3 + x 4 ) + (y 1 + y 3 + y 4 )

(x 2 + x 3 + x 4 ) + (y 2 + y 3 + y 4 ) ) =

(x 1 + y 1

x2 + y 2

x 3 + y3

x4 +y4

(x 1 + y 1) + (x 2 + y 2) + (x 3 + y 3)
(x 1 + y 1) + (x 3 + y 3) + (x 4 + y 4)

(x 2 + y 2) + (x 3 + y 3) + (x 4 + y 4) ) =

((x + y ) 1

(x + y ) 2

(x + y ) 3

(x + y) 4

(x + y ) 5

(x + y) 6

(x + y ) 7 ) C

so C  .
16d

19.
20.

15

c, d C : c d : d(c, d) = d(c d, d d) = d(c d, 0) = wt ((c d) 0 ) = wt(c d) , where c d is some element of


C.
m
d = m +1
d

21.

m
d = 2m + 1
d

22.
23.

24.
25.

From Exercise 19, the minimum nonzero weight of code words is the minimum distance between code words.
Then we can detect 2t + 1 = m + 1 m = 2t and correct 2t + 1 = 2m + 1 m = t errors.
For there to be a minimum distance of 3 between code words, changing one bit in each of two code words may
map those two code words into the same coset. The number of cosets is thus the number of ways of changing 0 or
1 bits in a code word, so 2 n k 1 + n .
Similarly, the number of cosets is the number of ways of changing 0, 1, or 2 bits in a code word, so
2 n k 1 + n + 12 n(n 1 ) .
Simply try the formula with increasing value of n:
47

k
a. 2
b. 4
c. 8
d. 2
e. 4
f. 8

26.

27.

m
3
3
3
5
5
5

nk
3
3
4
5
6
7
1

1 1

1
1
1

G =
1
1
1 . By Exercise 24, n k 4.

1
1 1

1
1
1
1

1
G =

1 1

1
1
1
1

1
1 1
. By Exercise 24, n k 4.
1 1 1

1
1
1 1

1
1
13

28.

1
1
1
a. wt(0) = 0 0 H (identity). h H : h = h wt(h ) = wt(h) x H (inverse). Finally, see that
wt(x + y ) = wt(x) + wt(y ) 2a , where a is the number of positions where x i = y i = 1 (closure).
b. A word is either even or odd. Let x G be odd, then h H : xh is odd, and because G is a group, xH is a coset
of G so G = H xH .

29.

3.1 Homomorphisms
1.

a,b  : ( a +  b ) = a +  b .

2.

3.

a,b  : ( ab ) = ab = a b = a b .

4.

a,b  6 : a = 2a 2 + a 0 , b = 2b2 + b 0

1
2

+

1
2

) = 1 = 1,

1
2

+ 

1
2

= 0 + 0 = 0 .

) (

a +  6 b = 2 ( a 2 + b 2 ) +  6 ( a0 + b 0 ) = (a 0 + b 0 ) mod 2 =
5.

a 0 +  2 b 0 = ( 2a 2 + a0 ) +  2 ( 2b 2 + b 0 ) = a +  2 b

8 +  9 1 = 0 = 0; 8 +  2 1 = 0 +  2 1 = 1 . is an even-odd calculator, but in 9, 8 and 8 + 1 are both even.

6.

a,b  : ( a +  b ) = 2 a + b = 2 a 2 b = a b .

7.

gi , g i G i : i ( g i gi ) = (e 1,K, g i g i ,K,e r ) = (e 1 ,K, g i ,K,e r ) (e 1,K, g i ,K,e r ) = i g i i g i .

def.

8.

If G is commutative, g, g G :
not generally an isomorphism.

9.

f , g F : ( f + g ) =

10.

f , g F : ( f + g ) =

d 2( f + g )
dx 2

( gg ) = ( gg )
=

d2 f
dx 2

f + g dx =

4
0

d2 g

= g 1 g 1 = g 1 g 1 = g g . If G is not commutative, f is

= f + g .

dx 2

f dx +

g dx = f g .

48

11.

f , g F : ( f + g ) = 3( f + g ) = 3 f + 3 g = f + g .

12.

1 0 0 0
1 0 1 0
1 0
0 0 1 0 0 0

= 1;
+
= 0 + 0 = 0.
+
=
=
+
=
0 1 0 1
0 0
0 1 0 0 0 1
0 0 0 1

13.

A, B M n : (A + B) = tr( A + B) =

14.

Since GL(n , R ) Mn , the proof of Exercise 13 holds.

15.

f (x ) = x

(f + f ) =
(f ) =

a
i

+bi =

( f f )(x ) d x =
1

a +b
i

= trA + trB.

x 2 d x = 13 x 3

f (x) d x = x d x = 12 x 2
0

(f ) + ( f ) = x 2

16.

Ker = A3 .

17.

4 :   : n a 4n is a homomorphism by Example 7, and 7 :   7 : n a n mod 7 is by Example 10, so


= 7 o 4 , 1 = 4 1 mod 7 = 4 mod 7 = 4 is a homomorphism. Then
Ker = Ker 7 o 4 = ( 7 o 4 )

18.

Let = 10 o 6 ;
Ker =

19.

inv

inv

0 = 4

inv

7 =

0 = ( 10 o 6 )

inv

0 = 6

inv

10 =

25 = ( 7 o 4 ) 25 = 7100 = 2 .

10
6

5

3

 =

 = 5;

18 = ( 10 o 6 )18 = 10 108 = 8 .

First, note that (1 4 2 6)(2 5 7) = (1 4 2 5 7 6) . Let :  S8 : n a (1 4 2 5 7 6) n , then

20 = (1 4 2 5 7 6)

= (1 2 7)(4 5 6) .

1 = 8 , then Ker = inv 0 =

20.

Let :  10  20 : n a 8 Z 20 n;

21.

Let :  24 S8 : n a ((2 5)(1 4 6 7) ) ;


n

20
8

  10 = 10 + 5;

3 = 8  20 3 = 4 .

1 = (2 5)(1 4 6 7) , then

Ker = inv 0 = 4  24 = {0, 4,8,12,16, 20};

23.

 = 7;

1 = 6 . Then

Ker = inv() = 6;

22.

7

4

14 = ((2 5)(1 4 6 7) ) = (1 6)(4 7) .


2

Let :    : (x , y ) a 3x 5y ; (1, 0) = (2, 3), (0,1) = (1,5) , then


Ker = inv 0 = {(x , y)   | 3x 5y = 0}; ( 3, 2 ) = 3 3 5 2 = 9 10 = 19 .
Let :     : (x , y ) a ( 2x y , 3x + 5y );

(1, 0) = (2, 3), (0,1) = (1,5) , then

Ker = 0 = {(x , y)   | 2x y = 0 3x + 5y = 0} = {(0, 0)} , because


2x y = 0
y = 2x
y = 0



, and (4, 6) = (2 4 6, 3 4 + 5 6) = (2,18) .
3x + 5y = 0
3x + 10x = 7x = 0
x = 0
inv

24.

:   S10 : (x , y) a ((3 5)(2 4) ) ((1 7)(6 10 8 9) ) ; (1, 0) = (3 5)(2 4), (0,1) = (1 7)(6 10 8 9) ,
x

then Ker = inv 0 = 2 4;

(3,10) = ((3 5)(2 4)) ((1 7)(6 10 8 9) ) = (3 5)(2 4)(6 8)(10 9) .


1

25.

There are two: 1 :   : i a i;

26.

n N :

27.

There are two: 1 :   2 : i a i mod 2; 0 :   2 : i a 0 .

28.

x , y G : g (xy ) = g x g y

g ( xy ) = gx gy

29.

x , y G : g ( xy ) = g x g y

g(xy) g 1 = gxg 1 gyg 1 = gxyg 1

30.
31.
32.

1 :   : i a i .

n :   : m a nm .
xy = xgy

y = gy

g=e.

g G .

A group homomorphism of a group G into a group G' is a map : G G such that for all x , y G
OK
a. true (odd times even equals even times odd)
b. true (the trivial homomorphism)
c. false (the trivial homomorphism)
d. true (Corollary 18)
e. false (there are 4 cosets in G, but 4 does not divide 6)
f. false ( is a function, so G G )
49

g. true (the trivial homomorphism)


h. true (the trivial homomorphism)
i. false ( e e = e e = e e Ker )
j. false ( :  2   2 : i a (0,i) )
33.
No, there must be 5 cosets in G, but 5 does not divide 12.
34.
:  12  4 : i a i mod 4.
35.
36.

:  2  4  2  5 : (i , j ) a (i, 0) .
No.

37.

:  3 S3 : i a (1 2 3 ) .

38.

:  S3 : i a (1 2 3 ) .

39.

:   2 : (x , y ) a 2x .

40.

: 2   : i a (i , 0 ) .

41.

, , , a (1 2) 0
: D4 S 3 : 0 1 2 3
.
1, 2, 1 , 2 a (1 2) 1

42.

: S3 S 4 : s a s .

43.
44.

s even a (1 2) 0
: S4 S 3 :
.
s odd a (1 2) 1
partitions G into |G| cosets (Theorem 15), so |G| divides |G|. Also, since is a function
G < G G < G < .

45.

G G

G < G < . Also, G G

G divides G .

46.

g G : g = i a k i

47.

By Exercise 44, G divides G , so G = 1 (trivial homomorphism) G = G (injective map) .

48.

Obvious. Ker = An .

g = i a ki = i a ki = i a ki = i a ki = g

49.

g, g G : ( gg ) = (g g ) = g g .

50.

G commutative

h, h G H : hh = h h

h = h hh 1

Theorem 12.2

g, g G : g gg 1 g 1

g, g G : g gg 1 g 1 Ker

51.

n,m  : (nm ) = a nm = a n a m = n m .  = a ,

52.
53.

m,n   : (mn ) = m n

= .

g g (g )

Ker = 

h hh 1h 1 = e .
(g )

=e

(where  E) .

((m1,m 2) (n 1 ,n 2)) = ((m1,m 2)) ((n1 ,n 2))


(m 1n1 ,m 2n2) = (m 1,m 2 ) (n 1 ,n 2)
h m 1 n 1 k m2 n 2 = h m 1 k m 2 h n1 k n 2
kh = hk, so

54.

h, k G : hk = kh

55.

i, j  n : (ij ) = i j

{h , k}

is commutative

G is commutative .

i + Zn j

= h ih

h = e

h = n

hn =e .

3.2 Factor Groups


50

1.

6

3 = 6

2.

 4  12

3.

 4  2 (2,1) = 8 2 = 4 .

4.

 3  5 {0}  5 =  3 {0}  5  5 = 3 1 = 3 .

5.

2 4

6.

 12  18 (4,3) =  12  18

7.

 2 S3 (1, 1) =  2 S3

8.

 11  15 (1,1) =  11  15

9.

5+ 4

10.

26 + 12

11.

(2,1) + (1,1)

12.

(3,1) + (1,1)

13.

(3,1) + (0,2)

14.

(3,3) + (1,2)

15.

(2,0) + (4,4)

16.

i 1 : S 3 S3 : a 11

17.

Replace for all h H with for all g G .

 12

3 = 6/2 = 3.

2 2 = 4

2  12

(1,1) =  2  4

4
 60

=1+ 4
12

(1,1) = 8 4 = 2 .

 12

(1, 1)

 4  4

(1, 1)

 4  8

(0, 2)

 4  8

(1, 2)

 6  8

(4,3) = 216 6 = 36 .

(1, 1) = 12 6 = 2 .

= 2 + 12

 3  6

2 = 2 2 = 4.

(1,1) = 161 161 = 1 .


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

 60

12

= {2,4, 6,8,10,12 = 0} + 12 = 6 .

= (1, 0) + (1,1)
= (2, 0) + (1,1)

 3  6

(1, 1)

= {(1,0), (2, 0),(3 = 0, 0)} + (1,1) = 3.

 4  4

(1, 1)

= {(2,0), (4 = 0, 0)} + (1,1) = 2 .

= {(3,1),(6 = 2, 2 = 0), (5 = 1,1), (4 = 0, 2 = 0)} + (0, 2) = 4 .


= (4 = 0,5) + (1, 2)

 4  8

(1, 2)

= (2 = 4, 4 = 4) + (4, 4)

(4, 4)

: i 1{ 0 , 1} = 1 0 1

= {(0,5), (0,10 = 2 = 0)} + (1, 2) = 2 .

 6  8

(4 , 4 )

= (0, 0) + (4, 4)

= 0 , 1 1 1

 6  8

(4, 4 )

=1.

= 2 .

The book definition says ghg 1 H , but this definition is equivalent.


Replace into with onto. This makes the homomorphism an isomorphism, which is what an automorphism is
supposed to be.
20.
A normal subgroup can be used to form a factor group.
21. a. This doesn't necessarily have to be nonsense, but apparently students that write a G H don't realize that
a = g a H , g a G . Since they don't realize that elements of G/H are sets (cosets of H), the proofs make no sense.
b. Let aH and bH be two elements of G/H.
18.
19.

c. aH,bH G H : (aH)(bH) = a(Hb)H


22. a. See Exercise 21a.
b. See Exercise 21b.

( gH ) n

c. gH G H : n N : g n = e
23.

H normal

G commutative

a(bH)H

Theorem 4

(ba)HH

H normal

b(Ha)H = (bH)(aH) .

g nH = eH = H , which is the identity element of G/H.

a. true (if N is not normal, the factor group does not exist Definition 6)
b. true (Example 8)
c. true ( i g : G G : x a gxg 1

commutative

xgg 1 = x )

d. true (G cannot have more cosets than elements)


e. true (Exercise 22)
f. false (  2 = 2 )
g. true (Exercise 21)
h. false (G/G is commutative)
i. true (Example 7)
51

j. false ( n = 

 n = E )

All permutations of A are even, and those of S\A are all odd. {A, S \ A} are the cosets of A in S. If Sn is even,
so are A = A ; similarly if is odd.
evenA evenA
Z2 0 1

24.

evenA evenA oddA


oddA oddA

evenA

0 0

1 1

The group is isomorphic to  2.


25.
26.

G is commutative, so the subgroup T is normal in G.


gT G T : n N * : ( gT ) = T
n

27.

H G :

g nT = T

i e (H ) = H (reflexive)

H,K G : H ~ K

g G : i gH = K

) ( )(

k K : h H : i g h = k

)( )

i g 1 k = i g 1i g h = i g 1 ghg 1 = g 1 ghg 1 g 1
H,K ,L H :

H ~ K,K ~ L

h H : i g g h = ( g g )h ( g g )
i g gH = L
28.

=h

i g 1K = H

(symmetric)
K ~H

g, g G : K = i gH , L = i g K

(transitive)

( )

= g ghg 1 g 1 = g ghg 1 g 1 = g i g h g 1 = i g i g h = i g i g h

H ~L

If H is normal to G, then by the discussion after Definition 9, the image of H under all the inner automorphisms is
H itself. So H is normal iff its cell of the partition under conjugacy contains only itself.

29.

{i

30.
31.

Let H,K G be normal. g G :

= i 3 ,

i 0 = i 2 ,

} {{

i 1 = i 1 =

0 , 1

}, { 0 , 2}, { 0 , 3}} .

g (H K) Hg g (H K ) Kg

From the converse, g(H K ) = (H K) g .

32.

g (H K ) Hg Kg = (H K ) g .

Suppose there were two distinct smallest normal subgroups containing S, then their intersection would be smaller,
contain S, and be (Exercise 31) normal.

33.
34.

If G has one subgroup H of order |H|, then H must be invariant under all inner automorphisms, so (by the
discussion after Definition 9) H is normal.
H N H by Exercise 1.5.54. h H : h (H K) H K h (H K ) = H K , and by the converse,

35.

(H K ) k = H K , so h(H K ) = (H K) h .

36.
37.

g n T , so G /T is indeed torsion-free.

a. g G : i e o i g : x G : i e o i g x = i e i g x = i g x

i e o i g = i e (identity)

g G : i g 1 o i g : x G : i g 1 o i g x = i g 1 gxg 1 = g 1 gxg 1 g = x

i g 1 o i g = i e (inverse)

g, h , k G : i g o i h o i k = i g o (i h o i k ) because function composition is associative (associative)


b.
38.
39.

Let * : G H G H : gH a (g )H . This is a homomorphism if gH, g H G H ,

*( gH g H ) = * gH * g H * (( gg )H ) = * gH * g H

( gg )H = (g H ) (g H ) = (g g ) H ( gg ) = g g
which holds because is an isomorphism.
40. a. H = {M GL (n , R) | det M = 1} is normal in G because
g G, h H : ghg 1 H

det ghg 1 = det g det h det g 1 = det g det h ( det g)

= det h = 1 .
52

b. H = M GL (n , R) | det M = 1 is normal in G because of a similar argument.

41.

(AB )C = {ab | a A,b B}C = {( ab)c | a A,b B,c C } = {a (bc ) | K} = A{bc | K} = A(BC )
(associativity). E G H G : EH = {eh | e E , h H} = {h | h H} = H (identity). Suppose G has an
inverse G' in its power set, then GG = E { gg | g G , g G } = {e } , but { gg | K} G E .

a. A,B,C G :

b.
c. Let M = {m G | m is a coset of N } = { gN | g G } . The operation is associative, as shown in (a).

gN M : ( gN )N = { g}NN = { g}N N , so N is an identity in M. Finally, because N is normal in G,

gN M : g 1N ( gN ) =

({ g }N ) ({ g}N) = { g }N{g}N = ({g }{ g})NN = {e }N = N (inverse).


1

3.3 Factor-Group Computations and Simple Groups


1.

 2  4 (0,1)  2 {0}  2 .

2.

 2  4 (0, 2)  2  2 .

3.

 2  4 (1, 2)  4 .

4.

 4  8 (1, 2)  8 .

5.

 4  4  8 (1, 2, 4) =  4  8 .

6.

  (0,1) =  .

7.

  (1, 2) =  .

8.

   (1,1,1) =   .

9.

   4 (3, 0, 0) =  3   4 .

10.

   8 (0, 4, 0) =   4  8 .

11.

  (2, 2) =  2  .

12.

   (3,3,3) =  3   .

13.

14.

15.

ZD4 = { 0 , 2} . Is the center a natural choice for a minimal normal subgroup? In any case,

D4 Z = { 0 Z , 1Z , 1Z , 1Z } is commutative, by manual verification, so by Theorem 20 Z C . But


D4 E D4 is not commutative, so E
/ C C =Z.
First, note that for any commutative group a,b G : aba 1b 1 = ab (ba ) = (ab ) ( ab ) = e , so CG = E . Then,
Z  3 =  3, and by Example 19 Z S3 = E . Further, C  3 = E , and by Example 21 C S3 = A 3. So
Z(  3 S 3) =  3 E , C (  3 S3 ) = E A3 .
1

Z( S 3 D4 ) = Z S3 ZD 4 = (Example 19, Exercise 13) E { 0 , 2}

C( S 3 D4 ) = C S3 CD 4 = (Example 21, Exercise 13) A3 { 0 , 2}


16.

Subgroups of  4  4 with one generator (cyclic):


||

4 4 < >

<>

{<>}

(0,0)

(0,0)

4 4

(0,1)

(0,0) (0,1) (0,2) (0,3)

4 1

(0,2)
(0,3) ~ (0,1)
(1,0)

(0,0) (0,2)

4 2

(0,0) (1,0) (2,0) (3,0)

1 4

(1,1)

(0,0) (1,1) (2,2) (3,3)

4

(1,2)

(0,0) (1,2) (2,0) (3,2)

 4 (figure left)

(1,3)

(0,0) (1,3) (2,2) (3,1)

 4 (figure center)
53

(2,0)

(0,0) (2,0)

2 4

(2,1)

(0,0) (2,1) (0,2) (2,3)

4

(2,2)
(0,0) (2,2)
2
2 4
(2,3) ~ (2,1)
(3,n) ~ (1,n)
Subgroups with two generators (not cyclic), with order less than or equal to 4:
(0,2) (2,0)
(0,0) (0,2) (2,0) (2,2)
4
V (figure right)
(0,2) (2,2) ~ (0,2) (2,0)
(2,0) (2,2) ~ (2,0) (0,2)
There are no subgroups with more than two generators with order less than or equal to 4.

17.
18.
19.

The center of a group G is a set containing all


The book uses aba 1b 1, but this definition is equivalent.
a. true (Theorem 9)
b. false (by Exercise 16, G G E )
c. false (

1
2 

d. true (

1
)
n  

e. false ( 1 12

 

1
2 

= 0  )

1
).
2 

f. true (Exercise 14)


g. false (not C H but H C )
h. false (when G is simple and commutative)
i. true (By Theorem 20, the commutator subgroup is normal to G, so if G is simple then C is trivial or nonproper.
But if C were trivial, then G E G would be commutative. So C is nonproper.)
j. false (by Theorem 15, A5 is nontrivial, finite, simple, and of 5! nonprime order)
20.
21.
22.

23.
24.

{f

{f

F | f 0 = 0} F .

F * | f 0 = 1 F * .

The cosets each represent a specific additive discontinuity, of the form a ( x b ) , where is the step function. An
element of order two would represent a discontinuity that is its own inverse, which under addition could only be the
identity discontinuity, which has order one.
See Exercise 22. Each discontinuity with a < 0 is its own inverse under multiplication, and has order two.
z 0U =U U z 0U E .

25.

1 U = {1,+1}; U 1 U .

26.

z n ; U z n U .

27.

  [ 0,1[ U .

28.

 has  2  2 .

29.
30.

Let G =  2  4 , then  1  2  2  1 but  2  4  2  1 =  2  2 /  2  4  2  1 =  1  4 .


a. The center of every commutative group consists of all the elements of that same group.
b. Suppose z ZG : g G : zg = gz

z n z , n  : z n g = gz n , so z is normal to G and, since G is

simple, wither trivial or nonproper. Since z is commutative but G is not, z E , so Z = E .


54

a. If G is commutative, then G G E is too, so by Theorem 20 C E C = E .


b. Since G is not commutative, by the same argument C
/ E C E . Since G is simple, C must then be
nonproper.
simple group G
not commutative
commutative
center ZG
E
G
commutator CG
G
E
32.
If H G such that G H exists, then H is normal to G. Since G : H > 1 H G , and since H is nontrivial
H E . So H is a proper nontrivial normal subgroup of G, so G is not simple.
33.
g G : g N = N g inv g N = N inv g inv g N = N inv g

31.

inv g N = N inv g
34.
35.

) (

g inv g : gN =Ng N is normal.

Suppose G ZG is cyclic, then g * ZG G ZG : g * ZG = G ZG , and


g G : n  : g ( g * ZG ) = g * n ZG
n

z ZG : g = g * n z . Then

)(

g, g G : z , z ZG , n ,n  : gg = g * n z g * n z = zg * n g * n z = z ( g *)

)(

n +n

z = zg * n g * n z =

g * n zz g * n = g * n z zg * n = g * n z g * n z = g g
so G is commutative. So if G is not commutative, G ZG is not cyclic.
36.

37.

Since G = pq , the order of any subgroup of G must (Lagrange) have order pq, p, q, or 1, and the resultant factor
group must therefore have order 1, q, p, or pq. By Exercise 35, the factor group G ZG is not cyclic. Since all
groups of prime order are cyclic, the factor group must have order pq, so ZG = 1 ZG = E .
a. (i j k ) = (i j ) ( j k ) , so every 3-cycle is the even product of transpositions and is therefore in An . Obviously An
only contains 3-cycles if n 3.
b. An consists of all products of even transpositions. Every type of even transposition

(a b )( a b) = ( a b c ) 0 ; ( a b ) (a d ) = (a d b ) ; ( a b ) (c d ) = ( a c b) ( a c d )
can be formed from 3-cycles, An is generated by the 3-cycles.
c. For any r, s: (r s i ) (r s k) (r s j
An and therefore An itself.
2

) 2 (r s i ) = (r i s )(r

s k ) (r j s) (r s i ) = (i j k ) , so

d. Let N be normal to An and (r s i ) N , then j : (r s) (i j

{i (r s i )} N

N = An .

{i (r s i )} generates every 3-cycle in

) ) (r s i ) 2 ( (r s ) (i j ) ) 1 = (r

sj

) N , so

e. First, canonicalize the elements of N into products of disjoint cycles. Then, one of the following cases must hold:
1 N contains a 3-cycle, so by (d.) N = An .
2 N contains a product in which at least one of the cycles has length greater than 3, = ( a 1 a 2 a3 a r ) . Then
N

( a 1 a 2 a3 ) (a 1 a2 a 3) = ( h N ) h N, and
1
1
1
1
( a 1 a 2 a3 ) (a 1 a2 a 3) = ( ( a1 a 2 a 3 a r ) ) ( a1 a 2 a 3 ) ( a 1 a 2 a 3 ) =

( a 1 a 2 a3 ar ) 1 1 ( a1 a 2 a 3 ) ( a 1 a 2 a 3 a r )( a 1 a 2 a3 ) 1
( a 1 a 2 a3 a r ) 1 (a 1 a2 a3 ) ( a 1 a 2 a 3 a r ) ( a 1 a 2 a 3 ) 1 =
( a 1 a 3 ar ) ( a2 ) ( 3<k <r a k ) = (a 1 a3 a r )

disjoint

so Case 1 applied.
3 N contains no single 3-cycle or products with cycles of length greater than 3, but contains a product of at least two
3-cycles, = ( a 1 a 2 a3 ) ( a4 a 5 a 6 ) . Then

( a 1 a 2 a4 ) (a 1 a2 a 4 )

= ( h N ) h
1

N, and
55

1( a 1 a 2 a4 ) (a 1 a2 a 4 )

disjoint

( a 4 a 5 a 6 ) 1 (a 1 a2 a 3) 1 ( a 1 a 2 a 4 ) (a 1 a2 a 3) ( a 4 a 5 a 6 ) (a 1 a2 a 4 ) 1 =
( a 1 a 4 a2 a3 a 6 ) ( a 5 )

so Case 2 applied.
4 N contains no products with cycles of length greater than 3, no products with more than one 3-cycle, and no 3cycles, but contains a product with one 3-cycle, = ( a 1 a 2 a3 ) , where is an even product of 2-cycles. Then
N

disjoint

2 N, and 2

2( a 1 a 2 a3 )

are
transpositions

( a1 a 2 a 3 ) 2 = (a 1 a3 a2 ) , so Case 1 applied.

5 N contains no products containing cycles of length greater than or equal to 3. Since N is nontrivial and consists
solely of products of even transpositions, it must contain an element = ( a 1 a 2 ) (a 3 a4 ) . Then

( a 1 a 2 a3 ) (a 1 a2 a 3)

1( a 1 a 2 a3 ) (a 1 a2 a 3)

Call this product .

= (h N) h

N, and

disjoint

( a 3 a 4 ) 1( a 1 a 2 ) 1 (a 1 a2 a 3) ( a 1 a 2 )( a 3 a 4 ) (a 1 a2 a 3) 1 = ( a 1 a 3 ) (a 2 a4 ) .
Since n 5, there is an a5 , and let = (a 3 a1 a5 ) . Then

( ) ( )

1 = 1 1

= ( N) N , and

1 = (a 3 a1 a5 ) ( a 1 a 3 )( a 2 a 4 ) (a 3 a1 a 5) ( a 1 a 3 )( a 2 a 4 ) = ( a1 a 5 a 3 )( a 2 )( a 4 ) = (a 1 a5 a 3) ,
so Case 1 applied.
So, Case 1 always applies, so N = A5 .
1

38.

(closure) hn, h n HN : ( hn ) (h n ) = hn h n

N normal

(identity) hn HN : (ee ) ( hn ) = eehn = ehn = hn .


(inverse) hn HN : (hn )

(n N , h H) hnn h = h (nn ) h =
(n N , h H ) hh n = ( hh )n HN

( hn) = n 1h 1hn = n 1n = e .

So HN G . A subgroup containing both N and H must contain at least


N H = {i n i} { i hi } ij n i hj = HN , so HN must be the smallest subgroup that does.

39.

M is normal to G

M G

NM G . Then

nm NM, g G : g (nm ) g 1 = gng 1 gmg 1


40.

N, M normal

(n N , m M) n m NM , so NM is normal in G.

K normal

hkh 1 k 1 = ( k K ) k k 1 K

h H, k K : hkh 1k 1 =
= e , so C = E ,
H normal

1 1
= ( h H ) hh H
h kh k
so E C
H K E H K is commutative.

3.4 Series of Groups


1.

The two series


{0} 10 

{0} 25 

(  10 ,
(  25 ,

10 {0}  10

)
25 {0}  25)

have isomorphic refinements


{0} 250  10 

{0} 250  25 

2.

The two series


{0} 60 20 

{0} 245  49 

(, 3 , 20 )
(, 5 , 49 )
56

have isomorphic refinements


{0} 14700  300  60 20 

{0} 14700  735  245 49 

3.

The two series


{0} 3 24

{0}

4.

( 8, 3 )

8  24 (  3 ,  8 )
are already isomorphic.
The two series
{0} 18 3 72 (4 , 6 , 3 )

{0} 24 12 72 (3 , 2 , 12 )


have isomorphic refinements
{0} 36 18 9 3 72
{0}

5.

24 12 6 2  72

The two series


( 0, 0) 60  10   

(  , 6 E , 10 E )
{ }
{( 0, 0)}  80  20   (  , E 4 , E 20 )

have isomorphic refinements


( 0, 0) 60 80 60 20 60  10   

{ }
{( 0, 0)} 60 80 10 80  80  20  

6.

(this is not the answer the book gives, but seems okay)
Because 60 = 2 2 3 5, the composition series are of the form
 60 2  2 2 = 4  2 2 3 = 12  2 2 3 5 = 60
60

7.

8.

60

60

 60

=E

where the series of generators are formed from the following 12 permutations of the factorization of 60:
2235
3225
2253
3252
2325
3522
2352
5223
2523
5232
2532
5322
The series that are thus constructed are obviously isomorphic.
As in Exercise 6, the series of generators are formed from the following 5 permutations of the factorization of
48 = 2 2 2 2 3:
22223
22232
22322
23222
32222
5 3 E 3 E E = E
5 3 5 E E E = E

9.

S3  2 A 3  2 E  2 E E = E
S3  2 S 3 E A3 E E E = E
Isn't the following a composition series too?
S3  2 A 3  2 A3 E E E = E

57

10.

2 5 7 E 5 7 E E 7 E E E = E
2 5 7 E 5 7 E 5 E E E E = E
2 5 7 2 E 7 E E 7 E E E = E
2 5 7 2 E 7 2 E E E E E = E
2 5 7 2 5 E E 5 E E E E = E
2 5 7 2 5 E 2 E E E E E = E

11.

Z( S 3  4 ) = Z S 3 Z  4 = E  4 .

12.

Z( S 3 D4 ) = Z S3 ZD 4 = E { 0 , 2} .

13.

E E, E  4 , .

14.

15.
16.
17.

E E, E { 0 , 2}, E D 4 ,

Since D4 { 0 , 2} is of order four, it is commutative, so Z D 4 { 0 , 2} = D 4 { 0 , 2} which maps to D4 under


the canonical homomorphism.
Okay.
A solvable group is one that has a composition series of which the factor groups are all commutative.
a. true ( G i < G G i < G i + 1)
b. false ( G i < G i + 1
/ Gi < G)
c. true
d. false
e. false ( E 3   15 and
15

E 5

 15

 15 )

f. true ( E G can be finitely refined)


g. false ( S7 is not solvable (h.), but E A7 S 7 is a composition series with simple factor groups)
h. false ( E A7 S 7 is a composition series, but A7 is not commutative)
i. true
j. true (Every finite group of prime order is cyclic and thus commutative. Every finite group has a composition series,
and each of the factor groups is commutative because each of the numerator groups is a commutative subgroup of a
commutative group.)
18.
S3 S 3 A3 A 3 E E; S 3 S3 A 3 A3  2  2 , A3 A 3 E E A3 A 3
is solvable because it has a composition series with commutative factors.
19.

D4 { 0 , 2} E ; D 4 { 0 , 2} = 4, { 0 , 2} E = 2
is solvable because it has a composition series with commutative factors.

20.

Hi + 1 = Hi + 1 Hi Hi
Hk =

21.
22.

Hk
H0

= i

Hi + 1 Hi = Hi + 1 Hi

Hi + 1
= s1 sk
Hi

Suppose i H i is such a composition series. Then k: H k < , H k + 1 = . But then H k + 1 H k is commutative


with infinite order, so by the Exercise it has a proper normal subgroup and is this not simple.
Concatenate the factor groups' composition series into a series for the product group:
E i E k 0 i >1 j <k H j , nj H k , i j >k H j , 0
i H i, n i

3.5 Group Action on a Set

It is simplest to think of a G-set as a group of functions, where the group operations is just function composition.
The functions operate as permutations on some set.

58

1.

g G

2
1

s1 , s 3
s2 , s 4
m1 ,m 2

{ 0 , 1}
P2 ,P4 { 0 , 2}
{1, 2,3, 4}, {s 1, s 2 , s 3, s 4 }, {m1 ,m 2}, {d 1, d 2}, {C }, {P1 ,P2 ,P3 ,P4 } .
2

9.

D4

2,4

7.
8.

1
3

4.
5.
6.

d1 , d2

1,3

{C }
{m 1 ,m2 , d 1, d 2 ,C}
{C }
{s 1 , s3 ,m 1 ,m2 ,C ,P1 ,P3}
{s 2 , s4 ,m 1 ,m2 ,C ,P2 ,P4 }
{2, 4, d1 , d2 ,C }
{1,3, d1 , d2 ,C }

Gx

{ 0 , 2}
{ 0 , 1}
{ 0 , 1}
{ 0 , 2}
{ 0 , 2 , 1, 2}
{ 0 , 2 , 1 , 2}

0 = e X
2

3.

2. x X

Xg

P1 ,P3

Insert x X .
Insert x X and delete other.
The G-set can be thought of as a direct product of its orbits. A sub-G-set consists of a subset of the orbits of the Gset.
A G-set is transitive iff it has exactly one orbit.
a. false (the elements of a G-set are not associative)
b. true (Definition 1, Condition 1)
c. false (G may not act faithfully)
d. true (g are permutations, which are injective)
e. false (any number of distinct permutations may operate on any particular element in the same way)
f. true (Exercise 7)
g. true ( H G automatically abides by the same Conditions of Definition 1)
h. true (they are the same orbits, but not necessarily all of them)
i. true (Example 2)
j. true (G consists of Gx cosets of G x elements, each coset of which permutes x in a different way in its orbit)
a. : {s 1 , s 2, s 3 , s 4} {P1 ,P2 ,P3 ,P4 } : s i a Pi .

b. 2 G leaves 1 and 3 fixed in their orbit, but leaves no elements of the orbit {s 1 , s 2, s 3 , s 4} fixed.

c. {m 1 ,m 2} and {d 1 , d2} are not isomorphic. But trivially, any direct product of the two isomorphic sets of (a.) with
any other orbit, is again isomorphic.
10. a. Yes, there is only e G that leaves all the elemens of X fixed.
b. {1, 2,3, 4}, {s 1, s 2 , s 3, s 4 }, {P1 ,P2 ,P3 ,P4 } .
11.
12.

(identity) e G : x X : ex = x

y Y X : ey = y

(closure) g, g G Y : y Y : ( gg ) y = gg y = gy = y

gg G Y

( g g) y = g

(inverse) g G Y : g 1 G : y Y : g 1 g y = ey = y
13.

a. (identity) 0 G = (,+ ) :

a. Let X =

( gy ) =

g 1y = y

g 1 G Y

x 2 : rot 0 x = x

(associativity) , G : x 2 : rot rot x = rot


b. The circle centered around the origin containing P.
c. GP = 2.
14.

GY

x.

UX .
i

(Condition 1) e G : x i X i : ex i = x i

x X : ex = x

(Condition 2) g, g G : x i X i : ( gg ) x i = g( g x i )

x X : ( gg )x = g ( g x )
59

b. By Theorem 14, any G-set X can be partitioned into its orbits.


15.
Let : L X : gG x 0 a gx 0 .
(well-defined) Let g, g gG x 0 be elements of the same coset of G x 0 . Then g G x 0

( )

g * G x 0 : g = gg * ,

so g = g x 0 = gg * x 0 = g g *x 0 = gx 0 = g .
(surjective) Because X is transitive, x X : g G : gx 0 = x
g 1G x 0 g G x 0 G x 0

(injective) g G x 0 gG x 0
g

( g x 0 ) x 0

16.
17.

gG

g 1 g G x 0

(g

g x0 x 0

g g

g x 0 gx 0

So is an isomorphism from

gG x 0 = gx 0 = x

gG x 0 X .

Every G-set is the union of its orbits (Exercise 14b). An orbit is a transitive G-set, so every G-set is (Exercise 15)
isometric to a union of G-sets of left cosets in G. By the Exercise, this union can be made disjoint.
a. G x 0 g0 are the actions g G that leave g0 x 0 fixed. If we move g0 x 0 into x 0 , act leaving x 0 fixed, and return
x 0 to g0 x 0 , we have actions that leave g0 x 0 fixed, so G g 0 x 0 g 0G x 0 g 0

. Conversely, any action that leaves

g0 x 0 fixed can be converted into one leaving x 0 fixed, so G x 0 g0 G g 0 x 0 g 0 , from which


g 0G x 0 g 0

G g0 x 0

G x 0 g 0 = g 0G x 0 g 0

b. It seems reasonable that H K if g G : K = gHg 1, that is K is inner automorphic to H, that is (Exercise 3.27)
K is conjugate to H.
c.

3.6 Applications of G-Sets to Counting


1.

The group has one permutation that leaves all 8 elements invariant, and 3 others that leave 4 invariant:
1
r=
+ g G X g = 1 8 + 3 4 = 20 = 5 .
4
4
G

2.

The group has one permutation that leaves all 8 elements invariant, one (1 3) that leaves 6 invariant, two (2 4 7)
and (2 7 4) that leave 5 invariant, and two more that leave only 3 elements invariant:
1
r=
+ g G X g = 1 8 + 6 + 2 5 + 2 3 = 30 = 5.
6
6
G

3.

G is the group of 12 rotations of the tetrahedron, and X is the set of 4! markings. The identity rotation leaves all
markings invariant; because every face has a different color, every other rotation none:
1
r = 12
(4!) = 2 .

4.

G is the group of rotations of the cube: there are six ways to fix one face, then four ways to fix a second, so G = 24 .
X is the set of 8! 2! markings. As in the previous exercise, there is only the identity rotation leaving all markings
invariant:
1 8!
r = 24
= 840 .
2!

5.

The identity rotation leaves all 8 6 markings invariant. The 9 rotations that leave a pair of faces invariant can be
divided in three groups (rotations along the x, y, and z-axis) of 3 rotations: one of which rotates the cube 180
along the axis, which leaves four independent choices of color for markings that remain invariant under the rotation;
and two which rotate the cube 90, and leave only three independent choices of color. The 8 rotations that leave a
pair of opposite vertices invariant are 120 rotations along the four diagonal axes that leave only two independent
choices of face coloring. The 6 rotations that leave a pair of opposite edges invariant are 180 rotations along axes
perpendicular to diagonally opposite edges, which leave three independent choices of face coloring:
1
r = 24
1 8 6 + 3 1 8 4 + 2 8 3 + 8 8 2 + 6 8 3 = 11712 .

6.

The identity rotation leaves all 4 markings invariant. The 3 groups of 9 face-invariant rotations each have one
180 rotation leaving four independent colors, and two 90 rotations leaving two. The 8 vertex-invariant
rotations leave four independent colors. The 6 edge-invariant rotations also leave four:
60

r=
7.

1
24

(1 4

+ 3 1 4 4 + 2 4 2 + 8 4 4 + 6 4 4 = 2916 .

The rotations are the fourth dihedral group.


a. Only the identity rotations leaves all markings invariant:
6!
r = 18 = 45 .
2!
b. 0 leaves 4 choices of color, 1, 3 leave one, 2 leaves two, 1, 2 leave three, and 1, 2 leave two:
r=

8.

1
8

(1 6

+ 2 6 1 + 1 6 2 + 2 6 3 + 2 6 2 = 231 .

The tetrahedron can be rotated by fixing one of four faces and then one of three remaining faces, so G = 12 . The
rotation that leaves the first and the second face invariant leaves six independent choices of color. The two
rotations that leave the first face invariant and rotates the second leaves two choices. In each of the two remaining
groups of rotation for the first face, one leaves the second face invariant and leaves four choices, and two rotate the
second face also and leave only once independent choice of color:
1
r = 12
1 2 6 + 2 2 2 + 3 1 2 4 + 2 2 1 = 11 .

9.

))

What is the shape of a prism?


r = 18 6 6 + 1 6 4 + 2 6 3 + 1 6 4 + 1 6 3 + 2 6 2 = 6246 is not correct.

4.1 Isomorphism Theorems


3.

Homomorphisms preserve normal subgroups. The Lemma states that, in factor groups at least, this preservation is
bijective: there are no more or fewer normal groups containing the factor, then there are in the factor group.
Let N < G , and : G G N the canonical homomorphism. Then, the canonical correspondence given by
: L L between normal groups containing N in G, and normal groups in G N is bijective.
Note the fact that we have two names and for essentially the same operation. operates on elements h to
produce (h) , but has an implicit extended interpretation in which it operates on sets H to produce
[H] h H (h) . is simply a name given to this interpretation. The book uses the special notation with square
brackets to indicate the extended interpertation.
First, show that is well-defined. If L < G ( L N is not really relevant here), and : G G N is a
homomorphism, then by Theorem 3.3.16 L = L < G N , so really does produce normal groups.
To show that is injective we need to be able to calculate inverses. By Theorem 3.1.15, the inverse of the forward
homomorphism of an element is the coset of its kernel containing that element, i.e. inverses of forward mappings of
elements g G under are of the form g Ker . Let L < G,L N . Since Ker = N L and L is a subgroup and
thus closed, g L : g Ker L, so L Ker L. Conversely, g L : g g Ker so L L Ker , so
L Ker = L L = 1L = 1L.
Now, show that is injective. Let L,M < G such that L = M. Then from the above, L =
M = 1M = 1L so L = M.

5.

L and

Finally, show that is surjective. Let H < G N , then 1H G : 1H = 1H = H , 1H N and normal in


G by Theorem 3.3.16.
Given a homomorphism, Theorem 2 allows us to generate isomorphisms between the image of that homomorphism
and a factor group. Applying this procedure twice, this allows us to generate isomorphisms between factor groups.
Note that it is not even necessary to consider the canonical homomorphism (the missing side of the triangles in
the diagram).

61

H
H N

HN

|H homomorphism

|HN homomorphism

Ker |H = H N

Ker |HN =N

HN
N

1 isomorphism

HN

2 isomorphism

H
H N

HN
N

1.

a. :  12  3 : i a 2i mod3;

1 = 2 . Ker =

b. {0,3, 6, 9}, {1, 4, 7,10}, {2,5,8,11} .

3

2

 12.

c. :  12 K  3 : Ker + i a i .
2.

a. :  18  12 : i a 10i mod12;

1 = 10 . Ker =

12
10

  18 = {0, 6,12} .

b. {0, 6,12}, {1,7,13}, {2,8,14}, {3,9,15}, {4,10,16}, {5,11,17} .


c. By Theorem 2 it is isomorphic to  18 Ker  18  3  6 .
d. :  18 18 : Ker + i a i .
H= 4

3.

 24

= {0,4,8,12,16, 20}, N = 6

a. HN = {0, 2, 4,, 22} , H N = {0,12} .

 24

= {0, 6,12,18} .

HN
= {0, 6,12,18}, {2,8,14, 20}, {4,10,16, 22} .
N
H
c.
= {0,12}, {4,16}, {8, 20} .
H N
HN
H
d. :

: N + i a (H N ) + 2i .
N
H N
Note that the book gives a different correspondence. This is possible because  3 is automorphic.

b.

H= 6

4.

 36

= {0,6,12,18, 24}; N = 9

a. HN = {0,3, 6,,33} , H N = {0,18} .

= {0, 9,18, 27}.

HN
= {0, 9,18, 27}, {3,12, 21,30}, {6,15, 24,33} .
N
H
c.
= {0,18}, {6,24}, {12,30} .
H N
HN
H
d. :

: N + i a (H N ) + 2i .
N
H N
H = 4  = {0,4,8,, 20}; K = 8  = {0,8,16} .
b.

5.

 36

24

24

62

G
= {0, 4,8,12,16, 20}, {1,5,9,13,17, 21}, {2,6,10,14,18, 22}, {3, 7,11,15,19, 23} .
H
G
b.
= {0,8,16}, {1, 9,17}, {2,10,18}, {3,11,19}, {4,12, 20}, {5,13, 21}, {6,14, 22}, {7,15, 23} .
K
H
c.
= {0,8,16}, {4,12, 20} .
K
G /K
d.
= {0,8,16}, {4,12, 20} , {1, 9,17}, {5,13, 21} , {2,10,18}, {6,14, 22} , {3,11,19},{7,15, 23} .
H /K

a.

{{

} {

} {

} {

}}

G
G /K

: H + i a (H /K ) + i .
H
H /K
Note that the book writes the correspondence as i a (H /K ) + (K + i ) . This gives the same sets using a different
computation.
H = 9  = {0,9,18, 27}; K = 18  = {0,18} .

e. :

6.

36

36

{0, 9,18, 27}, {1,10,19, 28}, {2,11, 20, 29},

G
a.
= {3,12, 21,30}, {4,13, 22,31}, {5,14, 23,32}, .
H

{6,15, 24,33}, {7,16, 25,34}, {8,17, 26,35}

G {0,18}, {1,19}, {2, 20}, {3, 21}, {4, 22}, {5, 23}, {6,24}, {7, 25}, {8, 26},
=
.
K {9, 27}, {10, 28}, {11, 29}, {12,30}, {13,31}, {14,32}, {15,33}, {16,34}, {17,35}
H
c.
= {0,18}, {9,27} .
K
{0,18}, {9, 27} , {1,19}, {10, 28} , {2, 20}, {11, 29} ,

G /K
d.
= {3, 21},{12,30} , {4,22}, {13,31} , {5, 23}, {14,32} , .
H /K

{6, 24},{15,33} , {7,25}, {16,34} , {8, 26}, {17,35}


G
G /K
e. :

: H + i a (H /K ) + i .
H
H /K
b.

{
{
{

7.

} {
} {
} {

H : {0} 12 3  36 ;
H0

H1

H2

H3

} {
} {
} {

}
}
}

K : {0} 18  36 .
K0

K1

K2

(
)
H 01 = H 0 (H 1 K1 ) = E ( 12 18 ) = EE = E
H 02 = H 0 (H 1 K2 ) = E ( 12  36 ) = E 12 = 12
H 10 = H 1(H 2 K0 ) = 12 ( 3 E ) = 12 E = 12
H 11 = H 1(H 3 K1 ) = 12 ( 3 18 ) = 12 18 = 6
H 12 = H 2(H 3 K2 ) = 12 ( 3  36 ) = 12 3 = 3
H 00 = H 0 (H 1 K0 ) = E 12 E = EE = E

H 20 = H 2(H 3 K0 ) = 3 (  36 E ) = 3 E = 3

H 21 = H 2(H 3 K1 ) = 3  36 18

)=

3 18 = 3

H 22 = H 3(H 3 K2 ) = 3 (  36  36 ) = 3  36 =  36
and

63

(
K01 = K 0 (K 1 H 1 ) = E ( 18
K02 = K 0 (K 1 H 2 ) = E ( 18
K03 = K 0 (K 1 H 3 ) = E ( 18

K00 = K 0 (K 1 H 0 ) = E 18 E = EE = E

) = EE = E

12

) = E 18 = 18
 36 ) = E 18 = 18
3

K10 = K 1(K 2 H 0 ) = 18 (  36 E ) = 18 E = 18

(
) = 18 12
K12 = K 2(K 2 H 2 ) = 18 (  36 3 ) = 18 3 =
K11 = K 1(K 2 H 1 ) = 18  36 12

= 6
3

K13 = K 2(K 2 H 3 ) = 18 (  36  36 ) = 18  36 =  36
This gives the chains
E

=E

12

12
3

6
3

3
 36

and E

=E
18

E
6

18
3

18
 36

,
=  36

=  36

or E 12 6 3  36 ;

E 18 6 3  36 .

The factor group isomorphisms are:


A:
12 E 6 18  3

8.

B:

12 18 E  2

C:

6 3

D:

 36

3  36 3  3

6 2

H : {0} 12 4  24 ;
H0

H1

H2

H3

K : {0} 6 3  24
K0

K1

K2

K3

(
)
H 01 = H 0 (H 1 K1 ) = E ( 12 6 ) = E 12 = 12
H 02 = H 0 (H 1 K2 ) = E ( 12 3 ) = E 12 = 12
H 03 = H 0 (H 1 K3 ) = E ( 12  24 ) = E 12 = 12
H 10 = H 1(H 2 K0 ) = 12 ( 4 E ) = 12 E = 12
H 11 = H 1(H 2 K1 ) = 12 ( 4 6 ) = 12 12 = 12
H 12 = H 1(H 3 K2 ) = 12 ( 4 3 ) = 12 12 = 12
H 13 = H 1(H 4 K3 ) = 12 ( 4  24 ) = 12 4 = 4
H 00 = H 0 (H 1 K0 ) = E 12 E = EE = E

H 20 = H 2(H 3 K0 ) = 4 (  24 E ) = 4 E = 4

(
)=
H 22 = H 3(H 4 K2 ) = 4 (  24 3 ) =
H 21 = H 3(H 4 K1 ) = 4  24 6

4 6 = 2
4 3 = 1 =  24

H 23 = H 3(H 4 K3 ) = 4 (  24  24 ) = 4  24 =  24
and
K00 = K 0 (K 1 H 0 ) = E 6 E = EE = E

(
K01 = K 0 (K 1 H 1 ) = E ( 6
K02 = K 0 (K 1 H 2 ) = E ( 6
K03 = K 0 (K 1 H 3 ) = E ( 6

12

) = E 12

) = E 12
 24 ) = E 6
4

= 12
= 12
= 6

64

(
K11 = K 1(K 2 H 1 ) = 6 ( 3
K12 = K 2(K 3 H 2 ) = 6 ( 3
K13 = K 3(K 3 H 3 ) = 6 ( 3

K10 = K 1(K 2 H 0 ) = 6 3 E = 6 E = 6
12

)=

6 12 = 6

) = 6 12 = 6
 24 ) = 6 3 = 3
4

K20 = K 2(K 3 H 0 ) = 3 (  24 E ) = 3 E = 3

(
) = 3 12 = 3
K22 = K 2(K 3 H 2 ) = 3 (  24 4 ) = 3 4 = 1 =  24
K21 = K 2(K 3 H 1 ) = 3  24 12

K23 = K 3(K 3 H 3 ) = 3 (  24  24 ) = 3  24 =  24
This gives the chains
E

=E

12

12

12

12
4

12
2

12
 24

4
 24

or E 12 4 2  24 ;

and E
=  24

=E

12

12

6
3

6
3

6
 24

3
 24

,
=  24

E 12 6 3  24 .

The factor group isomorphisms are:


A:

12 E 12 E  2

B:

12  4

3 3

C:

4 6

12  2

D:

 24

2 3

6 2
H

H*
gN = Ng

hH* = H* h

N<G

9.

10.

(left figure) Because H is a group, h H : h(H N ) H . Also, obviously h(N \ H N ) H , so

h(H N ) = hN H . Similarly, (H N ) h = Nh H . Because N < G , hN = Nh, so h(H N ) = (H N ) h and


so H N < H .
(right figure) Let h H K . Then h(H* K) hH* K . Also, obviously
h H \ H K : h (H* K) hH* K , so h(H* K) = hH* K . Similarly, (H* K )h = H* h K .

Because H* < H ,
11.

h (H* K ) = (H* K ) h , so H* K < H K .

a. Prove that K /H < G /H . Now, this is the case if gH G /H, kH K /H : gH kH ( gH )

K /H . Since

coset multiplication is well-defined by H < G , this is true if gkg 1 H K /H or gkg 1 K , which is just to say
that K < G . The same argument proves L /H < G /H .
Inclusion follows immediately from K L l L \ K : lH L /H , lH K /H K /H L /H .
A /B
b. Because B,C < A, B C , by the Third Isomorphism Theorem
A /C , or writing the synonyms out,
C /B

65

G /H
K /H
L /H
K /H
12.

G /H
G /L . This exercise proves a sort of transitivity of the Third Isomorphism Theorem.
L /H

By Lemma 4, K L = KL = G , so by the Second Isomorphism Theorem


mutandis G /K .

13.

14.

Since G is solvable, there is a maximal

KL
K

L
K L

( i G i ) such that G i < G i + 1 and G i + 1

G K

= K . Mutatis
L
E

G i commutative. By Exercise 10,

K G i < K G i + 1, so ( i K G i ) forms a subnormal series. I don't know by what argument the factor groups
are simple, so that this is also a composition series. G i are commutative, and thus so are K G i and
K G i + 1 K G i . So K G is solvable.
(See figure)

( i HiN ) is a composition series iff it is a subnormal series with simple factors.

Obviously

H i <Hi+1

N<G

H 0N = E < H 1N . For all other subgroups in the series, hn H i + 1N : ( hn )(H iN ) = H i (hn )N = (H iN ) (hn )
so H i < H i + 1 and the series is subnormal. To see that the factor groups are simple, we evaluate
H i + 1N H i + 1 H iN
=
H iN
H iN

2 Iso Th

Hi +1
H i + 1 H iN

3 Iso Th

H i + 1 /H i

(Hi + 1 H iN ) /Hi

. Now H i + 1 /H i is simple, so the

denominator must be either trivial or nonproper. Obviously H i H i + 1 H i + 1 H iN H i + 1, so the


denominator is proper and must therefore be trivial. So the fraction as a whole is isomorphic to just H i + 1 /H i , and
thus the factor groups of our series are simple also.
H 3N
H3

H 2N

H 1N

H2

H1

N<G

15.

( i Hi /N ) is a composition series iff it is a


hN H i + 1 /N : hN (H i /N ) = (H i /N ) hN .

(See figure relating to Exercise 14, repacing H iN with H i /N )


subnormal series with simple factors. H i /N < H i + 1 /N iff
Hi < H i + 1

hH i = H i h
coset multiplication
well-defined

hH i N = H i h N

N<G

hN H i = H i hN

canonical
homomorphism

hN H i H i hN
=
N
N

hN H i /N = H i /N hN

so the series is indeed subnormal. To see that the factor groups are simple, we first find that

H i + 1 /N H i + 1N

H i /N
H iN

(*) which (we saw in Exercise 14) is simple. The isomorphism follows from the fact that
66

H i + 1 /N
: hn a hN (H i /N ) is a homomorphism:
H i /N
hn, h n H i + 1N : i ( hn ) i ( h n ) = hN (H i /N ) h N (H i /N ) = hN h N (H i /N ) =

i : H i + 1N

16.

) (

hh N (H i /N ) = i ( hn h n )
By the First Isomorphism Theorem, the range of the homomorphism is isomorphic to the kernel factor group of the
range, which is the beforementioned (*) isomorphism above.
Let G be solvable by ( i G i ) , and be a homomorphism. By the First Isomorphism Theorem, G G / Ker .
Then by Exercise 15, G / Ker has a composition series also in the distinct groups of ( i G i / Ker ) . Since
G i + 1 /G i are commutative, then so are G i + 1 / Ker G i / Ker . So G / Ker and G are solvable.

4.2 Sylow Theorems

1.

The normalizer N G H is the largest subgroup of G in which H is normal.


This theorem applies the obvious fact that when a G-set X is stripped of its irrelevant part, the remainder reflects
something of the structure of the group. In particular, if G is a p-group, the important part of X has a multiple of p
elements.

6.

orbits of X
This is the important part of X that
actually says something about G
X XG p

p-group G
3.

XG
This is the irrelevant part of X that is
invariant for all g and says nothing about
G

G-set X

This applies the previous theorem. The entire X also has a multiple of p elements, so we can conclude that the
irrelevant part does too. That irrelevant part happens to consist of p-tuples of one single element, and because there
is at least one, there have to be at least p.
Now the theorem leads to conclude that there is a multiple of p cosets of H outside of the normalizer.

G
NGH

8.

Since there is a multiple of p cosets of H in the whole of G, and (by Lemma 6) a multiple outside of the normalizer,
there must be a multiple of p inside it as well. Inside the normalizer, we can then find one that has exactly p, and if
H is of order p i , this new one will form a subgroup of order p i +1 .
67

1.

12 = 2 2 31 , so by the remark following Definition 9 the maximal 3-subgroups have order 31 = 3 .

2.

54 = 2 33 , so the maximal 3-subgroups have order 33 = 27 .

3.

By the Third Sylow Theorem, the number must be in 2 + 1 1, 2, 3, 4, 6, 8, 12 = 1, 3 .

4.

The number must be in 3 + 1 1, 3, 5, 17, 3 5 = 15, 3 17 = 51, 5 17 = 85, 3 5 17 = 255 = 1, 85 .

5.

S 4 = 4! = 24 = 2331, so a maximal 3-subgroup has order 31 = 3 . There are 3 + 1 1, 2, 3, 4, 6, 8, 12 = 1, 4 of

) {

) {

} { }
) {

} { }

} { }

(1 2 3) = {( ), (1 2 3), (1 3 2)} is a 3-subgroup and maximal, and so are the other three 3-cycles.
example, (1 2 4) is conjugate by (3 4) :
(3 4) ( )(3 4) = ( ); (3 4) (1 2 3)(3 4) = (1 2 4); (3 4) (1 3 2)(3 4) = (1 4 2) .
them. Now

By

The rest follow similarly.


6.

The order of a maximal 2-subgroup of S 4 is (Exercise 5) 23 = 8 , and there are (Exercise 3) either 1 or 3 of them.
There are 4! 0!4 = 6 4-cycles, 4! 1!3 = 8 3-cycles, 4! 2!2 = 6 2-cycles, 4! 4 2 = 3 22-cycles, and 1 1-cycle. The
3-cycles have order 3 and cannot participate in 2-subgroups. Every subgroup must contain the 1-cycle identity.
Conjecture that the remaining 7 elements of each of the three 2-subgroups result from some symmetric
distribution of the 4-, 2-, and 22-cycles. One such distribution is to assign all 3 22-cycles, and one-third each of
the 4- and 2-cycles to each 2-subgroup. Since the 1- and 22-cycles are the only even permutations, they are closed
in each subgroup. It remains to be shown that the product of any odd and even permutation results in one of the
four odd 4- and 2-cycles from its distribution. Assign to a 2-subgroups the two component 2-cycles from one of
the 22-cycles, for example, (1 2) and (3 4):
(1 2)(3 4) (1 2) = (3 4), (1 2)(3 4) (3 4) = (1 2)
(1 3)(2 4) (1 2) = (1 4 2 3), (1 3)(2 4) (3 4) = (1 3 2 4)
(1 4)(2 3) (1 2) = (1 3 2 4), (1 4)(2 3) (3 4) = (1 4 2 3)
Hence the two 4-cycles that need to be distributed to the 2-subgroup follow naturally. Note that the two 2- and 4cycles are each others' inverses, so the entire 2-subgroup is closed and thus well-defined.
The other two 2-subgroups follow directly from mechanical substitution of letters in the permutations.
To show conjugacy, note first that the subgroup of even cycles (which is contained by each 2-subgroup) is normal.
Finally, verify that the odd cycles of one 2-subgroup are conjugate to those in another under one of the 3-cycles:
(1 2 3) 1 (1 2) (1 2 3) = (1 3);

(1 2 3) 1 (1 3 2 4) (1 2 3) = (1 4 3 2);

(1 2 3) 1 (3 4) (1 2 3) = (2 4); (1 2 3) 1 (1 4 2 3) (1 2 3) = (1 2 3 4).
7.
order power of p
8.
the maximal set of elements by whose inner automorphisms
9.
Correct this uses Corollary 4.
10. a. true (by the Third Sylow Theorem)
b. true (by Example 13)
c. true (by Corollary 4)
d. false (a 2-subgroup of a group of order 2 2 could have order 21 )
e. true (any subgroup of a commutative group is invariant under conjugation)
f. false?
g. true (Definition 5)
h. true (by the Second Sylow Theorem all maximal p-subgroups are conjugate and thus not invariant)
i. false (for a commutative group N G H = G )
j. false (but it is true that it has no proper p-subgroup)
11.

( ) ( )

(closure) g , g G H : gg H gg
(identity) e G :

eHe 1 = H

(inverse) g G H :

= gg Hg 1 g 1

e G H .

gHg 1 = H

Hg 1 = g 1H

g GH

g GH

gHg 1 = H

gg G H .

( )

H = g 1Hg = g 1H g 1

g 1 G H .
68

12.

By the Second Sylow Theorem, all maximal p-subgroups are conjugate. If G has only one such subgroup, then it
must therrefore be invariant under conjugacy, which is to say, it is a normal subgroup. Assuming that G p  , this
subgroup is proper; and assuming that p > 1, it is not trivial. Then G is not simple.

13.
14.

15.

45 = 3 251, so the maximal 3-subgroups of such a group have order 3 2 = 9 and their number is in
3 + 1 1, 3, 5, 9, 15, 45 = 1 . So by Exercise 12 the subgroup is normal.

) {

} {}

If a group is divisible by a prime other than p, then by Cauchy it has a subgroup of that order, which is cyclic and
thus has an element of order of that prime, so the group is not a p-group. Conversely, suppose that a p-group has
an element of order of a power of some other prime. Then that would generate a subgroup of other prime power
order which would hence not divide the order of the group, which is impossible by Lagrange.
P < N G P g N G P : i g P = P so by the Second Sylow Theorem, N G P has only the p-subgroup P. Now,
suppose N GN G P N G P

g N GN G P \ N G P : i g P P

of N G P . However, g N GN G P

i g P / N G P , so there is another p-subgroup outside


i g N G P = N G P and P N G P i g P i g N G P = N G P so this other p-

16.

subgroup would have to be inside of N G P . This is a contradiction, so N GN G P / N G P . Therefore N GN G P = N G P .


By Cauchy, H is contained in some maximal p-subgroup P of G. By the Second Sylow Theorem,
g G : gP g 1 = P gHg 1 P .

17.

G = 353 = 537 3 , so the 5-subgroups in G have order 53 = 125. The only divisors of 125 that can be in 5 + 1

} {}

cannot contain powers of 5, and 5 + 1 7 0 = 1, 7 1 = 7, 7 2 = 49, 7 3 = 343 = 1 , so the only 5-subgroup is


normal.
18.

The only divisors of G that can be in 17  + 1 cannot contain powers of 17. The largest remaining divisor of G is
3 5 = 15 < 18 also cannot possibly be in 17  + 1. Therefore there is one normal 17-subgroup.

19.

The number of p-subgroups divides p r m and is in p + 1, so the divisors p s r m t 1 cannot contain any powers of p.
The only possible divisors therefore are m 0, 1 , but since m < p it cannot be in p + 1. So there is one normal psubgroup.

20.

} {

a. G G = g G | x G : i x g = xgx 1 = g = g G | x G : xg = gx = Z G .
b. By Theorem 1, G G G is divisible by p, and because G is a p-group and thus divisible by p, so is G G . Because G is
nontrivial, p > 1. Since e G G , G G > 1 so G G = Z G is nontrivial.

21.

By the First Sylow Theorem, we know that a group G with the given characteristics has a subnormal series. The
Exercise asks us to prove that it has a normal series. We will prove this by showing that any subnormal series is itself
a normal series.
Let

0 i n

Hi be a subnormal series of G; we show that these are the only subgroups of G. Let H be a subgroup of

G. Since G is a p-group, H is a p-subgroup and i : 0 i n : H = p i , so we may reasonably refer to this subgroup


as Hi . By the First Sylow Theorem, this group is contained in an Hi+ 1 and so on. Obviously for some k, Hk = Hk
. By the First and Second Sylow Theorem, Hk 1 , Hk 1 are normal conjugate maximal p-subgroups of Hk , so
Hk 1 = Hk 1 , and so forth.
Now we show by induction that every Hi < G . Obviously H0 = E < G . Consider ZG . By Exercise 20,
ZG = Hk 0 for some 0 < k0 n . Now for i : 0 < i k0 : Hi Z G so h H : g G : hg = gh Hi < G . If
k0 = n we are done. Otherwise, consider G Hk 0 , and since G Hk 0 = G Hk 0 = p n p k 0 = p n k 0 it is again a p-

subgroup. The same argument shows that Z G Hk 0 = Hk1 Hk 0 for some k0 < k1 n . For

i : k0 < i k1 : Hk1 Hk 0 Z G Hk 0

Hk1 Hk 0 < G Hk 0 . If : G G Hk 0 is the canonical

homomorphism, then inv Hk1 Hk 0 = Hk1 < G . Since ki > ki 1 , this procedure terminates under induction.
22.

Let H be a normal p-subgroup of G, so H is invariant under conjugation by G. By the First Sylow Theorem, H is
contained in at least one maximal p-subgroup. Since by the Second Sylow Theorem every other maximal p69

subgroup is conjugate to this one, and since H is invariant under conjugation, H is also contained in every
conjugate.

4.3 Applications of the Sylow Theory


1.

a. The table lists the conjugations i g x = gxg 1 :


x

1
2
3
1
2
1
2

g
g 1

1
2
3
0
1
2
2
1

3
1
2
3
2
1
2
1

3
0
1
2
1
2
1

2
1
2
3
1
2
1
2

{ }{

3
0
1
2
2
1
1
2

1
1
2
3
2
1
2
1

}{ }{

1
2
2
1
0
2
3
1

}{

1
3
2
1
1
2
2
1

1
1
2
2
0
1
3

2
3
2
1
1
2
2
1

1
3
2
1
2
1
1
2

1
2
2
1
3
0
2

2
2
1
1
3
1
2
0

2
3
2
1
2
1
1
2

So the conjugate classes are 0 , 1 , 3 , 2 , 1 , 2 , 1 , 2 .


2.

3.

b. 8 = 2 + 2 + 2 + 2 .
0
1
2
3
4
5
6
00

10

20

30
E12

14
40
40

45

50

prime

Example 9

Theorem 7

Exercise 2.19
40
40 = 235 has one 5-subgroup
45
45 = 3 25 has one 5-subgroup
a. true ( 159 = 53 3 , and 53 = 27 3 + 2 so cyclic by Theorem 7)
b. true ( 102 = 2 3 17 , not simple by Exercise 2.19)

13

c. false (Example 3.4.17 shows S3 is solvable, and S3 = 3! = 6 )


d. true (Theorem 1)
e. true
f. true (Theorem 7)
g. true ( 125 = 53 , by Exercise 21 has a normal subgroup of order 51 , i.e. commutes with every element)
h. true ( 42 = 2 3 7 , by Exercise 2.19)
i. false ( 42 = 2 3 7 cannot by Lagrange even have any subgroup of that order)
j. false (trivially, A5 is simple)
4.

Let G be a group of order 5 7 47 . By familiar reasoning, it has one 5-subgroup H5 and one 7-subgroup H 7 .
Then G / H5 = 7 47 and G / H 7 = 5 47 so both factor groups are cyclic by Theorem 7. Then by Theorem

{ } { }

3.3.20, H5 , H 7 C G contain the commutator subgroup of G, so C G 1, 5 1, 7 so CG = E . Therefore


G / C G = G / E G is commutative, and each of its subgroups is normal.
5.

Let G be a group of order 96 = 253 . The number of 2-subgroups of order 25 = 32 must be 1 or 3. Suppose it has
3, and let H and K be two distinct ones. H K is again a 2-subgroup of order a power of 2. If H K = 23 then
by Lemma 8 HK =

25 25
3

= 2 7 = 128 > 96 = G which is impossible. Since H K

2
H K = 2 4 . Then H H K = 25 2 4 = 2 , so H K < G .

6.

H K < G , so

Let G be a group of order 160 = 255. The number of 2-subgroups of order 25 = 32 must be 1 or 5. Suppose it has
70

5, and let H and K be two distinct ones. H K is again a 2-subgroup of order a power of 2. If H K = 2 2 then
by Lemma 8 HK =

25 25
2

= 28 = 256 > 160 = G which is impossible. Since H K

H K 23 , 2 4 . By Exercise 2.21, H K < H , K

N G H K H , K and (why?)

N G H K = n 25 , n > 1 and divides the order of G, so N G


7.

(
)
(H K ) = G

H K < G , so

N G H K = G , so H K < G .

a. 1 = a0 a 2 am 1 1 , so the only letters affected by are the ai and all other letters are invariant under the

)( ) (

) ( )

( )
), = (b

entire product. i : 1 ai = 1 ai = ai = ai = a(i + 1) mod m , so 1 = a0 a1 am 1 .0

b. For any two cycles of the same length = a0 a1 am 1

b1 bm 1 , let = a0 a1 an 1 be any cycle

i : x = ai x a bi
of all the letters not in , and similarly. Then define :
, which is a bijection and a
i : x = ai x a bi
permutation. By (a.), 1 = i ai 1 = i ai = i bi = , so ~ .

( )

c. Write the products of cycles as = i < s i = i < s

) ( )
( a ),
j < ri

ij

= i < s i = i < s

j < ri

bij , and let s =

j <rs

a j be any

cycle of all the letters not in any i < s , and similarly. Then define : i s , j < rs : x = aij a bij , which is a

bijection and a permutation. Then 1 = i s i 1

8.

9.

i disjoint

i s

) (
(b.)

i 1 = i s i = .

d. Differently factored disjoint products cannot be conjugate. Any disjoint factoring into cycles is unique: disjoint
factors cannot be combined into a cycle, and a cycle cannot be split into disjoint factors. For any disjoint
permutation, every letter must be in exactly one cycle (perhaps a 1-cycle). So pn as described gives the number of
ways permutations of Sn are factored into disjoint cycles, which are (by c.) the conjugate classes.
e. p1 = 1 1
p2 = 2 1 1, 2
p3 = 3 1 1 1, 2 1, 3
p4 = 5 1 1 1 1, 1 1 2, 2 2, 3 1, 4
p5 = 7 1 1 1 1 1, 1 1 1 2, 1 2 2, 1 1 3, 2 3, 1 4, 5
p6 = 11 1 1 1 1 1 1, 1 1 1 1 2, 1 1 2 2, 2 2 2, 1 1 1 3, 1 2 3, 3 3, 1 1 4, 2 4, 1 5, 6
p7 = 15 1 1 1 1 1 1 1, 1 1 1 1 1 2, 1 1 1 2 2, 1 2 2 2, 1 1 1 1 3, 1 1 2 3, 2 2 3, 1 3 3, 1 1 1 4,
1 2 4, 3 4, 1 1 5, 2 5, 1 6, 7
By Exercise 7, S 4 has 5 conjugate classes:
4!
=1
(1)(2)(3)(4)
4!
4!
=6
(1 2), (1 3), (1 4), (2 3), (2 4), (3 4)
2!2!
4!
=3
(1 2)(3 4), (1 3)(2 4), (1 4)(2 3)
2!2 2
4!
=8
(1 2 3), (1 3 2), (1 2 4), (1 4 2), (1 3 4), (1 4 3), (2 3 4), (2 4 3)
3
4!
=6
(1 2 3 4), (1 2 4 3), (1 3 2 4), (1 3 4 2), (1 4 2 3), (1 4 3 2)
4
24 = 1 + 6 + 3 + 8 + 6
The class equation can be found as follows. First, find the structure of each of the conjugate classes as in Exercise
7e. To find the number of distinct permutations in each conjugate class, imagine listed in a table the n! different
ways of writing the letters of Sn , and draw dividing lines between the columns of this table so as to separate each
row into cycles according to the partition of the conjugate class. This surely represents every possible element of the
class, although each element may be overrepresented. In particular, if the conjugate class has m l cycles of a certain
71

length l, the m l ! rearrangements of these cycles within a permutation are equivalent. Also, every cycle of length p
can itself be written in p different ways by rotating its letters. So the number of cycles of a conjugate class is
n!
. With the help of the partitioning found in Exercise 7,
l m l ! i pi
5!
5!
5!
5!
5!
5! 5!
+
+
+
+
+ +
120 = 1 + 10 + 15 + 20 + 20 + 30 + 24
5! 3!2 2!2 2 2!3 2 3 4 5
6!
6!
6!
6!
6!
6!
6!
6!
6!
6! 6!
6! = +
+
+
+
+
+
+
+
+ +

6! 4!2 2! 2!2 2 3!2 2 2 3!3 2 3 2!3 3 2!4 2 4 5 6


720 = 1 + 15 + 45 + 15 + 40 + 120 + 40 + 90 + 90 + 144 + 120
5! =

S5 :
S6 :

10.

By Theorem 2.4.12 the commutative groups of order p n are isomorphic to i p ni , where


i p ni = p n + i ni = n . Therefore the commutative groups of order p n differ only (up to isomorphism) in the
distribution of n, which can be done in pn ways.

11.

Z Sn = Sn | Sn : = = Sn | = 1 , so the center of Sn is the permutations that are invariant


under conjugation of all Sn , which is the conjugate classes of Sn that have exactly one element. By Exercise 7 the
conjugate class consisting only of 1-cycles contains only the identity. Also, any permutation containing an n-cycle
will be conjugate to every other permutation with an n-cycle. If n > 2 two distinct n-cycles can always be found, so
that the conjugate class has more than one element. Therefore, for n > 2 ZSn = E .

4.4 Free Abelian Groups


1.
2.

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

(2, 1), (3, 1) = (2, 1), (1, 0) = (0, 1), (1, 0) =  


2 + 3 = 0

1 + 1 = 0

(2, 1) + (3, 1) = (0, 0)

( )

2 + 3 = = 0

= 0

= 0

So this does form a basis.


3.
4.

(2, 1), (4, 1) = (2, 1), (2, 0) = (0, 1), (2, 0)   does not form a basis.

{(a, b), (c, d)} is a basis for   iff (Theorem 1, Condition 2)

(a , b) + (c , d) = (0, 0)
(0,1).

(a , b), (c , d) =   and

, = 0. Show that these conditions are equivalent to being able to generate (1,0) and

(a , b) + 1 (c , d) = (1, 0)
Suppose that 1, 2 , 1, 2 : 1
. Prove that this implies (a , b), (c , d) is a basis by showing that
2 (a , b) + 2 (c , d) = (0, 1)
it satisfies Condition 2 of Theorem 1. For any (e , f )   ,

) (

) (

(e , f ) = e (1, 0) + f (0, 1) = e 1 (a , b) + 1 (c , d) + f 2 (a , b) + 2 (c , d) = e 1 + f 2 (a , b) + e1 + f 2 (c , d)
so (a , b), (c , d) =   . Next,

(a , b) + (c , d) = (0, 0)

) (
)
(a + c )(1, 0) + (b + d )(0, 1) = (0, 0)

a (1, 0) + b (0, 1) + c (1, 0) + d (0, 1) = (0, 0)


a + c = 0 b + d = 0

) (

) (
) (
)
Suppose 0, then a = 0 and b = 0 , but then {(a , b), (c , d)} cannot possibly generate   .
= 0 a = 0 = 0 c = 0 = 0 b = 0 = 0 d = 0

Similarly 0 is

impossible. So , = 0 .

If (a , b), (c , d) is a basis, then obviously they can generate (1,0) and (0,1).
(a , b) + 1 (c , d) = (1, 0)
Now, find conditions on a, b, c, d such that 1, 2 , 1, 2 : 1
, that is
2 (a , b) + 2 (c , d) = (0, 1)
72

1a + 1c = 1 2a + 2c = 0

1b + 1d = 0 2b + 2d = 1
[2] 2c = 0 2 = 0
[1] 1c = 1 1 , c = 1
, so (0, 1), (1, d) are possible

First, suppose a = 0. Then


[4] 1 d = 0 d = 1 [3] 2b = 1 b = 1
bases. Similarly for b, c, or d = 0.
The remainder of the cases have a , b , c , d 0. Then
[2] 1 d b a + 1c = 1 ad b + c 1 = 1
ad b + c 0
ad bc

[1] 1b = 1d 1 = 1 d b
b ad c 0

This is the familiar condition of linear independence that the determinant formed by a basis be nonzero.
Replace generating set with basis.
Correct.
2  both have rank 1.
a. true (Exercise 10)
b. true (any minimal generating set is a basis)

5.
6.
7.
8.

c. true ( n )
d. true (the condition implies that the group is torsion-free)
e. true
f. false (if Y X the expression of elements in terms of Y is not unique)

{ }

g. false (  has only 1 as bases)


h. true (Theorem 9)
i. true (why?)
j. false (  2 =  2 is not free commutative)
9.

(injective) g , g G : g = g ,

( n ) 

(surjective)

: g G : g = + i ni x i , g =

ni , ni : g = + i ni x i , g = + i ni x i

( n ).
i

( n ) = ( n )
i

g = g.

(associative) For all g , g G : ni , ni : g = + i ni x i , g = + i ni x i ,

g + g =
10.

) (

ni x i +

) ( n ) + ( n ) = (

ni x i =

) ( (n

If G had an element g G of order n, and g = + i i x i

) (

)(

) ) (( n x ) + (
for some basis { x } , then

ni + ni =

+ ni x i =
i

i i

ni x i

)) = ( g + g ) .

g = ng + g = n + 1 g = n + 1 + i i x i = + i (n + 1) i x i , contradicting the uniqueness of the expression of g in


11.

terms of its basis elements.


Let X and X be bases for G and G , respectively. Show that Condition 2 of Theorem 1 holds:

g , g G G : g G , g G

) (

i x i , 0 i 0, x i

ni , ni : g = + i ni x i , g = + i ni x i

( g , g ) = (+

ni x i , + i ni x i so

= G G.

+ i ni x i , + i ni x i = (0, 0)

+ i ni x i = 0 + i ni x i = 0

ni = 0 ni = 0 .

12. If G is free commutative of finite rank, then by Condition 2 of Theorem 1 the finite basis generates it. By Exercise
10 it has no elements of finite order.
Let X be a minimal generating set of G. We just have to prove the uniqueness of zero. Suppose + i ni x i = 0, and
let K partition the coefficients such that nk K 0, nk K = 0 , and + i K ni x i = 0. Suppose there is k K , and thus
nk x k = + i K , i k ni x i . If + i K , i k ni x i 0 then it and nk x k are different expressions of the same element so X could
not have been minimal. If + i K , i k ni x i = 0 then x k is an element of finite order nk . So K = .
13.

Since for any prime p, 1 p n cannot be formed from 1 q m for any other prime q, or from 1 p n for n < n, a basis
for  would have to contain at least p  lim p n , but no element can have a definite expression in terms of such
n

a basis.
73

14.

Clearly the torsion subgroup is finite. By the First Sylow Theorem, T has a p-subgroup Tp of elements of some
power of p, and p does not divide T Tp . So T Tp has no elements of order p, which must therefore all be in Tp .

15.

16.

Since T is isomorphic to its prime-power decomposition, the subgroup Tp of all elements of power of p has a
corresponding subgroup of all elements of power of p in the decomposition, which is exactly the direct product of
the cyclic factors of order some power of p.
G [n] G follows from:
(identity) 0 G : n0 = 0 0 G [n] ;
(inverse) g G [n] : ng = 0

( )

ng

(closure) g , g G [n] : ng = 0, ng = 0
17.

g p r [p]

pg = 0

( )

commutative

=0

n g 1 = 0

( )( )

commutative

ng ng

n : pg = n p r

g = np r 1

g 1 G [n] ;

( )

n gg = 0

gg G [n] .

p r [p] = p , and p r [p] p r is commutative, so

p r [p] p .

18.
19.

p i [p] = i p i [p] i p .

a. If i p ri i p si

i p ri [p] i p si [p]

i p ri i p si

b. Let j 0 such that i < j : ri = si ; r j < s j . Certainly p


r

for any q r j , p j p q = E , so i < j E


r

p j p s j E
20.

rj

i p ri p

rj

i > j p j p ri i < j E

m = n.
r

i p si i p j p ri i p j p si . Now,
r

i = j E

p m p n

i = j p j p si

i > j p j p si with

r j < s j , but this is impossible by (a).


q

p j , q j + , then G = i j 

Factorize each of the torsion coefficients m i = j p j j ,


m 0 = 2333 , m 1 = 2 2

T =  216  4

q ij

pj

qj

. For example,

G =  23 33  2 2 .

21.

From Exercise 2.4.42, m j = i : j < ni pi

m 0 = i pi

22.

From Exercise 2.4.42 (not really proved there).

qi 0

where qi0 is the highest power of pi in the decomposition.

4.5 Free Groups


1.

a. a 2b 2a 3c 3b 2 ;

2.
3.

a. a 5c 3 ; a 5c 3 ; b. a 4b 3c 6 ; a 4b 3c 6 .
By Theorem 12, there is exactly one homomorphism for each selection of 2 elements in the range G , so there are
G

4.

b 2c 3a 3b 2a 2 ; b. a 1b 3a 4c 6a 1 ;

homomorphisms: a. G

= 4 2 = 16 ; b. G

a 1c 6a 4b 3a 1 .

= 6 2 = 36 ; c. G

( ( ) )

= 3!2 = 6 2 = 36 .

( ( ) ) ( )

In this case, the ai must also generate G . a. 2 2 4 2 2 = 12 ; b. 2 2 6 2 2 + 2 2 1 = 20 + 4 = 24 ;

( )

( )

c. 2 2 3 + 2 3 2 = 24 .
5.
6.
7.
8.
9.

Correct.
Insert free before generators. I don't think it's been proved that there are no other generators.
It would seem obvious that this operation of multiplication is well-defined and associative. I think this is obvious
too. Can't think of anything that might throw a spoke in the wheel.
10. a. false (E is not free by definition)
b. false (a subgroup of a commutative group is commutative and thus not generally free)
c. false (the image of the trivial homomorphism is not free)
d. true (by Definition 4.2)
74

e. false (torsion groups are finitely generated but not free commutative)
f. false (is this a trick question?)
g. false (a free group generated by one element is free commutative on the basis of that element)
h. true (a free commutative group of rank greater than one must have more than one generator, but any free group
with more than one generator is not commutative)
i. false (Theorem 9)
j. true (Theorem 4.5)
11.

a. 1 2 + 2 3 = 0,

1 2 = 1, 2 3 = 2 . {1} is a basis for  4 , certainly 1 =  4 and any one-element generating set is a

basis under the definition.


b. {1} is a one-element generating set and hence a basis. Also, {2, 3} = 6 .
+ i m i bi = 0 m 0 2 + m 1 3 = 0 2m 0 = 3m 1 so m 0 is a multiple of 3 and m 1 is a multiple of 2, so 2m 0 and
3m 1 are multiples of 6, so 2m 0 = 0, 3m 1 = 0 .
c. No, because a basis of a free commutative group induces unique expressions in terms of it.
d. A finite commutative group has an expression in terms of torsion coefficients, one dividing the next, where each
factor in the direct product has an element of the order of its coefficient.

G
12.

a. G

2 = 21

G1

G2

1 = 1 2

G2

G1

isomorphism

1
G Ker

G K

G 1 , G 2 can each be factored in terms of the other, so 1 = 1 2 , 2 = 21 . Then


1 = 1 2 = 1 21 i = 1 2 ; 2 = 21 = 21 2 i = 21 . Now
*
*
*

ker 1 = g G 2 | 1 g = e 21 g = e ig = e = E ; g G 1 : 2 g G 2 : 1 2 g = 1 2 g = ig = g

( ) (

so 1 is injective and surjective, so is an isomorphism, so G 1 G 2 .


b. Consider all possible homomorphisms of G into commutative G , and let K be a minimal set contained by the
kernels of all these homomorphisms. If K1, 2 < G are kernels of two homomorphisms such that G K1, 2 are
commutative, then K1 K 2 < G must be the kernel of a homomorphism with G K1 K 2 commutative, so K is the
minimal kernel of all commutative homomorphisms.
Refer to the figure on the right. By Fundamental Homomorphism Theorem, any homomorphism can be
factored into a homomorphism onto its kernel factor group and an isomorphism from this group. By Exercise
3.3.35. K < G ker K = K < K ker , so there are canonical homomorphisms G G K and G K G ker
. G K is thus a blip group.
c. The blip group of G is its commutator subgroup.
13.

a.

S
f1

f 2 = f 2 f1
G1

G2

f 2

S
f2

f 1 = f1 f 2
G2

G1

f1

Refer to the figure on the left for the adjusted naming. Suppose f is not injective. Then
s1 , s 2 S , s1 s 2 : fs1 = fs 2 . Then there is a group G and f : S G such that g 1 = f s1 , g 2 = f s 2 and g 1 g 2 .
But then there cannot be a homomorphism f such that f = f f , because then
f s1 = f fs1 f g 1 = g 1 ;

f s 2 = f fs 2 f g 2 = f g 1 = g 2 and f would not even be a function.

Now, suppose fS does not generate G. Then there is a g G that is not generated by fS, and then for any G , f
and f : f = f f we can let f g equal any element of G without affecting f f , contradicting the uniqueness
of f .
75

Now let G 1, 2 be blop groups. The figures on the right illustrate how f 1, 2 can each be factored in terms of the
other, so f 1 = f 1 f 2 ,

f 2 = f 2 f 1 . Then f 1 = f 1 f 2 f 1 f 1 f 2 = i ;

ker f 1 = g 2 G 2 | f 1 g 2 = e

f 2f1 g 2 = f 2 e

g 1 G 1 : g 2 G 2 : f 1 g 2 = g 1

ig 2 = e

f 2 f1 g 2 = f 2 g1

f 2 = f 2 f 1 f 2 f 2 f 1 = i . Then

g2 = e = E ;

ig 2 = f 2 g 1

g 2 = f 2 g1,

so f 1 is an isomorphism and G 1 G 2 .
b. Let F [S] be the free group on S =

{ s }.
i

Then by Theorem 12, for any group G and f : S G there is a

unique homomorphism f such that f fsi = f si Since fS = G , it follows that f f = f , so F [S] is a blop
group on S.
c. A blop group on S is the free group on S.
14.

A group G is a free commutative group if it is isomorphic to n for some n + .

4.6 Group Presentations


1.

4 a : a 4

Trivially,  4 a , b : a 4 , b is akin to saying that b does not generate anything at all. Also  4 a , b : a 4 , a 2b 1

which implies a = 1, b = a = 2 .

Trivially,  4 a , b , c : a 4 , b , c . Also,  4 a , b , c : a 4 , a 2b 1 , ac implies a = 1, b = 2, c = a 1 = 3 .


?

2.

3
2
2
1

S3 1 , 1 , 2 : 1 , 1 , 2 , 1 2 1 .

3.

a
a 2 a3
b
ab a 2b a 3b
1
a
a 2 a3
ab a 2b a 3b
b
1
2
3
2
3
a
a
a a b a b
b
ab
1
a3
a
a 2 a 3b
b
ab a 2b
1
b
a 3b a 2b ab
a3 a 2
a
1
3
2
3
ab
b
a b a b a
a
a2
1
2
3
2
a b ab
b
a b a
a
a3
1
a 3b a 2b ab
b
a3 a 2
a
1

4.

The commutative groups of order 14 are isomorphic to 14  2  7 . Suppose G is a noncommutative group of
order 14. Then G contains normal subgroups G 2, 7 of order 2 and 7 respectively, and both cyclic so
a G : a = G 2 , a 2 = 1;

b G : b = G 7 , b 7 = 1. Since G 7 < G , ia is an automorphism of G 7 so ia b must also be

an element of order 7, so ia b = aba 1


aba

a
a 2 a3
b
ab a 2b a 3b
1
a
a 2 a3
ab a 2b a 3b
b
1
2
3
2
3
a
a
a a b a b
b
ab
1
a3
a
a 2 a 3b
b
ab a 2b
1
b
a 3b a 2b ab a 2
a
a3
1
ab
b
a 3b a 2b a 3 a 2
a
1
2
3
3
2
a b ab
b
a b
a
a
a
1
a 3b a 2b ab
b
a
a3 a 2
1

2i 6

b i . i = 1 is not possible, because this would imply

= b ab = ba that G was commutative. By Exercise 13b. this gives a group of order 14 iff i 2 = 7 1 i = 6 .

So this leaves a , b : a 2 , b 7 , aba 1b 6 .


5.

The commutative groups of order 21 are isomorphic to  21 3  7 . Suppose G is a noncommutative group of


order 21. Then G contains normal subgroups G 3, 7 of order 3 and 7 respectively, and both cyclic so
a G : a = G 3 , a 3 = 1;

b G : b = G 7 , b 7 = 1. Since G 7 < G , ia is an automorphism of G 7 so ia b must also be

an element of order 7, so ia b = aba 1

2i 6

b i . i = 1 is not possible, because this would imply

aba 1 = b 1 ab = ba that G was commutative. By Exercise 13b. this gives a group of order 21 iff
i 3 = 7 1 i 2, 4 . Why are these isomorphic?

{ }

6.

Raised to powers is redundant.


76

7.

8.

This appears to be completely incorrect. Example 3 shows that presentations with different numbers of generators
(between which hence no one-to-one correspondence can exist) can still give isomorphic groups. Rewrite the
definition as: Group presentations are isomorphic iff they give isomorphic groups.
a. true (remark before Example 4; by Theorem 5.13 every group is homomorphic to a free group, and the generators
of the kernel are the relators of its presentation)

b. false (depends how you define different; E has only a : a = 1 )


c. false (if the presentations are not isomorphic then neither are the groups)
d. false (the question is unsolvable by the remark after Example 3)

( )

e. false ( a :  has a finite presentation)

f. true (every cyclic group is isomorphic to n a : a n )


g. true (the relators form a normal subgroup that is thus invariant under conjugation)

h. false ( a : a 2  2 , a : a 3 3 )

9.

i. true ( F [A] R is isomorphic to the group and thus commutative, so R contains the commutator subgroup)
j. true (I think so)
A noncommutative group G of order 15 would have normal subgroups G 3, 5 of order 3 and 5 respectively, and both
cyclic so a G : a = G 3 , a 3 = 1;

b G : b = G 5 , b 5 = 1. Since G 5 < G , ia is an automorphism of G 7 so ia b must

also be an element of order 5, so ia b = aba 1


i 2 = 5 1, but this is not so for any i.
10.

2i 4

b i . By Exercise 13b. this gives a group of order 15 iff

By Exercise 13b, a , b : a 3 , b 2 , ba = a 2b has 2 2 = 4 = 3 1 so is a group of order 2 3 = 6 . If this group were


2

commutative, then ab = ba ab = a b 1 = a , but then the group would have one generator of order 1 and one
of order 2, which cannot possibly generate a group of order 6.

} and from Exercise 10 we know i = 2 yields S . So this must be the only

11.

By familiar reasoning, aba 1

12.

noncommutative group of order 6.


A4 consists of the even permutations on 4 letters, so disjoint products of 1111-cycles (order 1), 22-cycles
(order 2), and 31-cycles (order 3), and no elements of order 6, so cannot be isomorphic to 6 .

2i 2

bi

S3 has two elements ( 1, 2 ) of order 3 and three elements ( 1, 2, 3 ) of order 2. Suppose A4 has two elements of
order 3, that is two 3-cycles. To form a group, these elements have to be each other's inverse. Without loss of
generality, let (1 2 3), (1 3 2) be these two elements. A4 would have to contain three elements of order 2, that is
all three 22-cycles. But then (1 2)(3 4) (1 2 3) = (2 4 3) and A4 would have to contain at least three elements of
order 3, so cannot be isomorphic to S3 either.
13.
14.

5.1 Rings and Fields


1.

12  24 16 = 192 mod 24 = 0 .

2.

16 32 3 = 48 mod 32 = 16 .

3.

11 15 4 = 44 mod 15 = 1 .

4.

20  26 8 = 160 mod 26 = 22 .

5.

(2, 3) 5  9 (3, 5) = (2 5 3, 3  9 5) = (6 mod 5, 15 mod 9) = (1, 6) .

6.

(3, 5)  4 11 (2, 4) = (3  4 2, 5 11 4) = (6 mod 4, 20 mod 11) = (2, 2) .

7.

n are commutative groups. Check multiplication:


77

(closed) na , nb n : na nb = n 2ab n

(associative) na , nb , nc n : na nb nc = n 2ab nc = n3abc = na n 2bc = na nb nc

(commutative) na , nb n : na nb = n 2ab = n 2ba = nb na


So they are also commutative fields. Do they have a multiplicative identity?
(multiplicative identity) na n : nb n : na nb = nb n 2ab = nb
So only 1 =  has unity. Which elements have a multiplicative inverse?
(multiplicative inverse) a  : b  : ab = 1 a = 1, b = a
So not even  is a division ring.
8.
9.

na = 1 n = 1, a = 1

+ under addition is not even a group.


  is a commutative group. Checking the multiplication:

)(

(a , a ) (b , b ) = (a b , a b )  
(associative) (a , a ), (b , b ), (c , c )   :
((a , a ) (b , b )) (c , c ) = (a b , a b ) (c , c ) = (a b c , a b c ) = (a , a ) (b c , b c ) = (a , a ) ((b , b ) (c , c ))
(commutative) (a , a ), (b , b )   : (a , a ) (b , b ) = (a b , a b ) = (b a , b a ) = (b , b ) (a , a )
(identity) (a , a )   : (b , b )   :

(a , a ) (b , b ) = (b , b ) (a b , a b ) = (b , b ) aa bb == bb aa == 11

(inverse) (a , a )   : (b , b )   : (a , a ) (b , b ) = (1, 1) (a b , a b ) = (1, 1) (a , a ) = (1, 1)


(closed) a0 , a1 , b0 , b1   :
0

0 0

0 0

1 1

0 0 0

1 1 1

0 0

0 0

1 1

1 1

0 0

1 1

0 0

1 1

0 0

1 1

1 1

0 0

1 1

10.

So it is a commutative ring with unity, but not a division ring.


2  is a commutative group. 2,  are both commutative rings by Exercise 7, so 2  is a commutative ring
by Example 7. 2 does not have a unity by Exercise 7, so neither does 2  .

11.

G = a0 + a1 2 | a0, 1  . It is obvious that X = 1, 2 is a generating set for G. Now

g G , g = a + b 2 , a , b  : g = 0

{ }

a +b 2 =0

a, b = 0

since there is no common multiple of 1 and 2 , so G is free commutative on X. Check multiplication:


(closed) a0 + a1 2 , b0 + b1 2 G :
a + a 2 b + b 2 = a b + a b + a b
2 + 2a1b1 = a0b0 + 2a1b1 + a0b1 + a1b0 2 G
1
1
0 0
0 1
1 0
0
0

) (

Multiplicative associativity and commutativity follows from the operation in . Since  is a commutative group
under addition, G is a commutative ring. Obviously 1G = 1 is the multiplicative identity.
(inverse) a = a0 , a1 2 G * : b = b0 , b1 2 G * :
a b + 2a1b1 = 1
2 =1 0 0
a0b1 + a1b0 = 0
From the first equation, a0b0 must be odd, but if a0 is even this is not possible, so G is not a division ring.
From Exercise 11, G is a commutative ring with multiplicative inverse. Also from that exercise,
(inverse) a = a0 , a1 2 G * : b = b0 , b1 2 G * :
1

2
a0b0 + 2a1b1 = 1 a 0 0a0b0 + 2a1 a1b0 a0 = 1
a0 2a1 a0 b0 = 1
b0 = a0 2a1 a0

b1 = a1b0 a0
a0b1 + a1b0 = 0

b = a b a

0
1 0
1

) (

ab = 1 a0 , a1 2 b0 , b1 2 = a0b0 + 2a1b1 + a0b1 + a1b0

12.

So a has inverse b if a0 0 a0 2a1 a0 0

a0 = 2a1 a0

2a1 a0

a1 0 , that is, for all G*, so G


78

is a field.

13.

G = ri | r  is not closed under multiplication because i i = 1 G , so G is not a ring.

14.

The identity of * is 1. a  : b  : ab = 1 a = 1.

15.

From Exercise 9, 1, 1 have inverses.

16.

From Example 17, 1, 2, 3, 4 = 5 * have inverses.

17.

The identity of  * is 1. a b * : b a  :

18.
19.
20.

{1}  * {1} have inverses.


{1, 3} have inverses.

a. M 2  2 =  2

22

a b b a = 1, so all of  * have inverses.

= 2 4 = 16 .

b. Under matrix multiplication, the identity is obviously the identity matrix, and all matrices with nonzero determinant
have an inverse:
0 1 1 0 1 0, 1 0, 1 1

,
,
,

1 0, 1 0, 1 1 0 1 1 0

( )

( )

21.

:    : n a n, 0 is obviously a homomorphism, and has 1 = 1, 0 0', 1'.

22.

For det to be a ring homomorphism, it must preserve addition as well, but


1 0 1 0 2 0

+
=
, 1 + 1 = 2 4 ,
0 1 0 1 0 2
so it is not even a group homomorphism.

23.

0 a 0
: a a ia are all the group homomorphisms.
By Theorem 4.5.12, since  is free on 1 , i :   :
1 a i
a , b  : ab = a b i ab = ia ib = i 2ab i = 0 i = 1, so the only ring homomorphisms are trivial or

{ }

( )

( )

the identity.

( ) (

24.

From Exercise 23, the ring homomorphisms are ij :    : a , b a ia , jb ,

25.
26.

The projection maps i (Example 25) or the trivial homomorphism.


From Exercise 25, there are 3 + 1 = 4 .

0

1

The problem is that ab = 0


/ a = 0 b = 0 . For example,
= 0 .

1
0

Is there some more effective way to do this?


( x 2 ) ( x + 3) = 0
x = 2 x = 3 = 11
x 2 + x 6 = 0 (x 2 ) ( x + 3) 14

( x 2 ) ( x + 3) = n 14

27.
28.

i , j = 0, 1.

31.

x = 2,11

x {2, 9} {1,3,5,,11}
x = 9
x {2, 4,9,11}

x = 4

x {0, 2,4,,12} {4,11}


That is the definition for a division ring. A field also needs commutative multiplication.
The concept of magnitude has not been defined in the context of a ring. A unit in a ring is an element with a
multiplicative inverse.
2,3  6 : 2 3 = 0 .

32.

 6 has multiplicative identity 1, 3  6 has identity 3.

x 2 7 x + 3 2
x 2 2 x + 3 7

29.
30.

33.

a. true (a field is a commutative division ring)


b. false ( 2, by Exercise 7)
79

c. false (E)
d. false ()
e. true ( 2  )
f. false (they relate its two operations)
g. true (by Definition 16)
h. true (the operation is associateive by definition of ring, the identity exists and is nonzero by definition of field, and
every nonzero element has an inverse by definition of division ring)
i. true (by Definition 1)
j. true (because a ring is an additive group)
34. (associative) f , g , h F : x  :
f ( gh ) x = fx ( gh ) x = fx gx hx = ( fg ) x hx = ( fg ) h x f ( gh ) = ( fg ) h

(distributive) Left distributivity follows from f , g , h F : x  :

( f ( g + h ) )x

and right distributivity be a similar evaluation.


f , g F : x  : x ( fg ) = ( fg )x = fx gx = x f x g

35.
36.

= fx ( g + h ) x = fx ( gx + hx ) = fx gx + fx hx = ( fg ) x + ( fh ) x = ( fg + fh )x f ( g + h ) = fg + fh

(reflexivity) Obviously under the identity isomorphism a,b R :

i (ab ) = ab = ia ib .

(symmetry) Let : R R be a ring isomorphism. a ,b R : a,b R : a = a ,b = b :

( ab ) = a b = a b

) (

inv ( a b ) = inv ( ab ) = ab = inv a inv b = a b

inv

so : R R is a ring isomorphism also.


(transitivity) : R R, : R R : a,b R :
( ab ) = a b ( ) (ab ) = (ab ) = (a b) = (a ) (b ) = ( )a ( )b .

37.

( ab ) (b a ) = abb a = aa = 1

(closure) a,b U : a ,b R : aa = 1,b b = 1

ab U .

(associativity) by definition of a ring


(identity) 1 R : 1 1 = 1 1 U .
(inverse) a U : a U : aa = 1 a U .
so U is a group.

(a + b ) ( a b) = ( a + b) ( a + ( b) ) = ( a + b ) a + (a + b ) ( b ) = a a + b a + a ( b ) + b (b ) = a 2 b 2 + b a a b

38.

= a2 b 2 b a a b = 0 a b = b a
Clearly this multiplication is associative and distributive, and hence forms a ring.
2 has an element such that a a = a + a (for a = 2), while 3 does not.  has an element of multiplicative order
4 (i), while  does not.
n
n
Since  p is distributive and commutative, the binomial expansion holds: (a + b ) = + i a ib n i . So
i

(a + b ) p = + 0 i p p a ib p i = a p + b p + 0 <i <p a ib p i . Now since p is prime, p = p p !i !i! is always a multiple of p


)
i
i (
for 0 < i < p , so that any such term is always zero.
A field is some closed collection of units of a ring, and by Exercise 37 forms a group under multiplication, so the
identity of any of its subgroups is its identity.
By Exercise 37, U , is a group, which has unique inverses.

39.
40.
41.

42.
43.
44.

a. a,b R : a 2 = 1,b 2 = 1 :
b.

({0,1,3,4}, {0,1, 4, 9}) .

(ab ) 2 = ( ab )( ab ) = abab = aabb = a 2b 2 = 1 .

45.
46.

a,b R : n,m  + : a n ,b m = 0

( ab )

nm

commutative

a nmb nm = 0 m0 n = 0 is easy. How about a + b ?

47. x 0 : x 2 = 0 x is nilpotent
80

x 0 : x is nilpotent

minimal n 0, x n = 0

48. (additive identity) 0 S


(additive inverse) a,b S : a b S
(additive closure) a,b S

49.

1
1n
n
n even : x 2 = 0, x 2 0

2
1

( n + 1) = 0,x 21 ( n + 1) 0
n +1
= x n x = 0 x 2
n odd : x

0 b = 0 + ( b ) = b S
a ( b) = a + b S

b S

So S is a subgroup.
(multiplicative closure) a,b S : ab = S
(multiplicative associativity, distributivity) follow because R is a ring.
So S is a subring.
a. Let R1, 2 R be subrings. From Exercise 48,
0 R1, 0 R2 0 R1 R2
a,b R1 R2 a,b R1 a,b R2

a b R1 a b R2
ab R1 ab R2

a b R1 R2

ab R1 R2

so R1 R2 is a subring.
b. If R is a field, then it is multiplicatively commutative and every element has a multiplicative inverse. Obviously,
multiplication remains commutative in R1 R2 and becuase it is closed, every element has an inverse in R1 R2.
So it is a subfield.
50.
Using Exercise 48. x , y I a ,
a0 = 0 0 Ia
ax, ay = 0 a (x y ) = ax ay = 0 x y Ia
a( xy ) = ( ax ) y = 0y = 0
so Ia R is a subring.

51.
52.

53.

xy Ia

Consider the isomorphism from Example 15 :  rs  r  s : x a x (1,1 ) . Obviously


0 :  rs  r : x a x modr and 1 :  rs  s : x a x mod s . So the problem amounts to finding x such that
0x = m modr , 1x = n mod s , i.e. x = (m modr ,n mod s )  r  s . Since is an isomorphism and thus
surjective, such an x exists.
a. For a set S = {i s i} of relatively prime positive integers. By the Fundamental Theorem of commutative groups,
 i

si

:  i

i  s i are group isomorphic. Since they are generated by 1 and


si

( i 1) respectively, with Theorem 4.5.12

i  s i : x a x ( i 1 ) is a group isomorphism. Multiplicative isomorphism follows from

x , y  i

si

: ( xy ) = ( xy ) ( i 1) = ( i xy ) =

( i x ) ( i y ) = x ( i 1) y ( i 1) = x y .

So is a ring isomorphism.

b. Let ri , si  * with ri relatively prime, show that x  + : i : x = s i r i . Consider the isomorphism of (a.),

:  i

si

i  s i . Obviously i :  i

ix = ri mod s i , i.e. x =
54.

si

 s i : x a x mod si , so the problem amounts to finding x such that

( i ri mod s i ) i s

. Since is an isomorphism and thus surjective, such an x exists.

(additively commutative) a,b S :


(1 + 1 ) a + (1 + 1 )b = a + a + b + b
a + a +b +b = a +b + a +b a +b = b + a
(1 + 1) ( a + b) =
1 ( a + b ) + 1( a + b ) = a + b + a + b
so S is a commutative group.
(multiplicative associativity) Even though we haven't shown S is a ring, the proof of Theorem 8 shows that
multiplication is associative when either of the operands is 0, so multiplication is associative over all of S * (that is,
including the additive identity).
Distributivity holds by axiom, so S is a ring.
(multiplicative identity and inverse) Since S * is a group, it has an identity not the same as the additive identity, and
81

55.

56.

each element has an inverse.


So S is a division ring.
Since every element is idempotent, a R:
a + 1
(a + 1) ( a + 1) = 2
a + 1 = a + a + a + 1 a + a = 0 a = a
a + a + a + 1 = a + a + a + 1
so then a,b R
a + b
(a + b ) ( a + b) = 2
a + b = a + ab + ba + b ab + ba = 0 ab = (ba ) = ba
a + ab + ba + b 2 = a + ab + ba + b
so R is commutative.
S = {a,b }, PS = {,{a},{b},{a,b}} .
a. +

a
b
ab

a
b
ab

a
a

ab
b

b
b
ab

a
b
ab

ab
ab
b
a

b. We show that PS 

b
b

ab

a
b
ab

by :  PS : ( i b i ) a {s i S | b i = 1} .
S

(additively homomorphic) x , y  :

( x + y ) = si S | ( x + y ) i = 1

xi +yi =1

x i =1 y i = 1 x i yi

= {si S | x i = 1} {si S | y i = 1} \ {si S | x i = 1} {si S | y i = 1}


= {si S | x i = 1} + {si S | y i = 1} = x + y
S

(multiplicatively homomorphic) x , y  :

( xy ) = si S | ( xy ) i = 1

x iy i = 1

x i =1 y i = 1

= {si S | x i = 1} {si S | y i = 1}

= {si S | x i = 1} {si S | y i = 1} = x y
S

So is a ring homomorphism. Clearly x = x = 0 ker = E and S PS : x  : x = S , so is


injective and surjective, so is a ring isomorphism.
Now b  n : b 2 = b b =

( i bi b i ) = ( i bi ) = b

5.2 Integral Domains


1.

so  , PS are boolean rings.

x 3 2x 2 3x = x x 2 2x 3 = x ( x 3 )( x + 1 )  12 . This holds if any of the factors is 0, or the product

contains the factors of 12 = 2 2 3. It seems easier to just try x {0,3,5,8,9,11} .


2.

3x = 7 2
3x = 23 2

x = 7 2 3 = 7 2 3 1 = 7 2 5 = 7 10 = 7 3 . Since 3 does not divide 7, there are no other solutions.


x = 23 2 3 = 23 2 3 1 = 23 2 8 = 23 16 . Since 3 does not divide 23, there are no other solutions.
2 4 4 2 1 2 2 i
=
has no integer solutions.
2
2

3.

x 2 + 2x + 2 = 6 0

4.

x 2 + 2x + 4 = ( x + 2 ) = 6 0

5.

x =
2

x = 6 2 = 6 4 . There are no other solutions.

char 2 = 0 ; 6. char  = 0; 7. char 3 3 = 0 ; 8. char 3  3 = 3; 9. char 3  4 = 12;


10. char 6  15 = 30.

11.

(a + b ) 4

= a 4 + 4a 3b + 6a 2b 2 + 4ab 3 + b 4 = a 4 + 2a 2b 2 + b 4 .

12.

(a + b ) 9

3
= ( a + b ) = a 3 + 3a 2b + 3ab 2 + b 3

) = (a
3

+b3

) = (a )
3

( ) (b ) + 3( a ) (b ) + (b )

+ 3 a3

= a9 + b 9 .
82

13.

(a + b ) 6

3
= ( a + b ) = a 3 + 3a 2b + 3ab 2 + b 3

) = (a
2

+b3

) = (a )
2

) ( )

+ 2 a 3b 3 + b 3

= a 6 + 2a 3b 3 + b 6 .

1 2
1 2 2 2 0 0
det
= 0 so the row vectors are linearly dependent:

=
.
2 4
2 4 1 1 0 0
15.
If a,b R are elements of a ring R
16.
If n is the least positive integer
17. a. false ( n does not have a multiplicative identity for n > 1)
b. true (Theorem 9)
c. false (they all have characteristic 0)
d. false ( has multiplicative inverse but 2 doesn't)
e. true (Definition 6 and Theorem 5)
f. true (if it was finite n a = 0 for some n )
g. false (Example 7)
h. true ( a : b : ab = 0 : c : ac = 1 ab + ac = 1 a(b + c ) = 1 and because the inverse is unique, b + c = c b = 0
so a would not be a divisor of 0)
i. false ( n does not have a multiplicative identity for n > 1)
j. false ( is not a division ring or a field)
18.
ring
14.

M n 2
+ multiplicative identity
+ commutative
multiplication

ring with unity


Mn 
+ multiplicative inverses

commutative ring
2

division ring
M n | det 0
+ not multiplicatively
commutative

+ no divisors of 0
integral domain

19.

field


strictly skew field

The matrix is not invertible, has a zero determinant, linearly dependent row or column vectors. (Book says
something about eigenvalues.)

20.

integral domain
commutative
ring

21.
22.

field

ring with unity


strictly skew
field

a R* : a 2 = a aa = a aaa1 = aa1 a1 = 1 a = 1 . Also, for 0 R : 0 2 = 0. So the additive


and multiplicative identities are the only idempotent elements of a division ring.
By Exercise 1.49a, an intersection of rings is a ring, and therefore an intersection of commutative rings is again a
commutative ring. Since the multiplicative identity is unique, it is contained in each of the domains and hence in
83

their intersection. Finally, none of the domains have divisors of zero so neither does the intersection. Therefore,
the intersection is itself an integral domain.
It remains to be shown that each element has a multiplicative inverse. Let R* = {i 1, ai } , and consider

23.

aR* = {i a1,aai } . Each of these elements is distinct, because aai = aaj


a1 = aai

cancellation

a. Suppose a R* : b R : ab = 0
*

b. a R : b R : aba = a
b=0

a i = aj

a j aai = a i

aba = 0a = 0 a , so R cannot have divisors of 0.

abab = ab

abab ab = 0

cancellation

a(bab b ) = 0

aba a = a (ba 1) = 0

ba = 1

no divisors
of 0

bab = b . If

a 1 = b .

Using Theorem 15, the smallest n such that n 1 = 0 must be the same in any subdomain.
{n  n 1} D is a commutative ring with unity and no divisors of 0, so is itself an integral domain. Since any
subdomain of D contains unity and is closed under addition, it must certainly contain

{n  n 1} .

We know that char = 0. Suppose that D : n,m  + : charD = n m . Then

(n 1) (m 1) = ( + i <n 1) ( + i <m 1)

28.

a j = a i , so the left

a = aba = a0a = 0 , so b R .

d. a R* : b R* : aba = a

27.

i = j and

c.
25.
26.

a i = 1 . Now R has no divisors of zero, so / ai : aai = 0 . Thus aR = R, and either

a1 = 1 a = 1 a 1 = 1 or aai = 1 a 1 = a i . Suppose aj : a j a = 1
multiplicative inverse is also the right multiplicative inverse.
24.

cancellation

distributive

distributive

= + i <n (1 + i <m 1 )
= + i <nm 1 1 = + i <nm 1 = 0
and (Theorem 15) n 1,m 1 0 , which would show that D has divisors of 0. So the characteristic has to be prime
or 0.
a. It is fairly obvious that multiplication is closed on S, and we know that S is a commutative group because R and  n
are. Multiplicative associativity follows directly from the definition by observing that swapping indices yields the
same expression. Multiplicative distributivity obviously holds for the second component. For the first,
(r 1,n 1 ) , (r 2 ,n2 ) , (r3 ,n 3) R  n :

(r 1 ,n1 ) ( (r2 ,n 2) + (r 3,n 3 ) ) = (r 1 ,n1 ) (r2 + r 3 ,n2 + n 3 )


= (r 1 (r 2 + r3 ) + n 1 (r2 + r 3) + (n 2 + n 3) r 1 ,n1 (n 2 + n3 ) )
= (r 1r2 + r 1r3 + n 1r2 + n1r3 + n2r1 + n 3r1,n 1n2 + n 1n3 )
= (r 1r2 + n 1r2 + n2r1 ,n1n2 ) + (r1r3 + n 1r3 + n3r1 ,n1n3 )
= (r 1 ,n1 ) (r2 ,n 2 ) + (r 1 ,n1 ) (r3 ,n 3)

Surely right distributivity follows similarly. So S is a ring.


b. (r 1,n 1 ) : (r 2 ,n2 ) : (r1 ,n 1 ) (r 2 ,n2 ) = (r2 ,n 2 ) (r1r2 + n 1r2 + n 2r1,n 1n2) = (r 2,n 2 )
n 1 = 1

r1r2 + r2 + n 2r1 = r2 r 1r2 + n 2r1 = r 1(r 2 + n21 ) = 0 r1 = 0

so 1 S = (0,1 ) .

c. The characteristic of S is the minimal n such that n 1 S = 0 n (0,1 ) = 0 n 1  n = 0 , which is the characteristic
of  n by axiom.
d. Show that is a ring isomorphism so that R S is a ring. r1 ,r2 :
r1 = r2 (r1 , 0 ) = (r2 , 0 ) r 1 = r2 (injective)

(r 1 + r2 ) = (r1 + r 2, 0 ) = (r1 , 0 ) + (r2 , 0 ) = r1 + r 2 (group homomorphism)

(r 1 r2 ) = (r1 r 2, 0 ) = (r1 r 2 + 0 r2 + 0 r 1 , 0) = (r 1, 0 ) (r2 , 0 ) = r1 r 2 (ring homomorphism)


29.

There are  3 = 3 4 = 81 code words and  3 = 9 message words. (Note that the terminology used in this
exercise appears to be inconsistent with that in 2.5).
84

30.

There are F

= 16 4 = 2 16 = 65536 code words and F

= 16 2 = 2 8 = 256 message words.

5.3 Fermat's and Euler's Theorems


1.
2.

3.

{
4 = {4 =
3 = {3 =
2 = {2 =
2 = {2 =
3 = {3 =

3 = 30 = 7 1, 31 = 7 3, 3 2 = 7 2, 33 = 7 6, 3 4 = 7 4, 35 = 7 5, 36 = 7 1 =  7

}
1} 

11 1,

41 = 11 4, 4 2 = 11 5, 43 = 11 9, 4 4 = 11 3, 45 = 11 1 11

11 1,

31 = 11 3, 3 2 = 11 9, 33 = 11 5, 3 4 = 11 4, 35 = 11

11 1,

21 = 11 2, 2 2 = 11 4, 23 = 11 8, 2 4 = 11 5, 25 = 11 10, 26 = 9, 2 7 = 7, 28 = 11 3, 29 = 11 6, 210 = 11 1 = 11

*
11

17

1, 21 = 17 2, 2 2 = 17 4, 23 = 17 8, 2 4 = 17 16, 25 = 17 15, 26 = 17 13, 2 7 = 17 9, 28 = 17 1 17

17

1, 31 = 17 3, 3 2 = 17 9, 33 = 17 10, 3 4 = 17 13, 35 = 17 5, 36 = 17 15, 3 7 = 17 11,

38 = 17 16, 39 = 17 14, 310 = 17 8, 311 = 17 7, 312 = 17 4, 313 = 17 12, 314 = 17 2, 315 = 17 6, 316 = 17 1 = 17

( )
= (37 ) 37

4.

3 47 = 23 3 22 33 = 23 12 33 = 23 27 = 23 4 .

5.

37 49

6.

22

17

= 7 18 37 1 = 7 37 = 7 2 .

( )2

= 19 2n 18 + 14 = 19 218

14

( )

( )

= 19 1n 214 = 19 214 = 19 2 4 2 2 = 19 3 4 = 19 27 4 = 19 11 4 = 19 44 = 19 6

217 = a 18 + b 216 = a 9 + 1 b

( )

216 = 9 26

2 4 = 9 12 2 4 = 9 7 217 = 18 14

17

2 2 + 1 = 19 6 + 1 = 19 7
7.
0x
x0
x1
x2
x3
x4
x5
x6
x7
x8
x9

1
1
2
21 = 2
4
2
6
4
32 = 6

1x
2x
4
8
10
12
4
10
12
22
6
8
8 54 = 20
8
12
16
18
6
12
18
28

8.

p 2 = p p 1 .

9.

pq = pq p q + 1 = p 1 q 1 .

( )

( )

125

( )

= 24 7 24

)(

125

3x
8

10.

7 1000 = 24 7 8

11.

2x = 4 6 x = 2 3 x

12.

22x = 15 5 7 x = 15 5 x = 15 7 15 = 15 55 = 15 10

= 24 1125 = 1 .

}.

{ } (3 + k 2) + 4

k 0, 1

7 1, 2, 4, 7, 8, 11, 13, 14 = 15 7, 14, 13, 4, 11, 2, 1, 8 7 1 = 15 11


10 + 15
85

13.

36x = 24 15; gcd 36, 12 =15 12, 15 =12 3 0 .

14.

45x 24 15 15x =8 5 7 x =8 5 x =8 7 15 =8 7 5 = 35 =8 3

) (

) (

7 1, 3, 5, 7 = 7, 21, 35, 49 =8 7, 5, 3, 1 7 1 =8 7

x 3 + 3 + 8

15.

39x =9 125; gcd 39, 9 = 3, 125 =3 1 0 .

16.

41x =9 125 5x =9 8

) (

) (

5 1, 2, 4, 5, 7, 8 = 5, 10, 20, 25, 35, 40 =9 5, 1, 2, 7, 8, 4 51 =9 2


1

x =9 5 8 =9 2 8 = 16 =9 7

x 7 + 9
17.

155x =65 75 31x =13 15 5x =13 2


5 8 = 40 =13 1 51 =13 8
x =13 512 =13 8 2 = 16 =13 3

x 3 + 13
18.

39x =130 52 3x =10 4


3 7 = 21 =10 1 31 =10 7
x =10 314 =10 7 4 = 28 =10 8

x 8 + 10

)(

19.

By Exercise 26, p 1 ! =p 1 p 1 p 2 ! = 1 = p 1 p 2 ! =p 1.

20.

(37 2)!=

37

( )

35! =37 35 34! =37 1 34! =37 351 = 5 7

= 517 1 =37 15 16 = 240 =37 18

5 15 = 75 =37 1 51 =37 15
7 16 = 112 =37 1 7 1 =37 16
(Ex 19)

21.

51!= 51 50 49! =53 1 49!=53 51 50

=53 61 = 9

51 50 = 2550 =53 430 =53 6


9 6 = 54 =53 1 61 = 9
22.

(Ex 19)

29 2 != 27! = 27 26 25 24! = 29 1 24! = 29 27 26 25

= 29 17 1 = 29 8

27 26 25 = 27 650 = 27 27 2 = 54 = 27 17
8 17 = 136 =53 1 17 1 = 8
23.

a. false ( a p 0 p does not divide a )


b. true
c. true (by definition)
d. false ( 1 = 1 / 1 1 = 0 )
e. true
f. true (a product of two relatively prime numbers is still relatively prime)
g. false (the product will not be relatively prime)
h. true
i. false ( if a = p ax =p px =p 0 b )

24.

j. true (what is an incongruent solution?)


The units of 12 are 1, 5, 7, and 11.
86

1
5
7 11
5
1 11
7
7 11
1
5
11
7
5
1
Its multiplicative group is isomorphic to the Klein 4-group.
(ring)

25.

) (

(ring)

)(

Let x p : x 2 = 1 x 2 1 = 0 x 2 + x x 1 = x x + 1 x x + 1 = x 1 x + 1 = 0 . By Corollary 2.4, p


has no divisors of 0, so x =p 1 x =p 1 = p 1.

26.

Since p is odd, p 1 is even, so

{2,, p 2}

is also even. Since by Exercise 25, 1 and p 1 are the only elements

who are their own inverses, the even number elements in 2,, p 2 each have their inverses in that same subset,
so

pi = 2 i

) (

)(

) (

= p!=p 1 (cf. Exercise 19), so p 1 ! = p 1 p 2 ! = p 1 =p 1 .

383838 = 37 19 13 7 3 2

27.

1
n37 n = n36 n n =p n n = 0

n37 n = n18 n n =p n n = 0

37
12
n n = n
n n =p n n = 0
n37 n =383838 0 .
6

n37 n = n6 n n =p n n = 0

18

n37 n = n 2 n n =p n n = 0

36
n37 n = n1 n n =p n n = 0

28.

n37

( )
( )
( )
( )
( )
( )
n = (n ) n n = (n ) n n =
4

51

n n = 0; 383838 5 = 1919190 n37 n =1919190 0 .

5.4 The Field of Quotients of an Integral Domain


1.
2.
3.
4.

In the same way that the field of quotients   was reinterpreted as  , this field of quotients D D can be
interpreted as  i .
 + 2?
A field is a division ring, in which by definition every nonzero element is a unit. Since the zero of D is the zero of F,
that last part of the definition is redundant.
a. true
b. false ( 2 is not a quotient of )
c. true (  *  )
d. false (i is not a quotient of )
e. true
f. true (otherwise + and could not be defined)
g. false (see h.)
h. true (every nonzero element of a division ring is a unit, and a field is a division ring)
i. true
j. true (Corollary 9)

5.

[(2, 4)] , and


[(2, 4)] + [(2, 4)] = [(2 4 + 4 2, 4 4)] = [(16, 16)] = [(16 1, 16 1)] = [(1, 1)], so its field of quotients is at least a subset
of   . Similarly, for any element [(a , b )], a , b  in the field of quotients of , [( 2a , 2)], [( 2, 2b )] are in the
2  is an integral domain. Its field of quotients includes

87

field of quotients of 2 and


6.

[(2a, 2)] [(2, 2b )] = [(2a 2, 2 2b )] = [(4a, 4b )] = [(a, b )] , so the fields of quotients of 

and 2 are equal.


Prove that addition in F is associative.

([(a, b )] + [(c, d)]) + [(e, f )] = [(ad + bc, bd)] + [(e, f )]

[( ) ( ) ( ) ]
= [(adf + bcf + bde , bdf )]
= [(a (df ) + b (cf + de ), b (df ))]
= [(a , b )] + [(cf + de , df )]
= [(a , b )] + ([(c , d )] + [(e , f )])
[(0, 1)] is an additive identity in F: [(a, b )] F : [(a, b )] + [(0, 1)] = [(a 1 + b 0, b 1)] = [(a, b )].
[(a, b )] is an additive inverse in F. [(a, b )] F :
[(a, b )] + [(a, b )] = [(a b + b a, b b )] = [((a + a) b , b b )] = [(0 b , b b )] = [(0, b b )] = [(0, 1)]
= ad + bc f + bd e , bd f

7.
8.

( )

0 1 = b b 0 0 = 0

9.

10.

11.

12.

13.
14.
15.
16.

[( )] [( )] [( )] F :
([(a, b )] [(c, d)]) [(e, f )] = [(ac, bd)] [(e, f )] = [((ac)e, (bd) f )]
= [(a (ce ), b (df ))] = [(a , b )] [(ce , df )] = [(a , b )] ([(c , d )] [(e , f )])
Multiplication in F is commutative. [(a , b ), (c , d )] F :
[(a, b )] [(c, d)] = [(ac, bd)] = [(ca, db )] = [(c, d)] [(a, d)].
Distribution laws hold in F. [(a , b )], [(c , d )], [(e , f )] F :
[(a, b )] ([(c, d)] + [(e, f )]) = [(a, b )] + [(cf + de, df )] = [(a(cf + de ), b (df ))] = [(ba(cf + de ), bb (df ))]
= [(bacf + bade , bbdf )] = [(ac bf + bd ae , bd bf )] = [(ac , bd )] + [(ae , bf )] = [(a , b )] [(c , d )] + [(a , b )] [(e , f )]
a. t T : [(t , t )] is unity.
b. t , t T : [(t , t )] [(t , t )] = [(tt , t t )] = 1.
By Exercise 12, Q (R, a ) is a commutative ring with unity.
Q (  , {1, 3}) = { = , = , = = , = = } has 4 elements.
Q , {
2 } are all {
n 2 }.

Q 3, {
6 } are {3n 6 } = {3n 6 6 } = { n 6 } = all fractions n 2 + o 3 .

Multiplication in F is associative. a , b , c , d , e , f

0
1

n  +

n  +

0
3

1
1

3
3

2
1

6
3

n , m  +

2
3

3
1

9
3

1
3

n 1

1
2

n 1

17.

5.5 Rings of Polynomials

[]

R x is the set of formal polynomials with coefficients in R and indeterminate x. Formal means that the
indeterminate is to be seen as purely a symbol with no algebraic interpretation. A polynomial is an infinite sum
88

[]

[]

f x = + i f i x i with a finite number of nonzero coefficients. R R x are the constant polynomials. The

1.

finiteness enables or simplifies some kinds of operations (see for example Phigh in the section on ordered rings) but
isn't necessary for the polynomial concept itself. In fact, that same section defines power series rings and
Laurent series fields which modify this restriction in different ways.
The evaluation homomorphism assigns a value from some superfield E to the indeterminate:
a + i f i x i = + i f i a i .

[]
[]
[]
f [ x ] + g [ x ] = ( 4x 5) + ( 2x 4x + 2) = 2x + ( 4 4)x + ( 5 + 2) = 2x + 5 .
f [ x ] g [ x ] = ( 4x 5)( 2x 4x + 2) = 4x ( 2x 4x + 2) 5( 2x 4x + 2)
f x = 4x 5, g x = 2x 2 4x + 2 in 8 x :
2

= 8x 3 16x 2 + 8x 10x 2 + 20x 10 = 8x 3 26x 2 + 28x 10 = 8 6x 2 + 4x + 6

2.

[]
[]
[]
f [ x ] + g [ x ] = ( x + 1) + ( x + 1) = 2x + 2 = 0
f [ x ] g [ x ] = ( x + 1)( x + 1) = x ( x + 1) + 1( x + 1) = x + 2x + 1 =
f [ x ] = 2x + 3x + 4, g [ x ] = 3x + 2x + 3 in  [ x ] :
f [ x ] + g [ x ] = ( 2x + 3x + 4) + (3x + 2x + 3) = 5x + 5x + 7 =
f [ x ] g [ x ] = ( 2x + 3x + 4) (3x + 2x + 3)
f x = x + 1, g x = x + 1 in  2 x :
2

3.

x 2 + 1.

5x 2 + 5x + 1

) (

) (

= 2x 2 3x 2 + 2x + 3 + 3x 3x 2 + 2x + 3 + 4 3x 2 + 2x + 3
= 6x 4 + 4x 3 + 6x 2 + 9x 3 + 6x 2 + 9x + 11x 2 + 8x + 12
= 6x 4 + 13x 3 + 24x 2 + 17 x + 12
= 6 x 3 + 5x

4.

[]
[]
[]
f [ x ] + g [ x ] = ( 2x + 4x + 3x + 2) + (3x + 2x + 4) = 3x
f [ x ] g [ x ] = ( 2x + 4x + 3x + 2) (3x + 2x + 4)

f x = 2x 3 + 4x 2 + 3x 2, g x = 3x 4 + 2x + 4 in 5 x :
3

+ 2x 3 + 4x 2 + 5x + 6 = 5 3x 4 + 2x 3 + 4x 2 + 1

) (

) (

= 2x 3 3x 4 + 2x + 4 + 4x 2 3x 4 + 2x + 4 + 3x 3x 4 + 2x + 4 + 2 3x 4 + 2x + 4
7

= 6x + 4x + 8x + 12x + 8x + 16x + 9x + 6x + 12x + 6x + 4x + 8


= 6x 7 + 12x 6 + 9x 5 + 10x 4 + 16x 3 + 22x 2 + 16x + 8
= 5 x 7 + 2x 6 + 4x 5 + x 3 + 2x 2 + x + 3
3+1

= 2 4 = 16 .

5.

2

6.

5

7.

2 x 2 + 3 = 22 + 3 = 4 + 3 = 7 = 7 0 .

8.

2+1

= 53 = 225.

( )
( 2x x
3

+ 3x + 2 = 2 03 0 2 + 3 01 + 2 = 2 = 7 2 .

89

9.

( [ ] [ ])

3 f x g x

homomorphism

[]

[]

3 f x 3 g x

) (
)
= (3 + 2 3) (3 3 3 + 3) = (81 + 6) ( 27 27 + 3)
= 3 x 4 + 2 x 3 x 3 3 x 2 + 3
4

10.

)(

=7 3 3 = 9 =7 2

)(

) (

) (

5 x 3 + 2 4x 2 + 3 x 7 + 3x 2 + 1 = 5 x 3 + 2 5 4x 2 + 3 5 x 7 + 3x 2 + 1

)(

)(

= 53 + 2 4 52 + 3 57 + 3 52 + 1

= 5 127 103 5 + 75 + 1 = 5 2 3 + 1 = 6 = 5 1
11.

4 3x 106 + 5x 99 + 2x 53 = 4 4 3x 4 + 5x 3 + 2x 5
4

= 3 4 + 5 4 + 2 4 = 3 256 + 5 64 + 2 1024
=4 3 0 + 5 0 + 2 0 = 0

) ( ) ( ) ( ) {}
) ( x + 2x + 2) = ( 2, 5, 14, 35, 74, 137, 230) = ( 2, 5, 0, 0, 4, 4, 6) Ker( x + 2x + 2) = {2, 3} .
) ( x + 3x + x + 2x ) = (
) (3x + x + 3x ) = (0, 7, 34, 99, 220) = (0, 2, 4, 4, 0)
Ker(3x + x + 3x ) = {0, 4}

12.

(0, 1) x 2 + 1 = 1, 2 = 2 = 1, 0 Ker x 2 + 1 = 1 .

13.

(0,, 6

14.

(0,, 4

15.

0, , 4

( [ ] [ ])

[]

[] (
= (5 0, 1 5, 0 2, 1 5, 3 0, 6 1, 5 5)
= (0, 5, 0, 5, 0, 6, 25) = (0, 5, 0, 5, 0, 6, 4)
Ker( f [ x ] g [ x ]) = {0, 2, 4}

(0,6) f x g x = i (0,6) i f x i g x = 5 0, 8 5, 21 16, 50 33, 101 56, 230 85, 293 110
7

) (

16.

3 x 231 + 3x 117 2x 53 + 1 = 5 3 x 3 + 3x 1 2x 1 + 1 = 3 x 3 + x + 1 = 27 + 3 + 1 = 31 = 5 1 .

17.

2x 219 + 3x 74 + 2x 57 + 3x 44 = 5 2x 3 + 3x 2 + 2x 1 + 3x 0

) (
+ 3x ) = {0, 1, 2, 3}

) (

(0,, 4) 2x 3 + 3x 2 + 2x 1 + 3x 0 = 0, 10, 35, 90, 187 = 5 0, 0, 0, 0, 2

Ker 2x 219 + 3x 74 + 2x 57
18.
19.
20.

44

Replace coefficients ai and ai 0 for a finite number of i.


Seems okay.

(3x

) (

) (

) (

+ 2x y 3 + x 2 6x + 1 y 2 + x 4 2x y + x 4 3x 2 + 2

= 3x 3 y 3 + 2xy 3 + x 2 y 2 6xy 2 + y 2 + x 4 y 2xy + x 4 3x 2 + 2

( ) (

) (

) (

= y + 1 x 4 + 3y 3 x 3 + y 2 3 x 2 + 2y 3 6y 2 2y x + y 2 + 2
21.

22.

1 is unity in  4 1 is unity in  4 x . 1 + 2x  4 x has 1 + 2x

23.

1 i +1
i  5 x

[]

x i Ker 5 :  x  .

a. true
b. true (Theorem 2)

[]

[]

= 1 2x .

[]

c. true (If D has no divisors of zero, then D x cannot possibly have them either)
d. true (if d D is a divisor of zero with d D : dd = 0 , then dx d x = 0x 2 = 0 )
90

e. false
f. false ( 2x 3 2x 4 =  4 [ x ] 4x 7 = 0 )

( [ ] [ ])

[]

[]

[]

g. true (because f x g x = a f x g x = 0 a g x = 0)
h. true (if F is a field it has no divisors of zero, so a product with a polynomial of degree > 0 can never have degree 0)
i. true (because 1 R is never a divisor of zero)

[]

j. false ( 2x  4 x is a divisor of zero, because 2x 2x = 4x 2 = 0 )


24.

[]
of zero, neither does D[ x ] . f [ x ] D[ x ], f [ x ] 0 f [ x ] = +
f x : i  a 0 . For any
g D[ x ], g [ x ] = +
g x , let f , g be the first coefficients of f [ x ], g [ x ] such that f , g 0. Since D is an
integral domain, f g 0, and since this the only term of degree i + j of f [ x ] g [ x ], f [ x ] g [ x ] 0.
a. Since an integral domain has no divisors of zero, suppose f [ x ] is of degree 1 and f [ x ] = + f x , g [ x ] = + g x ,
let f , g be the highest coefficients f 0, g 0; i , j 1. Then f [ x ] g [ x ] will contain a term
f g 0 f [ x ] g [ x ] 1. For every f [ x ] of degree 0 f [ x ] = f , so f [ x ] g [ x ] = 1[ x ] iff f D is a unit. If the
degree of f [ x ] / 0, then f [ x ] = 0 f [ x ] g [ x ] 1[ x ] . So the units of D[ x ] are exactly the units of D.
b. The only units of  are 1 and 1, so by (a.) 1[ x ], 1[ x ] are the only units of [ x ] .
c. By Corollary 2.12,  is a field so all i  are units, so by (a.) all {
i[ x ]} are units.
f [ x ], g [ x ], h [ x ] R[ x ] :
f [ x ] ( g [ x ] h [ x ]) = + f x ( + g x + + h x )
= + f x + ( g + h )x
f ( g h ))x
= + (+
Theorem 2 says that if F is a commutative ring then F x is also. Remaining to be proved that if D has no divisors
i  i

25.

i 
i

26.

definition

i i

i
j =0

Def 5.1R3

definition

i  7

ij ij

+ i + ij = 0 f j g i j + + ij = 0 f j h i j x i

) (

= + i + ij = 0 f j g i j x i + + ij = 0 f j h i j x i

= + i + ij = 0 f j g i j x i + + i + ij = 0 f j h i j x i
definition

+i f i x i +i g i x i + +i f i x i +i hi x i

[] [] [] []

= f x g x + f x h x
27.

[] [] []
D( f [ x ] + g [ x ]) = D( + f x + + g x ) = D + ( f + g )x
= + i ( f + g )x
= + (i f + i g )x
= + (i f )x
+ + (i g )x
=D+ f x
= Df [ x ] + Dg [ x ]

a. f x , g x F x :

i +1

i 1

definition

i +1

5.1R3

i 1

i +1

i 1

i 1

+ D +i g i x i

So D is a group homomorphism. But Dx Dx = 1 1 = 0 Dx 2 = 2x , so D is not a ring homomorphism.

[] []

[]

b. f x F x : Df x = 0 D + i f i x i = + i + 1 if i x i 1 = 0 i > 0 : f i = 0 , so Ker D =

f 0 F

f 0x 0 = F .
91

[] [] []
() f [ x ] F [ x ] : f [ x ] = + f x

[]

c. () f x F x : Df x = D + i f i x i = + i 1 if i x i 1 F x
i

28.

a.

) : F [ i x i ] E : f [ i x i ] = + ( p ) f ( p ) i
n

n
i i

[]

f
f
: D + i i x i + 1 = + i 1 i i 1 x i 1 = + i 1 f i 1x i 1 = + i f i x i = f x .
i +1
i

n
i

n
i

( )

xi

pi

a+

( p ) f ( p ) i i
n

n
i

n
i

pi

( )

2
4
4
2 2

b. 3, 2 x 1 x 2 + 3x 1 x 2 = 3 2 2 + 3 3 21 = 9 4 + 3 81 2 = 36 + 486 = 519 .

c. A zero of a polynomial f
29.

RR =

f :R R

[ x ] is an n-tuple ( ) such that (


n
i

n
i

) f [ i i ] = 0.
n

n
i i

f .

(associativity) , , RR , r R :

(( + ) + )r = ( + )r + r = r + y + r = r + ( + )r = ( + ( + ))r .
(

RR : + 0 r = r + 0r = r .

(additive identity) 0 RR : r a 0;

( )

(additive inverse) RR : 1 RR : r R : r a r + 1 r = r + 1r = r + r = 0 .
(multiplicative associativity) , , RR : r R:

(( ) )r = ( )r r = r r r = r ( )r = ( ( ))r

(left distributivity) , , RR : r R:

( (

))

+ r = r + r = r r + r

R a ring

) ( ) (

r r + r r = r + r = + r

(right distributivity) id.


Ex. 29

30.

[]

(additive closure) , PF : PF ,

( )

f , g F x : a F : a = fa , a = ga

a F : a = a a = fa ga = f g a PF .
(additive identity) 0F F F F : a a 0 . 0PF PF : + i 0 x i : a a 0

[] []

[]

0PF = 0F F .

(additive inverse) PF : f x F x : a F : a = f x a :

1 F F : a F : 1a = a

( )

a F : + 1 a = a + 1a = a + a = 0

a = fa a F : a = fa
[]
(multiplicative closure) , P : f , g F [ x ] : a F : a = fa , a = ga
a F : ( )a = a a = ( f g )a , f g F [ x ] P .
f

F x : a F : f

PF

(left, right distributivity)


b. It seems obvious that every polynomial can be interpreted as an element of PF under the evaluation
homomorphism, and conversely. So they can be not isomorphic only considering tricks such as letting x 2 and

( x )

31.

a.  2

2

b.  2

2

be different functions in F F .
= 2

2

= 2 2 = 4 ; 3

3

= 3

3

= 33 = 27 .

, +  2  2 ; 3 3 , + 3 3 3 .

c. It remains to be shown that F F PF .

[] []

)(

) (

)(

f i x F x : a a1 a 2 ai 1 ai + 1 a F so that
92

[]
[] (

ai : f i x = 0

= ai : f i x = ai a1 ai a 2 ai ai 1 ai ai + 1 ai a F

fi x
F
Let f F x : a + i x i
(exists because F is a field), so
f i x i

)(

[]

ai F : fai = + j a j

[]
[]

[]
f [ x ]a

f j x ai
j

) (

= ai

)(

[]
f [ x ]a

f i x ai
i

= ai and f PF F F PF F F = PF .

5.6 Factorization of Polynomials over a Field


1.

i
Let f [ x ] = + ni f i x i , g[ x ] = + m
i g i x F [ x] ;

f n , g m 0 . Then f [ x ] = q[x ] g[ x] + r [ x ] , where q[ x ] ,r[ x ] F [ x]

are unique and r[ x ] = 0 deg r [x ] < deg g[ x ] . Roughly:

f [x ]
r[x ]
= q[x ] +
g[ x]
g[ x ]

1.

x 2 + 2x 3

x 6 + 3x 5 + 4x 2 3x + 2 x 4 + x 3 + x 2 + x + 5
x 6 + 2x 5 3x 4
x 5 + 3x 4 + 4x 2 3x + 2
x 5 + 2x 4 3x 3
x 4 + 3x 3 + 4x 2 3x + 2
x 4 + 2x 3 3x 2
x 3 + 7x 2 3x + 2
x 3 + 2x 2 3x
5x 2 + 2
5x 2 + 10x 15
10x + 17 = 4x + 3

2.

3x 2 + 2x 3 x 6 + 3x 5 + 4x 2 3x + 2 5x 4 + 5x 2 + 6x
x 6 + 3x 5 + 6x 4
x 4 + 4x 2 3x + 2
x 4 + 3x 3 + 6x 2
4x 3 + 5x 2 3x + 2
4x 3 + 5x 2 + 3x
x +2

3.

2x + 1

x 5 2x 4 + 3x 5 6x 4 + 7x 3 + 2x 2 + 10x + 2
x 5 + 6x 4
3x 4 + 3x 5
3x 4 + 7x 3
4x 3 + 3x 5
4x 3 + 2x 2
9x 2 + 3x 5
9x 2 + 10x
4x 5
4x + 2
4

93

4.

5x 2 x + 2

x 4 + 5x 3 3x 2 9x 2 + 5x + 10
x 4 + 2x 3 + 7x 2
3x 3 + x 2
3x 3 + 6x 2 + 10x
6x 2 + x
6x 2 + x + 9
2

5.

2 = 5 2 0 = 1,2 1 = 2, 2 2 = 4, 2 3 = 3 =  5 , so  5 = 2 1 = 5 2 = 2 3 = 5 3 ; {2,3} .

6.

3 =7

7.

3 = 17

{
{3 = 1,3 = 3,3
{3 = 1,3 =
0

= 2,3 3 = 6,3 4 = 4,3 5 = 5 =  7 , so  7 = 3 1 = 7 3 = 3 5 = 7 5 ; {3,5} .

17

3,3 2 = 17 9,3 3 = 17 10,3 4 = 17 13,3 5 = 17 5,3 6 = 17 15,3 7 = 17 11,

17

, so

3 8 = 17 16,3 9 = 17 14,3 10 = 17 8,3 11 = 17 7,3 12 = 17 4,3 13 = 17 12,3 14 = 17 2,3 15 = 17 6 =  17

 17 = 3 1 = 17 3 , 3 3 = 17 10 , 3 5 = 17 5 , 3 7 = 17 11 , 3 9 = 17 14 , 3 11 = 17 7 , 3 13 = 17 12 , 3 15 = 17 6 ;

{3,10,5,11,14, 7,12, 6} .
8.

5 = 23 5 0 = 23 1,5 1 = 23 5,5 2 = 23 2,5 3 = 23 10,5 4 = 23 4,5 5 = 23 20,5 6 = 23 8,5 7 = 23 17,5 8 = 2316,


5 9 = 23 11,5 10 = 23 9,5 11 = 23 22,5 12 = 23 18,5 13 = 23 21,5 14 = 23 13,5 15 = 23 19,5 16 = 23 3,

5 17 = 23 15,5 18 = 23 6,5 19 = 23 7,5 20 = 23 12,5 21 = 23 14


*

, so  23 = 5 = 23 5 = 5 = 23 15 = 5 = 23 20 = 5 = 23 17 = 5 9 = 23 11

= 5 13 = 23 21 = 5 15 = 23 19 = 5 17 = 23 15 = 5 19 = 23 7 = 5 21 = 23 14

{5,10, 20,17,11, 21,19,15, 7,14} .


9.

10.

(
(
(

x4 + 4
= x3 + x 2 + x +1
x 1
x3 + x 2 + x +1
2 x 3 + x 2 + x + 1 = 5 3 + 4 + 2 + 1 = 5 0;
= x 2 + 3x + 2
x 2
x 2 + 3x + 2
3 x 2 + 3x + 2 = 5 4 + 4 + 2 = 5 0;
= x +1
x 3
x 4 + 4 =5 x 1 x 2 x 3 x 4

1 x 4 + 4 = 5 0;

(
(x

)(

)(

x 3 + 2x 2 + 2x + 1
= x2 + x +1
x +1

1 x 3 + 2x 2 + 2x + 1 = 7 1 + 2 = 2 + 1 = 7 0;
2

x + 2x + 2x + 1 = 7

x2 + x +1
= x +3
x 2
x +1 x 2 x + 3

+ x + 1 = 7 4 + 2 + 1 = 7 0;

11.

)(

)(

)(

2x 3 + 3x 2 7 x 5
= 2x 2 + 9x + 9
x 3
2x 2 + 9x + 9
= 2x + 3
x +3

3 2x 3 + 3x 2 7 x 5 = 11 1 + 5 + 1 5 = 11 0;

3 2x 2 + 9x + 9 = 11 18 27 + 9 = 11 0;

4 2x + 3 = 11 8 + 3 = 11 0;
2x 3 + 3x 2 7 x 5 = 11

2x + 3
=2
x 4
2 x 3 x +3 x 4

)(

)(

94

12.

(
(x

)
+ 2x + 1) = 1 2 + 1 =

x 3 + 2x + 3
= x 2 + 2x + 1
x 2
x 2 + 2x + 1
= x +1
x +1

2 x 3 + 2x + 3 = 5 3 + 4 + 3 = 5 0;
1

)(

)
(2, 2, 1, 1, 4) is irreducible.

x 3 + 2x + 3 = 5 x 2 x + 1

0;
2

13.

(0,, 4) 2x 3 + x 2 + 2x + 2 = 5

14.

If f x is reducible over , then by Theorem 10 it has a zero in , and by Corollary 12 it has a zero in  that

[]

[] {

[]

divides 2, which should therefore be one of 1, 2 . But {1, 2} f x = 7, 9, 18, 16 . The roots of f x are
8 8 2 4 1 2
= 4
2 1
15.
16.

17.

[]

64 + 8 , so f x is reducible over  and .

1
2

[]

g x is an Eisenstein polynomial with p = 3, so it is irreducible over . Since D = 6 2 4 1 12 = 36 48 it is


irreducible over  but reducible over .
By Corollary 12, if it is reducible over  then it has a zero in  that divides 8, which should therefore be one of

{1, 2, 4} .

) {

But {1, 2, 4} x 3 + 3x 2 8 = 4, 6, 12, 4, 104, 24 .

Likewise, it should have a zero that divides 1, which should therefore itself be 1. But

) {

{1} x 4 22x 2 + 1 = 20, 22 .


18.

Yes for p = 3, a 2 = 1 3 0,

a1 = 3 0,

19.

Yes for p = 3, a3 = 8 3 0,

a 2 = 6 = 3 0,

20.

No. Because a3 = 9 = 3 2 , the only possibility is p = 3, but a0 = 18 = 3 2 0 .

21.

Yes for p = 5, a10 = 2 5 0,

22.

x 5 , 2 ,

6x + 17 x + 7 x 2 + x 10
x+

5
2

6x 3 + 2x 2 + 2x 4
x

23.
24.
25.

2
3

a0 = 12 3 2 0 .
a1 = 9 = 3 0,

a3 = 25 = 5 0,

a0 = 24 3 2 0 .

a 2 = 10 = 5 0,

a0 = 30 52 0 .

= 6x 3 + 2x 2 + 2x 4

= 6x 2 + 6x + 6

6x 2 + 6x + 6 = 0 x 2 + x + 1 = 0;
12 4 1 1 = 3 , so there are no other roots in .
nonconstant polynomial. Insert and g, h both of lower degree than f.
a. true (of degree 1, so both factors can't have degree less than 1)
b. true (same reason)
c. true (both roots are in  \  )
d. false (because 2 is a zero, so by Theorem 10 is reducible)
e. true (The degree of a product of nonzero polynomials is always the sum of the degrees of the factors, so a nonzero
polynomial can only have an inverse if it is of degree 1. The zero polynomial has no inverse.)

()

f. ? (what is F x ?)
g. true (Corollary 3)
h. true (Corollary 3)
i. true
j. false (because of the zero polynomial; however, the book gives true)
26.
x + 2 is a factor if 2 is a zero, so

2 x 4 + x 3 + x 2 x + 1 = 0

(2) + (2) + (2) (2) + 1 = 16 8 + 4 + 2 + 1 = 14 =


4

{ }

0 , so p 2, 7 .
95

27.

x 2 + x + 1.

28.

x 3 + x 2 + 1,

29.

x2 +1
2x 2 + 2
x2 + x + 2
2x 2 + x + 1
2
x + 2x + 2 2x 2 + 2x + 1

30.

x 3 + 2x + 1
x 3 + 2x + 2
x3 + x 2 + 2
x3 + x 2 + x + 2
x 3 + x 2 + 2x + 1
x 3 + 2x 2 + 1
x 3 + 2x 2 + x + 1
x 3 + 2x 2 + 2x + 2

x 3 + x + 1.

2x 3 + x + 2
2x 3 + x + 1
2x 3 + 2x 2 + 1
2x 3 + 2x 2 + 2x + 1
2x 3 + 2x 2 + x + 2
2x 3 + x 2 + 2
2x 3 + x 2 + 2x + 2
2x 3 + x 2 + x + 1

31.
32.

By Euler's Theorem, x p 1 = p 1

x p + a = p x + a . Thus, for any a p , a is a zero of x p + a so by the Factor

Theorem x + a is a factor of x p + a , so it is not irreducible.


33.
34.
35.

( )

a 0

a n 1 a + ni = 0 an i x i = a n + ni = 0 an i a i = + ni = 0 an i a n i = + ni = 0 ai a i = a ai x i = 0

( )
[ ] [ ]( ) [ ]
a. : [ x ]  [ x ] : + a x a + a x . For any f [ x ] = + f x , g [ x ] = + g x
( f [ x ] g [ x ]) = ( + f x + g x ) = ( + + f g x ) = + ( + f g )x

1 a + ni = 0 an i x i = 0.

f x = q x x + r x . Then obviously, for x = : f = q + r = r .


i

m i

i
j

ij

i
j

ij

[]

 x :

= m + i + ij m f i g i j x i = m + i + ij m f i m g i j x i = + i m f i x i + i m g i x i = m + i f i x i m + i g i x i

[] []
b. f [ x ] [ x ] : deg f [ x ] = deg f [ x ] . Suppose f [ x ] = g [ x ] h [ x ], deg g [ x ], deg h [ x ] < deg f [ x ] .
Since is a homomorphism, g [ x ], h [ x ] [ x ] : g [ x ] = g [ x ], h [ x ] = h [ x ], so
f [x ] = g [x ] h [x ]
=
g [ x ]h [ x ] . Suppose f [ x ] is reducible in [ x ] and by Theorem 11 then in
[ x ] as f [ x ] = g [ x ] h [ x ] . Then f [ x ] = g [ x ]h [ x ] = g [ x ] h [ x ] would also be reducible in  [ x ] (
= m f x m g x

homomorphism

does not affect the degree).

c. Consider 3 x 3 + 17 x + 36 = x 3 + 2x = x x 2 + 2

5.7 Noncommutative Examples


1.
2.
3.

(2e + 3a + 0b ) + (4e + 2a + 3b ) = 1e + 0a + 3b .
(2e + 3a + 0b )(4e + 2a + 3b ) = (2 4 + 3 3 + 0 2)e + (2 2 + 3 4 + 0 3)a + (2 3 + 3 2 + 0 3)b = 2e + 1a + 2b .
(3e + 3a + 3b ) = (3 3 + 3 3 + 3 3)e +a +b = 2e + 2a + 2b
(3e + 3a + 3b ) = (2e + 2a + 2b ) = (2 2 + 2 2 + 2 2)e +a +b = 2e + 2a + 2b .
(i + 3 j )(4 + 2 j k ) = (0 4 1 0 3 2 0 1) + (0 0 + 1 4 + 3 1 0 2)i
+(0 2 1 1 + 3 4 + 0 0) j + (0 1 + 1 2 3 0 + 0 4)k = 6 + 1i + 13 j + 2k
2

4.

96

5.
6.
7.

( )

( )

i 2 j 3kji 5 = i 2 j 2 jkj i 2 i = 1 1 jkj 1 i = jkji = i k = j .

(i + j ) = i i + i j + j i + j j = 1 + k k 1 = 2 (i + j ) = i +1 j ii ++ jj = i +2j = (i + j ) .
(1 + 3i )(4 j + 3k ) = 1 4 j + 1 3k + 3i 4 j + 3i 3k = 4 j + 3k + 12k 9 j = 5j + 15k = 5( j + 3k )
( j + 5k ) = j j j 3k + 3k j + 3k 3k = 1 3i + 3i 9 = 10
1
1
j + 3k
j + 3k
((1 + 3i )(3 j + 3k )) =
=

=
= ( j 3k )
j + 3k j + 3k
10
5( j + 5k )
1

1
2

8.

1
5

1
50

(0 + 1 + 0 + 0 + 1 + 1 )(1 + 1 + 0 + 1 + 0 + 1 )
= (0 + 0 + 0 + 0 + 0 + 0 ) + (1 + 1 + 0 + 1 + 0 + 1 ) + (0 + 0 + 0 + 0 + 0 + 0 )
+(0 + 0 + 0 + 0 + 0 + 0 ) + (1 + 1 + 0 + 1 + 0 + 1 ) + (1 + 1 + 0 + 1 + 0 + 1 )
0

9.

1
5

= 10 + 11 + 0 2 + 11 + 0 2 + 13

is commutative, so Z =  . Now consider
\  , that is all the quaternions that are nonzero in at least one
of i, j, or k. Considering just these three components, we can show that they form a group isomorphic with 3
under vector cross product: g , h
\  :

( ) ((
)(
))
= (( g h 1 + g h k g h j ) + ( g h k g h 1 + g h i ) + ( g h j g h i g h 1))
= (()1 + ( g h g h )i + ( g h + g h ) j + ( g h g h )k )
= [g h g h
g h gh
gh g h ]
= [g
g
g ] [h h h ]

gh = g i i g j j g k k h i i h j j h k k
i i

i k

j k

j k

k i

j i

i k

i k

j k

k i

k i

k k

j i

j i

= g h
This shows that for any g
\  we can find an h which is noncolinear under its vector interpretation. Since

g h h g for g, h 0 and not colinear, we have that for any g 1

gi

gj

] [

g k , g 1 hi

hj

h k will not

commute. So Z
\  = E , and Z
* = * .
10.

Let
1 j ,
1k
such that
1 j =

hi , hk = 0 h 11 + h i i

+ h j j + h k k and
1k =

hi , hj = 0 h 11 + h i i

+ h j j + h k k . In the

following, consider j, k as quaternions but let i be the complex root of 1. Show that the field of complex numbers
is isomorphic to one of these subsets of the quaternions under a simple projection
1 j :
1 j  : h 11 + 0i
+ h j j + 0k a h 1 + h j i . Then g , h  : g = g 1 + g i i , h = h 1 + h i i :

1j

(h + h i )
= ( g 1 + 0i + g j + 0k ) + (h 1 + 0i + h j + 0k )
= (( g + h )1 + 0i + ( g + h ) j + 0k )
=
(( g + h ) + ( g + h )i ) = ( g + h )
h =
( g + g i ) (h + h i )
= ( g 1 + 0 g + g j + 0 g ) (h 1 + 0h + h j + 0h )
= ( g h 1 + g h j + g h j g h 1)
= (( g h g h )1 + 0i + ( g h + g h ) j + 0k )
=
(( g h g h ) + ( g h + g h )i ) = ( g h )

g + 1j h = 1j

(g

+ g i i + 1j

1j

1j

g 1j

1j
1

1 1

1 1

1j

1j

1 i

i 1

i i

i i

1 i

i 1

1 i

i 1

1j

1 1

i i

1j

97

so 1 j

is a ring homomorphism. Obviously Ker 1 j

= 0  and 1 j  =
1 j , so it is an isomorphism and


1j . Similarly, 
1k . Obviously
1 j
1k .
11.

a. false (Example 2.8)


b. false (for A M 2  2 to have an inverse, A 0)
c. false ( EndE has only one element and can therefore not have a nonzero multiplicative identity)
d. false ( End has nonzero multiplicative identity)
e. false (isomorphisms under addition are generally not again isomorphisms, e.g. f :   : x a x , f f = 0  )
f. false ( R , + as a group ring has elements that are formal sums that can't be combined under +  and is therefore
infinite-dimensional)

g. true (by the definition, + i ai g i + i bi g i = + i + j , k , g i : g i = g j g k a j bk g i = + i + j , k , g i : g i = g j g k bk a j g i = + i bi g i + i ai g i iff


R is commutative)
h. false (
is not commutative)
i. true (
* is associative by the definition of a ring, generates inverses because the field of quaternions is strictly skew
by Theorem 9, and thus commutative with multiplicative identity 1 by definition; and thus meets all of the
requirements for a group)
j. false ( 
is a field)
12.

a. In
, x 2 + 1 = 0 has solutions i 2 + 1 = 0, j 2 + 1 = 0, k 2 + 1 = 0 .
b. Consider the multiplicative subgroup of
. This is indeed a group because it is associative by definition of a ring,
and each element has an inverse because it is strictly skew. None of the elements of this group are generators:

(1)

13.

( ) = 1, (k ) = 1.
End   : (m , n) = (m + n, 0) , and let End   : (m , n) = (m , m ) .
( )(m , n) = ( (m , n)) = (m , m ) = m + (m ) = 0, so is a left divisor also.

14.

Since F is a field, 0, 1 F . An element of M 2 F has a multiplicative inverse iff its determinant is nonzero, which

15.

1 0 1 0 1 1 0 1 1 1 0 1
includes
,
,
,
,
,
.
0 1 1 1 0 1 1 0 1 0 1 1
Characterize all the endomorphisms of . First, 0 = 0 . Second, let 1 = n , then i = n i and this fully

( )

= 1, i

= 1, j

Then

determines . So n , n  are all the endomorphisms. Also, if n * then n is an automorphism. Now consider
the map : End   : n a n . Then n , m End  and i :

( )
)i = ( i ) = (m i ) = (n m ) i =

n i + m i = n i + m i = n + m i = n + m i

n m

17.

n m = n m n m = n m = n m = n m .

n m i

n End  so is surjective, and Ker = 1 so is injective

so is a homomorphism. Furthermore, n * :
and bijective, so is an isomorphism.
16.

n + m = n + m n + m = n + m = n + m = n + m .

[]

+ i ai x i F x :

(YX XY )(+ a x ) = (YX )(+ a x ) (XY )(+ a x )


i

= Y X + i ai x i X Y + i ai x i

= Y + i ai x

( )

i +1

) X (+ ia x )
i

i 1

= + i i + 1 ai x i + i iai x i = + i ai x i
so YX XY = 1.
18.

{}

If G = E = e , then by definition RE =

re R re e

}.

Let : RE R : re e a re , then re , r e RE :
98

) ((

))

( ) ( )

) (( ) )

( ) ( )

re + r e = r + r e = r + r = re + r e and re r e = rr e = rr = re r e ,
so is a homomorphism. Since r R :

( )

rg = r = 0R
RE R .
a , b , c
:

19.

re RE : rg = r , is surjective and because

{ }

rg = 0RG we have that Ker = 0RG so is injective and bijective, so is an isomorphism and

a = a11 + ai i + a j j + ak k , b = , c = :

(a b )c = (a 11 + a i i + a j j + a k k ) (b11 + bi i + b j j + bk k ) (c 11 + c i i + c j j + c k k )
= (a 1b 1 )1 + (a 1b i )i + (a 1b j ) j + (a 1b k )k + (a i b 1 )i (a i b i )1 + (a i b j )k (a i b k ) j
+(a j b 1 ) j (a j b i )k (a j b j )1 + (a j b k )i + (a k b 1 )k + (a k b i ) j (a k b j )i (a k b k )1 (c 11 + c i i + c j j + c k k )

) (

) (
) (
) (
)
(
) (
) (
) (
)
+(a 1b i + a i b 1 + a j b k a k b j )c 1i (a 1b i + a i b 1 + a j b k a k b j )c i 1 + (a 1b i + a i b 1 + a j b k a k b j )c j k (a 1b i + a i b 1 + a j b k a k b j )c k j
+(a 1b j a i b k + a j b 1 + a k b i )c 1 j (a 1b j a i b k + a j b 1 + a k b i )c i k (a 1b j a i b k + a j b 1 + a k b i )c j 1 + (a 1b j a i b k + a j b 1 + a k b i )c k i
+(a 1b k + a i b j a j b i + a k b 1 )c 1k + (a 1b k + a i b j a j b i + a k b 1 )c i j (a 1b k + a i b j a j b i + a k b 1 )c j i (a 1b k + a i b j a j b i + a k b 1 )c k 1
= (a 1b 1c 1 a i b i c 1 a j b j c 1 a k b k c 1 a 1b i c i a i b 1c i a j b k c i + a k b j c i a 1b j c j + a i b k c j a j b 1c j a k b i c j a 1b k c k a i b j c k + a j b i c k a k b 1c k )1
+(a 1b 1c i a i b i c i a j b j c i a k b k c i + a 1b i c 1 + a i b 1c 1 + a j b k c 1 a k b j c 1 + a 1b j c k a i b k c k + a j b 1c k + a k b i c k a 1b k c j a i b j c j + a j b i c j a k b 1c j )i
+(a 1b 1c j a i b i c j a j b j c j a k b k c j a 1b i c k a i b 1c k a j b k c k + a k b j c k + a 1b j c 1 a i b k c 1 + a j b 1c 1 + a k b i c 1 + a 1b k c i + a i b j c i a j b i c i + a k b 1c i ) j
+(a 1b 1c k a i b i c k a j b j c k a k b k c k + a 1b i c j + a i b 1c j + a j b k c j a k b j c j a 1b j c i + a i b k c i a j b 1c i a k b i c i + a 1b k c 1 + a i b j c 1 a j b i c 1 + a k b 1c 1 )k
= a 1 (b 1c 1 b i c i b j c j b k c k )1 + a i ( b i c i + b 1c 1 b k c k b j c j )i + a j ( b j c j b k c k + b 1c 1 b i c i ) j + a k ( b k c k b j c j b i c i + b 1c 1 )k
+a 1 (b 1c i + b i c 1 + b j c k b k c j )i a i (b i c 1 + b 1c i b k c j + b j c k )1 a j (b j c k b k c j + b 1c i + b i c 1 )k + a k ( b k c j + b j c k + b i c 1 + b 1c i ) j
+a 1 (b 1c j b i c k + b j c 1 + b k c i ) j + a i ( b i c k + b 1c j + b k c i + b j c 1 )k a j (b j c 1 + b k c i + b 1c j b i c k )1 a k (b k c i + b j c 1 b i c k + b 1c j )i
+a 1 (b 1c k + b i c j b j c i + b k c 1 )k a i (b i c j + b 1c k + b k c 1 b j c i ) j + a j ( b j c i + b k c 1 + b 1c k + b i c j )i a k (b k c 1 b j c i + b i c j + b 1c k )1
= (a 11 + a i i + a j j + a k k ) (b 1c 1 b i c i b j c j b k c k )1 + (b 1c i + b i c 1 + b j c k b k c j )i + (b 1c j b i c k + b j c 1 + b k c i ) j + (b 1c k + b i c j b j c i + b k c 1 )k
(b c )1 + (b c )i + b c j + (b c )k + (b c )i (b c )1 + b c k (b c ) j
= (a 11 + a i i + a j j + a k k )
( 1 j) 1 k 1 1 i i ( i j) i k
1i
11
+(b j c i ) j (b j c i )k (b j c j )1 + (b j c k )i + (b k c 1 )k + (b k c i ) j (b k c j )i (b k c k )1
= (a 11 + a i i + a j j + a k k ) (b 11 + b i i + b j j + b k k ) (c 11 + c i i + c j j + c k k ) = a (b c )
= a 1b 1 a i b i a j b j a k b k 1 + a 1b i + a i b 1 + a j b k a k b j i + a 1b j a i b k + a j b i + a k b i j + a 1b k + a i b j a j b i + a k b 1 k c 11 + c i i + c j j + c k k
= a 1b 1 a i b i a j b j a k b k c 11 + a 1b 1 a i b i a j b j a k b k c i i + a 1b 1 a i b i a j b j a k b k c j j + a 1b 1 a i b i a j b j a k b k c k k

5.8 Ordered Rings and Fields


x 2 x 1 Phigh

1.

x a Phigh

2.

x i x i + 1 Plow

3.

All the positive elements of  2 :



n + m 2 P

4.
5.
6.
7.
8.
9.
10.
11.

i. a c d e b
i. a c e d b
i. c a b e d
i. d a b c e
i. e a c b d
i. c a e d b
bdace
a:

a < x;

x 1 < x 2 ; so a < x 1 < x 2 < .

x i + 1 < x i ; so x 3 < x 2 < x 1 < x 0 = 1 < x 1 < x 2 < x 3 .

inv

n + m 2 P

n m 2 P

n m 2 > 0

n > 0 m < 0

n > 0 2m 2 < n 2 .

2
2
m > 0 n < 2m

ii. d b a e c
ii. e c b a d
ii. e c a b d
ii. d c e a b
ii. c d a e b
ii. e c b a d

1
x2
1
x
= 1 + x + x 2 + ; b:
= x 2 x 3 + x 4 + ; c:
= x 1 + 1 + x + ; d:
= x + x 3 x 5 + ;
2
2
1 x
1+ x
x x
1+ x
99

e:
12.

13.

14.

a:

3 2x
4x + x 3

5 7x

3
4

x 1

1
2

+ . d b a e c.

= 5x 2 22x 1 + ; b:

2 + 4x
7 + 2x
= 1 + 5 x + ; c:
=
2
8
4 3x
4 3x

x 2 + 3x 3
3 5x
e:
= 1 + 4 x + . a b e c d.
2
6
6 + 2x
1 x
3 5x
1
a:
= 1 10 x + ; c:
=
= 1 2x + ; b:
3
1+ x
3 + 5x
4x + x 2
4x + x 2
e:
= 4x + 5x 2 + . d e b a c.
1 x

1
4

x 1

The smallest subfield of the field of complex numbers containing

1
16

2 is

7
4

+ ; d:

1
3x + x

9
2

+ ;

= 1 x 1 1 + ;
3

2  and hence has the induced

ordering from the field of real numbers. By Theorem 10, a subfield of  containing
15.

9 3x 2
=
2 + 6x

+ ; d:

1 + i 3
has an ordering
2

induced from the isomorphism.


a. true (discussion after Example 2)
b. true (id.)
c. false (?)
d. true
e. true (both Plow and Phigh )
f. false (even in , if a < 0 there is no such n)
g. true (if b 0 it's always true; if b > 0 it's a restatement of Definition 7)

( )

h. false ( 1 is positive)

16.
17.

i. false (neither 0, 0 are positive)


j. true (Theorem 3)
With the ordering Phigh , q  : q < .
m , n, m , n  : m + n 2 , m + n 2  2 :

m + n 2 + m + n 2 = m + m + n + n

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

= m n 2 + m n 2
= m + n 2 + m + n 2

) (
)
= (mm + 2nn ) (mn + m n) 2

m + n 2 m + n 2 = mm + 2nn + mn + m n

= m n 2 m n 2

= m + n 2 m + n 2
Theorem 5

18.

19.

a P

a 0 P

Lemma: a , c P ; b R;

0 < a.
Definition 1

b P : a b P
c P

ab = c : b P : a b P ab P c P (contradiction)
b = 0 :
a b = 0 P c P (contradiction)

b P .

100

c =0

ac = 0

b P b 0

bd = 0

d = 0.

Lemma

c P

ac P

bd P

( )

c P
a<b

b a P

21.

a < 0 0 a P
0 < b b 0 P
b b 1 = 1; b , 1 P

23.

a < 1 1 a P

( ) b ( d ) P
( a ) ( b ) P b < a .

ac P

bd P

a P
ab P
b P
Lemma
from Ex.19

22.

cd P .
Lemma

( )
( b ) a P

a c P

20.

d P

b 1 P ;

0 ab P

a b = ab 1

0<a
24.

a 0 = a P

1 < a

( )

0 a = a P

( )(a + 1) P

a
25.

( )

a + 1 P

a<0

a 1

a a a

ab < 0.

c d = c d P .

Definition 1

1
a 1 a P
1
a P

Lemma

d P

= 1 a 1 P

a 1 P

1< a

a 1 < 1 1 a < 1

First, show that P defines positive numbers as per Definition 1:

) ( )

(closure) a , b R : a , b R : a = a , b = b . Because P is positive, a + b , ab P so a + b , ab P .

( )

Because is a ring isomorphism, a + b = a + b = a + b P and ab = a b = a b P .


(trichotomy) a R : a R : a = a . Because P is positive:
a P
a = a P

a P a = a = a P
a =
a = a = 0
0

Then, show that the ordering induced by P is the same as <:


a , b R : a , b R : a = a , b = b :

( )

a<b
26.

b a P

b a = b a = b a P

a < b .

(closure) a , b S : a , b P :
ring

a +b S

Definition 1

a +b

ring

a b S

27.

b a P

a + b P S

Definition 1

a b

a b P S

(trichotomy) a S :
a P a P S

group

a R a P a S a P S
a = 0

Let P be such that p P if and only if 0 < p . Show that P is a well-defined set of positive numbers:
(closure) a , b P :
isotonicity

0 < a, b

transitivity

b <a +b

0<b <a +b

0<a +b

isotonicity

0 < a, b

a + b P

transitivity

a 0 < a b

a < ab

0 < a < ab

0 < ab

ab P
101

(trichotomy) a P :
0 a P
a < 0

a = 0 a = 0
a = 0 P
0 < a

Now a , b R :
28.

For all a, b:
a = b
a b = 0


a = b
a + b = 0

a b 0

a + b 0

a b

a b

so a 2 = b 2

29.

a<b

a P

a = 0
a P

0<b a

a <P b , so P implies the same relation.

)(

a2 = b2

)(

a2 b2

a + b a b = a2 b2 = 0
a + b a b = a2 b2 0

a = b . So a 2n + 1 = b 2n + 1

Ordering the following elements of R[x , y ] :

[ ][ ]

b a P

( ) ( )
n

a2 a = b2 b
x 1y 1

y 1

x 1
x 1y

1
y

a =b.

xy 1
x :
xy

Rx y
low

low

xy < x < xy 1 < y < 1 < y 1 < x 1y < x 1 < x 1y 1

low

high

xy 1 < x < xy < y 1 < 1 < y < x 1y 1 < x 1 < x 1y

high

low

x 1y < x 1 < x 1y 1 < y < 1 < y 1 < xy < x < xy 1

high

high

x 1y 1 < x 1 < x 1y < y 1 < 1 < y < xy 1 < x < xy

low

low

xy < y < x 1y < x < 1 < x 1 < xy 1 < y 1 < x 1y 1

low

high

x 1y < y < xy < x 1 < 1 < x < x 1y 1 < y 1 < xy 1

high

low

xy 1 < y 1 < x 1y 1 < x < 1 < x 1 < xy < y < x 1y

high

high

x 1y 1 < y 1 < xy 1 < x 1 < 1 < x < x 1y < y < xy

[ ][ ]

Ry x

6.1 Homomorphisms and Factor Rings

1.

The concepts of normal and ideal didn't accidentally result in factor groups and rings their requirements were
defined precisely so that the resulting groups and rings would be well-defined:
N a normal group: g G : g + N = N + g (Definition 3.1.19)
N an ideal ring: r R : r + N N , N + r N (Definition 6.1.10)
A ring endomorphism of  by Theorem 3 has to have 0 = 0 , and 1 = 1 iff R has unity 1 1 or else 1 = 0 . So

( ) ( )

0, 0 = 0, 0 and:

102

( ) ( )
(1, 0) = (1, 0)
(1, 0) = (0, 1)
(1, 0) = (0, 1)
(1, 0) = (1, 1)
(1, 0) = (0, 0)
(1, 0) = (0, 0)
(1, 0) = (0, 0)
(1, 0) = (0, 0)

( ) ( )
(0, 1) = (0, 0)
(0, 1) = (1, 0)
(0, 1) = (0, 0)
(0, 1) = (0, 0)
(0, 1) = (0, 1)
(0, 1) = (1, 0)
(0, 1) = (1, 1)
(0, 1) = (0, 0)

1, 0 = 1, 0

0, 1 = 0, 1

completely define the only possibilities.


2.

For all even n there is a n m n m =  2 , whereas for all odd n there is no element i such that i 2 = 0 and so

will never have a coset such that i + H

= H or a subring isomorphic to  2 .

3.

The ideals and their isomorphic subrings are:


121 12
12 121 12
6 2 12
12 6 2 6
43 12
12 43  4
3 4 12
12 3 4 3
26 12
12 26  2
112 12
12 112 1

4.

2 = , 4, 2, 0, 2, 4, ,

8 = , 16, 8, 0, 8, 16, ;

0 + 8

2 + 8

4 + 8

6 + 8

0 + 8 0 + 8

2 + 8

4 + 8

2 + 8 2 + 8

4 + 8

4 + 8 4 + 8
6 + 8 6 + 8

2 8 = 0 + 8, 2 + 8, 4 + 8, 6 + 8


0 + 8

2 + 8

4 + 8

6 + 8

6 + 8

0 + 8 0 + 8

0 + 8

0 + 8

0 + 8

6 + 8

0 + 8

2 + 8 0 + 8

4 + 8

0 + 8

4 + 8

6 + 8

0 + 8

2 + 8

4 + 8 0 + 8

0 + 8

0 + 8

0 + 8

0 + 8

2 + 8

4 + 8

6 + 8 0 + 8

4 + 8

0 + 8

4 + 8

5.
6.

2 8 /  4 because while  4 has a multiplicative identity, 2 8 does not.


Insert is a ring homomorphism.
Change additive subgroup to subring.

7.

Change to r R r = 0 .

8.

f , g F : f + g = f + g = f + g ; f g = f g + g f = fg + gf , so is a group but not a ring

isomorphism. The subring C of Example 12 is the kernel of . If would have been an homomorphism, then C
would have been an ideal in F.
9.

( )

( ) ( )

( )

( )( ) ( ) ( )

Let :    : n a n, 0 , then 1 = 1, 0 1, 1 = 1  but m, 0 : m , 0 1, 0 = m , 0 so 1, 0 is the

multiplicative identity of .
10. a. true (Theorem 17)
b. false (cf. last paragraph of the section)
c. true (Corollary 6)

d. false ( q  : q 2  )
e. true (Definition 10)
f. false (Example 12)
g. true (because multiplication is defined by means of multiplication of representatives, which is commutative)
103

11.
12.
13.
14.

h. true (Example 8)
i. true (Obviously, if N = R then 1 N . Conversely, if 1 N then r R : r = r 1 N
j. true
No. (See discussion after Example 2.4)
See Example 2.1.
See Example 2.4.
See Example 2.2.

(1, 1)

15.
16.

( )( ) ( ) ( )

  , but from 1, 0 1, 1 = 1, 0 1, 1

R=N)

is not ideal.

a. Because the expression rs = sr is a statement about the ring R and not about the quotient ring.

)(

) (

)(

b. Then r + N s + N = s + N r + N for all r , s R


c. Suppose R N is commutative. r , s R :

(r + N )(s + N ) = (s + N )(r + N )

so nr , ns , nr , ns N :

)(

) (

)(

)(

) (

)(

r + N s + N s + N r + N = 0R N = N ,

n N :

) (
) (
rs sr = n (rn + n s + n n ) + ( sn + n r + n n ) N

r + nr s + ns s + ns r + nr = rs + rns + nr s + nr ns sr + snr + ns r + ns nr = n
s

s r

because N is ideal. Conversely, suppose that r , s R :

(r + n )(s + n ) (s + n )(r + n ) = = (rs sr ) + (rn


r

so
r + N s + N s + N r + N = N = 0R N

)(

) (

)(

and R N is commutative.
17.

First, show that R =

a , a 

rs sr N . Then nr , ns , nr , ns N :

) (

+ nr s + nr ns snr + ns r + ns nr N

)(

) (

)(

r +N s +N = s +N r +N

a + a 2 is well-defined as a ring. Additive closure, associativity, identity, and

inverse follow fairly obviously and directly from those properties in , so R is a group. Multiplicative closure and
additive commutativity are similarly obvious. Additive associativity follows from a , a , b , b , c , c :
a + a 2 b + b 2 c + c 2 = ab + 2a b + ab + a b 2 c + c 2

) (

) (

= abc + 2a b c + 2ab c + 2a bc + ab c + a bc + abc + 2a b c

) (

= a + a 2 bc + 2b c + bc + b c

= a + a 2 b + b 2 c + c 2

Left distributivity follows from a , a , b , b , c , c :


a + a 2 b + b 2 + c + c 2 = a + a 2 b + c + b + c 2

) (

= a b + c + b + c

) (

) (

2 + a 2 b + c + b + c

= a + a 2 b + b 2 + a + a 2 c + c 2

a 2a
Right distributivity follows similarly. Therefore R is a ring. Now, showing that R = a , a 
is a ring.

a a
Again, additive associativity, additive identity, and the additive inverse follow fairly directly from their corresponding
properties in M 2  and , so R is a group. Additive closure follows from a , a , b , b  :

(
(

) (
) (

2 a + b
a 2a b 2b a + b

+
=
a +b
a a b b a + b
and multiplicative closure a , a , b , b  :

) R

104

(
(

) (
) (

)
)

a 2a b 2b ab + 2a b 2ab + 2a b ab + 2a b 2 a b + ab
R .

=
=
2a b + ab a b + ab
ab + 2a b
a a b b a b + ab

Additive commutativity is again similarly obvious, and although multiplicative associativity follows directly from
M 2  and , it is derived in analogy to the additive property: a , a , b , b , c , c :
a 2a b 2b c 2c a 2a bc + 2b c 2bc + 2b c

2b c + bc
a a b b c c a a b c + bc
a bc + 2b c + 2a b c + bc a 2bc + 2b c + 2a 2b c + bc

=
a bc + 2b c + a b c + bc
a 2bc + 2b c + a 2b c + bc

abc + 2ab c + 2a b c + 2a bc 2abc + 2ab c + 4a b c + 2a bc


=

2a bc + 2a b c + 2ab c + abc
a bc + 2a b c + ab c + abc

(
) (

) (
) (

) (
) (

) (

)
)

(
) (

)
)

(
(

ab + 2a b c + 2a b + 2ab c 2a b + ab 2c + 2ab + 2a b c

=
a b + ab c + ab + 2a b c
ab + a b 2c + 2a b + ab c

ab + 2a b 2ab + 2a b c 2c
=

2a b + ab c c
a b + ab
a 2a b 2b c 2c
=

a a b b c c
a 2a
Distributivity follows directly from M 2 , so R is a ring. Let : R R : a + a 2 a
. Then:
a a
(additive homomorphy) a , a , b , b  :
a + a 2 + b + b 2 = a + b + a + b 2

( ) ( )
(a + b ) 2(a + b ) a
=
=
(a + b ) (a + b ) a

2a b 2b
+

a b b

= a + a 2 + b + b 2
(multiplicative homomorphy) a , a , b , b  :
a + a 2 b + b 2 = ab + 2a b + ab + a b

) (
) 2
(ab + 2a b ) 2(ab + a b ) a
=
=
(ab + a b ) (ab + 2a b ) a

2a b 2b

a b b

= a + a 2 b + b 2
(isomorphy) a , a : a + a 2 = 0R
18.

19.

a, a = 0

Ker = 0R

so is a ring isomorphism, and R R .


Following Theorem 2.5, if N R is ideal and contains any nonzero element of R, it contains a unit and therefore
unity, and then N = R . So a field contains no proper nontrivial ideals, and by the Fundamental Homomorphism
Theorem any field homomorphism is either trivial or identity.
Exercise 3.1.49 already shows that is a group homomorphism. g , h R :

( ) ( ( ))

gh = gh
20.

a 2a 0 0

a a 0 0

homomorphism

g h

homomorphism

g h .

a , b R :

105

) (

a +b = a +b

p
p!
= +0i p a p i b i = +0i p
a p i b i
i
p i ! i!

( )
( )

p 1 ! p i i
p! p 0
p! 0 p
a b + +0 <i <p p
a b +
a b

0! p!
p! 0!
p i ! i!

= a p + +0 <i <p 0 + b p = a p + b p
The middle terms vanish because p is the characteristic of the ring.

( ) ( )

ab = ab

commutative

a p b p = a b .

Why does it matter that p is prime?


21.
22.

Suppose that 1 1. Then r R* : 1 1 r = 1 r 1 r = r r = 0 , where r , 1 1 0 so that R has


divisors of zero. Consequently, if R has no divisors of zero, then 1 = 1. (Due to Doug Rosenberg)
a. a R, n N : a R, n N : a = a , n = n :
an N

( )

an N

a n N

a n N

Similarly, N a N , so N < R .

( )

b. 2 <  . Let :    : n a n, n . Then 2 =

a N N .

{ (n, n)} , but (1, 0) (1, 1) = (1, 0) 2 so 2 </   .


n 

c. If N < R , then also N R < R , so we only need to consider the case of N < R . Consider the isomorphism

: R N R from the Fundamental Homomorphism Theorem. By isomorphism invN < inv R R M , so any
element r + M R M multiplied by invN is again in invN . Then obviously any element r R multiplied by

inv invN = invN is again in invN , so invN < R .

[ x ] : s S :

( )

s fg = s f s g = 0 s g = 0

[ x ].

23.

f N S , g F

24.

By Exercise 18, any homomorphism from a field is either an isomorphism or trivial. Since every ideal subring gives
rise to a homomorphism, the only ideals of a field are the field itself or the trivial field, so the only factor rings of a
field are trivial or the field itself.
If N R then R N E has more than one element. Since R has multiplicative identity, 1 multiplies any such
element to itself, and since in the factor ring multiplication happens by representatives, 1 must be a representative of
a multiplicative identity in the factor ring.

25.

NS < F

commutative

26.

a R : I a =

27.

Any element multiplied by either ideal is again that same ideal, so the subset must multiply to itself:
r R, n N N : n N n N rn N rn N rn N N .
Lemma. A ring homomorphism/isomorphism induces a ring homomorphism/isomorphism on any of its quotient
rings. Let R, R be rings, N < R , and : R R. By Exercise 22a, N < R . Let , be the canonical

28.

x R

x ax = 0 . So x I a , r R : a rx

fg N S

r ax = r 0 = 0

rx I a , and I a < R .

homomorphisms : R R N , : R N . Then * = inv : R N R N is a homomorphism. Furthermore,

29.

if is an isomorphism then Ker * = Ker inv = inv Ker = inv N = N = N = ER N and * is an


isomorphism.
Back to the Exercise. : R R induces a homomorphism * : R N R N . Because N < R , under the
N
R
<
so there exists a canonical homomorphism
canonical homomorphism : R R N , N < R
N N
R
R N R
* :

by the Third Isomorphism for rings (proved in Exercise 38). So


N
N N N
* = * o * : R N R N is a ring homomorphism.
Suppose there is a unit of R in the kernel of , then 0 would have a multiplicative inverse in R , but then the
multiplicative identity in R would be 0 , which is counter to the definition of unity of Definition 1.16 and u
cannot therefore have a multiplicative inverse in R . Conversely, suppose no unit of R is in the kernel of . Since
106

( )

( )

u, u 1 , 1 are units of R, u, u 1 , 1 0 and 1 = uu 1 = u u 1, so u is a unit in R .


30.

Let A be the set of all nilpotent elements of R. First, A is a subring because 0 is obviously nilpotent and
a , b A : n, m + : a n = 0, b m = 0:
n + m (n + m ) i i
= +0i n +m
b . Since the sum of the powers of a and b in
a
i
each of the terms is always n + m , either the power of a is at least n or that of b is at least m, so that the terms all
vanish and a + b is nilpotent.

(additive closure) Consider a + b

( )

(multiplicative closure) ab

n +m

n +m

commutative

a n + mb n + m = a n a mb n b m = 0a m + b n 0 = 0 , so ab is nilpotent.

( )

Then a A, r R : n + : a n = 0, so ar
31.

commutative

a n r n = 0r n = 0 and ar is nilpotent. So A is ideal.

The elements in the nilradical of n are those that contain all the prime factors of n:

{ }

12 : 12 = 2 23 has 0, 6

{
has {0}

32 : 32 = 2 has 0, 2, 4, , 30
 
32.

Obviously, 0 + N is nilpotent in R N . Since multiplication in the factor ring occurs by representatives in N, and no
elements in R \ N are nilpotent, it is also the only nilpotent element of R N .

33.

Let r R . Since the nilradical of R N is itself, there is an rN such that r rN + N and rN nilpotent, and there is
an n N such that r = rN + n and n nilpotent. By the proof of additive closure of nilpotents in a commutative ring
in Exercise 30, r is also nilpotent. Therefore R is its own nilradical.

34.

First, show that the radical in fact forms a subring. a , b N : n, m + : a n , b m N :


(identity) Since N is an ideal and a subring, 0 N and because 0 = 0 , 0 N .
n + m (n + m ) i i
= +0i n +m
b . Since the sum of the powers of a and b in
a
i
each of the terms is always n + m , either the power of a is at least n or that of b is at least m, so each of the terms is

(additive closure) Consider a + b

of the form a j a n b k b m = a j na b k nb

n +m

commutative

a j b k na nb = a j b k nab , where na , nb , nab N . Because nab is an element

of the ideal, a j b k nab N so each of the terms is as well. Because the ideal is a subring and closed under addition,
the entire sum is in the ideal.

( )

(multiplicative closure) ab

n +m

= a n + mb n + m = a n a m + b n b m = na a m + b n nb , where na , nb N . Similarly,

na a m , b mnb and the sum is in the ideal.


So
35.

a.  < ;
b. 2 < ,

36.

( )

N is a subring. n N : i + : ni = nn N . Then r R: rn

commutative

r i ni = r i nn N , so

N <N.

i  , i  .
2 = 2 .

The radical of N is the set of all the elements that by some power end up in N. The nilradical of R N is the cosets
of N that by some power equal the coset 0 + N . So
element of the nilradical of R N .

N is precisely the elements of R that are representatives of an

107

37.

M +N
M +N
N

M
M
M N

M N

First, show that M + N is a ring. m + n, m + n M + N :


(identity) 0 = 0 + 0 M + N ;

) (

) (

) (

(additive closure) m + n + m + n = m + m + n + n M + N ;

)(

(multiplicative closure) m + n m + n = mm + mn + nm + nn = m m + m n + nm + nn = m mn + nmn M + N ,


where m m , m n , m mn M , nm , nn , nmn N .

Then show that M + N < R: m + n M + N , r R : r m + n = rm + rn = m r + nr M + N , where


m r M , nr N .
Now, follow the proof of Theorem 4.1.5. Let : R R N be the canonical homomorphism. Under , then,
M R
Ker

M R N . First, consider the restriction

M +N

: M M which is a homomorphism with

M
M .
N M
: M + N M + N . Now n N : n = N = 1R N , so M + N = M and

= N M . By the Fundamental Homomorphism Theorem there exists an isomorphism 1 :

Second, consider the restriction

: M + N M with Ker

M +N

M +N

= N and there similarly exists an isomorphism 2 :

M +N
M . Therefore,
N

M
M +N

.
N M
N
38.

Follow the proof of Theorem 4.1.7. Let : R

RM
: r a r + M + N M . First, show that is a ring
N M

homomorphism. a , b R :

) ((

((

) (

(multiplication) ab = ab + M +

((

)(

))

(*)
(*)
N
N
N
N
= a +M + b +M +
= a + M + + b + M + = a + b , where
M
M
M
M
(*) holds because coset addition in a ring is well-defined.

(addition) a + b = a + b + M +

))

N
N
N
N
= a +M b +M +
= a + M + b + M + = a b , where
M
M
M
M
(*) holds because coset multiplication in a ring is well-defined.
N
RM
The identity element in
and Ker = N , so by the Fundamental Homomorphism Theorem
is 0 + M +
M
N M
R RM

.
N N M

( ) (

39.

a a'
Show that :  M 2  : a + a i a
is an isomorphism. a + a i , b + b i :
a a
(addition)
108

a +b
a + b a a b b
a + a i + b + b i = a + b + a + b i =
=
+
= a + a i + b + b i
a b a + b a a b b
(multiplication)
a + a i b + b i = ab a b + ab + a b i

((

) (

)) ((

((

)(

)) ((

))

) (

ab a b ab + a b a a b b
=
=

= a + a i b + b i
ab a b ab a b a a b b

))

) (

40.

) (

) (

ring

a. x , y R, + : a x + y = a x + y = ax + ay = a x + b x .
b. Show that it is a ring: a , b R :
(identity) 1 : R R : x a 1 x = x is the identity of End R, + ;
(additive closure): x R :

(multiplicative closure) x R :

+ b x = a x + b x = a x + b x = a + b x = a + b x , so a + b = a + b R ;

( )

( )

( ) ( )

b x = a b x = a b x = a b x = a b x = a b x so a b = a b R .

c. Let : R R : a a a . a , b R :

(b.)

(addition) a + b = a + b = a + b = a + b ;
(b.)

( )

(multiplication) a b = a b = a b = a b ;

{}

(isomorphy) The identity of R is 1 , so Ker = 1 .


So R R . a is a permutation, and every ring R is thus isomorphic to a ring of permutations.

6.2 Prime and Maximal Ideals


9

As a proof of concept, restate Theorem 9 in a format that shows the hierarchial top-down structure of the proof.
This is truer to the 1 1 2 -dimensional nature of a proof than the flattened linear text, and more consistent than the
alternating bidirectional imply/infer logic of the text stream. Because it obviates mentally reconstructing the true
structure of the proof and permits the reader to selectively ignore details of the proof, it should theoretically be
easier to understand.
M is a maximal ideal of R R M is a field
( R M is a field M is a maximal ideal) (
R M is a commutative ring with unity (
R is a commutative ring with unity
)
R M has multiplicative inverses (
Let a R : N a =

r R, m M

ra + m .

N a is a group under addition (


N a is closed (

) (

) (

) (

ra + m , r a + m N a : ra + m + r a + m = r + r a + m + m N a
)
N a has identity (

) (

) (

) (

ra + m : 0a + 0 + ra + m = 0 + r a + 0 + m = ra + m
)
N a has inverses (
ra + m N a :

((r )a + (m )) + (ra + m ) = (r + r )a + (m + m ) = 0a + 0 = 0

Na

)
),
109

Na < R (
ra + m N a :

) ( )

g R : g ra + m = gr a + gm N a (
gr R ,
gm R (
M <R
)
)

),
Na = R (
Na M (
Na M (
m M : m = 0a + m N a
)
Na M (
a N a (
a = 1a + 0 N a
)
a M (
a + M 0R M
)
)
)
M maximal
)
1 N a
ba + m N a : ba + m = 1

)(

ba + M = b + M a + M = 1 + M
)
)
(M is a maximal ideal R M is a field) (
Suppose M is not maximal: N < R : R N M
R M is not a field (
R M contains a proper nontrivial ideal (
Let : R R M be the canonical homomorphism:
N <R
N < R = R M

R M N 0 + M

)
)
)
We can compact the presentation with a few simple heuristics. Roughly, let or be implied between two
lines at the same indentation level, and at increasing indentation:
M is a maximal ideal of R R M is a field
( R M is a field M is a maximal ideal)
R M is a commutative ring with unity
R is a commutative ring with unity
R M has multiplicative inverses
Let a R : N a =

r R, m M

ra + m .

N a is a group under addition


110

) (
) ( ) (
)
identity: ra + m : (0a + 0) + (ra + m ) = (0 + r )a + (0 + m ) = ra + m
inverses: ra + m N : (( r )a + ( m )) + (ra + m ) = ( r + r )a + ( m + m ) = 0a + 0 = 0
closed: ra + m , r a + m N a : ra + m + r a + m = r + r a + m + m N a

Na < R

Na

) ( )

ra + m N a : g R : g ra + m = gr a + gm N a
gr R , gm R M < R
Na = R
Na M
Na M
m M : m = 0a + m N a
Na M
a N a a = 1a + 0 N a
a M a + M 0R M
M maximal
ba + m N a :

1 N a

)(

ba + m = 1

ba + M = b + M a + M = 1 + M
(M is a maximal ideal R M is a field)
Suppose M is not maximal: N < R : R N M
R M is not a field
R M contains a proper nontrivial ideal
Let : R R M be the canonical homomorphism:
N <R
11

N < R = R M

R M N 0 + M

R is a field R has no proper nontrivial ideals


R is a field R has no proper nontrivial ideals
Corollary 6
R is a field R has no proper nontrivial ideals
Theorem 9

R R E is a field
15

N is prime ab N

E < R maximal R has no proper nontrivial ideals

a N b N , so a prime ideal is such that the corresponding factor ring has no divisors

of 0. In other words, N < R prime iff R N is an integral domain.


16 Maximal and prime in factor rings correspond to field and integral domain.
18 Let :  R be the homomorphism from Theorem 17.

(R contains a subring isomorphic to 

char R > 1

Ker = n
Ker < 
N <  s  : N = s
n is the smallest integer such that n 1 = 0 Theorem 5.2.15
  n n

(R contains a subring isomorphic to 

char R = 0

Ker = E

m * : m 1 0
  E 
24

char R = 0

[]

If F is a field, then every ideal in F x is principal

(N principal

[ ])

N <F x

111

{} [ ]

N principal N = E = 0 F x

N principal N E
Let g N be of minimal degree.

(deg g = 0

[ ])

N = 1F x

Theorem 5

deg g = 0

(deg g > 0

25

[]

g F

[ ])

[]

[]

N = F x = 1F x

N = gF x

f N
f = gq + r where r = 0 deg r < deg g Theorem 5.6.1
r =0
r N
gq N
f N
g N gq N gq + r N
g N is of minimal degree deg g > 0
f = gq

[]

pF x maximal p irreducible over F x

[]

f , g F x : p = fg

[]

deg f deg p deg g deg p

[]

f pF x g pF x

[]

[]

pF x maximal

pF x prime

[]

p pF x

[]

[]
g N : N = gF [ x ]

[]

Let N < F x : F x N pF x

Theorem 24

N principal

[]

q F x : p = gq
p N

[]

p F x N

deg g = 0 deg q = 0

(deg g = 0

[]

N <F x

p irreducible

[ ])
N = gF [ x ] = 1 F [ x ] g F is a unit of F [ x ]
(deg q = 0 N = pF [x ])
g pF [ x ] gF [ x ] = pF [ x ]
N =F x

g = p q , q F

27

rs = p 0

contradiction.
r =p 0 s =p 0

[]

rs pF x

[]

rs = r 0

pF x is prime
1.

1
3
4
5
6

[]

pF x is maximal

By Example 2.7.
Not a division ring because it doesn't have a multiplicative inverse.
By Theorem 2.11.
Not an integral domain because it has a divisor of zero
By Theorem 2.9.
112


and isomorphic subrings
subring
factor ring
16
6 16 1

2.

3.

4.

5.

23

6 23  2

int dom?
no3
yes1

3 2

6 3 2 3

yes1

yes4

yes

yes

61
subring
112

6 61 6
factor ring
12 112 1

no
normal?
no

12 26  2

no6
field?
no6
yes4

no
prime?
no

26

no1
int dom?
no3
yes1

yes

yes

3 4

12 3 4 3

yes1

yes4

yes

yes

43

12 43  4

no

12 6 2 6

no6
no6

no

6 2

no1
no1

no

no

no1
int dom?

no6
field?

no
prime?

no
normal?

no3

no3

no

no

yes1

yes4

yes

yes

no5

no6

no

no

field?
no6
yes4

prime?
no

normal?
no

yes

yes

121
subring

12 121 12


factor ring
2 2
1 2 1 2
1
1 2 1 2
2 2
1  2 2  1  2
2
1 2 21
2 2
21 21 1
2 2
21 21
subring
factor ring
1 1 1
1 2 1 4

field?
no6
yes4

prime?
no

normal?
no

yes

yes

1 2 2 2

1  2  2

int dom?
no3
yes1

1 2 41

1  4  4

yes1

yes4

yes

yes

21 1 4

 2 1  2

yes1

yes4

yes

yes

21 2 2

2 2

no

2 4

no6
no6

no

21 41

no5
no5

no

no

[]

3 x

[]

subring

{0, 2, 4}
{0, 3}
subring

{0, 2, 4, , 10}
{0, 3, 6, 9}

subring

{0, 1} {0}
subring

{0, 1} {0, 2}
{0, 1} {0}
{0} {0, 1, 2, 3}

[]

x + c is a field iff x + c < 3 x is maximal iff x + c is irreducible in 3 x . If x 2 + c is reducible,

then it has at least one (i.c., actually two) factors of degree one x a and by the Factor Theorem then has a zero
for x = a . By calculation, the sets Ac of zeroes a for given c are: A0 = 0 , A1 = , A2 = 1, 2 . So the
polynomial is irreducible and the factor ring a field for c = 1.

{}

{ }

6.

Following the procedure of Exercise 5 if x 3 + x 2 + c is reducible, it has to have at least one factor of degree one
and a corresponding zero: A0 = 0, 2 , A1 = 1 , A2 = . So the factor ring is a field for c = 2 .

7.

A0 = 2 , A1 = 1 , A2 = ; c = 2 .

8.

A0

9.

A0

10.
11.
12.

is a proper ideal
The given definition is valid only if R =  because prime elements have not been defined elsewhere.
Comparing to Definition 20, p and  can contain no nontrivial proper subfields, and any other field properly
contains either of these fields so the definition is indeed equivalent.
Since a principal ideal consists of all products of the field with the geneator, it is certainly the smallest ideal
containing the generator. Since this defines minimal ideals for every element, all minimal ideals are principal

13.

{ }

{}

{}
{}
= {0, 4} , A = , A = , A = {1, 3} , A = {2} ; c = 1, 2 .
= {2, 3} , A = , A = {4} , A = {1} , A = ; c = 1, 4 .
1

113

14.

therefore the definition is equivalent.


a. false (should find a counterexample of a factor ring that is an integral domain but not a field)
b. true (Corollary 16)
c. true (by Theorem 19 because the characteristic of  is zero)
d. false (by Theorem 19 the characteristic of  is zero, so  is the prime subfield)
e. true (Theorem 19)
f. true (    )
g. true (Theorem 19)

[]

[]

h. true (if F has no divisors of zero then F x and F x N don't either, so N is prime)

15.

i. true (Theorem 24)


j. false (by Theorem 25, only if the generating polynomial is maximal)
2  <   .   2   2 E  2 is a field, so 2  is maximal.

17.

 E <   .    E E  is not a field, so  E  is not maximal. Since  has no divisors of zero,


 E is prime.
4  <   .   4   4 E  4 has divisors of zero, so  4 E is not prime.

18.

x

16.

[]

[]

x 2 5x + 6 is a field iff x 2 5x + 6 is maximal iff x 2 5x + 6 is irreducible in  x . By the Factor

Theorem it is irreducible iff it has no zeroes in . The roots are

21.
22.
23.
24.
25.

4 1 6

( ) (6)

4 1 6

6 12
= 3 3 
2 1
2
and the factor ring is therefore a field.
Since R is finite, so is N R and R N . Since R is prime, R N is an integral domain. By Theorem 5.2.11, R N
is a field, therefore N is maximal.
n m is a ring with multiplicative identity containing n E n and E m m as subrings.
Idem.
If a ring contains subrings isomorphic to p , q , then it should contain a subring isomorphic to pq , which is not
an integral domain. So any containing ring cannot be an integral domain either.
x=

20.

5 1
=
= 2, 3 ,
2 1
2
so the factor ring is not a field.
Following the procedure of Exercise 18:
x=

19.

( ) (5)

N < R maximal

R N simple

(suppose N not maximal

R N not simple

M : R M N , M < R
Let : R R N be the canonical homomorphism

:M R N
M <R

M < R N and M R N

(suppose R N not simple

N not maximal

Let be some canonical homomorphism.


R N not simple M < R N M 0 + N

26.
27.
28.
29.
30.

A+B =

a A, b B

inv M < R and R inv M N .

a +b .
114

a. Show that A + B R is a subring:


(additive identity) 0 A, B 0 = 0 + 0 A + B .

( ) ( )

(additive inverse) a + b A + B : a A, b B

a + b A + B :

(a + b ) + ((a ) + (b )) = (a + (a )) + (b + (b )) = 0 + 0 = 0
(additive closure) a , a A; b , b B : (a + b ) + (a + b ) = (a + a ) + (b + b ) A + B , where a + a A, b + b B
(multiplicative closure) a , a A; b , b B :

(a + b ) (a + b ) = aa + ab + ba + bb

A, B <R

= a + a + b + b = a + b A + B where a = aa A, a = ab A

and similarly in B.

A, B <R

Now show that A + B < R is an ideal: a + b A + B : r R: r a + b = ra + rb

= a + b , where a A, b B

.
b. Because A, B are ideals they are subrings and contain the additive identity. Then a A:
a = a + 0 A + B A A + B and similarly B A + B .
31.

AB =

+ ni , a

n  +

A, b i B

ai bi .

a. Show that AB R is a subring.


(additive identity) 0 = + 0i ai bi AB .
(additive inverse) n + , ai A, bi B :

+ ni ai bi

)+(

+ ni

) (a b + (a b )) =

ai bi =

+ in

i i

i i

+ in ai bi AB : ai A
+ in 0

( )

+ ni ai bi = + in ai bi AB :

=0

(additive closure) The sum of both sums of terms is just a larger single sum of terms.
(multiplicative closure) n, n + ; ai , ai A; bi , bi B :

(+ a b ) (+
n
i i i

n
i aibi

)=+

n
i

+ nj ai bi a j b j

A, B <R

= + ni + nj ai b j = + kn n akbk AB , where a = ai bi and b j = a j b j .

Then show that AB < R is ideal. n + ; ai A; bi B : r R :


ai = r ai .
A <R

A <R

r + ni ai bi = + in r ai bi = + ni aibi AB , where

B <R

b. For any + ni ai bi AB , + ni ai bi = + ni ai A and + ni ai bi = + ni bi A where ai, bi= ai bi . So + ni ai bi A B and


AB A B .
32.

A:B =

r R

r b B : rb A .

a. First, show that A : B R is a subring.


(additive identity) 0 R : b B : 0 b = 0 A

0 A : B .

(r ) b = (rb ) A r A : B .
(additive closure) r , r A : B : b B : (r + r )b = rb + r b = a + a A , where a = rb A , a = r b A .
(additive inverse) r A : B , b B :

So

r + r A : B .

(multiplicative closure) r , r A : B : b B :
r r A : B .

( )

Show that A : B < R is ideal. r A : B : s R: b B :


33.

a = rb A and a = a s .
Show that S M 2 F is a subring:

( )

A <R

rr b = r r b = ra = a , where a = r b A and a = ra . So

(rs )b

commutative

(rb )s = a s = a A

rs A : B , where

115

0 0
(additive identity)
S .
0 0
a b
a b
a b a
(additive inverse)
S : a , b F :
S :
+
0
0 0
0
0 0 0
(additive closure) follows directly from the closure of F and M 2 F .
a b a b
(multiplicative closure)
,
S :
0 0 0 0

b 0 0
=
.
0 0 0

a b a b aa + b 0 ab + b 0 aa bb

=
=
S .
0
0 0 0 0 0a + 00 0b + 00 0

a b
f 00 f 01
Now
S;
M2 F :
0 0
f 10 f 11
a b f 00 f 01 af 00 + bf 10 af 01 + bf 11 af 00 + bf 10 af 01 + bf 11

=
=
S
0
0

0 0 f 10 f 11 0 f 00 + 0 f 10 0 f 01 + 0 f 11
but
f 00 f 01 a b f 00a + f 01 0 f 00b + f 01 0 af 00 bf 00

=
=
,
f 10 f 11 0 0 f 10a + f 11 0 f 10b + f 11 0 af 10 bf 10
which is not necessarily in S.
34.

Enumerate all the possible elements that could be contained in an ideal of M 2  2 .

0 0

= E M 2  2 is the trivial
0 0

ideal. Consider a matrix with one non-zero element:


1 0 n00 n01 1n00 + 0n10 1n01 + 0n11 n00 n01

=
=

0
0 0 n10 n11 0n00 + 0n10 0n01 + 0n11 0
which we know from Exercise 33 is not an ideal. By symmetry we know that neither are any of the other principals
generated by matrices with one non-zero component or with two non-zero components along a row or column. As
1 0
0 1
to the other two matrices with two non-zero components, obviously
=
= M 2  2 , and this
0 1
1 0
implies that neither are the ideals generated by matrices with three non-zero elements proper.

6.3 Grbner Bases for Ideals

The discussion after Example 2 states in essence that


generators, and that they are

r
i

n
i

fi =

ri R

+ i ri f i

} are the principal ideals with multiple

f i = ri f i the intersection of the individually-generated ideals.

{ f } are the common zeros of

The common zeros of

This is just a generalization of Theorem 2.24 to multiple indeterminates: every ideal of F x is principal

Let f 1 : x + y 3z 8 = 0 and f 2 : 2x + y + z + 5 = 0 , then f 3 : y + 7z + 21 = 0 can be formed from

fi .

[]

fi .

116

f = gq + r

f3 = f 2 2f1

f3 = f 2 2f1

f 2 = 2f1 + f3
f3 f 2
=
2
f1
f1

remainder product quotient


divisor
11

Keep in mind that the algebraic variety of an ideal is equal to that of any basis. In the left figure are plotted the
zeros of the two original polynomials of the Example. Disregarding some plotting artifacts, it can be seen that they
intersect in one point. The right figure shows the zeros of the Grbner basis calculated in Example 13, and it can
be seen that they intersect in the same point.
3

-1

-1

-2

-2

-3

-3
-3

-2

-1

-3

-2

-1

1.

Write out the exponents of the power products, and sort them lexicographically like words:
135, 213, 221, 300 300, 221, 213,135 : 3x 3 + 7 x 2 y 2z 5x 2 yz 3 + 2xy 3z 5 .

2.

025,100, 033, 007

3.

010,100, 003,122, 212

4.
5.

000,101, 011,110, 013 110,101, 013, 011, 000 : 8xy 4xz + 3yz 3 + 2yz + 38 .
Write out the exponents in reverse order:
531, 312,122, 003 531, 312,122, 003 : 2z 5y 3 x 5z 3 yx 2 + 7zy 2 x 2 3x 3 .

6.

520, 001, 330, 700

7.

010, 001, 300, 221, 212

300, 221, 212, 010, 001 : 10z 3 2z 2 y 2 x + 2z 2 yx 2 + 3y 7 x .

8.
9.

000,101,110, 011, 310

310,110,101, 011, 000 :

100, 033, 025, 007 :

4x + 5y 3z 3 + 3y 2z 5 8z 7 .

212,122,100, 010, 003 :

700, 520, 330, 001 :

2x 2 yz 2 2xy 2z 2 7 x + 3y + 10z 3 .

8z 7 + 3z 5y 2 + 5z 3 y 3 4x .
3z 3 y + 2zy 4zx 8yx + 38 .

1< z < y < x


< z 2 < yz < y 2 < xz < xy < x 2

10.
11.

< z 3 < yz 2 < y 2z < y 3 < xz 2 < xyz < xy 2 < x 2z < x 2 y < x 3
<
Write the sum of the exponents as an exponent and sort by degree first:
1359 , 2136 , 2215 , 3003 1359 , 2136 , 2215 , 3003 : 2xy 3z 5 5x 2 yz 3 + 7 x 2 y 2z 3x 3 .
0257 ,1001 , 0336 , 007 7

0257 , 007 7 , 0336 ,1001 :

3y 2z 5 8z 7 + 5y 3z 3 4x .
117

12.

0101 ,1001 , 0033 ,1225 , 2125

2125 ,1225 , 0033 ,1001 , 0101 :

13.
14.

0000 ,1012 , 0112 ,110 2 , 013 4

013 4 ,110 2 ,1012 , 0112 , 0000 :

2x 2 yz 2 2xy 2z 2 + 10z 3 7 x + 3y .
3yz 3 8xy 4xz + 2yz + 38 .

maximum-order term
xy 2 2x

x 2 y + 4xy

xy y 2

x 2y y 2x

4xy + y 2 x
leaving xy 2 2x , 4xy + y 2 x , xy y 2 .
15.

xy + y 3

y3 +z

x y4

xy y 5

y5 + y3
leaving y 5 + y 3 , y 3 + z , x y 4 .

( )

16.

Can't be reduced as required, because x 3 can't be divided by any of the 1p f i .

17.

y 2z 3 + 3

y 3z 2 2z

y 2z 2 + 3

y 3z 2 + 3 y

3y 2z
leaving y 2z 3 + 3, 3y 2z , y 2z 2 + 3 .
18.

w + x y + 4z 3 2w + x + y 2z + 4 w + 3x 3y + z 5
2w + 2x 2y + 8z 6
2
x + 3y = 10z + 10
1

w + x y + 4z 3
2

2x 2y 3z 2
2x 6y + 20z 20
4y 23z + 18

x 3y + 10z 10

w + 2y 6z + 7
2y 23 z + 9
w+

2
11
z
2

1
2

2
3y 69 z +
x +

2w + 11z 4

4
29
z
4

27
2
7
2

4x + 29z 14

leaving 2w + 11z 4, 4x + 29z 14, 4y 23z + 18 . Every Sij has a leading term containing at least a nonzero
power of w, x, or z and can thus be divided by the leading term of one of the basis polynomials. We have thus
found a Grbner basis.

118

19.

w 4x + 3y z + 2 2w 2x + y 2z + 5 w 10x + 8y z 1
2
2w 8x + 6y 2z + 4
6x 5y + 1
1

w 4x + 3y z + 2
6x + 5y 3

leaving w 4x + 3y z + 2, 6x 5y + 1, 6x + 5y 3 . Since the second and third polynomials have no common


zeros, the Grbner basis is 1 .
20.

x 4 + x 3 3x 2 4x 4 x 3 + x 2 4x 4
x
x 4 + x 3 4x 2 4x
x2 4
x

x 3 4x
x2 4

leaving x 2 4 , which is a single-element basis and thus a Grbner basis.


21.

x 4 4x 3 + 5x 2 2x
x 4 x 3 4x 2 + 4x

x 3 x 2 4x + 4
x

x 3 3x + 2

3x 2 + 9x 2 6x
3x 3 + 3x 2 + 12x 12

6x 18x + 12
x 2 3x + 2
x 3 3x + 2

x x + 2
x

x 3 + x 2 2x
x 2 x + 2

x2 + x 2
4x + 4

x 1
x

x 2 + x
2x + 2
x 1

leaving just x 1 .

119

22.

x 5 + x 2 + 2x 5
x5 x 4 + x3 x 2

x3 x 2 + x 1
x2

x 4 x 3 + 2x 2 + 2x 5
x 4 x3 + x 2 x

x + 3x 5
x 3 + 3x 2 5x

4x 2 + 6x 1
4

4x 2 12x + 20
18x 21
6x 7

x2 7 x

1x

25
x 5
6
36

25

6x 36
5

1
5

leaving 1 .
x 2y x 2

23.

xy + 2y 9

x y + 2xy 9x

2xy + 8x 2
xy 4x + 1
1

xy 4x + 1
4x + 2y 10
2x + y 5

xy + 1 y 2 5 y
2

4x 1 y + 5 y + 1
2

4x 2y + 10

1 2
y
2
2

+ y 9
9
2

y 9y + 18
y has zeros y =

+9 9 2 4 1 18 9 81 72 9 9 9 3
=
=
=
= 3, 6 and from
2 1
2
2
2

2x + y 5 = 0

2x = y + 5 x =

1
2

( y + 5) the corresponding algebraic variety is {(1, 3), (

)} . In the left

1
,6
2

figure are plotted the zeros of the two original polynomials, in the right figure the zeros of the corresponding
Grbner basis. Again, the common zeros of the Grbner basis are the same as those of the original, but the basis is
as simple as it could possibly be.

120

7.5

7.5

2.5

2.5

-2.5

-2.5

-5

-5

-7.5

-7.5
-7.5

24.

x 2y + x

-5

-2.5

2.5

xy 2 y

7.5

-7.5

) (

S12 = y x 2 y + x x xy 2 y
2 2

-5

-2.5

2.5

7.5

2 2

= x y + xy x y + xy
= 2xy
x 2y

1 x
2

x
xy 2

1 y
2

y
2 y

2xy
0

leaving x , y which is obviously a Grbner basis. The corresponding algebraic variety is


25.

x 2y + x + 1

xy 2 + y 1

) (

{(0, 0)} .

S12 = y x 2 y + x + 1 x xy 2 + y 1

= x 2 y 2 + xy + y x 2 y 2 xy + x
=x +y
x 2 y + xy 2

xy

xy + x + 1
xy 2 y 3

y 2

x + y3 +1
x +y

y y +1
xy 2 + y 3

y 2

y 3 + y 1
y3 y +1
leaving y 3 y + 1, x + y .

121

-1

-1

-2

-2

-3

-3
-3

26.

-2

x 2 y + xy 2
2

x y x

-1

-3

-2

-1

xy x
x

) (

S12 = y x 2 + xy 2 x xy x
x 2 + xy 2

= x 2 y + xy 3 x 2 y + x 2
= xy 3 + x 2
y 2

xy 3 xy 2
x 2 + xy 2
x 2 + xy 2

0
2

xy xy

x + xy
xy x

x +x
leaving x 2 + x , xy x .
27.

a. true (Theorem 5, the Hilbert Basis Theorem)


b. false (a fractal for example has infinite complexity, or   has infinitely many disjoint subsets, and neither can be
described by a finite-basis ideal)
c. true ( V 1 )
d. true (every point in 2 is the intersection of 2 and a line perpendicular to 2 )
e. true (corresponding to the intersection of two planes in 3 )
f. true (every line is the intersection of 2 and a plane perpendicular to 2 in 3 )
g. true
h. true (finding solutions to systems of linear equations)
i. false
j. false (the algebraic variety is only a property of the basis, not the basis itself notably, x x , y but x x 2 , y 2 )

28.

y < x but y x 0.

29.

+ i ci f i I :

r R :

r + i ci f i = + i rci f i = + i ci f i I where ci = rci . So why does the ring need to be

commutative with unity?

[]

30. Let s F x be a common divisor of f and g, f = sf , g = sg . Then


f = gq + r

r = f gq = sf sg q = s f g q so s is also a divisor of r.
122

[]

Let s F x be a common divisor of g and r, g = sg , r = sr . Then f = gq + r = sg q + sr = s gq + r so s is also a


divisor of f.
31.

y2 y

xy

( )

S = y xy x y 2 y

= xy xy + y = y
x

xy
0
y2

y
y

0
Since the only possible S is reducible to 0, the given basis must be a Grbner basis.
32.

First, show that I S =

[]

f F x

(additive identity) 0F [ x ] :

[]

f s S : fs = 0 F x is a subring:

s S :

0F [ x ] s = 0

0F [ x ] I S ;

( f )s = ( fs) = 0 = 0 f I ;
(additive closure) f , g I : s S : ( f + g )s = fs + gs = 0 + 0 = 0 f + g I ;
(multiplicative closure) f , g I : s S : ( fg )s = fs gs = 0 0 = 0 fg I .
Next, show that I < F [ x] :
f I : g F [ x] : s S : ( fg )s = fs gs = 0 gs = 0 fg I .
(additive inverse) f I S :

s S :

33.

x F n :

34.

Let S =

x V IS :

(x , y ) x

x , y 

f I S :
2

fx = 0 and this is obviously true by definition for all s S , so S V I S .

} {( )} be the unit circle about the origin except for the single point on the

+ y 2 = 1 \ 1, 0

positive x-axis. Then I S is the ideal generated by x 2 + y 2 1 of all polynomials intersecting that circle. Because of

( )

the continuity of , obviously 1, 0 V I S .


The following figures demonstrate some elements (polynomials) in that ideal and how they each intersect the unit
circle:

1
1

0
-1

-1
0

-1

1
-0.5
-1

-1

-1
0

x2 + y 2 1

-1

-1

-1

(x

)( )

+ y 2 1 y x

(x

)( )

+ y 2 1 x 2

35.

Obviously any polynomial in N is zero-valued for any element of VN, so N I V N .

36.

Let N = x 2 , so N is every polynomial in  x , y in which every term is divisible by x 2 . Obviously the yaxis

[ ]

{ (0, y )} V N . Also, any point (a, b ) not on the y-axis cannot be in VN because x
y 

N and (a , b ) x 2 = a 2 0

, so VN is precisely the y-axis. Now I V N are all the polynomials that are zero-valued for the y-axis, which obviously
123

includes N. But also x I V N and x x 2 .

7.1 Unique Factorization Domains

By definition, a reducible can be factored into irreducibles but not vice versa. The key characteristic of a Prinicpal
Ideal Domain is that every element can be identified with an ideal. Since the infinite union of an infinite sequence
of properly contained ideals is 1 , this terminates the sequence of ideals, and correspondingly therefore every
element has a finite factorization. In a Prinicpal Ideal Domain an irreducible is prime, so the factorization is unique.

7.1.16

D PID

D UFD
7.1.28

[]

[]

F x PID

D x UFD
5.6.20

6.2.24
F field
28

[]

D UFD

D x UFD
recover a factorization of
f D x from one in F x

[]

5.6.20

F field of quotients

[]

[]

F x UFD

3.
4.

Since 5 is prime, the only factorization up to associates of 5  is 5 = 1 5 where 1 is a unit, so 5 is irreducible.


Since 17 is prime, the only factorization up to associates of 17  is 17 = 1 17 where 1 is a unit, so 17 is
irreducible.
14 = 2 7 is reducible.
Is a primitive polynomial and irreducible.

5.

2x 10 = 2 x 5 is reducible.

6.

2x 3 is of degree 1 and a 0 : 2x 3 =

7.
8.

Idem, irreducible.
Irreducible.

9.

x :

1.
2.

2x 3
a where a is a unit, so irreducible.
a

[ ] {2x 7, 2x + 7}
[ x ] : {2x 7, 4x 14, 6x 21, 8x 28}
 [ x ] : {2x 7, 4x 3, 6x + 1, 10x 2}
11

10.

The roots of the polynomial are


x=

( ) (4)

[]

So in  x :

[]

11 x :

2 4

4 48

4 16 128 4 112 4 4 7
=
=
=
8
8
8

[]

2 2 x 2 x + 2 , in  x :

)(

1
2

1
2

7 .

4x 2 4x + 8 . In 11 the polynomial has roots x = 5, 7 so in

2 2 x 5 x 7 .

124

11.

12.

234 = 21 3 2 131

3250 = 21 53 131 gcd = 21 131 = 26

1690 = 21 51 13 2

4
2

784 = 2 7

3
1
2
1960 = 2 5 7 gcd = 23 7 1 = 56

448 = 26 7 1

13.

2178 = 21 3 2 112

396 = 2 2 3 2 111
1
2
1
gcd = 2 3 11 = 198
3
2
1
792 = 2 3 11

594 = 21 33 111

14.

6 3x 2 2x + 8 .

15.

18x 2 12x + 48 .

16.

2x 2 3x + 6 .

17.

2x 2 3x + 6 .

a b = ab 1 is only well-defined if b has an inverse. But even elements without an inverse can be associates, e.g.
26 = 26 1. So if and only if a = bu , where u is a unit.
19.
Insert without one of the factors being a unit.
20.
Smaller is not defined. if and only if any divisor divides at least one of the factors in any factorization.
21. a. true (a field does not have any nonzero nonunit elements)
b. true (by Corollary 6.2.6 a field has only the trivial and nonproper ideals, which are both principal)
c. true (Theorem 16)
d. false (Example 30)

18.

[]

e. true (by Corollary 17  is a UFD, and by Theorem 28  x is also)


f. false ( 5, 7  are irreducible but not associates)

[]

[ ][ ] [ ]

g. false (  x is a PID but  x y =  x , y is not)


h. true (Theorem 28)
i. false (an associate of p could appear)
j. true (by Definition 5 a UFD is only defined for an integral domain, which cannot have divisors of zero)
22.

[]

[]

By Lemma 26. The irreducibles of D x are the irreducibles of D and the irreducibles in F x that are primitive in

[]

Dx .
23.

[]

[]

[]

Again following Lemma 26, a nonprimitive polynomial in D x is reducible in F x but irreducible in D x , for

[]

[]

example 2x + 2 = 2 x + 1 is reducible in  x but irreducible in  x .

( ) ( )( ) ( )( )

24.

With divisors of zero, factorizations are no longer unique. For example, 1, 0 = 1, 0 1, 3 = 1, 0 1, 5 .

25.

Suppose p = ab is reducible where a , b are not units. Then a , b p 0 are not divisible by p, for suppose without
loss of generality that a = p 0 then a = ab 0 c : a = abc . Since an integral domain has no divisors of zero and

26.
27.

p 0, then a 0. Since cancellation holds in an integral domain we have 1 = bc c = b 1 but b is not a unit. So p
is not prime. So if p is prime, it is irreducible.
Let p = ab be a factorization of an irreducible p. Then without loss of generality, a is a unit. Since this
factorization is unique up to associates and ab = p 0 b = p 0, so p is prime.
(reflexive) a = a 1 so a ~ a ;
125

(symmetric) a , b D : a ~ b

u D : a = bu

(transitive) a , b , c D : a ~ b , b ~ c

b = au 1

b ~ a , where u, u 1 are units;

u, v D : a = bu, b = cv

( )

( )

a = bu = cv u = c vu

a ~ c , where vu is

also a unit.
28.

Let a , b D * U be two nonzero nonunits. Since D is an integral domain, it has no divisors of zero so ab 0.
1
1
1

Suppose ab was a unit, then ab ab


= 1 a b ab = 1 a 1 = b ab
and a would be a unit. So

( )( )

( )

( )

D * U is closed. It is not a group because it does not contain the identity 1.


29.

[]

[]

[]

Let f D x be primitive, and let g , h D x : f = gh . Suppose g = cg is not primitive. Because D x is a UFD,


f = cg h and f is thus not primitive.

30.

Lemma 9 shows that every principal ideal is contained in a finite chain of ideals that terminates in D = 1 .

31.

x 3 y 3 has a root for x = y , so x 3 y 3 = x y x 2 + xy + y 2 . The quotient has roots

x=

32.

y y 2 4 1 y 2

y 3y 2

)(

y y 3
that are not in .
2

2 1
2
ACC MC
By ACC, any chain of strictly increasing ideals is finite, therefore there is a last ideal in this chain that is not properly
contained in any other ideal.

MC FBC FBC MC

Suppose there is an ideal N that has no finite basis set. Surely it has at least an infinite one. Then we can construct
an infinite set of ideals by iteratively adjoining one element from theis basis set, with each new ideal containing the
pervious ideal. This set therefore does not satsify MC.
FBC ACC
If every ideal has a finite basis, then we can construct a finite chain of ideals by iteratively adjoining an element from
the basis set to the previous ideal. Since every ideal in any chain can be constructed from a finite chain of ideals, the
chain must be of finite length. (shaky)
33.

DCC mC mC DCC

Suppose S was a set of ideals in which every ideal contains some other ideal of S. Then an infinitely long decreasing
sequence of ideals would exist.
mC DCC
If any strictly decreasing sequence of ideals has an ideal that does not properly contain any other ideal in that
sequence, the sequence must be finite.
34.

ACC holds in , but for any finite-basis ideal n <  there is always another relative prime that can be adjoined to
the basis to construct a new ideal properly contained in it.

7.2 Euclidean Domains

The valuation gives a measure by which we can guarantee that a factorization will at some point terminate.

Let D be a Euclidean domain with valuation . Then for r0 , r1 D * :


r0 = r1q 2 + r2
r1 = r2q3 + r3

1.

r2 = 0 r2 < r1
r3 = 0 r3 < r2

M
ri 1 = ri qi + 1 + ri + 1 ri + 1 = 0 ri + 1 < ri
If ri 1 = d 0, ri = d 0 ri + 1 = ri 1 ri qi + 1 = d 0 and if ri + 1 = d 0, ri = d 0 ri 1 = ri qi + 1 + ri + 1 = d 0 , so the
common divisors of ri 1 , ri are the same as those of ri , ri +1 . So when rs is the first remainder equal to zero, a
greatest common divisor of rs 2 , rs 1 is also one of r0 and r1 . And since rs 2 = rs 1q s + rs = rs 1q s , a greatesst
common divisor of rs 2 , rs 1 is rs 1 .
On , the q and r of Condition 1 do exist by Theorem 1.5.3 and 0 r < b . From r < b and r , b 0 we have
126

( )

r 2 < b 2 r < b . Then a , b D * :


2.

a 2 ab

= a 2b 2

1 b

b 1.

We know by Theorem 5.6.1 that on the ring of polynomials over a field the quotient and remainder q and r in
a = bq + r are unique, so a solution may not exist in  x and Condition 1 is not satisfied. For example,

[]

(2x 1) = (2) (x ) + (0) .


1
2

3.

4.

Again, the quotient and remainder are unique but if the remainder is nonzero we cannot guarantee that r < b and
satisfy Condition 1. For example, 1x + 7 = 1x 1 + 7 where 7 </ 1x
7 </ 1 . The problem is that the

) ( )() ( )

( )

process of division does not necessarily reduce .


In a field, for any a , b F , b 0, a = bq always has a solution so Condition 1 is satisfied. But in , for 0 < b < 1

( )

( )

and any a * : a / ab

( )

ab < a

a 2b 2 = ab

< a2

b2 <1

( )

5.

From Exercise 4, Condition 1 is satisfied. Also, a , b * : a = ab .

6.

23 = 3 138 391

0 < b < 1.

= 3 3266 391 8 391 = 3 3266 24 391 = 3 3266 25 391

= 3 3266 25 7 3266 1 22471 = 3 3266 175 3266 + 25 22471


7.

= 172 3266 + 25 22471


49349 = 15555 3 + 2684

15555 = 2684 6 549


gcd 49349, 15555 = 61
2684 = 549 5 61

549 = 61 9

61 = 5 549 1 2684

8.

= 5 6 2684 1 15555 1 2684 = 29 2684 5 15555

= 29 1 49349 3 15555 5 15555 = 29 29349 92 15555


9.

By polynomial long division:

(x

10

3x 9 + 3x 8 11x 7 + 11x 6 11x 5 + 19x 4 13x 3 + 8x 2 9x + 3

)(

) (

= x 4 2x x 6 3x 5 + 3x 4 9x 3 + 5x 2 5x + 2 + x 4 3x 3 2x 2 5x + 3

(x 3x + 3x 9x + 5x 5x + 2) = (x + 6x 19) (x 3x 2x
(x 3x 2x 5x + 3) = ( x + ) (59x 118x + 59)
so 59 ( x 2x + 1) and x + 2x 1 are greatest common divisors.
6

1
59

) (

5x + 3 + 59x 3 118x + 59

3
59

10.

Calculate di + 1 = gcd ai , di , where d0 = a0 .

11.

2178 = 396 5 + 198


d1 = gcd 2178, 396 = 198
396 = 198 2

792 = 198 4

d2

726 = 198 4 66
198 = 66 3
12.

)
= gcd(792, 198) = 198
(

d3 = gcd 726, 198 = 66

[]

a. Yes, because  is a UFD by the Fundamental Theorem of Arithmetic and  x is a UFD by Theorem 1.28.

[]

b. This is the subset of  x with even constant term. It is fairly obvious that it is in fact closed and a subring. Now

[]

consider any g  x : g = xg + g 0 + 1 ,
f = xf + f 0 ,

[]

[]

g  x , g 0 2 polynomial with odd constant term and any

f  x , f 0 2 in the subring. Then


127

)) (
f g + f xg + ( g
f g + ( f g + ( g
(

f g = xg + g 0 + 1 xf + f 0
= x2
= x2

0
0

) ( )
+ 1) f )x + ( g f + f )
+ 1 xf + g 0 + 1 f 0
0 0

It is obvious that the constant term g 0 f 0 + f 0 is again even, so the subring is indeed ideal.
c. No. Any generator of the ideal in (b.) would have to have even constant term, but this wouldn't then generate
polynomials with odd coefficients on nonconstant terms. For example, there is no polynomial that will generate
both 2 and x.
d. No, by Theorem 4.
13. a. true (Theorem 4)
b. false (by the discussion after Corollary 5)
c. true (Corollary 5)
d. false (by the discussion after Corollary 5 and Exercise 12)
e. true (in a field, every nonzero element is a greatest common divisor of any set of nonzero elements)
f. true
g. true (Theorem 6)
h. false (by Theorem 6 every unit u has u = 1, not only the multiplicative identity)
i. true (Theorem 6)
j. true (Example 3)
14.
No, because the arithmetic structure of a domain D is defined by its operations and is independent of any particular
choice of valuation.
15.
If a and b are associates then there exists a unit u such that a = bu . By Condition 2 of Definition 1,

( )

a = bu bu u 1 = b , and conversely, so a = b .

( )

16.

If b is a unit, then a and ab are associates and by Exercise 15 a = ab . Conversely

17.

This is the set of all elements with valuation greater than that of a unit. Condition 2 shows that the set is closed
under multiplication, but it is not closed under addition and hence not a group. For example, for 3, 2  ,

( )

( ( ))

3, 2 > 1 and 3 + 1 = 1.
18.

In any field, Condition 1 holds with zero remainder always. If is the identity i D * , Condition 2 holds as well.

19.

a. Since is minimal for 1, has minimum value 1 = 1 + s > 0 , so : D * + . Also, if Condition 1 holds for
then it also holds for because r < b r < b , and if Condition 2 holds for it also holds for because
a ab a ab .

( )

( )

b. Since is minimal for 1 and r > 0, l has minimum value r 1 0 , and since maps to integers and r + ,
: D * + . If Condition 1 holds for then it also holds for because r < b r < b , and if Condition 2
holds for it also holds for because a ab a ab .

( )

( )

c. Let be any valuation. Then : D * + : a a a 1 100 + 1 is a Euclidean valuation by (a.) and (b.), with

)
a ((1 + 1) 1) 100 + 1 = 101 > 100.

1 = 1 1 100 + 1 = 1 . Since has minimum value 1, a 1 + 1 for any nonzero nonunit, and

20.

For any a , b D * , a , b are their principal ideals, that is, all their multiples. Then a b is an ideal by
Exercise 6.1.27 of all the common multiples of a and b. Since D is a Euclidean domain it is a PID, so
c D : a b = c . Since ab 0 and ab c we know that c E and c 0. Since c divides every element
of c , it is a least common multiple.

21.

( )

If a and b are relatively prime, then by Theorem 9 , D : a + b = gcd a , b = 1, so a and b generate 1 = .

128

a b
Conversely, let d = gcd a , b . Every element generated by a and b is of the form a + b = d
+
d , so to
d
d
generate  we must have d = 1, which is to say that a and b must be relatively prime.
If a and n are relatively prime, then by Theorem 9 ,  :

( )

22.

( ) ( )

a + n = 1 b a + n = b
23.

( )

( )

b a = n b

x = b .

Let d = gcd a , n . By Theorem 9 ,  : a + n = d . Since d divides b,  : b = d . Then

a + n = d = b
 : ax + n = b .

24.

b a + b n = b

( )a = b x = . Conversely, if ax = b then
( ) ( )
Now d = gcd(a , n) obviously divides ax + n , so if it does not also divide b the equation
a + n = b

cannot possibly have a solution. In other words, ax = b has a solution for x in n iff the greatest common divisor
of a and n divides b.
b
Find by the procedure outlined in Exercise 6, and let d = gcd a , n . Verify that d divides b, then x = =
. So
d
42 = 22 2 2, 22 = 2 11 so d = gcd 42, 22 = 2 . We see that 2 indeed divides 18, so there is a solution
x=

b 2 18
=
= 18.
d
2

( )

7.3 Gaussian Integers and Norms

5 a bi
5
5a 5bi 5 10i
=
=
=
= 1 2i , where a = 1, b = 2 , so 5 = 1 + 2i 1 2i .
a + bi
5
a + bi a bi
a2 + b2

)(

1.

a + bi  :

2.

N7 = 49 has to be factored into two factors, so we are looking for a + bi with norm 7 but that doesn't exist by
Theorem 10. Irreducible.

3.

4.

5.

6.

)(

( )( ) (
) (
) (
) (
)
( )( )
4 + 3i = (1 + 2i )( 2 i ) .
6 7i (6 7i )(a bi ) (6a 7b ) + ( 7a 6b )i (6 4 7 1) + ( 7 4 6 1)i
=
=
=
= 1 2i , where a = 4, b = 1 , so
a + bi (a + bi )(a bi )
1 +4
a +b
6 7i = ( 4 + i )(1 2i ) .
4 + 3i a bi
4a + 3b + 3a 4b i
4 1 + 3 2 + 3 1 4 2 i
4 + 3i
=
=
= 2 i , where a = 1, b = 2 , so
=
2
2
a + bi
a + bi a bi
a +b
12 + 2 2

6 = 23
= 1 + i 5 1 i 5

( )( ) ( ) ( ) + i = 1 + i
( )( )
= = (7 + 2i ) (3 4i )(1 + i ) = (7 + 2i ) ((3 + 4) + (3 4)i ) = (7 + 2i ) (7 i ) = i .
8 + 6i
(8 + 6i )(5 + 15i ) = (8 5 6 15) + (8 15 + 6 5)i = 50 + 150i = + i

=
5 15i (5 15i )(5 + 15i )
250
5 + 15

i (5 15i ) = 15 + 5i

(8 + 6i ) (15 + 5i ) = 7 + i

8 + 6i = i (5 15i ) + ( 7 + i )
7 + 2i 3 + 4i
21 8 + 28 + 6 i
7 + 2i
=
=
=
=
3 4i
3 4i 3 + 4i
32 + 42

7.

13
25

34
25

1
5

3
5

129

)(

) ( ( ) ( ) ( )) ( ( ) ( ) ( ))

)(

5 7 15 1 + 5 1 + 15 7 i 50 + 100i
5 15i 7 i
5 15i =
=
= 1 + 2i
=
7 + i
50
50
7 2 + 12

5 15i = 1 + 2i 7 + i

gcd 8 + 6i , 5 15i = 7 + i
8.

a. true (Theorem 4 and Theorem 2.4)


b. true (Theorem 4)
c. true (Definition 1)
d. false (

1
2

[]

 i )

e. true (the Euclidean algorithm holds in any Euclidean domain, Theorem 9)


f. true (in the case of Theorem 7, a prime multiplicative norm corresponds to an irreducible)
g. true (Theorem 7)
h. false ( deg 0 < 0 so Condition 1 of Definition 6 doesn't hold, deg 1 = 0 so Condition 2 doesn't hold, and

9.

10.

deg x 2 deg x deg x 2 1 1 so Condition 3 doesn't hold either)


i. true (all three conditions of Definition 6 hold)
j. true (Example 9)
If D such that N is minimal, then if was reducible there would be , D : = with neither , a
unit and thus N , N 1, but then either N , N < N which is a contradiction. So is irreducible.

( )

a. 2 = i 2i = i 1 + i .

)(

[]

b. By Theorem 10, p = 4 1 p = a 2 + b 2 = a + bi a bi reducible in  i . Conversely, if p = 4 3 was reducible then


2

p = a b and Np = p = Na Nb Na , Nb = p and Na = p N a + a i = a 2 + a 2 = p but there is no such


expression by Theorem 10.
11. 1. N = a 2 + a 2 0 , where a , a .
2. N = a 2 + a 2 = 0 a 2 = 0 a 2 = 0 a = 0 a = 0 a = 0 .

( )

((

)(
)) ((
) (
))
= ( ) + ( + )
= ( ) 2 + ( ) + ( ) + 2 + ( )

3. N = N + i + i = N + + i
2

)(

= 2 + 2 2 + 2

) (

= N + i N + i = N N
12.

= + i 5, = + i 5 :
N = N + i 5 + i 5 = N 5 + + i 5

( )

(
= ( )

= 5

+ 5 +

)(

10 + 25 + 5

= 2 + 5 2 2 + 5 2

) (

+ 10 + 5

= N + i 5 N + i 5 = N N
13.

Let d be a nonzero nonunit. Suppose a , b : d = ab where a , b nonunit. Because Nd = Na Nb and Na , Nb > 1 it


must be that Na , Nb < Nd . Otherwise, if / a , b : d = ab where a , b nonunit, d is irreducible. Because Nd has a
finite factorization in , repeating this procedure will at some point terminate.
130

(
(

15.

)(
)(

) (
)
( ) ) (

) (

16 + 7i 10 + 5i
16 10 7 5 + 16 5 + 7 10 i 125 + 150i
16 + 7i =
=
=
= 1+11i
5
10 5i
125
125
10 5i 10 + 5i

1 + i 10 5i = 10 1 5 1 + 10 1 5 1 i = 15 + 5i

16 + 7i 15 + 5i = 1 + 2i

16 + 7i = 1 + i 10 5i + 1 + 2i

10 1 5 2 + 10 2 + 5 1 i 25i
10 5i 1 2i
10 5i
=
=
= 5i
=

5
5
1 + 2i 1 2i
1 + 2i

gcd 16 + 7i , 10 5i = 1 + 2i

14.

( )(
) (
)
(
) (
)
( )(
) ( )
) ( ( ) ( ) )
(
)( ) (
( )( )
(
)
a. Since [i ] is a Euclidean domain, there exists a valuation on [i ] . Then D : , D : = +
[i ]
[i ]
where = 0 or < . So : [i ]
: a + is the canonical homomorphism onto
and the
*

conditions on 0 show that there are a finite number of them.

[]
is maximal and [i ]

b. If were not maximal then there would be :  i so : = where not a unit (else

= ), but then would be reducible. So

a field.

c. I verified these by plotting on graph paper. The characteristic is pretty simple to find, the order seems always to be
equal to the norm.
i.

ii.

iii.
16.

[ ] ={

i
3

[]

i

1+ i

0 , < 3

= 2;

[]

i

[]

i

char

= 5;

1 + 2i

i
} 3[ ] = 9;

+ i + 3 .

1+ i

char

char

[ ] = 3;

i
3

= 2;

[]

i

1 + 2i

= 5.

I don't think n needs to be square free in this exercise but in the next one.
a. Obviously  n : N 0. Also, N = 0 a 2 + nb 2 = 0 a 2 , b 2 = 0 a , b = 0 . Finally,

,  n : = + i n , = + i n :

N = N + i n + i n

) (
)
= ( n ) + n( + )
= ( ) 2n + n + n( )
= N n + + i n
2

)(

= 2 + n 2 2 + n 2

+ 2n + n

= N + i n N + i n = N N
2 = 1 n 2 = 0
= 1 = 0

b.  n : N = 1 N + i n = 2 + n 2 = 1

2
2
= 0 n = 1
= 0 = 1 n = 1
Since these elements are also the only possible units, this describes precisely all the units.
131

c. We show that  n is an integral domain by showing it has no divisors of zero:

= 0 N = N N = 0 N = 0 N = 0 = 0 = 0

( )

17.

Then by Exercise 13 and (b.) every nonzero nonunit has a factorization into irreducibles.
I think n needs to be square free in this exercise but not in the previous one.
*

a. Obviously  n : N 0. Also, : N = 0 2 n 2 = 0 2 = n 2 , = 0 , where (*)



holds only if n is square free for example, if n = 3 : 2 = 3 2 , = 0 but if n = 4 :
2 = 4 2 = 2, = 1 . Then, ,  n :

N = N + n + n

( )

) (
)
= ( + n ) n( + )
= N + n + +
2

+ 2n + n

)(

= 2 n 2 2 n 2

2n n

= 2 n 2 2 n 2
= N + n N + n = N N

( )

b. This can only hold if one of 2 , 2 is even and the other odd. Since 1
2 = 1 n 2 = 0

2 = 0 n 2 = 1

is the only odd integer square,

= 1 = 0

= 0 = 1 n = 1

c.  n is an integral domain because it has no divisors of zero because it has a multiplicative norm, so by Exercise

13 and (b.) every nonzero nonunit has a factorization into irreducibles.
18.
Let ,  2 : = + i , = + i , 0 and let q = = q + iq , q , q  , and let

= + i , ,  as close as possible to q so that q , q 1 . Then

((

))

) (

((

2
2

) (
) ) ( ) + 2 ( ) = + = and
= N = N ( ) = N ( ( )) = N N (q ) N < N = , so  n is a Euclidean domain.

(

N q = N q + iq + i = N q + q i

1
2

1
2

1
4

2
4

3
4

3
4

Similarly for  n , N q = n

Euclidean domains.

()
1
2

n = 2, 3

= 1 n and = N 1 n < N = so  2 ,  3 are


4
4

8.1 Introduction to Extension Fields

So extension field is just superfield.

1.

x = 1+ 2

2.

x= 2+ 3

x 1 = 2

x 1

= x 2 2x + 1 = 2

x 2 2x 1 = 0 .

x 2 = 2 + 3 = 2 + 2 6 + 3 = 5 + 2 6

x 5= 2 6

x 5

= x 2 10x + 25 = 4 6 = 24

x 2 10x + 1 = 0
132

3.

x = 1+ i

4.

x = 1+ 3 2

x 1 = i

x 1

x=

2 i

= x 2 2x + 1 = 1 x 2 2x + 2 = 0 .

x 2 = 1+ 3 2

) = (x + 1)(x

x2 1

= x6 x 4 x 2 +1 = 2
5.

x2 = 3 2 i

2x 2 + 1 = x 6 2x 4 + x 2 + x 4 2x 2 + 1

x6 x 4 x 2 1 = 0
x2 +i = 3 2

(x + i ) = (x + i )(x + i ) = (x + i )(x + 2x i 1) = (x + 2x i x + x i 2x i ) = 2
(x 3x 2) = (3x + 1)i
(x 3x 2) = (3x + 1) (x 3x 2) + (3x + 1) = 0
3

x 12 3x 8 2x 6 3x 8 + 9x 4 + 6x 2 2x 6 + 6x 2 + 4 + 9x 8 6x 4 + 1 = 0
x 12 + 3x 8 4x 6 + 3x 4 + 12x 2 + 5 = 0
x2 = 3 6

x = 3 6

6.

x2 3

x2 3 = 6

= x 4 6x 2 + 9 = 6

x 4 6x 2 + 3 = 0

irr 3 6 ,  = x 4 6x 2 + 3;

x=

7.

1
3

x2 =

+ 7

irr

x = 2 +i

8.

1
3

1
3

x2

+ 7

deg 3 6 ,  = 4

1
3

+ 7 ,  = x 4 2 x 2 6 8 ;
3
9

x i = 2

x i

1
3

= x4 2 x2 +
3

1
9

=7

+ 7 ,  = 4

x 2 3 = 2ix

deg 2 + i ,  = 4

x2 3

= x 4 6x 2 + 9 = 4x 2

x 2 = 1 x 2 + 1 = 0; algebraic with degree 2.

9.

x =i

10.

x = 1+ i

11.

x = ; transcendental.

12.
13.

x=
idem.

14.

x = 2 ; transcendental.

15.

x =2

x 2 = 0 ; algebraic with degree 1.

16.

x =2

x 3 = 6 ; algebraic with degree 3.

18.

deg

1
3

= x 2 2ix 1 = 2

irr 2 + i ,  = x 4 2x 2 + 9;

17.

x2

= 7

x 1 = i

x2 =

x 1

= x 2 2x + 1 = 1 x 2 2x + 2 = 0 ; algebraic with degree 2.

x 2 = 0; algebraic with degree 2.

) (

x2 + x +1
= x + +1 x2 + x +1 = x x + +1 .
x
a. Since the polynomial has no zero for any element of 3 , it is irreducible by Theorem 5.6.10.

x2 +1

0
1
2

1
2
2

)(

b.

133

0
1
2

0
1
2

2
1+
1 + 2
2+
2 + 2

0
1
2

2
1+
1 + 2
2+
2 + 2

1
2
2
0
0
1
1+
2+
2
1 + 2 2 + 2

1+
2+
1 + 2 2 + 2 2

2+
1+
2 + 2
2
1 + 2
0
1
2
0
0
0
0
0
0
0
0
0

0
1
2

2
1+
1 + 2
2+
2 + 2

1+

1 + 2

2+

2 + 2

2
1 + 2
2 + 2
0

1+
2+

1 + 2
2 + 2
2
1
1+
2
2+
0

2+

2 + 2
2
1 + 2
2
2+
0

1+
2+
2
0
1 + 2
1
2 + 2
2

0
2
1
2

1
1+
2
2+
2

1 + 2

0
0
1+
1 + 2
2 + 2 2 +
1
2+
1+
2
1 + 2 2 + 2
1 + 2 2
2

2 + 2
2
1+
1
2

2+
1
0
2

2
2
1
2 + 2 2 +
2+
1+
1 + 2 2 + 2
1+
1 + 2

where + 1 = 0 = 1 = 2 .

1 + 2
1
2 + 2
2
2
0
1+

1+
2 + 2
2
2
0
1 + 2
1
1
1+
2+
2 + 2
0
2+
1 + 2
2 + 2
1+
1
2

0
2 + 2
1+
1 + 2
2+

1
2
2

19.
20.
21.
22.
23.

[]

of some nonzero polynomial in F x .


nonzero
are having the coefficient of the highest-degree term equal to 1.
Correct?
a. true (there is no polynomial over  having as a root)
b. true

[]

c. true ( f F : x f F x has f as a root)


d. true (   )
e. false (  /  2 because addition on  2 is not the one induced from  )
f. true (Definition 14)

[]

[]

g. false ( x 2 2  x has degree 2 but x 2  x has degree 1)


h. true (Kronecker's Theorem)
i. false (   2  is an extension field but x 2 + 1 has no zero in it)
j. true (as in the discussion after Example 19)
24.

( )

a. In F = 1, 3 , irr , F = x 3 3 with degree 3.

( )

b. In E = 1, e 10 , irr e 2 , E = x 5 e 10 with degree 5.


25.

) {}

a.  2 x 3 + x 2 + 1 = 1 has no zero in  2 , so no nonunit factors.


b.

) ( )

x3 + x 2 +1
= x 2 + 1 + x + 2 + , so x 3 + x 2 + 1 = x x 2 + 1 + x + 2 + . To factorize the second

x
2
2
2
factor, finding a zero by applying the elements of  2 : 0, 1, , + 1, , + 1, + , 2 + + 1 . Eventually we
find that 2 x 2 + 1 + x + 2 + = 2

(
()
+ (1 + )

+ 2 + = 3 + + 2 + 3 + 2 + = 0 , where
134

3 + 2 + 1 = 0 3 = 2 + 1 ;

x + 1+ x + +
2

) = x + (

4 = 3 = 2 + 1 = 3 + . Then

( ) ( )( ) (
)
that the solution in the text has a minus sign in x ( + + 1) ?
deg ( ,  ) = 3  ( ) : x = 0 ;  ( ) = {0, 1, , + 1, , + 1, + , + + 1} . The table gives
x

+ + 1 , so x 2 + 1 + x + + 2 = x x 2 x + 2 + + 1 . Note

26.

()

3 , + :
0
1

1
0
+1

2
2 +1

+1

+1
0
1

2 +1
2

2 +
2 + +1

2 + +1
2 +

2 +
2 + +1
0
1

2 + +1
2
2 +1
+1
2 +
2 +1
2
2
2
2
2
1
+1

+1
+
+ +1

2
2
2
2
0

+
1
+1

+ +1
+

0
1
+1
2 + 2 + +1
2
2 +1

+1
1
0
2 + +1 2 +
2 +1
2

By the Fundamental Theorem of Finitely-Generated Commutative Groups, this has to be isomorphic to either
 2  2  2 ,  4  2 , or 8 . Since  4 , 8 have elements of order 4 and 8, respectively, which  2 does not,
1
0

{ } { }

()

{ }

we must have  2  2  2  2 0, 1 0, 0, 2 .

27.
28.
29.
30.

()

()

 2 , has order 7 so it can only be isomorphic

to  7 .
Because it is (Theorem 13) of minimal degree.

}
a. x ( x + 1)( x + 2) = x ( x + 3x + 2) = x ( x + 2) = x
()

By Theorem 18, F =

)(

()

+ ni =01 bi i where each of the elements are unique, so F = F

b i F

= qn .

+ 2x evaulates to zero for x 0, 1, 2 = 3 so

x x + 1 x + 2 + 1 = x 3 + 2x + 1 evaluates to one over 3 and is therefore irreducible.

()

b. By Exercise 29, 3 = 33 = 27 .
31.

a. Since 12 = 1 and p 1

= p 2 2p + 1 = p 1 there must be at least one element in p that is not a square.

()

b. By (a.), there is an element a p that is not a square, so x 2 a has no zero in p so a is of degree 2 in p ,

()

and by Exercise 30b p a = p .


32.

()

()

()

Because if F is algebraic then p F : p = 0 and then p F : a p = a + p such that

p = p = 0 , and would be algebraic also. It is clear that p is in fact polynomial also.


33.

34.

It is clear that a , b , c  a + b 3 2 + c 3 2  is the simple extension  3 2 described by Theorem 18 and a

field, and that  3 2  .

i. Since 8 = 23 , we look for an irreducible polynomial of degree 3 in  2 : x x 1 x + 1 = x 3 x 2 + 1 . So

()

2

[]

2 x
3

x x +1

()

and  2 = 23 = 8 (Exercise 25).

135

ii.Since 16 = 2 4 , we look for an irreducible polynomial of degree 4 in  2 : x x 1 x 2 + 1 = x 4 x 3 + 1. So

[]

2 x

()

2

x x +1

()

and  2 = 2 4 = 16 .

) {

iii. Since 25 = 52 , we look for an irreducible polynomial of degree 5 in 5 . Since 5 x 2 + 3 = 2, 3, 4 that one is

()

irreducible and 5
35.

[]

5 x
2

x +3

()

and 5 = 52 = 25 (Exercise 31b).

Since F is finite it is of prime characteristic and contains a prime subfield p F (Theorem 6.2.19). By Theorem

5.3.1, a p : a p 1 = p 1 a x p 1 1 = 0 and algebraic over p .


36.

By Exercise 35 every finite field can be considered an extension of its prime subfield. Then by Exercise 29, the
order of the field is a prime power.

8.2 Vector Spaces


1
4

So, dimensionality does not even enter into the definition of a vector space the defining aspect is only scalar
multiplication with a field. It's almost like a G-set (Definition 3.5.1) except X has to be an actual group and G a
field. Note in particular that no relationship between V and F is implied.
This seems profound but is almost meaningless: any superfield can be regarded as a vector space in the same way
that any field is a vector space.

1.

{(0, 1), (1, 1)}, {(1, 0), (1, 1)}, {(1, 0), (1, 1)} .

2.

Since

[1
[0
[0

0
1
0

] ([1
0] = ([1
1] = ([0
0 =

] [
0] + [0
1] + [1

] [
1] [1
1] [1

])
1])
0])

1
2

1 0 + 1 0 1 0 1 1

1
2

1
2

1
0

0
1

and by Lemma 16, this set of vectors obviously spans 3 . Since dim 3 = 3 , by Theorem 17 this is a basis.

3.

4.
5.
6.
7.
8.
9.
10.

4 y + 2z = 0
10
y + 5z = 0

1 y + 1z = 0
5
5x 7 y = 0

10x 14y = 0
so not linearly independent by Definition 10, and hence not a basis.
1x + 1y + 2z = 0

2z 3y + 1z = 0
10x 14y = 0

{1, 2} .
{1}.

{2

, 21 3 , 2 2 3 .

{2

, 21 4 , 2 2 4 , 23

03

{1, i} .
{1, i}
0 4

}.

The same polynomial of Example 1.19 has a zero for 1 + :

) ( ) + (1 + ) + 1 = 1 +

1+ x 2 + x + 1 = 1 +
11.
12.
13.

+ = 1+ +1+ = 0.

Delete uniquely.
Correct.
Correct.
136

14.
15.

independent.
a. true ( , V : + V )
b. false ( a , b F : a + b F )
c. true ( a , b F : a b F )
d. true ( a F , V : a V )

[]

e. false ( F x has an infinite basis, Example 7)


f. false (Definition 15)
g. false (wouldn't be linearly independent)
h. true (Theorem 23)
i. true (idem)
j. true (discussion before Lemma 16)
16. a. A subspace of a vector space V over F is a vector space of a subgroup of V over F with the induced operations.
b. We have to show that the intersection is closed. Let U,V be subspaces over F:
, U V : + U , + V + U V ,
U V , a F : a U , a V a U V .
17.
18.
19.
20.
21.
If every vector in V can be generated by the i then they certainly span at least V. If the zero vector is the sum of
none of the i , and the zero vector can be expressed only uniquely as a linear combination of i , then they are
linearly independent. So i are a basis. Conversely, if i are a basis then they span V. Each vector is a unique
linear combincation of the i , otherwise the difference between two expressions of the same vector would give a
linear combination of the zero vector and i would not be linearly independent.
22.

a. Considering the vector space F m over F, we have i : + nj aij x j = bi

a j = a0 j am 1, j ,

+ j x j a j = b, where

{ a }.
can be expressed uniquely as a linear combination of the basis { a } .
)

b = b0 bm 1 F . The system has a solution iff b is in the span of

b. By the Exercise, every F m


23.

They are naturally isomorphic by their coefficients. Let

n 1
i =0

v i and

n 1
i = 0 fi

} be bases for V and F

, respectively.

Then :V F n : x = + i x i v i a + i x i fi is an isomorphism: x = + i x i v i , y = + i y i v i V :

) (
= + (x

) ( (

) )
= ( + x v ) + ( + x v ) = x + y

x + y = +i x i vi + +i y i vi = +i x i + y i vi
i

+ y i fi = + i x i fi + + i y i fi

and a F , x = + i x i v i V :

i i

i i

( ) (
) ( ( )) = (+ (ax )v )
= + (ax )f = + a ( x f ) = a + x f = a ( + x v ) = a x
a. v = + v : v = ( + v ) = + (v ) = + v .
ax = a + i x i v i = + i a x i v i
i

24.

i i i

i i

i i i

i i

i i

i i

i i

b. Since by (a.) a linear transformation is completely determined by its action on the basis vectors, the action required
for the basis vectors specified here suffices.
25. a. homomorphism.
b. The nullspace of is the set of vectors v V : v = 0 . To show that Ker V is a subspace we have to show that it
is closed under the induced operations from V:
v , w Ker : v + w = v + w = 0 + 0 = 0 v + w Ker .

c. When it is a homomorphism (linear transformation) with Ker = E .


26.

The quotient space V S over F is the vector space in the group of cosets of S in V over F, with scalar multiplication
by representatives in V. The coset group exists by Corollary 3.2.5 and is clearly commutative. Show that the five
137

conditions of a vector space hold: a , b F : + S , + S V S :

(1) a + S = a + S V S ;

( (

)) ( )
( )
(3) (a + b )( + S ) = (a + b ) + S = (a + b ) + S = ((a ) + S ) + ((b ) + S ) = a ( + S ) + b ( + S ) ;
(4) a (( + S ) + ( + S )) = a (( + ) + S ) = a ( + ) + S = (a + a ) + S = (a + S ) + (a + s ) ;
(5) 1( + S ) = 1 + S = + S .

(2) a b + S = a b + S = ab + S = ab + S ;

27.

a. We know that is a homomorphiosm so that operations under coincide with the ones induced from V . We
need to show that V V is closed:

, V : , V : = , = + = + = + V .
b. Let

{ } be a basis for Ker V .


i

To this basis can be adjoined dimV dim Ker vectors to form a basis for V.

Since :V Ker V is an isomorphism, we have dim V = dimV dim Ker .

8.3 Algebraic Extensions


1.
2.

3.

4.

5.

So we have two ways of determining the degree of an extension: by the order of the basis, and by the degree of the
irreducible polynomial.
deg  2 :  = 2 , so by Theorem 2.23 , 20 2 , 21 2 is a basis for  2 .

0 2
12
2 ,2
is a basis for  2 over . It is clear that 3 cannot be axpressed as a linear combination of this
basis, so 30 2 , 31 2 is a basis for  2 , 3 over  2 . By Theorem 4, 1, 31 2 , 21 2 , 61 2 is a basis for
 2 , 3 over  and  2 , 3 :  = 4 .

18 = 3 2  2 , 3 , so from Exercise 2,  2 , 3 , 18 :  = 4 and 1, 31 2 , 21 2 , 61 2 is a basis for

 2 , 3 , 18 over .

{2

, 21 3 , 2 2 3 is a basis for  3 2 over . Since deg 3 ,  = 2 does not divide deg 3 2 = 3 , 3  3 2


and by Theorem 4, 1, 21 3 , 2 2 3 , 31 2 , 21 3 31 2 , 2 2 3 31 2 spans  3 2 , 3 over . Since this set is linearly
independent, it forms a basis, and  3 2 , 3 :  = 4 .

3
0 2
12

2 ,2
is a basis for  2 over . deg 2 ,  = 3 does not divide deg 2 ,  = 2 , so 3 2  2 . So
03

{
{2

03

( )

, 21 3 , 2 2 3 is a basis for  21 3 , 21 2 over  21 2 and by Theorem 4, 20 6 , 2 2 6 , 2 4 6 , 23 6 , 25 6 , 2 7 6 is a

basis for  21 3 , 21 2 over . Try to simplify the basis. 2 7 6 = 2 21 6 , so 21 6  21 3 , 21 2 and

(
) ( ) ( )
( 2 ) , and ( 2 , 2 ) ( 2 ) . Then
 2 , 2
) :  = (2 , 2 ) : (2 ) (2 ) : 
(
6 = ( 2 , 2 ) : ( 2 ) 6

and ( 2 , 2 ) : ( 2 ) = 1, so by the discussion after Definition 2, ( 2

 21 3 , 21 2  21 6 . 1 6 x 6 2 = 0 is an Eisenstein polynomial for p = 2 and irreducible over , so


16

13

13

13

12

12

16

12

16

13

12

16

13

12

16

16

13

) ( )

, 21 2 :  21 6 . So

138

 21 3 , 21 2 ,  = 6 and

Exercise 8.1

x= 2+ 3

7.

{1, 6} is a basis for 

8.

Analogous to Exercise 4,

11.
12.

2i

} is a basis.

{1,

2 + 3 is a basis for  2 + 3 , and

 2 + 3 :  = 2 .

6 , and  6 :  = 2 .

} is a basis for 

2 , 3 5 , and  2 , 3 5 :  = 6 .

3
3
3
3
3
3
3
3
3
3
3
3
3
3

24 = 2 6 , so  2 , 6 , 24 =  2 , 6 . 6 = 2 3 , so  2 , 6 =  2 , 3 (why?). Then
i 3
3 j 3 is a basis for  3 2 , 3 6 , 3 24 over , and  3 2 , 3 6 , 3 24 :  = 9 .
0 i < 3, 0 j < 3 2

 2 , 6 =  2 , 3 , so 1, 2 is a basis for  2 , 6 over  3 and  2 , 6 :  = 2 .

0
2
1
2
2 + 3  3 , so 2  3 and 2 , 2
is a basis for  2 + 3 over ,  2 + 3 :  = 2 .

By Theorem 4,
 2 , 3 :  3 =  2 + 3 :  2 + 3  2 + 3 :  3





2 =  2 + 3 :  2 + 3 2

10.

0i <6

x 2 10x + 1 = 0 .

6.

9.

0 i < 2, 0 j < 3

2i

{ }

5j

13.

so  2 + 3 :  2 + 3 = 1 (?!?)

2  3 + 5 but 6 + 10  3 + 5, 2 , so 20 2 , 21 2 is a basis for  2 , 6 + 10 over


 3 + 5 .

14.
15.
16.
17.

is a field E where each element of E is


to a basis for F
Correct.
nonconstant polynomial over F

18.

 (the algebraic numbers) are real elements such as

[]

2 that have polynomials in  x such as x 2 2 with those

[]

elements as zeroes. However,  x has polynomials such as x + 1 with imaginary roots that are not in  .
19.

a. true (Theorem 3)

[]

b. false (the extension of  containing all powers of ,  , is algebraic and infinite)


c. true (Theorem 4)

[]
e. false ( x 2 [ x ], but the root 2  so  )
f. true (the only elements of ( x ) that are algebraic in  are  itself; for example, x 1 has no root in [ y ] )
g. false (the polynomial ( x + 1)y + x ( x )[ y ] does not have a root y )
d. false ( x 2 + 1  x has no zero in  )
2

h. false (Theorem 17)


i.
j. false (  is an algebraically closed extension of , but  shows that  is not an algebraic extension of )
20.

[(

) ]

Since a + bi  , 1, a + bi is a basis for  a + bi over  and  a + bi :  = 2 . Similarly,  :  = 2, so

[ :  ] = [ :  (a + bi )] [ (a + bi ) :  ]

)]

 :  a + bi =

[ :  ] = 2 = 1. So  =  (a + bi ).
[ :  (a + bi )] 2

139

21.

()

Since E F is a finite extension and E : F is prime, E is a simple extension of F , E \ F : E = F . Consider

{ }

()

()
E :F
[E : F ] = [E : F ( )] [F ( ) : F ] [E : F ( )] = F[( ) : F] = [E : F ] which is impossible because [E : F ] is prime, so
[
]
[E : F ] = 2 and [E : F ( )] = 1 so E = F ( ) is simple.

F for any E \ F . Then 1, is a basis for F over F and


1
2

22.

x 2 3 has roots 3  3 2 .

23.

Every root increases the degree of the field by a factor of 2, so  ni

24.

pi :  = 2n . Since a zero of x 14 3x 2 + 12

n
n

has degree 14 = 2 7 , which does not divide 2 , there is no element of  i pi that can be such a zero.
Since E is a finite extension, by Theorem 11 i : E = F
g = +i g i i ,

( ).
i

Let g , h D E ; g , h 0, so

h = + i h i i . Since g , h 0 and the extension is finite, there are maximal j , k : g j , h k 0 , so

g h = +i , j g i h j i + j .
25.

Since

{ }

3  7 , 1, 3 is a basis for  3 , 7 over  7 , so

x= 3+ 7

x 3= 7

x2 4

x 3 = 7

x2 2 3 +3 = 7

x 4 8x 2 + 16 = 12

= 4 3 = 12

 3 , 7 :  7 = 2 . Now consider:


x2 4 = 2 3

x 4 8x 2 + 28 = 0

so deg 3 + 7 ,  = 4 and  3 + 7 :  = 4 , so

 3 + 7 :  =  3 + 7 :  7  3 : 


4 =  3 + 7 :  7 2

2 =  3 + 7 :  7

Then:
 3 , 7 :  7 =  3 , 7 :  3 + 7  3 + 7 :  7


2 =  3 , 7 :  3 + 7 2

1 =  3 , 7 :  3 + 7

26.
27.

28.

( )

( ) [() ]

()

A zero of an irreducible p is of degree deg , F = deg p and deg , F = F : F . But E F F , and so

[F ( ) : F ] = deg p would not divide [E : F ], which is impossible by Theorem 14.


If is of degree 1, , F and F ( ) = F ( ) = F . Suppose n is of at least degree 3. By Theorem 2.23, F ( )
has basis {
} , where n = deg ( , F ) . Since 2 is relatively prime to n, 2 generates  by Corollary 1.5.18 and

) ={
} and F ( ) = F ( ) .

n 1
i =0

n 1
i =0

n 1
i =0

140

Оценить