Are you sure?
This action might not be possible to undo. Are you sure you want to continue?
J.S. Milne
S
3
r =
1 2 3
2 3 1
} =
1 2 3
1 3 2
Version 3.10
September 24, 2010
The ﬁrst version of these notes was written for a ﬁrstyear graduate algebra course. As in
most such courses, the notes concentrated on abstract groups and, in particular, on ﬁnite
groups. However, it is not as abstract groups that most mathematicians encounter groups,
but rather as algebraic groups, topological groups, or Lie groups, and it is not just the groups
themselves that are of interest, but also their linear representations. It is my intention (one
day) to expand the notes to take account of this, and to produce a volume that, while still
modest in size (c200 pages), will provide a more comprehensive introduction to group
theory for beginning graduate students in mathematics, physics, and related ﬁelds.
BibTeX information
@misc{milneGT,
author={Milne, James S.},
title={Group Theory (v3.10)},
year={2010},
note={Available at www.jmilne.org/math/},
pages={131}
}
Please send comments and corrections to me at the address on my website www.jmilne.
org/math/.
v2.01 (August 21, 1996). First version on the web; 57 pages.
v2.11 (August 29,2003). Fixed many minor errors; numbering unchanged; 85 pages.
v3.00 (September 1, 2007). Revised and expanded; 121 pages.
v3.01 (May 17, 2008). Minor ﬁxes and changes; 124 pages.
v3.02 (September 21, 2009). Minor ﬁxes; changed TeX styles; 127 pages.
v3.10 (September 24, 2010). Many minor improvements; 131 pages.
The multiplication table of S
3
on the front page was produced by Group Explorer.
Copyright c (1996, 2002, 2003, 2007, 2008, 2010 J.S. Milne.
Single paper copies for noncommercial personal use may be made without explicit permis
sion from the copyright holder.
Contents
Contents 3
1 Basic Deﬁnitions and Results 7
Deﬁnitions and examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Multiplication tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Subgroups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Groups of small order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Homomorphisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Cosets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Normal subgroups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Kernels and quotients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Theorems concerning homomorphisms . . . . . . . . . . . . . . . . . . . . . . . 21
Direct products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Commutative groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
The order of ab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2 Free Groups and Presentations; Coxeter Groups 31
Free monoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Free groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Generators and relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Finitely presented groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Coxeter groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3 Automorphisms and Extensions 41
Automorphisms of groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Characteristic subgroups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Semidirect products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Extensions of groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
The H¨ older program. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4 Groups Acting on Sets 53
Deﬁnition and examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Permutation groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
The ToddCoxeter algorithm. . . . . . . . . . . . . . . 66
Primitive actions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5 The Sylow Theorems; Applications 73
3
The Sylow theorems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Alternative approach to the Sylow theorems . . . . . . . . . . . . . . . . . . . . 77
Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
6 Subnormal Series; Solvable and Nilpotent Groups 81
Subnormal Series. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Solvable groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Nilpotent groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Groups with operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
KrullSchmidt theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
7 Representations of Finite Groups 95
Matrix representations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Roots of 1 in ﬁelds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Linear representations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Maschke’s theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
The group algebra; semisimplicity . . . . . . . . . . . . . . . . . . . . . . . . . 99
Semisimple modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Simple Jalgebras and their modules . . . . . . . . . . . . . . . . . . . . . . . . 101
Semisimple Jalgebras and their modules . . . . . . . . . . . . . . . . . . . . . 105
The representations of G . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
The characters of G . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
The character table of a group . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
A Additional Exercises 113
B Solutions to the Exercises 117
C TwoHour Examination 125
Bibliography 127
Index 129
4
NOTATIONS.
We use the standard (Bourbaki) notations: N = {0. 1. 2. . . .]; Z is the ring of integers; Q
is the ﬁeld of rational numbers; 1 is the ﬁeld of real numbers; C is the ﬁeld of complex
numbers; F
q
is a ﬁnite ﬁeld with q elements where q is a power of a prime number. In
particular, F
;
=Z¡]Z for ] a prime number.
For integers m and n, m[n means that m divides n, i.e., n ÷ mZ. Throughout the notes,
] is a prime number, i.e., ] =2. 3. 5. 7. 11. . . . . 1000000007. . . ..
Given an equivalence relation, Œ+ denotes the equivalence class containing +. The
empty set is denoted by 0. The cardinality of a set S is denoted by [S[ (so [S[ is the number
of elements in S when S is ﬁnite). Let 1 and · be sets; a family of elements of · indexed
by 1, denoted (a
i
)
i÷J
, is a function i ÷a
i
: 1 ÷·.
1
Rings are required to have an identity element 1, and homomorphisms of rings are
required to take 1 to 1. An element a of a ring is a unit if it has an inverse (element b such
that ab =1 =ba). The identity element of a ring is required to act as 1 on a module over
the ring.
X cY X is a subset of Y (not necessarily proper);
X
def
=Y X is deﬁned to be Y , or equals Y by deﬁnition;
X ~Y X is isomorphic to Y ;
X .Y X and Y are canonically isomorphic (or there is a given or unique isomorphism);
PREREQUISITES
An undergraduate “abstract algebra” course.
COMPUTER ALGEBRA PROGRAMS
GAP is an open source computer algebra program, emphasizing computational group the
ory. To get started with GAP, I recommend going to Alexander Hulpke’s page http://
www.math.colostate.edu/
~
hulpke/CGT/education.html where you will ﬁnd ver
sions of GAP for both Windows and Macs and a guide “Abstract Algebra in GAP”. The
Sage page http://www.sagemath.org/ provides a front end for GAP and other pro
grams. I also recommend N. Carter’s “Group Explorer” http://groupexplorer.sourceforge.
net for exploring the structure of groups of small order. Earlier versions of these notes
(v3.02) described how to use Maple for computations in group theory.
ACKNOWLEDGEMENTS
I thank the following for providing corrections and comments for earlier versions of these
notes: Dustin Clausen, Benoˆıt Claudon, Keith Conrad, Demetres Christoﬁdes, AdamGlesser,
Sylvan Jacques, Martin Klazar, Mark Meckes, Victor Petrov, Efthymios Sofos, Dave Simp
son, Robert Thompson, Michiel Vermeulen.
Also, I have beneﬁted from the posts to mathoverﬂow by Richard Borcherds, Robin
Chapman, Steve Dalton, Leonid Positselski, Noah Snyder, Richard Stanley, Qiaochu Yuan,
and others (a reference mo9990 means http://mathoverflow.net/questions/9990/).
1
A family should be distinguished from a set. For example, if } is the function Z ÷Z¡3Z sending an
integer to its equivalence class, then {}(i ) [ i ÷ Z] is a set with three elements whereas (}(i ))
i÷Z
is family
with an inﬁnite index set.
5
The theory of groups of ﬁnite order may be said to date from the time of Cauchy. To
him are due the ﬁrst attempts at classiﬁcation with a view to forming a theory from a
number of isolated facts. Galois introduced into the theory the exceedingly important
idea of a [normal] subgroup, and the corresponding division of groups into simple
and composite. Moreover, by shewing that to every equation of ﬁnite degree there
corresponds a group of ﬁnite order on which all the properties of the equation depend,
Galois indicated how far reaching the applications of the theory might be, and thereby
contributed greatly, if indirectly, to its subsequent developement.
Many additions were made, mainly by French mathematicians, during the middle
part of the [nineteenth] century. The ﬁrst connected exposition of the theory was given
in the third edition of M. Serret’s “Cours d’Alg` ebre Sup´ erieure,” which was published
in 1866. This was followed in 1870 by M. Jordan’s “Trait´ e des substitutions et des
´ equations alg´ ebriques.” The greater part of M. Jordan’s treatise is devoted to a devel
opement of the ideas of Galois and to their application to the theory of equations.
No considerable progress in the theory, as apart from its applications, was made
till the appearance in 1872 of Herr Sylow’s memoir “Th´ eor` emes sur les groupes de
substitutions” in the ﬁfth volume of the Mathematische Annalen. Since the date of this
memoir, but more especially in recent years, the theory has advanced continuously.
W. Burnside, Theory of Groups of Finite Order, 1897.
Galois introduced the concept of a normal subgroup in 1832, and Camille Jordan in the
preface to his Trait´ e. . . in 1870 ﬂagged Galois’ distinction between groupes simples
and groupes compos´ ees as the most important dichotomy in the theory of permutation
groups. Moreover, in the Trait´ e, Jordan began building a database of ﬁnite simple
groups — the alternating groups of degree at least 5 and most of the classical pro
jective linear groups over ﬁelds of prime cardinality. Finally, in 1872, Ludwig Sylow
published his famous theorems on subgroups of prime power order.
R. Solomon, Bull. Amer. Math. Soc., 2001.
Why are the ﬁnite simple groups classiﬁable?
It is unlikely that there is any easy reason why a classiﬁcation is possible, unless some
one comes up with a completely new way to classify groups. One problem, at least
with the current methods of classiﬁcation via centralizers of involutions, is that every
simple group has to be tested to see if it leads to new simple groups containing it in
the centralizer of an involution. For example, when the baby monster was discovered,
it had a double cover, which was a potential centralizer of an involution in a larger
simple group, which turned out to be the monster. The monster happens to have no
double cover so the process stopped there, but without checking every ﬁnite simple
group there seems no obvious reason why one cannot have an inﬁnite chain of larger
and larger sporadic groups, each of which has a double cover that is a centralizer of
an involution in the next one. Because of this problem (among others), it was unclear
until quite late in the classiﬁcation whether there would be a ﬁnite or inﬁnite number
of sporadic groups.
Richard Borcherds, mo38161.
CHAPTER 1
Basic Deﬁnitions and Results
The axioms for a group are short and natural. . . .
Yet somehow hidden behind these axioms is the
monster simple group, a huge and extraordinary
mathematical object, which appears to rely on nu
merous bizarre coincidences to exist. The axioms
for groups give no obvious hint that anything like
this exists.
Richard Borcherds, in Mathematicians 2009.
Deﬁnitions and examples
DEFINITION 1.1 A group is a set G together with a binary operation
(a. b) ÷a+b: G×G ÷G
satisfying the following conditions:
G1: (associativity) for all a. b. c ÷ G,
(a+b) +c =a+(b +c):
G2: (existence of a neutral element) there exists an element e ÷ G such that
a+e =a =e +a (1)
for all a ÷ G;
G3: (existence of inverses) for each a ÷ G, there exists an a
t
÷ G such that
a+a
t
=e =a
t
+a.
We usually abbreviate (G. +) to G. Also, we usually write ab for a + b and 1 for e; al
ternatively, we write a ÷b for a +b and 0 for e. In the ﬁrst case, the group is said to be
multiplicative, and in the second, it is said to be additive.
1.2 In the following, a. b. . . . are elements of a group G.
7
8 1. BASIC DEFINITIONS AND RESULTS
(a) An element e satisfying (1) is called a neutral element. If e
t
is a second such element,
then e
t
=e +e
t
=e. In fact, e is the unique element of G such that e +e =e (apply
G3).
(b) If b +a =e and a+c =e, then
b =b +e =b +(a+c) =(b +a) +c =e +c =c.
Hence the element a
t
in (G3) is uniquely determined by a. We call it the inverse of
a, and denote it a
1
(or the negative of a, and denote it ÷a).
(c) Note that (G1) shows that the product of any ordered triple a
1
, a
2
, a
3
of elements of
G is unambiguously deﬁned: whether we form a
1
a
2
ﬁrst and then (a
1
a
2
)a
3
, or a
2
a
3
ﬁrst and then a
1
(a
2
a
3
), the result is the same. In fact, (G1) implies that the product
of any ordered ntuple a
1
, a
2
,. . . , a
n
of elements of G is unambiguously deﬁned. We
prove this by induction on n. In one multiplication, we might end up with
(a
1
a
i
)(a
i÷1
a
n
) (2)
as the ﬁnal product, whereas in another we might end up with
(a
1
a
}
)(a
}÷1
a
n
). (3)
Note that the expression within each pair of parentheses is well deﬁned because of
the induction hypotheses. Thus, if i =; , (2) equals (3). If i =; , we may suppose
i < ; . Then
(a
1
a
i
)(a
i÷1
a
n
) =(a
1
a
i
)
(a
i÷1
a
}
)(a
}÷1
a
n
)
(a
1
a
}
)(a
}÷1
a
n
) =
(a
1
a
i
)(a
i÷1
a
}
)
(a
}÷1
a
n
)
and the expressions on the right are equal because of (G1).
(d) The inverse of a
1
a
2
a
n
is a
1
n
a
1
n1
a
1
1
, i.e., the inverse of a product is the
product of the inverses in the reverse order.
(e) (G3) implies that the cancellation laws hold in groups,
ab =ac == b =c. ba =ca == b =c
(multiply on left or right by a
1
). Conversely, if G is ﬁnite, then the cancellation laws
imply (G3): the map . ÷a.: G ÷G is injective, and hence (by counting) bijective;
in particular, e is in the image, and so a has a right inverse; similarly, it has a left
inverse, and the argument in (b) above shows that the two inverses are equal.
Two groups (G. +) and (G
t
. +
t
) are isomorphic if there exists a onetoone correspon
dence a a
t
, G G
t
, such that (a+b)
t
=a
t
+
t
b
t
for all a. b ÷ G.
The order [G[ of a group G is its cardinality. A ﬁnite group whose order is a power of
a prime ] is called a ]group.
For an element a of a group G, deﬁne
a
n
=
¸
¸
aa a n > 0 (n copies of a)
e n =0
a
1
a
1
a
1
n < 0 ([n[ copies of a
1
)
Deﬁnitions and examples 9
The usual rules hold:
a
n
a
n
=a
n÷n
. (a
n
)
n
=a
nn
. all m. n ÷ Z. (4)
It follows from (4) that the set
{n ÷ Z [ a
n
=e]
is an ideal in Z, and so equals mZ for some integer m _ 0. When m = 0, a
n
= e unless
n =0, and a is said to have inﬁnite order. When m =0, it is the smallest integer m > 0
such that a
n
=e, and a is said to have ﬁnite order m. In this case, a
1
=a
n1
, and
a
n
=e ” m[n.
EXAMPLES
1.3 Let C
o
be the group (Z. ÷), and, for an integer m_1, let C
n
be the group (Z¡mZ. ÷).
1.4 Permutation groups. Let S be a set and let Sym(S) be the set of bijections ˛: S ÷S.
We deﬁne the product of two elements of Sym(S) to be their composite:
˛ˇ =˛ ıˇ.
For any ˛. ˇ. , ÷ Sym(S) and s ÷ S,
((˛ ıˇ) ı,)(s) =(˛ ıˇ)(,(s)) =˛(ˇ(,(s))) =(˛ ı(ˇı,))(s). (5)
and so associativity holds. The identity map s ÷ s is an identity element for Sym(S),
and inverses exist because we required the elements of Sym(S) to be bijections. Therefore
Sym(S) is a group, called the group of symmetries of S. For example, the permutation
group on n letters S
n
is deﬁned to be the group of symmetries of the set {1. .... n] — it has
order nŠ.
1.5 When G and H are groups, we can construct a new group G×H, called the (direct)
product of G and H. As a set, it is the cartesian product of G and H, and multiplication is
deﬁned by
(g. h)(g
t
. h
t
) =(gg
t
. hh
t
).
1.6 A group G is commutative (or abelian)
1
if
ab =ba. all a. b ÷ G.
In a commutative group, the product of any ﬁnite (not necessarily ordered) family S of ele
ments is well deﬁned, for example, the empty product is e. Usually, we write commutative
groups additively. With this notation, Equation (4) becomes:
ma÷na =(m÷n)a. m(na) =mna.
When G is commutative,
m(a÷b) =ma÷mb for m÷ Z and a. b ÷ G,
1
“Abelian group” is more common than “commutative group”, but I prefer to use descriptive names where
possible.
10 1. BASIC DEFINITIONS AND RESULTS
and so the map
(m. a) ÷ma: Z×G ÷G
makes · into a Zmodule. In a commutative group G, the elements of ﬁnite order form a
subgroup G
tors
of G, called the torsion subgroup.
1.7 Let J be a ﬁeld. The n×n matrices with coefﬁcients in J and nonzero determinant
form a group GL
n
(J) called the general linear group of degree n. For a ﬁnite dimen
sional Jvector space V , the Jlinear automorphisms of V form a group GL(V ) called the
general linear group of V . Note that if V has dimension n, then the choice of a basis de
termines an isomorphism GL(V ) ÷GL
n
(J) sending an automorphism to its matrix with
respect to the basis.
1.8 Let V be a ﬁnite dimensional vector space over a ﬁeld J. A bilinear form on V is a
mapping ç: V ×V ÷J that is linear in each variable. An automorphism of such a ç is an
isomorphism ˛: V ÷V such that
ç(˛·. ˛n) =ç(·. n) for all ·. n ÷ V. (6)
The automorphisms of ç form a group Aut(ç). Let e
1
. . . . . e
n
be a basis for V , and let
1 =(ç(e
i
. e
}
))
1_i,}_n
be the matrix of ç. The choice of the basis identiﬁes Aut(ç) with the group of invertible
matrices · such that
2
·
T
1 · =1. (7)
When ç is symmetric, i.e.,
ç(·. n) =ç(n. ·) all ·. n ÷ V.
and nondegenerate, Aut(ç) is called the orthogonal group of ç.
When ç is skewsymmetric, i.e.,
ç(·. n) =÷ç(n. ·) all ·. n ÷ V.
and nondegenerate, Aut(ç) is called the symplectic group of ç. In this case, there exists a
basis for V for which the matrix of ç is
J
2n
=
0 1
n
÷1
n
0
. 2m=n.
2
When we use the basis to identify V with J
n
, the pairing ç becomes
o
1
.
.
.
o
n
.
¸
b
1
.
.
.
b
n
÷(a
1
. . . . . a
n
) 1
¸
b
1
.
.
.
b
n
.
If · is the matrix of ˛ with respect to the basis, then ˛ corresponds to the map
o
1
.
.
.
o
n
÷·
o
1
.
.
.
o
n
.Therefore,
(6) becomes the statement that
(a
1
. . . . . a
n
) ·
T
1 ·
¸
b
1
.
.
.
b
n
=(a
1
. . . . . a
n
) 1
¸
b
1
.
.
.
b
n
for all
o
1
.
.
.
o
n
.
¸
b
1
.
.
.
b
n
÷ J
n
.
On examining this statement on the standard basis vectors for J
n
, we see that it is equivalent to (7).
Multiplication tables 11
and the group of invertible matrices · such that
·
T
J
2n
· =J
2n
is called the symplectic group Sp
2n
.
REMARK 1.9 A set S together with a binary operation (a. b) ÷a b: S ×S ÷S is called a
magma. When the binary operation is associative, (S. ) is called a semigroup. The product
¸
·
def
=a
1
a
n
of any sequence · =(a
i
)
1_i_n
of elements in a semigroup S is welldeﬁned (see 1.2(c)),
and for any pair · and T of such sequences,
(
¸
·)(
¸
T) =
¸
(·LT). (8)
Let 0 be the empty sequence, i.e., the sequence of elements in S indexed by the empty set.
What should
¸
0 be? Clearly, we should have
(
¸
0)(
¸
·) =
¸
(0L·) =
¸
· =
¸
(·L0) =(
¸
·)(
¸
0).
In other words,
¸
0 should be a neutral element. A semigroup with a neutral element
is called a monoid. In a monoid, the product of any ﬁnite (possibly empty) sequence of
elements is welldeﬁned, and (8) holds.
ASIDE 1.10 (a) The group conditions (G2,G3) can be replaced by the following weaker conditions
(existence of a left neutral element and left inverses): (G2
t
) there exists an e such that e +a =a for
all a; (G3
t
) for each a ÷ G, there exists an a
t
÷ G such that a
t
+a =e. To see that these imply (G2)
and (G3), let a ÷ G, and apply (G3
t
) to ﬁnd a
t
and a
tt
such that a
t
+a =e and a
tt
+a
t
=e. Then
a+a
t
=e +(a+a
t
) =(a
tt
+a
t
) +(a+a
t
) =a
tt
+
(a
t
+a) +a
t
=a
tt
+a
t
=e.
whence (G3), and
a =e +a =(a+a
t
) +a =a+(a
t
+a) =a+e.
whence (G2).
(b) A group can be deﬁned to be a set G with a binary operation + satisfying the following
conditions: (g1) + is associative; (g2) G is nonempty; (g3) for each a ÷ G, there exists an a
t
÷ G
such that a
t
+a is neutral. As there is at most one neutral element in a set with an associative binary
operation, these conditions obviously imply those in (a). They are minimal in the sense that there
exist sets with a binary operation satisfying any two of them but not the third. For example, (N. ÷)
satisﬁes (g1) and (g2) but not (g3); the empty set satisﬁes (g1) and (g3) but not (g2); the set of 2×2
matrices with coefﬁcents in a ﬁeld and with ·+T =·T ÷T· satisﬁes (g2) and (g3) but not (g1).
Multiplication tables
A binary operation on a ﬁnite set can be described by its multiplication table:
e a b c . . .
e ee ea eb ec . . .
a ae a
2
ab ac . . .
b be ba b
2
bc . . .
c ce ca cb c
2
. . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
12 1. BASIC DEFINITIONS AND RESULTS
The element e is an identity element if and only if the ﬁrst row and column of the table
simply repeat the elements. Inverses exist if and only if each element occurs exactly once
in each row and in each column (see 1.2e). If there are n elements, then verifying the
associativity law requires checking n
3
equalities.
For the multiplication table of S
3
, see the front page. Note that each colour occurs
exactly once in each row and and each column.
This suggests an algorithm for ﬁnding all groups of a given ﬁnite order n, namely, list
all possible multiplication tables and check the axioms. Except for very small n, this is not
practical! The table has n
2
positions, and if we allow each position to hold any of the n
elements, then that gives a total of n
n
2
possible tables very few of which deﬁne groups. For
example, there are 8
64
=6277101735386680763835789423207666416102355444464
034512896 binary operations on a set with 8 elements, but only ﬁve isomorphism classes
of groups of order 8 (see 4.21).
Subgroups
PROPOSITION 1.11 Let S be a nonempty subset of a group G. If
S1: a. b ÷ S == ab ÷ S.and
S2: a ÷ S == a
1
÷ S.
then the binary operation on G makes S into a group.
PROOF. (S1) implies that the binary operation on G deﬁnes a binary operation S ×S ÷S
on S, which is automatically associative. By assumption S contains at least one element a,
its inverse a
1
, and the product e =aa
1
. Finally (S2) shows that the inverses of elements
in S lie in S.
2
A nonempty subset S satisfying (S1) and (S2) is called a subgroup of G. When S is
ﬁnite, condition (S1) implies (S2): let a ÷ S; then {a. a
2
. . . .] cS, and so a has ﬁnite order,
say a
n
=e; now a
1
=a
n1
÷ S. The example (N. ÷) c(Z. ÷) shows that (S1) does not
imply (S2) when S is inﬁnite.
EXAMPLE 1.12 The centre of a group G is the subset
7(G) ={g ÷ G [ g. =.g for all . ÷ G].
It is a subgroup of G.
PROPOSITION 1.13 An intersection of subgroups of G is a subgroup of G.
PROOF. It is nonempty because it contains e, and (S1) and (S2) obviously hold.
2
REMARK 1.14 It is generally true that an intersection of subobjects of an algebraic object
is a subobject. For example, an intersection of subrings of a ring is a subring, an intersection
of submodules of a module is a submodule, and so on.
PROPOSITION 1.15 For any subset X of a group G, there is a smallest subgroup of G con
taining X. It consists of all ﬁnite products of elements of X and their inverses (repetitions
allowed).
Subgroups 13
PROOF. The intersection S of all subgroups of G containing X is again a subgroup con
taining X, and it is evidently the smallest such group. Clearly S contains with X, all ﬁnite
products of elements of X and their inverses. But the set of such products satisﬁes (S1) and
(S2) and hence is a subgroup containing X. It therefore equals S.
2
The subgroup S given by the proposition is denoted 'X), and is called the subgroup
generated by X. For example, '0) = {e]. If every element of X has ﬁnite order, for
example, if G is ﬁnite, then the set of all ﬁnite products of elements of X is already a group
and so equals 'X).
We say that X generates G if G ='X), i.e., if every element of G can be written as a
ﬁnite product of elements from X and their inverses. Note that the order of an element a of
a group is the order of the subgroup 'a) it generates.
EXAMPLES
1.16 The cyclic groups. A group is said to be cyclic if it is generated by a single element,
i.e., if G ='r) for some r ÷ G. If r has ﬁnite order n, then
G ={e. r. r
2
. .... r
n1
] ~C
n
. r
i
i mod n.
and G can be thought of as the group of rotational symmetries about the centre of a regular
polygon with nsides. If r has inﬁnite order, then
G ={. . . . r
i
. . . . . r
1
. e. r. . . . . r
i
. . . .] ~C
o
. r
i
i.
Thus, up to isomorphism, there is exactly one cyclic group of order n for each n _ o. In
future, we shall loosely use C
n
to denote any cyclic group of order n (not necessarily Z¡nZ
or Z).
1.17 The dihedral groups D
n
.
3
For n _ 3, D
n
is the group of symmetries of a regular
polygon with nsides.
4
Number the vertices 1. . . . . n in the counterclockwise direction. Let
r be the rotation through 2¬¡n about the centre of polygon (so i ÷i ÷1 mod n), and let
s be the reﬂection in the line (= rotation about the line) through the vertex 1 and the centre
of the polygon (so i ÷n÷2÷i mod n). For example, the pictures
v
1
2 3
r
s
s =
1 1
2 3
r =1 ÷2 ÷3 ÷1
v
1
2
3
4
r
s
s =
¸
¸
1 1
2 4
3 3
r =1 ÷2 ÷3 ÷4 ÷1
3
This group is denoted D
n
or D
2n
depending on whether the author is viewing it concretely (as the
symmetries of an npolygon) or abstractly.
4
More formally, D
n
can be deﬁned to be the subgroup of S
n
generated by r: i ÷ i ÷1 (mod n) and
s: i ÷n÷2÷i (mod n). Then all the statements concerning D
n
can proved without appealing to geometry (or
illuminating the reader).
14 1. BASIC DEFINITIONS AND RESULTS
illustrate the groups D
3
and D
4
. In the general case
r
n
=e: s
2
=e: srs =r
1
(so sr =r
n1
s).
These equalites imply that
D
n
={e. r. .... r
n1
. s. rs. .... r
n1
s].
and it is clear from the geometry that the elements of the set are distinct, and so [D
n
[ =2n.
Let t be the reﬂection in the line through the midpoint of the side joining the vertices
1 and 2 and the centre of the polygon (so i ÷n ÷3 ÷i mod n). Then r = t s. Hence
D
n
='s. t ) and
s
2
=e. t
2
=e. (t s)
n
=e =(st )
n
.
We deﬁne D
1
to be C
2
= {1. r] and D
2
to be C
2
×C
2
= {1. r. s. rs]. The group D
2
is also called the Klein Viergruppe or, more simply, the 4group. Note that D
3
is the full
group of permutations of {1. 2. 3]. It is the smallest noncommutative group.
1.18 The quaternion group O: Let a =
0
1
1 0
and b =
0 1
1 0
. Then
a
4
=e. a
2
=b
2
. bab
1
=a
3
(so ba =a
3
b).
The subgroup of GL
2
(C) generated by a and b is
O={e. a. a
2
. a
3
. b. ab. a
2
b. a
3
b].
The group O can also be described as the subset {˙1. ˙i. ˙;. ˙k] of the quaternion alge
bra H. Recall that
H=11˚1i ˚1; ˚1k
with the multiplication determined by
i
2
=÷1 =;
2
. i; =k =÷;i.
The map i ÷a, ; ÷b extends uniquely to a homomorphism H÷M
2
(C) of 1algebras,
which maps the group 'i. ; ) isomorphically onto 'a. b).
1.19 Recall that S
n
is the permutation group on {1. 2. .... n]. A transposition is a permu
tation that interchanges two elements and leaves all other elements unchanged. It is not
difﬁcult to see that S
n
is generated by transpositions (see (4.26) below for a more precise
statement).
Groups of small order
[For] n =6, there are three groups, a group C
6
,
and two groups C
2
×C
3
and S
3
.
Cayley, American J. Math. 1 (1878), p. 51.
For each prime ], there is only one group of order ], namely C
;
(see 1.28 below). In
the following table, c ÷n =t means that there are c commutative groups and n noncom
mutative groups (up to isomorphism, of course).
Homomorphisms 15
[G[ c ÷n =t Groups Ref.
4 2÷0 =2 C
4
, C
2
×C
2
4.18
6 1÷1 =2 C
6
; S
3
4.23
8 3÷2 =5 C
S
, C
2
×C
4
, C
2
×C
2
×C
2
; O, D
4
4.21
9 2÷0 =2 C
9
, C
3
×C
3
4.18
10 1÷1 =2 C
10
; D
5
5.14
12 2÷3 =5 C
12
, C
2
×C
6
; C
2
×S
3
, ·
4
, C
4
C
3
5.16
14 1÷1 =2 C
14
; D
T
5.14
15 1÷0 =1 C
15
5.14
16 5÷9 =14 See Wild 2005
18 2÷3 =5 C
1S
, C
3
×C
6
; D
9
. S
3
×C
3
, (C
3
×C
3
) C
2
20 2÷3 =5 C
20
,C
2
×C
10
;D
10
,C
5
C
4
,'a. b [ a
4
=b
5
=1. ba =ab
2
)
21 1÷1 =2 C
21
; 'a. b [ a
3
=b
T
=1, ba =ab
2
)
22 1÷1 =2 C
22
; D
11
5.14
24 3÷12=15 See opensourcemath.org/gap/small groups.html
Here 'a. b [ a
4
= b
5
= 1. ba = ab
2
) is the group with generators a and b and relations
a
4
=b
5
=1 and ba =ab
2
(see Chapter 2).
Roughly speaking, the more high powers of primes divide n, the more groups of order
n there should be. In fact, if }(n) is the number of isomorphism classes of groups of order
n, then
}(n) _n
(
2
27
÷o(1))e(n)
2
where e(n) is the largest exponent of a prime dividing n and o(1) ÷0 as e(n) ÷o (see
Pyber 1993).
By 2001, a complete irredundant list of groups of order _ 2000 had been found — up
to isomorphism, there are exactly 49,910,529,484 (Besche et al. 2001).
Homomorphisms
DEFINITION 1.20 A homomorphism from a group G to a second G
t
is a map ˛: G ÷G
t
such that ˛(ab) = ˛(a)˛(b) for all a. b ÷ G. An isomorphism is a bijective homomor
phism.
For example, the determinant map det: GL
n
(J) ÷J
×
is a homomorphism.
1.21 Let ˛ be a homomorphism. For any elements a
1
. . . . . a
n
of G,
˛(a
1
a
n
) =˛(a
1
(a
2
a
n
))
=˛(a
1
)˛(a
2
a
n
)
=˛(a
1
) ˛(a
n
),
16 1. BASIC DEFINITIONS AND RESULTS
and so homomorphisms preserve all products. In particular, for m_1,
˛(a
n
) =˛(a)
n
. (9)
Moreover ˛(e) =˛(ee) =˛(e)˛(e), and so ˛(e) =e (apply 1.2a). Also
aa
1
=e =a
1
a == ˛(a)˛(a
1
) =e =˛(a
1
)˛(a).
and so ˛(a
1
) =˛(a)
1
. It follows that (9) holds for all m÷ Z, and so a homomorphism
of commutative groups is also a homomorphism of Zmodules.
As we noted above, each row of the multiplication table of a group is a permutation of
the elements of the group. As Cayley pointed out, this allows one to realize the group as a
group of permutations.
THEOREM 1.22 (CAYLEY) There is a canonical injective homomorphism
˛: G ÷Sym(G).
PROOF. For a ÷ G, deﬁne a
1
: G ÷G to be the map . ÷a. (left multiplication by a).
For . ÷ G,
(a
1
ıb
1
)(.) =a
1
(b
1
(.)) =a
1
(b.) =ab. =(ab)
1
(.).
and so (ab)
1
=a
1
ıb
1
. As e
1
=id, this implies that
a
1
ı(a
1
)
1
=id =(a
1
)
1
ıa
1
.
and so a
1
is a bijection, i.e., a
1
÷ Sym(G). Hence a ÷a
1
is a homomorphism G ÷
Sym(G), and it is injective because of the cancellation law.
2
COROLLARY 1.23 A ﬁnite group of order n can be realized as a subgroup of S
n
.
PROOF. List the elements of the group as a
1
. . . . . a
n
.
2
Unfortunately, unless n is small, S
n
is too large to be manageable. We shall see later
(4.22) that G can often be embedded in a permutation group of much smaller order than nŠ.
Cosets
For a subset S of a group G and an element a of G, we let
aS ={as [ s ÷ S]
Sa ={sa [ s ÷ S].
Because of the associativity law, a(bS) =(ab)S , and so we can denote this set unambigu
ously by abS.
When H is a subgroup of G, the sets of the form aH are called the left cosets of H
in G, and the sets of the form Ha are called the right cosets of H in G. Because e ÷ H,
aH =H if and only if a ÷ H.
Cosets 17
EXAMPLE 1.24 Let G =(1
2
. ÷), and let H be a subspace of dimension 1 (line through
the origin). Then the cosets (left or right) of H are the lines a÷H parallel to H.
PROPOSITION 1.25 Let H be a subgroup of a group G.
(a) An element a of G lies in a left coset C of H if and only if C =aH.
(b) Two left cosets are either disjoint or equal.
(c) aH =bH if and only if a
1
b ÷ H.
(d) Any two left cosets have the same number of elements (possibly inﬁnite).
PROOF. (a) Certainly a ÷ aH. Conversely, if a lies in the left coset bH, then a =bh for
some h, and so
aH =bhH =bH.
(b) If C and C
t
are not disjoint, then they have a common element a, and C =aH and
C
t
=aH by (a).
(c) If a
1
b ÷ H, then H = a
1
bH, and so aH = aa
1
bH = bH. Conversely, if
aH =bH, then H =a
1
bH, and so a
1
b ÷ H.
(d) The map (ba
1
)
1
: ah ÷bh is a bijection aH ÷bH.
2
The index (G : H) of H in G is deﬁned to be the number of left cosets of H in G.
5
For
example, (G : 1) is the order of G.
As the left cosets of H in G cover G, (1.25b) shows that they form a partition G. In
other words, the condition “a and b lie in the same left coset” is an equivalence relation on
G.
THEOREM 1.26 (LAGRANGE) If G is ﬁnite, then
(G : 1) =(G : H)(H : 1).
In particular, the order of every subgroup of a ﬁnite group divides the order of the group.
PROOF. The left cosets of H in G form a partition of G, there are (G : H) of them, and
each left coset has (H : 1) elements.
2
COROLLARY 1.27 The order of each element of a ﬁnite group divides the order of the
group.
PROOF. Apply Lagrange’s theorem to H ='g), recalling that (H : 1) =order(g).
2
EXAMPLE 1.28 If G has order ], a prime, then every element of G has order 1 or ]. But
only e has order 1, and so G is generated by any element a =e. In particular, G is cyclic
and so G ~C
;
. This shows, for example, that, up to isomorphism, there is only one group
of order 1. 000. 000. 007 (because this number is prime). In fact there are only two groups
of order 1. 000. 000. 014. 000. 000. 049 (see 4.18).
5
More formally, (G : H) is the cardinality of the set {aH [ a ÷ G].
18 1. BASIC DEFINITIONS AND RESULTS
1.29 For a subset S of G, let S
1
={g
1
[ g ÷ S]. Then (aH)
1
is the right coset Ha
1
,
and (Ha)
1
=a
1
H. Therefore S ÷S
1
deﬁnes a onetoone correspondence between
the set of left cosets and the set of right cosets under which aH Ha
1
. Hence (G : H)
is also the number of right cosets of H in G. But, in general, a left coset will not be a right
coset (see 1.34 below).
1.30 Lagrange’s theorem has a partial converse: if a prime ] divides m=(G : 1), then G
has an element of order ] (Cauchy’s theorem 4.13); if a prime power ]
n
divides m, then G
has a subgroup of order ]
n
(Sylow’s theorem 5.2). However, note that the 4group C
2
×C
2
has order 4, but has no element of order 4, and ·
4
has order 12, but has no subgroup of
order 6 (see Exercise 414).
More generally, we have the following result.
PROPOSITION 1.31 For any subgroups H ¯1 of G,
(G : 1) =(G : H)(H : 1)
(meaning either both are inﬁnite or both are ﬁnite and equal).
PROOF. Write G =
¸
i÷J
g
i
H (disjoint union), and H =
¸
}÷J
h
}
1 (disjoint union). On
multiplying the second equality by g
i
, we ﬁnd that g
i
H =
¸
}÷J
g
i
h
}
1 (disjoint union),
and so G =
¸
i,}÷J×J
g
i
h
}
1 (disjoint union). This shows that
(G : 1) =[1[[J[ =(G : H)(H : 1).
2
Normal subgroups
When S and T are two subsets of a group G, we let
ST ={st [ s ÷ S, t ÷ T ].
Because of the associativity law, 1(ST ) = (1S)T , and so we can denote this set unam
biguously as 1ST .
A subgroup N of G is normal, denoted N <G, if gNg
1
=N for all g ÷ G.
REMARK 1.32 To show that N is normal, it sufﬁces to check that gNg
1
c N for all g,
because multiplying this inclusion on the left and right with g
1
and g respectively gives
the inclusion N cg
1
Ng, and rewriting this with g
1
for g gives that N cgNg
1
for all
g. However, the next example shows that there can exist a subgroup N of a group G and
an element g of G such that gNg
1
cN but gNg
1
=N.
EXAMPLE 1.33 Let G =GL
2
(Q), and let H =
˚
1 n
0 1
ˇ
ˇ
n ÷ Z
¸
. Then H is a subgroup of
G; in fact H .Z. Let g =
5 0
0 1
. Then
g
1 n
0 1
g
1
=
5 0
0 1
1 n
0 1
5
1
0
0 1
=
1 5n
0 1
.
Hence gHg
1
´H (and g
1
Hg ,cH).
Normal subgroups 19
PROPOSITION 1.34 A subgroup N of G is normal if and only if every left coset of N in
G is also a right coset, in which case, gN =Ng for all g ÷ G.
PROOF. Clearly,
gNg
1
=N ” gN =Ng.
Thus, if N is normal, then every left coset is a right coset (in fact, gN =Ng). Conversely,
if the left coset gN is also a right coset, then it must be the right coset Ng by (1.25a). Hence
gN =Ng, and so gNg
1
=N.
2
1.35 The proposition says that, in order for N to be normal, we must have that for all
g ÷ G and n ÷ N, there exists an n
t
÷ N such that gn = n
t
g (equivalently, for all g ÷ G
and n ÷ N, there exists an n
t
such that ng =gn
t
). In other words, to say that N is normal
amounts to saying that an element of G can be moved past an element of N at the cost of
replacing the element of N by another element of N.
EXAMPLE 1.36 (a) Every subgroup of index two is normal. Indeed, let g ÷ GH. Then
G =H LgH (disjoint union). Hence gH is the complement of H in G. Similarly, Hg is
the complement of H in G, and so gH =Hg.
(b) Consider the dihedral group
D
n
={e. r. . . . . r
n1
. s. . . . . r
n1
s].
Then C
n
={e. r. . . . . r
n1
] has index 2, and hence is normal. For n _3 the subgroup {e. s]
is not normal because r
1
sr =r
n2
s … {e. s].
(c) Every subgroup of a commutative group is normal (obviously), but the converse
is false: the quaternion group O is not commutative, but every subgroup is normal (see
Exercise 11).
A group G is said to be simple if it has no normal subgroups other than G and {e]. Such
a group can still have lots of nonnormal subgroups — in fact, the Sylow theorems (Chapter
5) imply that every ﬁnite group has nontrivial subgroups unless it is cyclic of prime order.
PROPOSITION 1.37 If H and N are subgroups of G and N is normal, then HN is a
subgroup of G. If H is also normal, then HN is a normal subgroup of G.
PROOF. The set HN is nonempty, and
(h
1
n
1
)(h
2
n
2
)
1.35
= h
1
h
2
n
t
1
n
2
÷ HN.
and so it is closed under multiplication. Since
(hn)
1
=n
1
h
1
1.35
= h
1
n
t
÷ HN
it is also closed under the formation of inverses, and so HN is a subgroup. If both H and
N are normal, then
gHNg
1
=gHg
1
gNg
1
=HN
for all g ÷ G.
2
20 1. BASIC DEFINITIONS AND RESULTS
An intersection of normal subgroups of a group is again a normal subgroup (cf. 1.14).
Therefore, we can deﬁne the normal subgroup generated by a subset X of a group G to
be the intersection of the normal subgroups containing X. Its description in terms of X
is a little complicated. We say that a subset X of a group G is normal or closed under
conjugation if gXg
1
cX for all g ÷ G.
LEMMA 1.38 If X is normal, then the subgroup 'X) generated by it is normal.
PROOF. The map “conjugation by g”, a ÷gag
1
, is a homomorphismG÷G. If a ÷'X),
say, a =.
1
.
n
with each .
i
or its inverse in X, then
gag
1
=(g.
1
g
1
) (g.
n
g
1
).
As X is closed under conjugation, each g.
i
g
1
or its inverse lies in X, and so g'X)g
1
c
'X).
2
LEMMA 1.39 For any subset X of G, the subset
v÷G
gXg
1
is normal, and it is the
smallest normal set containing X.
PROOF. Obvious.
2
On combining these lemmas, we obtain the following proposition.
PROPOSITION 1.40 The normal subgroup generated by a subset X of G is '
v÷G
gXg
1
).
Kernels and quotients
The kernel of a homomorphism ˛: G ÷G
t
is
Ker(˛) ={g ÷ G[ ˛(g) =e].
If ˛ is injective, then Ker(˛) ={e]. Conversely, if Ker(˛) ={e], then ˛ is injective, because
˛(g) =˛(g
t
) == ˛(g
1
g
t
) =e == g
1
g
t
=e == g =g
t
.
PROPOSITION 1.41 The kernel of a homomorphism is a normal subgroup.
PROOF. It is obviously a subgroup, and if a ÷ Ker(˛), so that ˛(a) =e, and g ÷ G, then
˛(gag
1
) =˛(g)˛(a)˛(g)
1
=˛(g)˛(g)
1
=e.
Hence gag
1
÷ Ker (˛).
2
For example, the kernel of the homomorphism det: GL
n
(J) ÷J
×
is the group of n×n
matrices with determinant 1 — this group SL
n
(J) is called the special linear group of
degree n.
PROPOSITION 1.42 Every normal subgroup occurs as the kernel of a homomorphism.
More precisely, if N is a normal subgroup of G, then there is a unique group structure
on the set G¡N of cosets of N in G for which the natural map a ÷Œa: G ÷G¡N is a
homomorphism.
Theorems concerning homomorphisms 21
PROOF. Write the cosets as left cosets, and deﬁne (aN)(bN) =(ab)N. We have to check
(a) that this is welldeﬁned, and (b) that it gives a group structure on the set of cosets. It
will then be obvious that the map g ÷gN is a homomorphism with kernel N.
(a). Let aN =a
t
N and bN =b
t
N; we have to show that abN =a
t
b
t
N. But
abN =a(bN) =a(b
t
N)
1.34
= aNb
t
=a
t
Nb
t
1.34
= a
t
b
t
N.
(b). The product is certainly associative, the coset N is an identity element, and a
1
N
is an inverse for aN.
2
The group G¡N is called the
6
quotient of G by N.
Propositions 1.41 and 1.42 show that the normal subgroups are exactly the kernels of
homomorphisms.
PROPOSITION 1.43 The map a ÷ aN: G ÷ G¡N has
the following universal property: for any homomorphism
˛: G ÷G
t
of groups such that ˛(N) = {e], there exists
a unique homomorphism G¡N ÷G
t
making the diagram
at right commute:
G G¡N
G
t
.
o o1
˛
PROOF. Note that for n ÷ N, ˛(gn) =˛(g)˛(n) =˛(g), and so ˛ is constant on each left
coset gN of N in G. It therefore deﬁnes a map
¨ ˛: G¡N ÷G
t
. ¨ ˛(gN) =˛(g).
and ¨ ˛ is a homomorphism because
¨ ˛((gN) (g
t
N)) = ¨ ˛(gg
t
N) =˛(gg
t
) =˛(g)˛(g
t
) = ¨ ˛(gN) ¨ ˛(g
t
N).
The uniqueness of ¨ ˛ follows from the surjectivity of G ÷G¡N.
2
EXAMPLE 1.44 (a) Consider the subgroup mZ of Z. The quotient group Z¡mZ is a cyclic
group of order m.
(b) Let 1 be a line through the origin in 1
2
. Then 1
2
¡1 is isomorphic to 1 (because it
is a onedimensional vector space over 1).
(c) For n _2, the quotient D
n
¡'r) ={¨ e. ¨ s] (cyclic group of order 2).
Theorems concerning homomorphisms
The theorems in this subsection are sometimes called the isomorphism theorems (ﬁrst, sec
ond, . . . , or ﬁrst, third, . . . , or . . . ).
FACTORIZATION OF HOMOMORPHISMS
Recall that the image of a map ˛: S ÷T is ˛(S) ={˛(s) [ s ÷ S].
6
Some authors say “factor” instead of “quotient”, but this can be confused with “direct factor”.
22 1. BASIC DEFINITIONS AND RESULTS
THEOREM 1.45 (HOMOMORPHISM THEOREM) For any homomorphism ˛: G ÷ G
t
of
groups, the kernel N of ˛ is a normal subgroup of G, the image 1 of ˛ is a subgroup
of G
t
, and ˛ factors in a natural way into the composite of a surjection, an isomorphism,
and an injection:
G
˛
÷÷÷÷÷ G
t
surjective
vv1
¸
injective
G¡N
v1˛(v)
÷÷÷÷÷÷÷÷
isomorphism
1.
PROOF. We have already seen (1.41) that the kernel is a normal subgroup of G. If b =˛(a)
and b
t
=˛(a
t
), then bb
t
=˛(aa
t
) and b
1
=˛(a
1
), and so 1
def
=˛(G) is a subgroup of G
t
.
The universal property of quotients (1.43) shows that the map . ÷˛(.): G ÷1 deﬁnes
a homomorphism ¨ ˛: G¡N ÷1 with ¨ ˛(gN) = ˛(g). The homomorphism ¨ ˛ is certainly
surjective, and if ¨ ˛(gN) = e, then g ÷ Ker(˛) = N, and so ¨ ˛ has trivial kernel. This
implies that it is injective (p. 20).
2
THE ISOMORPHISM THEOREM
THEOREM 1.46 (ISOMORPHISM THEOREM) Let H be a subgroup of G and N a normal
subgroup of G. Then HN is a subgroup of G, H ¨N is a normal subgroup of H, and the
map
h(H ¨N) ÷hN: H¡H ¨N ÷HN¡N
is an isomorphism.
PROOF. We have already seen (1.37) that HN is a subgroup. Consider the map
H ÷G¡N. h ÷hN.
This is a homomorphism, and its kernel is H ¨N, which is therefore normal in H. Ac
cording to Theorem 1.45, the map induces an isomorphism H¡H ¨N ÷1 where 1 is its
image. But 1 is the set of cosets of the form hN with h ÷ H, i.e., 1 =HN¡N.
2
It is not necessary to assume that N be normal in G as long as hNh
1
= N for all
h ÷ H (i.e., H is contained in the normalizer of N — see later). Then H ¨N is still
normal in H, but it need not be a normal subgroup of G.
THE CORRESPONDENCE THEOREM
The next theorem shows that if
¨
G is a quotient group of G, then the lattice of subgroups in
¨
G captures the structure of the lattice of subgroups of G lying over the kernel of G ÷
¨
G.
THEOREM 1.47 (CORRESPONDENCE THEOREM) Let ˛: G
¨
G be a surjective homo
morphism, and let N =Ker(˛). Then there is a onetoone correspondence
{subgroups of G containing N]
1:1
{subgroups of
¨
G]
under which a subgroup H of G containing N corresponds to
¨
H =˛(H) and a subgroup
¨
H of
¨
G corresponds to H =˛
1
(
¨
H). Moreover, if H 
¨
H and H
t

¨
H
t
, then
Direct products 23
(a)
¨
H c
¨
H
t
” H cH
t
, in which case (
¨
H
t
:
¨
H) =(H
t
: H);
(b)
¨
H is normal in
¨
G if and only if H is normal in G, in which case, ˛ induces an
isomorphism
G¡H
:
÷
¨
G¡
¨
H.
PROOF. If
¨
H is a subgroup of
¨
G, then ˛
1
(
¨
H) is easily seen to be a subgroup of G con
taining N, and if H is a subgroup of G, then ˛(H) is a subgroup of
¨
G (see 1.45). Clearly,
˛
1
˛(H) =HN, which equals H if and only if H ¯N, and ˛˛
1
(
¨
H) =
¨
H. Therefore,
the two operations give the required bijection. The remaining statements are easily veriﬁed.
For example, a decomposition H
t
=
¸
i÷J
a
i
H of H
t
into a disjoint union of left cosets of
H gives a similar decomposition
¨
H
t
=
¸
i÷J
a
i
¨
H of
¨
H
t
.
2
COROLLARY 1.48 Let N be a normal subgroup of G; then there is a onetoone corre
spondence between the set of subgroups of G containing N and the set of subgroups of
G¡N, H H¡N. Moreover H is normal in G if and only if H¡N is normal in G¡N, in
which case the homomorphism g ÷gN: G ÷G¡N induces an isomorphism
G¡H
:
÷(G¡N)¡(H¡N).
PROOF. This is the special case of the theorem in which ˛ is g ÷gN: G ÷G¡N.
2
EXAMPLE 1.49 Let G=D
4
and let N be its subgroup 'r
2
). Recall (1.17) that srs
1
=r
3
,
and so sr
2
s
1
=
r
3
2
= r
2
. Therefore N is normal. The groups G and G¡N have the
following lattices of subgroups:
D
4
D
4
¡'r
2
)
'r
2
. s) 'r) 'r
2
. rs) '¨ s) '¨ r) '¨ r ¨ s)
's) 'r
2
s) 'r
2
) 'rs) 'r
3
s) 1
1
Direct products
Let G be a group, and let H
1
. . . . . H
k
be subgroups of G. We say that G is a direct product
of the subgroups H
i
if the map
(h
1
. h
2
. . . . . h
k
) ÷h
1
h
2
h
k
: H
1
×H
2
× ×H
k
÷G
is an isomorphism of groups. This means that each element g of G can be written uniquely
in the form g =h
1
h
2
h
k
, h
i
÷ H
i
, and that if g =h
1
h
2
h
k
and g
t
=h
t
1
h
t
2
h
t
k
, then
gg
t
=(h
1
h
t
1
)(h
2
h
t
2
) (h
k
h
t
k
).
24 1. BASIC DEFINITIONS AND RESULTS
The following propositions give criteria for a group to be a direct product of subgroups.
PROPOSITION 1.50 A group G is a direct product of subgroups H
1
, H
2
if and only if
(a) G =H
1
H
2
,
(b) H
1
¨H
2
={e], and
(c) every element of H
1
commutes with every element of H
2
.
PROOF. If G is the direct product of H
1
and H
2
, then certainly (a) and (c) hold, and (b)
holds because, for any g ÷ H
1
¨H
2
, the element (g. g
1
) maps to e under (h
1
. h
2
) ÷h
1
h
2
and so equals (e. e).
Conversely, (c) implies that (h
1
. h
2
) ÷h
1
h
2
is a homomorphism, and (b) implies that
it is injective:
h
1
h
2
=e == h
1
=h
1
2
÷ H
1
¨H
2
={e].
Finally, (a) implies that it is surjective.
2
PROPOSITION 1.51 A group G is a direct product of subgroups H
1
, H
2
if and only if
(a) G =H
1
H
2
,
(b) H
1
¨H
2
={e], and
(c) H
1
and H
2
are both normal in G.
PROOF. Certainly, these conditions are implied by those in the previous proposition, and so
it remains to show that they imply that each element h
1
of H
1
commutes with each element
h
2
of H
2
. Two elements h
1
. h
2
of a group commute if and only if their commutator
Œh
1
. h
2

def
=(h
1
h
2
)(h
2
h
1
)
1
is e. But
(h
1
h
2
)(h
2
h
1
)
1
=h
1
h
2
h
1
1
h
1
2
=
(h
1
h
2
h
1
1
) h
1
2
h
1
h
2
h
1
1
h
1
2
.
which is in H
2
because H
2
is normal, and is in H
1
because H
1
is normal. Therefore (b)
implies h
1
and h
2
commute.
2
PROPOSITION 1.52 A group G is a direct product of subgroups H
1
. H
2
. . . . . H
k
if and
only if
(a) G =H
1
H
2
H
k
.
(b) for each ; , H
}
¨(H
1
H
}1
H
}÷1
H
k
) ={e], and
(c) each of H
1
. H
2
. . . . . H
k
is normal in G,
PROOF. The necessity of the conditions being obvious, we shall prove only the sufﬁciency.
For k = 2, we have just done this, and so we argue by induction on k. An induction
argument using (1.37) shows that H
1
H
k1
is a normal subgroup of G. The conditions
(a,b,c) hold for the subgroups H
1
. . . . . H
k1
of H
1
H
k1
, and so the induction hypothesis
shows that
(h
1
. h
2
. . . . . h
k1
) ÷h
1
h
2
h
k1
: H
1
×H
2
× ×H
k1
÷H
1
H
2
H
k1
Commutative groups 25
is an isomorphism. The pair H
1
H
k1
, H
k
satisﬁes the hypotheses of (1.51), and so
(h. h
k
) ÷hh
k
: (H
1
H
k1
) ×H
k
÷G
is also an isomorphism. The composite of these isomorphisms
H
1
× ×H
k1
×H
k
(h
1
,...,h
k
)(h
1
···h
k1
,h
k
)
÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷H
1
H
k1
×H
k
(h,h
k
)hh
k
÷÷÷÷÷÷÷÷÷÷G
sends (h
1
. h
2
. . . . . h
k
) to h
1
h
2
h
k
.
2
Commutative groups
The classiﬁcation of ﬁnitely generated commutative groups is most naturally studied as part
of the theory of modules over a principal ideal domain, but, for the sake of completeness, I
include an elementary exposition here.
Let M be a commutative group, written additively. The subgroup '.
1
. . . . . .
k
) of M
generated by the elements .
1
. . . . . .
k
consists of the sums
¸
m
i
.
i
, m
i
÷ Z. A subset
{.
1
. . . . . .
k
] of M is a basis for M if it generates M and
m
1
.
1
÷ ÷m
k
.
k
=0. m
i
÷ Z == m
i
.
i
=0 for every i :
then
M ='.
1
) ˚ ˚'.
k
).
LEMMA 1.53 Suppose that M is generated by {.
1
. . . . . .
k
] and let c
1
. . . . . c
k
be integers
such that gcd(c
1
. . . . . c
k
) =1. Then there exist generators .
1
. . . . . .
k
for M such that .
1
=
c
1
.
1
÷ ÷c
k
.
k
.
PROOF. If c
i
< 0, we change the signs of both c
i
and .
i
. This allows us to assume that all
c
i
÷ N. We argue by induction on s =c
1
÷ ÷c
k
. The lemma certainly holds if s =1,
and so we assume s > 1. Then, at least two c
i
are nonzero, say, c
1
_c
2
> 0. Now
˘ {.
1
. .
2
÷.
1
. .
3
. . . . . .
k
] generates M,
˘ gcd(c
1
÷c
2
. c
2
. c
3
. . . . . c
k
) =1, and
˘ (c
1
÷c
2
) ÷c
2
÷ ÷c
k
< s,
and so, by induction, there exist generators .
1
. . . . . .
k
for M such that
.
1
=(c
1
÷c
2
).
1
÷c
2
(.
1
÷.
2
) ÷c
3
.
3
÷ ÷c
k
.
k
=c
1
.
1
÷ ÷c
k
.
k
.
2
THEOREM 1.54 Every ﬁnitely generated commutative group M has a basis; hence it is a
ﬁnite direct sum of cyclic groups.
PROOF.
7
We argue by induction on the number of generators of M. If M can be generated
by one element, the statement is trivial, and so we may assume that it requires at least k >1
7
John Stillwell tells me that, for ﬁnite commutative groups, this is similar to the ﬁrst proof of the theorem,
given by Kronecker in 1870.
26 1. BASIC DEFINITIONS AND RESULTS
generators. Among the generating sets {.
1
. . . . . .
k
] for M with k elements there is one for
which the order of .
1
is the smallest possible. We shall show that M is then the direct sum
of '.
1
) and '.
2
. . . . . .
k
). This will complete the proof, because the induction hypothesis
provides us with a basis for the second group, which together with .
1
forms a basis for M.
If M is not the direct sum of '.
1
) and '.
2
. . . . . .
k
), then there exists a relation
m
1
.
1
÷m
2
.
2
÷ ÷m
k
.
k
=0 (10)
with m
1
.
1
=0; we may suppose that m
1
÷Nand m
1
<order(.
1
). Let J =gcd(m
1
. . . . . m
k
) >
0, and let c
i
=m
i
¡J. According to the lemma, there exists a generating set .
1
. . . . . .
k
such
that .
1
=c
1
.
1
÷ ÷c
k
.
k
. But
J.
1
=m
1
.
1
÷m
2
.
2
÷ ÷m
k
.
k
=0
and J _m
1
< order(.
1
), and so this is a contradiction.
2
COROLLARY 1.55 A ﬁnite commutative group is cyclic if, for each n > 0, it contains at
most n elements of order dividing n.
PROOF. After the Theorem 1.54, we may suppose that G = C
n
1
× ×C
n
r
with n
i
÷ N.
If n divides n
i
and n
}
with i = ; , then G has more than n elements of order dividing n.
Therefore, the hypothesis implies that the n
i
are relatively prime. Let a
i
generate the i th
factor. Then (a
1
. . . . . a
i
) has order n
1
n
i
, and so generates G.
2
EXAMPLE 1.56 Let J be a ﬁeld. The elements of order dividing n in J
×
are the roots of
the polynomial X
n
÷1. Because unique factorization holds in JŒX, there are at most n of
these, and so the corollary shows that every ﬁnite subgroup of J
×
is cyclic.
THEOREM 1.57 A nonzero ﬁnitely generated commutative group M can be expressed
M ~C
n
1
× ×C
n
s
×C
i
o
(11)
for certain integers n
1
. . . . . n
x
_2 and r _0. Moreover,
(a) r is uniquely determined by M;
(b) the n
i
can be chosen so that n
1
_2 and n
1
[n
2
. . . . . n
x1
[n
x
, and then they are uniquely
determined by M;
(c) the n
i
can be chosen to be powers of prime numbers, and then they are uniquely
determined by M.
The number r is called the rank of M. By r being uniquely determined by M, we mean
that in any two decompositions of M of the form (11), the number of copies of C
o
will be
the same (and similarly for the n
i
in (b) and (c)). The integers n
1
. . . . . n
x
in (b) are called
the invariant factors of M. Statement (c) says that M can be expressed
M ~C
;
e
1
1
× ×C
;
e
t
t
×C
i
o
, e
i
_1, (12)
for certain prime powers ]
e
i
i
(repetitions of primes allowed), and that the integers ]
e
1
1
. . . . . ]
e
t
t
are uniquely determined by M; they are called the elementary divisors of M.
Commutative groups 27
PROOF. The ﬁrst assertion is a restatement of Theorem 1.54.
(a) For a prime ] not dividing any of the J
i
,
M¡]M ~(C
o
¡]C
o
)
i
~(Z¡]Z)
i
.
and so r is the dimension of M¡]M as an F
;
vector space.
(b,c) If gcd(m. n) =1, then C
n
×C
n
contains an element of order mn, and so
C
n
×C
n
~C
nn
. (13)
Use (13) to decompose the C
n
i
into products of cyclic groups of prime power order. Once
this has been achieved, (13) can be used to combine factors to achieve a decomposition as
in (b); for example, C
n
s
=
¸
C
;
e
i
i
where the product is over the distinct primes among the
]
i
and e
i
is the highest exponent for the prime ]
i
.
In proving the uniqueness statements in (b) and (c), we can replace M with its torsion
subgroup (and so assume r =0). A prime ] will occur as one of the primes ]
i
in (12) if
and only M has an element of order ], in which case ] will occur exact a times where ]
o
is the number of elements of order dividing ]. Similarly, ]
2
will divide some ]
e
i
i
in (12)
if and only if M has an element of order ]
2
, in which case it will divide exactly b of the
]
e
i
i
where ]
ob
]
2b
is the number of elements in M of order dividing ]
2
. Continuing in
this fashion, we ﬁnd that the elementary divisors of M can be read off from knowing the
numbers of elements of M of each prime power order.
The uniqueness of the invariant factors can be derived from that of the elementary divi
sors, or it can be proved directly: n
x
is the smallest integer > 0 such that n
x
M =0; n
x1
is the smallest integer >0 such that n
x1
M is cyclic; n
x2
is the smallest integer such that
n
x2
can be expressed as a product of two cyclic groups, and so on.
2
SUMMARY 1.58 Each ﬁnite commutative group is isomorphic to exactly one of the groups
C
n
1
× ×C
n
r
. n
1
[n
2
. . . . . n
i1
[n
i
.
The order of this group is n
1
n
i
. For example, each commutative group of order 90 is
isomorphic to exactly one of C
90
or C
3
×C
30
— to see this, note that the largest invariant
factor must be a factor of 90 divisible by all the prime factors of 90.
THE LINEAR CHARACTERS OF A COMMUTATIVE GROUP
Let u(C) ={: ÷ C [ [:[ =1]. This is an inﬁnite group. For any integer n, the set u
n
(C) of
elements of order dividing n is cyclic of order n; in fact,
u
n
(C) ={e
2tin¡n
[ 0 _m_n÷1] ={1. ¯. . . . . ¯
n1
]
where ¯ =e
2ti¡n
is a primitive nth root of 1.
A linear character (or just character) of a group G is a homomorphism G ÷u(C).
The homomorphism a ÷1 is called the trivial (or principal) character.
EXAMPLE 1.59 The quadratic residue modulo ] of an integer a not divisible by ] is
deﬁned by
a
]
=
1 if a is a square in Z¡]Z
÷1 otherwise.
28 1. BASIC DEFINITIONS AND RESULTS
Clearly, this depends only on a modulo ], and if neither a nor b is divisible by ], then
ob
;
=
o
;
b
;
(because (Z¡]Z)
×
is cyclic). Therefore Œa ÷
o
;
: (Z¡]Z)
×
÷{˙1] =
u
2
(C) is a character of (Z¡]Z)
×
.
The set of characters of a group G becomes a group G
·
under the addition,
(,÷,
t
)(g) =,(g),(g
t
).
called the dual group of G. For example, the dual group Z
·
of Z is isomorphic to u(C) by
the map , ÷,(1).
THEOREM 1.60 Let G be a ﬁnite commutative group.
(a) The dual of G
·
is isomorphic to G.
(b) The map G ÷G
··
sending an element a of G to the character , ÷,(a) of G
·
is
an isomorphism.
In other words, G ~G
·
and G .G
··
.
PROOF. The statements are obvious for cyclic groups, and (G×H)
·
.G
·
×H
·
.
2
ASIDE 1.61 The statement that the natural map G ÷G
··
is an isomorphism is a special case
of the Pontryagin theorem. For inﬁnite groups, it is necessary to consider groups together with a
topology. For example, as we observed above, Z
·
. u(C). Each m ÷ Z does deﬁne a character
¯ ÷¯
n
: u(C) ÷u(C), but there are many homomorphisms u(C) ÷u(C) not of this form, and
so the dual of u(C) is larger than Z. However, these are the only continuous homomorphisms. In
general, let G be a commutative group endowed with a locally compact topology for which the
group operations are continuous; then the group G
·
of continuous characters G ÷u(C) has a
natural topology for which it is locally compact, and the Pontryagin duality theorem says that the
natural map G ÷G
··
is an isomorphism.
THEOREM 1.62 (ORTHOGONALITY RELATIONS) Let G be a ﬁnite commutative group.
For any characters , and [ of G,
¸
o÷G
,(a)[(a
1
) =
[G[ if , =[
0 otherwise.
In particular,
¸
o÷G
,(a) =
[G[ if , is trivial
0 otherwise.
PROOF. If , =[, then ,(a)[(a
1
) =1, and so the sum is [G[. Otherwise there exists a
b ÷ G such that ,(b) =[(b). As a runs over G, so also does ab, and so
¸
o÷G
,(a)[(a
1
) =
¸
o÷G
,(ab)[((ab)
1
) =,(b)[(b)
1
¸
o÷G
,(a)[(a
1
).
Because ,(b)[(b)
1
=1, this implies that
¸
o÷G
,(a)[(a
1
) =0.
2
COROLLARY 1.63 For any a ÷ G,
¸
;÷G
_
,(a) =
[G[ if a =e
0 otherwise.
PROOF. Apply the theorem to G
·
, noting that (G
·
)
·
.G.
2
The order of ab 29
The order of ab
Let a and b be elements of a group G. If a has order m and b has order n, what can we say
about the order of ab? The next theorem shows that we can say nothing at all.
THEOREM 1.64 For any integers m. n. r >1, there exists a ﬁnite group G with elements a
and b such that a has order m, b has order n, and ab has order r.
PROOF. We shall show that, for a suitable prime power q, there exist elements a and b of
SL
2
(F
q
) such that a, b, and ab have orders 2m, 2n, and 2r respectively. As ÷1 is the
unique element of order 2 in SL
2
(F
q
), the images of a, b, ab in SL
2
(F
q
)¡{˙1] will then
have orders m, n, and r as required.
Let ] be a prime number not dividing 2mnr. Then ] is a unit in the ﬁnite ring
Z¡2mnrZ, and so some power of it, q say, is 1 in the ring. This means that 2mnr di
vides q ÷1. As the group F
×
q
has order q ÷1 and is cyclic (see 1.56), there exist elements
u, ·, and n of F
×
q
having orders 2m, 2n, and 2r respectively. Let
a =
u 1
0 u
1
and b =
· 0
t ·
1
(elements of SL
2
(F
q
)).
where t has been chosen so that
u· ÷t ÷u
1
·
1
=n÷n
1
.
The characteristic polynomial of a is (X÷u)(X÷u
1
), and so a is similar to diag(u. u
1
).
Therefore a has order 2m. Similarly b has order 2n. The matrix
ab =
u· ÷t ·
1
u
1
t u
1
·
1
.
has characteristic polynomial
X
2
÷(u· ÷t ÷u
1
·
1
)X ÷1 =(X ÷n)(X ÷n
1
),
and so ab is similar to diag(n. n
1
). Therefore ab has order 2r.
2
Exercises
11 Show that the quaternion group has only one element of order 2, and that it commutes
with all elements of O. Deduce that O is not isomorphic to D
4
, and that every subgroup of
O is normal.
12 Consider the elements
a =
0 ÷1
1 0
b =
0 1
÷1 ÷1
in GL
2
(Z). Show that a
4
=1 and b
3
=1, but that ab has inﬁnite order, and hence that the
group 'a. b) is inﬁnite.
13 Show that every ﬁnite group of even order contains an element of order 2.
30 1. BASIC DEFINITIONS AND RESULTS
14 Let n =n
1
÷ ÷n
i
be a partition of the positive integer n. Use Lagrange’s theorem
to show that nŠ is divisible by
¸
i
i=1
n
i
Š.
15 Let N be a normal subgroup of G of index n. Show that if g ÷ G, then g
n
÷ N. Give
an example to show that this may be false when the subgroup is not normal.
16 A group G is said to have ﬁnite exponent if there exists an m > 0 such that a
n
=e
for every a in G; the smallest such m is then called the exponent of G.
(a) Show that every group of exponent 2 is commutative.
(b) Show that, for an odd prime ], the group of matrices
¸
¸
¸
1 a b
0 1 c
0 0 1
ˇ
ˇ
ˇ
ˇ
ˇ
ˇ
a. b. c ÷ F
;
¸
¸
has exponent ], but is not commutative.
17 Show that any set of representatives for the conjugacy classes in a ﬁnite group gener
ates the group.
18 Two subgroups H and H
t
of a group G are said to be commensurable if H ¨H
t
is
of ﬁnite index in both H and H
t
. Show that commensurability is an equivalence relation
on the subgroups of G.
19 Show that a nonempty ﬁnite set with an associative binary operation satisfying the
cancellation laws is a group.
CHAPTER 2
Free Groups and Presentations;
Coxeter Groups
It is frequently useful to describe a group by giving a set of generators for the group and
a set of relations for the generators from which every other relation in the group can be
deduced. For example, D
n
can be described as the group with generators r. s and relations
r
n
=e. s
2
=e. srsr =e.
In this chapter, we make precise what this means. First we need to deﬁne the free group on a
set X of generators — this is a group generated by X and with no relations except for those
implied by the group axioms. Because inverses cause problems, we ﬁrst do this for monoids.
Recall that a monoid is a set S with an associative binary operation having an identity
element e. A homomorphism ˛: S ÷S
t
of monoids is a map such that ˛(ab) =˛(a)˛(b)
for all a. b ÷ S and ˛(e) = e — unlike the case of groups, the second condition is not
automatic. A homomorphism of monoids preserves all ﬁnite products.
Free monoids
Let X ={a. b. c. . . .] be a (possibly inﬁnite) set of symbols. A word is a ﬁnite sequence of
symbols from X in which repetition is allowed. For example,
aa. aabac. b
are distinct words. Two words can be multiplied by juxtaposition, for example,
aaaa+aabac =aaaaaabac.
This deﬁnes on the set of all words an associative binary operation. The empty sequence
is allowed, and we denote it by 1. (In the unfortunate case that the symbol 1 is already an
element of X, we denote it by a different symbol.) Then 1 serves as an identity element.
Write SX for the set of words together with this binary operation. Then SX is a monoid,
called the free monoid on X.
31
32 2. FREE GROUPS AND PRESENTATIONS; COXETER GROUPS
When we identify an element a of X with the word a, X becomes
a subset of SX and generates it (i.e., no proper submonoid of SX
contains X). Moreover, the map X ÷SX has the following uni
versal property: for any map of sets ˛: X ÷S fromX to a monoid
S, there exists a unique homomorphism SX ÷S making the dia
gram at right commute:
X SX
S.
o o
˛
Free groups
We want to construct a group JX containing X and having the same universal property as
SX with “monoid” replaced by “group”. Deﬁne X
t
to be the set consisting of the symbols
in X and also one additional symbol, denoted a
1
, for each a ÷ X; thus
X
t
={a. a
1
. b. b
1
. . . .].
Let W
t
be the set of words using symbols from X
t
. This becomes a monoid under juxtapo
sition, but it is not a group because a
1
is not yet the inverse of a, and we can’t cancel out
the obvious terms in words of the following form:
aa
1
or a
1
a
A word is said to be reduced if it contains no pairs of the form aa
1
or a
1
a. Starting with
a word n, we can perform a ﬁnite sequence of cancellations to arrive at a reduced word
(possibly empty), which will be called the reduced form n
0
of n. There may be many
different ways of performing the cancellations, for example,
cabb
1
a
1
c
1
ca ÷ caa
1
c
1
ca ÷ cc
1
ca ÷ ca.
cabb
1
a
1
c
1
ca ÷ cabb
1
a
1
a ÷ cabb
1
÷ ca.
We have underlined the pair we are cancelling. Note that the middle a
1
is cancelled with
different a’s, and that different terms survive in the two cases (the ca at the right in the ﬁrst
cancellation, and the ca at left in the second). Nevertheless we ended up with the same
answer, and the next result says that this always happens.
PROPOSITION 2.1 There is only one reduced form of a word.
PROOF. We use induction on the length of the word n. If n is reduced, there is nothing to
prove. Otherwise a pair of the form a
0
a
1
0
or a
1
0
a
0
occurs — assume the ﬁrst, since the
argument is the same in both cases.
Observe that any two reduced forms of n obtained by a sequence of cancellations in
which a
0
a
1
0
is cancelled ﬁrst are equal, because the induction hypothesis can be applied
to the (shorter) word obtained by cancelling a
0
a
1
0
.
Next observe that any two reduced forms of n obtained by a sequence of cancellations
in which a
0
a
1
0
is cancelled at some point are equal, because the result of such a sequence
of cancellations will not be affected if a
0
a
1
0
is cancelled ﬁrst.
Finally, consider a reduced form n
0
obtained by a sequence in which no cancellation
cancels a
0
a
1
0
directly. Since a
0
a
1
0
does not remain in n
0
, at least one of a
0
or a
1
0
must
Free groups 33
be cancelled at some point. If the pair itself is not cancelled, then the ﬁrst cancellation
involving the pair must look like
,a
1
0
,a
0
a
1
0
or a
0
,a
1
0
,a
0
where our original pair is underlined. But the word obtained after this cancellation is the
same as if our original pair were cancelled, and so we may cancel the original pair instead.
Thus we are back in the case just proved.
2
We say two words n. n
t
are equivalent, denoted n n
t
, if they have the same reduced
form. This is an equivalence relation (obviously).
PROPOSITION 2.2 Products of equivalent words are equivalent, i.e.,
n n
t
. · ·
t
== n· n
t
·
t
.
PROOF. Let n
0
and ·
0
be the reduced forms of n and of ·. To obtain the reduced form
of n·, we can ﬁrst cancel as much as possible in n and · separately, to obtain n
0
·
0
and
then continue cancelling. Thus the reduced form of n· is the reduced form of n
0
·
0
. A
similar statement holds for n
t
·
t
, but (by assumption) the reduced forms of n and · equal
the reduced forms of n
t
and ·
t
, and so we obtain the same result in the two cases.
2
Let JX be the set of equivalence classes of words. Proposition 2.2 shows that the
binary operation on W
t
deﬁnes a binary operation on JX, which obviously makes it into a
monoid. It also has inverses, because
(ab gh)
h
1
g
1
b
1
a
1
1.
Thus JX is a group, called the free group on X. To summarize: the elements of JX are
represented by words in X
t
; two words represent the same element of JX if and only if
they have the same reduced forms; multiplication is deﬁned by juxtaposition; the empty
word represents 1; inverses are obtained in the obvious way. Alternatively, each element of
JX is represented by a unique reduced word; multiplication is deﬁned by juxtaposition and
passage to the reduced form.
When we identify a ÷ X with the equivalence class of the (reduced) word a, then X
becomes identiﬁed with a subset of JX — clearly it generates JX. The next proposition
is a precise statement of the fact that there are no relations among the elements of X when
regarded as elements of JX except those imposed by the group axioms.
PROPOSITION 2.3 For any map of sets ˛: X ÷G from X to a group G, there exists a
unique homomorphism JX ÷G making the following diagram commute:
X JX
G.
o o
˛
34 2. FREE GROUPS AND PRESENTATIONS; COXETER GROUPS
PROOF. Consider a map ˛: X ÷G. We extend it to a map of sets X
t
÷G by setting
˛(a
1
) =˛(a)
1
. Because G is, in particular, a monoid, ˛ extends to a homomorphism of
monoids SX
t
÷G. This map will send equivalent words to the same element of G, and so
will factor through JX =SX
t
¡. The resulting map JX ÷G is a group homomorphism.
It is unique because we know it on a set of generators for JX.
2
REMARK 2.4 The universal property of the map : X ÷JX, . ÷., characterizes it: if

t
: X ÷J
t
is a second map with the same universal property, then there is a unique isomor
phism ˛: JX ÷J
t
such that ˛ ı =
t
,
JX
X
J
t
.
t
t
0
˛
We recall the proof: by the universality of , there exists a unique homomorphism ˛: JX ÷
J
t
such that ˛ ı  = 
t
; by the universality of 
t
, there exists a unique homomorphism
ˇ: J
t
÷JX such that ˇ ı 
t
= ; now (ˇ ı ˛) ı  = , but by the universality of , id
TA
is the unique homomorphism JX ÷ JX such that id
TA
ı = , and so ˇ ı ˛ = id
TA
;
similarly, ˛ ıˇ =id
T
0 , and so ˛ and ˇ are inverse isomorphisms.
COROLLARY 2.5 Every group is a quotient of a free group.
PROOF. Choose a set X of generators for G (e.g., X = G), and let J be the free group
generated by X. According to (2.3), the map a ÷a: X ÷G extends to a homomorphism
J ÷G, and the image, being a subgroup containing X, must equal G.
2
The free group on the set X ={a] is simply the inﬁnite cyclic group C
o
generated by
a, but the free group on a set consisting of two elements is already very complicated.
I now discuss, without proof, some important results on free groups.
THEOREM 2.6 (NIELSENSCHREIER)
1
Subgroups of free groups are free.
The best proof uses topology, and in particular covering spaces—see Serre 1980 or
Rotman 1995, Theorem 11.44.
Two free groups JX and JY are isomorphic if and only if X and Y have the same
cardinality. Thus we can deﬁne the rank of a free group G to be the cardinality of any free
generating set (subset X of G for which the homomorphism JX ÷G given by (2.3) is
an isomorphism). Let H be a ﬁnitely generated subgroup of a free group G. Then there
is an algorithm for constructing from any ﬁnite set of generators for H a free ﬁnite set of
generators. If G has ﬁnite rank n and (G : H) =i <o, then H is free of rank
ni ÷i ÷1.
In particular, H may have rank greater than that of J (or even inﬁnite rank
2
). For proofs,
see Rotman 1995, Chapter 11, and Hall 1959, Chapter 7.
1
Nielsen (1921) proved this for ﬁnitely generated subgroups, and in fact gave an algorithm for deciding
whether a word lies in the subgroup; Schreier (1927) proved the general case.
2
For example, the commutator subgroup of the free group on two generators has inﬁnite rank.
Generators and relations 35
Generators and relations
Consider a set X and a set 1 of words made up of symbols in X
t
. Each element of 1
represents an element of the free group JX, and the quotient G of JX by the normal
subgroup generated by these elements (1.40) is said to have X as generators and 1 as
relations (or as a set of deﬁning relations). One also says that (X. 1) is a presentation for
G, and denotes G by 'X [ 1).
EXAMPLE 2.7 (a) The dihedral group D
n
has generators r. s and deﬁning relations
r
n
. s
2
. srsr.
(See 2.9 below for a proof.)
(b) The generalized quaternion group O
n
, n _3, has generators a. b and relations
3
a
2
n1
=1. a
2
n2
=b
2
. bab
1
=a
1
.
For n = 3 this is the group O of (1.18). In general, it has order 2
n
(for more on it, see
Exercise 25).
(c) Two elements a and b in a group commute if and only if their commutator Œa. b
def
=
aba
1
b
1
is 1. The free abelian group on generators a
1
. . . . . a
n
has generators a
1
. a
2
. . . . . a
n
and relations
Œa
i
. a
}
. i =;.
(d) Let G ='s. t [ s
3
t. t
3
. s
4
). Then G ={1] because
s =ss
3
t =s
4
t =t
1 =s
3
t t
3
=s
3
ss
3
=s.
For the remaining examples, see Massey 1967, which contains a good account of the
interplay between group theory and topology. For example, for many types of topological
spaces, there is an algorithm for obtaining a presentation for the fundamental group.
(e) The fundamental group of the open disk with one point removed is the free group
on o where o is any loop around the point (ibid. II 5.1).
(f) The fundamental group of the sphere with r points removed has generators o
1
. .... o
i
(o
i
is a loop around the i th point) and a single relation
o
1
o
i
=1.
(g) The fundamental group of a compact Riemann surface of genus g has 2g generators
u
1
. ·
1
. .... u
v
. ·
v
and a single relation
u
1
·
1
u
1
1
·
1
1
u
v
·
v
u
1
v
·
1
v
=1
(ibid. IV Exercise 5.7).
3
Strictly speaking, I should say the relations a
2
n1
, a
2
n2
b
2
, bab
1
a.
36 2. FREE GROUPS AND PRESENTATIONS; COXETER GROUPS
PROPOSITION 2.8 Let G be the group deﬁned by the presentation (X. 1). For any group
H and map of sets ˛: X ÷H sending each element of 1 to 1 (in the obvious sense
4
), there
exists a unique homomorphism G ÷H making the following diagram commute:
X G
H.
o o
˛
PROOF. From the universal property of free groups (2.3), we know that ˛ extends to a
homomorphism JX ÷H, which we again denote ˛. Let 1 be the image of 1 in JX.
By assumption 1 c Ker(˛), and therefore the normal subgroup N generated by 1 is
contained in Ker(˛). By the universal property of quotients (see 1.43), ˛ factors through
JX¡N =G. This proves the existence, and the uniqueness follows from the fact that we
know the map on a set of generators for X.
2
EXAMPLE 2.9 Let G ='a. b [ a
n
. b
2
. baba). We prove that G is isomorphic to the dihedral
group D
n
(see 1.17). Because the elements r. s ÷ D
n
satisfy these relations, the map
{a. b] ÷D
n
. a ÷r. b ÷s
extends uniquely to a homomorphism G ÷D
n
. This homomorphism is surjective because
r and s generate D
n
. The equalities
a
n
=1. b
2
=1. ba =a
n1
b
imply that each element of G is represented by one of the following elements,
1. . . . . a
n1
. b. ab. . . . . a
n1
b.
and so [G[ _ 2n = [D
n
[. Therefore the homomorphism is bijective (and these symbols
represent distinct elements of G).
Similarly,
'a. b [ a
2
. b
2
. (ab)
n
) .D
n
by a ÷s, b ÷t .
EXAMPLE 2.10 Let G ='.. . [ .
n
. .
n
) where m. n >1. Then . has order m, . has order
n, and .. has inﬁnite order in G. To see this, recall that for any integers m. n. r > 1, there
exists a group H with elements a and b such that a, b, and ab have orders m, n, and r
respectively (Theorem 1.64). According to (2.8), there exists a homomorphism ˛: G ÷H
such that ˛(.) =a and ˛(.) =b. The order of . certainly divides m, and the fact that ˛(.)
has order m shows that . has order exactly m. Similarly, . has order n. As ˛(..) =ab,
the element .. must have order at least r. As this is true for all r > 1, the element .. has
inﬁnite order.
4
Each element of 1 represents an element of JX, and the condition requires that the unique extension of
˛ to JX sends each of these elements to 1.
Finitely presented groups 37
Finitely presented groups
A group is said to be ﬁnitely presented if it admits a presentation (X. 1) with both X and
1 ﬁnite.
EXAMPLE 2.11 Consider a ﬁnite group G. Let X =G, and let 1 be the set of words
{abc
1
[ ab =c in G].
I claim that (X. 1) is a presentation of G, and so G is ﬁnitely presented. Let G
t
= 'X [
1). The extension of a ÷a: X ÷G to JX sends each element of 1 to 1, and therefore
deﬁnes a homomorphism G
t
÷G, which is obviously surjective. But every element of G
t
is represented by an element of X, and so [G
t
[ _ [G[. Therefore the homomorphism is
bijective.
Although it is easy to deﬁne a group by a ﬁnite presentation, calculating the properties
of the group can be very difﬁcult — note that we are deﬁning the group, which may be quite
small, as the quotient of a huge free group by a huge subgroup. I list some negative results.
THE WORD PROBLEM
Let G be the group deﬁned by a ﬁnite presentation (X. 1). The word problem for G asks
whether there exists an algorithm (decision procedure) for deciding whether a word on X
t
represents 1 in G. The answer is negative: Novikov and Boone showed that there exist
ﬁnitely presented groups G for which no such algorithm exists. Of course, there do exist
other groups for which there is an algorithm.
The same ideas lead to the following result: there does not exist an algorithm that
will determine for an arbitrary ﬁnite presentation whether or not the corresponding group
is trivial, ﬁnite, abelian, solvable, nilpotent, simple, torsion, torsionfree, free, or has a
solvable word problem.
See Rotman 1995, Chapter 12, for proofs of these statements.
THE BURNSIDE PROBLEM
Recall that a group is said to have exponent e if g
e
=1 for all g ÷ G and e is the smallest
natural number with this property. It is easy to write down examples of inﬁnite groups
generated by a ﬁnite number of elements of ﬁnite order (see Exercise 12 or Example 2.10),
but does there exist such a group with ﬁnite exponent? (Burnside problem). In 1968, Adjan
and Novikov showed the answer is yes: there do exist inﬁnite ﬁnitely generated groups of
ﬁnite exponent.
THE RESTRICTED BURNSIDE PROBLEM
The Burnside group of exponent e on r generators T(r. e) is the quotient of the free group
on r generators by the subgroup generated by all eth powers. The Burnside problem asked
whether T(r. e) is ﬁnite, and it is known to be inﬁnite except some small values of r and
e. The restricted Burnside problem asks whether T(r. e) has only ﬁnitely many ﬁnite quo
tients; equivalently, it asks whether there is one ﬁnite quotient of T(r. e) having all other
ﬁnite quotients as quotients. The classiﬁcation of the ﬁnite simple groups (see p. 50) showed
that in order prove that T(r. e) always has only ﬁnitely many ﬁnite quotients, it sufﬁces to
38 2. FREE GROUPS AND PRESENTATIONS; COXETER GROUPS
prove it for e equal to a prime power. This was shown by Eﬁm Zelmanov in 1989 after
earlier work of Kostrikin. See Feit 1995.
TODDCOXETER ALGORITHM
There are some quite innocuous looking ﬁnite presentations that are known to deﬁne quite
small groups, but for which this is very difﬁcult to prove. The standard approach to these
questions is to use the ToddCoxeter algorithm (see Chapter 4 below).
We shall develop various methods for recognizing groups from their presentations (see
also the exercises).
Coxeter groups
A Coxeter system is a pair (G. S) consisting of a group G and a set of generators S for G
subject only to relations of the form (st )
n(x,t)
=1, where
¸
¸
m(s. s) = 1 all s.
m(s. t ) _ 2
m(s. t ) = m(t. s).
(14)
When no relation occurs between s and t , we set m(s. t ) = o. Thus a Coxeter system is
deﬁned by a set S and a mapping
m: S ×S ÷NL{o]
satisfying (14); then G ='S [ 1) where
1 ={(st )
n(x,t)
[ m(s. t ) <o].
The Coxeter groups are those that arise as part of a Coxeter system. The cardinality of S is
called the rank of the Coxeter system.
EXAMPLES
2.12 Up to isomorphism, the only Coxeter system of rank 1 is (C
2
. {s]).
2.13 The Coxeter systems of rank 2 are indexed by m(s. t ) _2.
(a) If m(s. t ) is an integer n, then the Coxeter system is (G. {s. t ]) where
G ='s. t [ s
2
. t
2
. (st )
n
).
According to (2.9), G .D
n
. In particular, s =t and st has order n.
(b) If m(s. t ) =o, then the Coxeter system is (G. {s. t ]) where G ='s. t [ s
2
. t
2
). Ac
cording to (2.10), s and t each have order 2, and st has inﬁnite order.
2.14 Let V =1
n
endowed with the standard positive deﬁnite symmetric bilinear form
((.
i
)
1_i_n
. (.
i
)
1_i_n
) =
¸
.
i
.
i
.
Coxeter groups 39
A reﬂection is a linear map s: V ÷V sending some nonzero vector ˛ to ÷˛ and ﬁxing the
points of the hyperplane H
˛
orthogonal to ˛. We write s
˛
for the reﬂection deﬁned by ˛; it
is given by the formula
s
˛
· =· ÷
2(·. ˛)
(˛. ˛)
˛.
because this is certainly correct for · =˛ and for · ÷ H
˛
, and hence (by linearity) on the
whole of V ='˛)˚H
˛
. A ﬁnite reﬂection group is a ﬁnite group generated by reﬂections.
For such a group G, it is possible to choose a set S of generating reﬂections for which
(G. S) is a Coxeter system (Humphreys 1990, 1.9). Thus, the ﬁnite reﬂection groups are all
Coxeter groups (in fact, they are precisely the ﬁnite Coxeter groups, ibid., 6.4).
2.15 Let S
n
act on 1
n
by permuting the coordinates,
o(a
1
. . . . . a
n
) =(a
c(1)
. . . . . a
c(n)
).
The transposition (i; ) interchanging i and ; , sends the vector
˛ =(0. . . . . 0.
i
1. 0. . . . . 0.
}
÷1. 0. . . .)
to its negative, and leaves the points of the hyperplane
H
˛
=(a
1
. . . . .
i
a
i
. . . . .
}
a
i
. . . . . a
n
)
ﬁxed. Therefore, (i; ) is a reﬂection. As S
n
is generated by the transpositions, this shows
that it is a ﬁnite reﬂection group (hence also a Coxeter group).
THE STRUCTURE OF COXETER GROUPS
THEOREM 2.16 Let (G. S) be the the Coxeter system deﬁned by a map m: S ×S ÷NL
{o] satisfying (14).
(a) The natural map S ÷G is injective.
(b) Each s ÷ S has order 2 in G.
(c) For each s =t in S, st has order m(s. t ) in G.
PROOF. Note that the order of s is 1 or 2, and the order of st divides m(s. t ), and so the
theorem says that the elements of S remain distinct in G and that each s and each st has
the largest possible order.
If S has only a single element, then G .C
2
(see 2.12), and so the statements are obvi
ous. Otherwise, let s and t be distinct elements of S, and let G
t
='s. t [ s
2
. t
2
. (st )
n(x,t)
).
The map S ÷G
t
sending s to s, t to t , and all other elements of S to 1 extends to a homo
morphism G ÷G
t
. We know that s and t are distinct elements of order 2 in G
t
and that st
has order m(s. t ) in G
t
(see 2.13), and it follows that the same is true in G.
2
REMARK 2.17 Let V be the 1vector space with basis a family (e
x
)
x÷S
indexed by S.
The standard proof of Theorem 2.16 deﬁnes a “geometry” on V for which there exist “re
ﬂections” o
x
, s ÷ S, such that o
x
o
t
has order m(s. t ). According to (2.8), the map s ÷o
x
extends to homomorphism of group G ÷GL(V ). This proves the theorem, and it realizes
G as a group of automorphisms of a “geometry”. See Humphreys 1990, Chapter 5, or v3.02
of these notes.
40 2. FREE GROUPS AND PRESENTATIONS; COXETER GROUPS
Exercises
21 Let D
n
='a. b[a
n
. b
2
. abab) be the nth dihedral group. If n is odd, prove that D
2n
~
'a
n
) ×'a
2
. b), and hence that D
2n
~C
2
×D
n
.
22 Prove that the group with generators a
1
. . . . . a
n
and relations Œa
i
. a
}
 = 1, i = ; , is
the free abelian group on a
1
. . . . . a
n
. [Hint: Use universal properties.]
23 Let a and b be elements of an arbitrary free group J. Prove:
(a) If a
n
=b
n
with n > 1, then a =b.
(b) If a
n
b
n
=b
n
a
n
with mn ,=0, then ab =ba.
(c) If the equation .
n
=a has a solution . for every n, then a =1.
24 Let J
n
denote the free group on n generators. Prove:
(a) If n < m, then J
n
is isomorphic to both a subgroup and a quotient group of J
n
.
(b) Prove that J
1
×J
1
is not a free group.
(c) Prove that the centre 7(J
n
) =1 provided n > 1.
25 Prove that O
n
(see 2.7b) has a unique subgroup of order 2, which is 7(O
n
). Prove
that O
n
¡7(O
n
) is isomorphic to D
2
n1.
26 (a) Prove that 'a. b [ a
2
. b
2
. (ab)
n
) .D
n
(cf. 2.9).
(b) Prove that G = 'a. b [ a
2
. abab) is an inﬁnite group. (This is usually known as the
inﬁnite dihedral group.)
27 Let G = 'a. b. c [ a
3
. b
3
. c
4
. acac
1
. aba
1
bc
1
b
1
). Prove that G is the trivial
group {1]. [Hint: Expand (aba
1
)
3
=(bcb
1
)
3
.]
28 Let J be the free group on the set {.. .] and let G =C
2
, with generator a =1. Let ˛
be the homomorphism J ÷G such that ˛(.) =a =˛(.). Find a minimal generating set
for the kernel of ˛. Is the kernel a free group?
29 Let G ='s. t [ t
1
s
3
t =s
5
). Prove that the element
g =s
1
t
1
s
1
t st
1
st
is in the kernel of every map from G to a ﬁnite group.
Coxeter came to Cambridge and gave a lecture [in which he stated a] problem for which he gave
proofs for selected examples, and he asked for a uniﬁed proof. I left the lecture room thinking.
As I was walking through Cambridge, suddenly the idea hit me, but it hit me while I was in
the middle of the road. When the idea hit me I stopped and a large truck ran into me. . . . So I
pretended that Coxeter had calculated the difﬁculty of this problem so precisely that he knew
that I would get the solution just in the middle of the road. . . . Ever since, I’ve called that theorem
“the murder weapon”. One consequence of it is that in a group if a
2
= b
3
= c
5
= (abc)
1
,
then c
610
=1.
John Conway, Math. Intelligencer 23 (2001), no. 2, pp. 8–9.
CHAPTER 3
Automorphisms and Extensions
Automorphisms of groups
An automorphism of a group G is an isomorphism of the group with itself. The set Aut(G)
of automorphisms of G becomes a group under composition: the composite of two auto
morphisms is again an automorphism; composition of maps is always associative (see (5),
p. 9); the identity map g ÷g is an identity element; an automorphism is a bijection, and
therefore has an inverse, which is again an automorphism.
For g ÷ G, the map i
v
“conjugation by g”,
. ÷g.g
1
: G ÷G
is an automorphism of G. An automorphism of this form is called an inner automorphism,
and the remaining automorphisms are said to be outer.
Note that
(gh).(gh)
1
=g(h.h
1
)g
1
, i.e., i
vh
(.) =(i
v
ıi
h
)(.).
and so the map g ÷i
v
: G ÷Aut(G) is a homomorphism. Its image is denoted by Inn(G).
Its kernel is the centre of G,
7(G) ={g ÷ G [ g. =.g all . ÷ G].
and so we obtain from (1.45) an isomorphism
G¡7(G) ÷Inn(G).
In fact, Inn(G) is a normal subgroup of Aut(G): for g ÷ G and ˛ ÷ Aut(G),
(˛ ıi
v
ı˛
1
)(.) =˛(g ˛
1
(.) g
1
) =˛(g) . ˛(g)
1
=i
˛(v)
(.).
EXAMPLE 3.1 (a) Let G =F
n
;
. The automorphisms of G as a commutative group are just
the automorphisms of G as a vector space over F
;
; thus Aut(G) =GL
n
(F
;
). Because G
is commutative, all nontrivial automorphisms of G are outer.
(b) As a particular case of (a), we see that
Aut(C
2
×C
2
) =GL
2
(F
2
).
41
42 3. AUTOMORPHISMS AND EXTENSIONS
(c) Since the centre of the quaternion group O is 'a
2
), we have that
Inn(O) .O¡'a
2
) ~C
2
×C
2
.
In fact, Aut(O) ~S
4
. See Exercise 34.
ASIDE 3.2 The inner automorphisms of a group are the only automorphisms that extend to every
overgroup (Schupp 1987).
COMPLETE GROUPS
DEFINITION 3.3 A group G is complete if the map g ÷i
v
: G ÷Aut(G) is an isomor
phism.
Thus, a group G is complete if and only if (a) the centre 7(G) of G is trivial, and (b)
every automorphism of G is inner.
EXAMPLE 3.4 (a) For n =2. 6, S
n
is complete. The group S
2
is commutative and hence
fails (a); Aut(S
6
)¡Inn(S
6
) ~C
2
and hence S
6
fails (b). See Rotman 1995, Theorems 7.5,
7.10.
(b) If G is a simple noncommutative group, then Aut(G) is complete. See Rotman
1995, Theorem 7.14.
According to Exercise 33, GL
2
(F
2
) ~S
3
, and so the nonisomorphic groups C
2
×C
2
and S
3
have isomorphic automorphism groups.
AUTOMORPHISMS OF CYCLIC GROUPS
Let G be a cyclic group of order n, say G = 'a). Let m be an integer _ 1. The smallest
multiple of m divisible by n is m
n
gcd(n,n)
. Therefore, a
n
has order
n
gcd(n,n)
, and so the
generators of G are exactly the elements a
n
with gcd(m. n) =1. An automorphism ˛ of G
must send a to another generator of G, and so ˛(a) =a
n
for some m relatively prime to n.
The map ˛ ÷m deﬁnes an isomorphism
Aut(C
n
) ÷(Z¡nZ)
×
where
(Z¡nZ)
×
={units in the ring Z¡nZ] ={m÷nZ [ gcd(m. n) =1].
This isomorphism is independent of the choice of a generator a for G: if ˛(a) =a
n
, then
for any other element b =a
i
of G,
˛(b) =˛(a
i
) =˛(a)
i
=a
ni
=(a
i
)
n
=(b)
n
.
It remains to determine (Z¡nZ)
×
. If n = ]
i
1
1
]
i
s
x
is the factorization of n into a
product of powers of distinct primes, then
Z¡nZ .Z¡]
i
1
1
Z× ×Z¡]
i
s
x
Z. m mod n (m mod ]
i
1
. . . .)
by the Chinese remainder theorem. This is an isomorphism of rings, and so
(Z¡nZ)
×
.(Z¡]
i
1
1
Z)
×
× ×(Z¡]
i
s
x
Z)
×
.
Characteristic subgroups 43
It remains to consider the case n =]
i
, ] prime.
Suppose ﬁrst that ] is odd. The set {0. 1. . . . . ]
i
÷1] is a complete set of representatives
for Z¡]
i
Z, and
1
;
of these elements are divisible by ]. Hence (Z¡]
i
Z)
×
has order ]
i
÷
;
r
;
=]
i1
(]÷1). The homomorphism
(Z¡]
i
Z)
×
÷(Z¡]Z)
×
is surjective with kernel of order ]
i1
, and we know that (Z¡]Z)
×
is cyclic. Let a ÷
(Z¡]
i
Z)
×
map to a generator of (Z¡]Z)
×
. Then a
;
r
(;1)
=1 and a
;
r
again maps to a
generator of (Z¡]Z)
×
. Therefore (Z¡]
i
Z)
×
contains an element ¯
def
=a
;
r
of order ] ÷1.
Using the binomial theorem, one ﬁnds that 1÷] has order ]
i1
in (Z¡]
i
Z)
×
. Therefore
(Z¡]
i
Z)
×
is cyclic with generator ¯ (1 ÷]) (cf. (13), p. 27), and every element can be
written uniquely in the form
¯
i
(1÷])
}
. 0 _i < ]÷1. 0 _; < ]
i1
.
On the other hand,
(Z¡8Z)
×
={
¨
1.
¨
3.
¨
5.
¨
7] ='
¨
3.
¨
5) ~C
2
×C
2
is not cyclic.
SUMMARY 3.5 (a) For a cyclic group of G of order n, Aut(G) .(Z¡nZ)
×
. The automor
phism of G corresponding to Œm ÷ (Z¡nZ)
×
sends an element a of G to a
n
.
(b) If n =]
i
1
1
]
i
s
x
is the factorization of n into a product of powers of distinct primes
]
i
, then
(Z¡nZ)
×
.(Z¡]
i
1
1
Z)
×
× ×(Z¡]
i
s
x
Z)
×
. m mod n (m mod ]
i
1
. . . .).
(c) For a prime ],
(Z¡]
i
Z)
×
~
ˆ
¸
ˆ
¸
C
(;1);
r1 ] odd,
C
2
]
i
=2
2
C
2
× C
2
r2 ] =2, r > 2.
Characteristic subgroups
DEFINITION 3.6 A characteristic subgroup of a group G is a subgroup H such that
˛(H) =H for all automorphisms ˛ of G.
The same argument as in (1.32) shows that it sufﬁces to check that ˛(H) cH for all ˛ ÷
Aut(G). Thus, a subgroup H of G is normal if it is stable under all inner automorphisms of
G, and it is characteristic if it stable under all automorphisms. In particular, a characteristic
subgroup is normal.
REMARK 3.7 (a) Consider a group G and a normal subgroup N. An inner automorphism
of G restricts to an automorphism of N, which may be outer (for an example, see 3.16
below). Thus a normal subgroup of N need not be a normal subgroup of G. However, a
characteristic subgroup of N will be a normal subgroup of G. Also a characteristic sub
group of a characteristic subgroup is a characteristic subgroup.
44 3. AUTOMORPHISMS AND EXTENSIONS
(b) The centre 7(G) of G is a characteristic subgroup, because
:g =g: all g ÷ G == ˛(:)˛(g) =˛(g)˛(:) all g ÷ G.
and as g runs over G, ˛(g) also runs over G. Expect subgroups with a general group
theoretic deﬁnition to be characteristic.
(c) If H is the only subgroup of G of order m, then it must be characteristic, because
˛(H) is again a subgroup of G of order m.
(d) Every subgroup of a commutative group is normal but not necessarily characteris
tic. For example, every subspace of dimension 1 in F
2
;
is subgroup of F
2
;
, but it is not
characteristic because it is not stable under Aut(F
2
;
) =GL
2
(F
;
).
Semidirect products
Let N be a normal subgroup of G. Each element g of G deﬁnes an automorphism of N,
n ÷gng
1
, and this deﬁnes a homomorphism
0: G ÷Aut(N). g ÷i
v
[N.
If there exists a subgroup O of G such that G ÷G¡N maps O isomorphically onto G¡N,
then I claim that we can reconstruct G from N, O, and the restriction of 0 to O. Indeed,
an element g of G can be written uniquely in the form
g =nq. n ÷ N. q ÷ O;
— q must be the unique element of O mapping to gN ÷ G¡N, and n must be gq
1
. Thus,
we have a onetoone correspondence of sets
G
11
÷÷N ×O.
If g =nq and g
t
=n
t
q
t
, then
gg
t
=(nq)
n
t
q
t
=n(qn
t
q
1
)qq
t
=n 0(q)(n
t
) qq
t
.
DEFINITION 3.8 A group G is a semidirect product of its subgroups N and O if N is
normal and the homomorphism G ÷G¡N induces an isomorphism O÷G¡N.
Equivalently, G is a semidirect product of subgroup N and O if
N <G; NO=G; N ¨O={1]. (15)
Note that O need not be a normal subgroup of G. When G is the semidirect product of
subgroups N and O, we write G =N O (or N
0
O where 0: O ÷Aut(N) gives the
action of O on N by inner automorphisms).
EXAMPLE 3.9 (a) In D
n
, n _2, let C
n
='r) and C
2
='s); then
D
n
='r)
0
's) =C
n
0
C
2
where 0(s)(r
i
) =r
i
(see 1.17).
Semidirect products 45
(b) The alternating subgroup ·
n
is a normal subgroup of S
n
(because it has index 2),
and C
2
={(12)] maps isomorphically onto S
n
¡·
n
. Therefore S
n
=·
n
C
2
.
(c) The quaternion group can not be written as a semidirect product in any nontrivial
fashion (see Exercise 31).
(d) A cyclic group of order ]
2
, ] prime, is not a semidirect product (because it has only
one subgroup of order ]).
(e) Let G =GL
n
(J). Let T be the subgroup of upper triangular matrices in G, T the
subgroup of diagonal matrices in G, and U the subgroup of upper triangular matrices with
all their diagonal coefﬁcients equal to 1. Thus, when n =2,
T =
+ +
0 +
¸
. T =
+ 0
0 +
¸
. U =
1 +
0 1
¸
.
Then, U is a normal subgroup of T, UT =T, and U ¨T ={1]. Therefore,
T =U T .
Note that, when n _2, the action of T on U is not trivial, for example,
a 0
0 b
1 c
0 1
a
1
0
0 b
1
=
1 ac¡b
0 1
.
and so T is not the direct product of T and U.
We have seen that, from a semidirect product G =N O, we obtain a triple
(N. O. 0: O÷Aut(N)).
and that the triple determines G. We now prove that every triple (N. O. 0) consisting of two
groups N and O and a homomorphism 0: O÷Aut(N) arises from a semidirect product.
As a set, let G =N ×O, and deﬁne
(n. q)(n
t
. q
t
) =(n 0(q)(n
t
). qq
t
).
PROPOSITION 3.10 The composition law above makes G into a group, in fact, the semidi
rect product of N and O.
PROOF. Write
q
n for 0(q)(n), so that the composition law becomes
(n. q)(n
t
. q
t
) =(n
q
n
t
. qq
t
).
Then
((n. q). (n
t
. q
t
))(n
tt
. q
tt
) =(n
q
n
t
qq
0
n
tt
. qq
t
q
tt
) =(n. q)((n
t
. q
t
)(n
tt
. q
tt
))
and so the associative law holds. Because 0(1) =1 and 0(q)(1) =1,
(1. 1)(n. q) =(n. q) =(n. q)(1. 1),
and so (1. 1) is an identity element. Next
(n. q)(
q
1
n
1
. q
1
) =(1. 1) =(
q
1
n
1
. q
1
)(n. q).
and so (
q
1
n
1
. q
1
) is an inverse for (n. q). Thus G is a group, and it is obvious that
N <G, NO =G, and N ¨O ={1], and so G =N O. Moreover, when N and O are
regarded as subgroups of G, the action of O on N is that given by 0.
2
46 3. AUTOMORPHISMS AND EXTENSIONS
EXAMPLES
3.11 A group of order 12. Let 0 be the (unique) nontrivial homomorphism
C
4
÷Aut(C
3
) .C
2
.
namely, that sending a generator of C
4
to the map a ÷ a
2
. Then G
def
= C
3
0
C
4
is a
noncommutative group of order 12, not isomorphic to ·
4
. If we denote the generators of
C
3
and C
4
by a and b, then a and b generate G, and have the deﬁning relations
a
3
=1. b
4
=1. bab
1
=a
2
.
3.12 Direct products. The bijection of sets
(n. q) ÷(n. q): N ×O÷N
0
O
is an isomorphism of groups if and only if 0 is the trivial homomorphism O ÷Aut(N),
i.e., 0(q)(n) =n for all q ÷ O, n ÷ N.
3.13 Groups of order 6. Both S
3
and C
6
are semidirect products of C
3
by C
2
— they
correspond to the two distinct homomorphisms C
2
÷C
2
.Aut(C
3
).
3.14 Groups of order ]
3
(element of order ]
2
). Let N = 'a) be cyclic of order ]
2
,
and let O='b) be cyclic of order ], where ] is an odd prime. Then Aut N ~C
;1
×C
;
(see 3.5), and C
;
is generated by ˛: a ÷ a
1÷;
(note that ˛
2
(a) = a
1÷2;
. . . .). Deﬁne
O÷Aut N by b ÷˛. The group G
def
=N
0
O has generators a. b and deﬁning relations
a
;
2
=1. b
;
=1. bab
1
=a
1÷;
.
It is a noncommutative group of order ]
3
, and possesses an element of order ]
2
.
3.15 Groups of order ]
3
(no element of order ]
2
). Let N = 'a. b) be the product of
two cyclic groups 'a) and 'b) of order ], and let O = 'c) be a cyclic group of order ].
Deﬁne 0: O÷Aut(N) to be the homomorphism such that
0(c
i
)(a) =ab
i
. 0(c
i
)(b) =b.
(If we regard N as the additive group N =F
2
;
with a and b the standard basis elements, then
0(c
i
) is the automorphism of N deﬁned by the matrix
1 0
i 1
.) The group G
def
=N
0
O
is a group of order ]
3
, with generators a. b. c and deﬁning relations
a
;
=b
;
=c
;
=1. ab =cac
1
. Œb. a =1 =Œb. c.
Because b = 1, the middle equality shows that the group is not commutative. When ] is
odd, all elements except 1 have order ]. When ] =2, G ~D
4
, which does have an element
of order 2
2
. Note that this shows that a group can have quite different representations as a
semidirect product:
D
4
(3.9a)
~ C
4
C
2
~(C
2
×C
2
) C
2
.
Semidirect products 47
For an odd prime ], a noncommutative group of order ]
3
is isomorphic to the group in
(3.14) if it has an element of order ]
2
and to the group in (3.15) if it doesn’t (see Exercise
43). In particular, up to isomorphism, there are exactly two noncommutative groups of
order ]
3
.
3.16 Making outer automorphisms inner. Let ˛ be an automorphism, possibly outer,
of a group N. We can realize N as a normal subgroup of a group G in such a way that
˛ becomes the restriction to N of an inner automorphism of G. To see this, let 0: C
o
÷
Aut(N) be the homomorphism sending a generator a of C
o
to ˛ ÷ Aut(N), and let G =
N
0
C
o
. The element g =(1. a) of G has the property that g(n. 1)g
1
=(˛(n). 1) for all
n ÷ N.
CRITERIA FOR SEMIDIRECT PRODUCTS TO BE ISOMORPHIC
It will be useful to have criteria for when two triples (N. O. 0) and (N. O. 0
t
) determine
isomorphic groups.
LEMMA 3.17 If there exists an ˛ ÷ Aut(N) such that
0
t
(q) =˛ ı0(q) ı˛
1
. all q ÷ O.
then the map
(n. q) ÷(˛(n). q): N
0
O÷N
0
0 O
is an isomorphism.
PROOF. For (n. q) ÷ N
0
O, let ,(n. q) =(˛(n). q). Then
,(n. q) ,(n
t
. q
t
) =(˛(n). q) (˛(n
t
). q
t
)
=(˛(n) 0
t
(q)(˛(n
t
)). qq
t
)
=(˛(n) (˛ ı0(q) ı˛
1
)(˛(n
t
)). qq
t
)
=(˛(n) ˛(0(q)(n
t
)). qq
t
).
and
,((n. q) (n
t
. q
t
)) =,(n 0(q)(n
t
). qq
t
)
=(˛(n) ˛
0(q)(n
t
)
. qq
t
).
Therefore , is a homomorphism. The map
(n. q) ÷(˛
1
(n). q): N
0
0 O÷N
0
O
is also a homomorphism, and it is inverse to ,, and so both are isomorphisms.
2
LEMMA 3.18 If 0 =0
t
ı˛ with ˛ ÷ Aut(O), then the map
(n. q) ÷(n. ˛(q)): N
0
O~N
0
0 O
is an isomorphism.
48 3. AUTOMORPHISMS AND EXTENSIONS
PROOF. Routine veriﬁcation.
2
LEMMA 3.19 If O is cyclic and the subgroup 0(O) of Aut(N) is conjugate to 0
t
(O), then
N
0
O~N
0
0 O.
PROOF. Let a generate O. By assumption, there exists an a
t
÷ O and an ˛ ÷ Aut(N) such
that
0
t
(a
t
) =˛ 0(a) ˛
1
.
The element 0
t
(a
t
) generates 0
t
(O), and so we can choose a
t
to generate O, say a
t
=a
i
with i relatively prime to the order of O. Now the map (n. q) ÷(˛(n). q
i
) is an isomor
phism N
0
O÷N
0
0 O.
2
SUMMARY 3.20 Let G be a group with subgroups H
1
and H
2
such that G =H
1
H
2
and
H
1
¨H
2
= {e], so that each element g of G can be written uniquely as g = h
1
h
2
with
h
1
÷ H
1
and h
2
÷ H
2
.
(a) If H
1
and H
2
are both normal, then G is the direct product of H
1
and H
2
, G =
H
1
×H
2
(1.51).
(b) If H
1
is normal in G, then G is the semidirect product of H
1
and H
2
, G =H
1
H
2
((15), p. 44).
(c) If neither H
1
nor H
2
is normal, then G is the ZappaSz´ ep (or knit) product of H
1
and H
2
(see http://en.wikipedia.org/wiki/ZappaSzep_product).
Extensions of groups
A sequence of groups and homomorphisms
1 ÷N
t
÷G
t
÷O÷1 (16)
is exact if  is injective, ¬ is surjective, and Ker(¬) =Im(). Thus (N) is a normal sub
group of G (isomorphic by  to N) and G¡(N)
:
÷÷O. We often identify N with the
subgroup (N) of G and O with the quotient G¡N.
An exact sequence (16) is also called an extension of O by N.
1
An extension is central
if (N) c 7(G). For example, a semidirect product N
0
O gives rise to an extension of
O by N,
1 ÷N ÷N
0
O÷O÷1.
which is central if and only if 0 is the trivial homomorphism.
Two extensions of O by N are said to be isomorphic if there exists a commutative
diagram
1 ÷÷÷÷÷ N ÷÷÷÷÷ G ÷÷÷÷÷ O ÷÷÷÷÷ 1
¸
¸
¸
~
¸
¸
¸
1 ÷÷÷÷÷ N ÷÷÷÷÷ G
t
÷÷÷÷÷ O ÷÷÷÷÷ 1.
1
This is Bourbaki’s terminology (Alg` ebre, I :6); some authors call (16) an extension of N by O.
Extensions of groups 49
An extension of O by N,
1 ÷N
t
÷G
t
÷O÷1.
is said to be split if it is isomorphic to the extension deﬁned by a semidirect product N
0
O.
Equivalent conditions:
(a) there exists a subgroup O
t
cG such that ¬ induces an isomorphism O
t
÷O; or
(b) there exists a homomorphism s: O÷G such that ¬ ıs =id.
In general, an extension will not split. For example, the extensions
1 ÷N ÷O÷O¡N ÷1 (17)
with N any subgroup of order 4 in the quaternion group O, and
1 ÷C
;
÷C
;
2 ÷C
;
÷1
do not split. We give two criteria for an extension to split.
THEOREM 3.21 (SCHURZASSENHAUS) An extension of ﬁnite groups of relatively prime
order is split.
PROOF. Rotman 1995, 7.41.
2
PROPOSITION 3.22 An extension (17) splits if N is complete. In fact, G is then direct
product of N with the centralizer of N in G,
C
G
(N)
def
={g ÷ G [ gn =ng all n ÷ N].
PROOF. Let O=C
G
(N). We shall check that N and O satisfy the conditions of Proposi
tion 1.51.
Observe ﬁrst that, for any g ÷ G, n ÷gng
1
: N ÷N is an automorphism of N, and
(because N is complete), it must be the inner automorphism deﬁned by an element , of N;
thus
gng
1
=,n,
1
all n ÷ N.
This equation shows that ,
1
g ÷ O, and hence g =,(,
1
g) ÷ NO. Since g was arbitrary,
we have shown that G =NO.
Next note that every element of N ¨O is in the centre of N, which (because N is
complete) is trivial; hence N ¨O=1.
Finally, for any element g =nq ÷ G,
gOg
1
=n(qOq
1
)n
1
=nOn
1
=O
(recall that every element of N commutes with every element of O). Therefore O is normal
in G.
2
An extension
1 ÷N ÷G ÷O÷1
gives rise to a homomorphism 0
t
: G ÷Aut(N), namely,
0
t
(g)(n) =gng
1
.
50 3. AUTOMORPHISMS AND EXTENSIONS
Let ¯ q ÷ G map to q in O; then the image of 0
t
( ¯ q) in Aut(N)¡Inn(N) depends only on q;
therefore we get a homomorphism
0 : O÷Out(N)
def
=Aut(N)¡Inn(N).
This map 0 depends only on the isomorphismclass of the extension, and we write Ext
1
(O. N)
0
for the set of isomorphism classes of extensions with a given 0. These sets have been ex
tensively studied.
When O and N are commutative and 0 is trivial, the group G is also commutative, and
there is a commutative group structure on the set Ext
1
(O. N). Moreover, endomorphisms
of O and N act as endomorphisms on Ext
1
(O. N). In particular, multiplication by m on
O or N induces multiplication by m on Ext
1
(O. N). Thus, if O and N are killed by m
and n respectively, then Ext
1
(O. N) is killed by m and by n, and hence by gcd(m. n). This
proves the SchurZassenhaus theorem in this case.
The H¨ older program.
It would be of the greatest interest if it were possible
to give an overview of the entire collection of ﬁnite
simple groups.
Otto H¨ older, Math. Ann., 1892
Recall that a group G is simple if it contains no normal subgroup except 1 and G. In
other words, a group is simple if it can’t be realized as an extension of smaller groups.
Every ﬁnite group can be obtained by taking repeated extensions of simple groups. Thus
the simple ﬁnite groups can be regarded as the basic building blocks for all ﬁnite groups.
The problem of classifying all simple groups falls into two parts:
A. Classify all ﬁnite simple groups;
B. Classify all extensions of ﬁnite groups.
A. THE CLASSIFICATION OF FINITE SIMPLE GROUPS
There is a complete list of ﬁnite simple groups. They are
(a) the cyclic groups of prime order,
(b) the alternating groups ·
n
for n _5 (see the next chapter),
(c) certain inﬁnite families of matrix groups, and
(d) the 26 “sporadic groups”.
By far the largest class is (c), but the 26 sporadic groups are of more interest than
their small number might suggest. Some have even speculated that the largest of them, the
FischerGriess monster, is built into the fabric of the universe.
As an example of a matrix group, consider
SL
n
(F
q
)
def
={m×m matrices · with entries in F
q
such that det · =1].
Exercises 51
Here q =]
n
, ] prime, and F
q
is “the” ﬁeld with q elements. This group is not simple if
q = 2, because the scalar matrices
¸
¸
¸
¸
¯ 0 0
0 ¯ 0
.
.
.
0 0 ¯
, ¯
n
= 1, are in the centre for any m
dividing q ÷1, but these are the only matrices in the centre, and the groups
PSL
n
(F
q
)
def
=SL
n
(F
q
)¡{centre]
are simple when m_3 (Rotman 1995, 8.23) and when m=2 and q >3 (ibid. 8.13). Other
ﬁnite simple groups can be obtained from the groups in (1.8). The smallest noncommutative
group is ·
5
, and the second smallest is PSL
3
(F
2
), which as order 168 (see Exercise 47).
B THE CLASSIFICATION OF ALL EXTENSIONS OF FINITE GROUPS
Much is known about the extensions of ﬁnite groups, for example, about the extensions of
one simple group by another. However, as Solomon writes (2001, p. 347):
. . . the classiﬁcation of all ﬁnite groups is completely infeasible. Neverthe
less experience shows that most of the ﬁnite groups which occur in “nature”
. . . are “close” either to simple groups or to groups such as dihedral groups,
Heisenberg groups, etc., which arise naturally in the study of simple groups.
As we noted earlier, by the year 2001, a complete irredundant list of ﬁnite groups was
available only for those up to an order of about 2000, and the number of groups on the list
is overwhelming.
NOTES The dream of classifying the ﬁnite simple groups goes back at least to H¨ older 1892. How
ever a clear strategy for accomplishing this did not begin to emerge until the 1950s, when work of
Brauer and others suggested that the key was to study the centralizers of elements of order 2 (the
involution centralizers). For example, Brauer and Fowler (1955) showed that, for any ﬁnite group
H, the determination of the ﬁnite simple groups with an involution centralizer isomorphic to H is a
ﬁnite problem. Later work showed that the problem is even tractable, and so the strategy became: (a)
list the groups H that are candidates for being an involution centralizer in some ﬁnite simple group,
and (b) for each H in (a) list the ﬁnite simple groups for which H occurs as an involution centralizer.
Of course, this approach applies only to the ﬁnite simple groups containing an element of order 2,
but an old conjecture said that, except for the cyclic groups of prime order, every ﬁnite simple group
has even order and hence contains an element of order 2 by Cauchy’s theorem (4.13). With the
proof of this conjecture by Feit and Thompson (1963), the effort to complete the classiﬁcation of the
ﬁnite simple groups began in earnest. A complete classiﬁcation was announced in 1982, but there
remained sceptics, because the proof depended on thousands of pages of rarely read journal articles,
and, in fact, in reworking the proof, gaps were discovered. However, these have been closed, and
with the publication of Aschbacher and Smith 2004 it has become generally accepted that the proof
of the classiﬁcation is indeed complete.
For a popular account of the history of the classiﬁcation, see the book Ronan 2006, and for a
more technical account, see the expository article Solomon 2001.
Exercises
31 Let G be the quaternion group (1.18). Prove that G can’t be written as a semidirect
product in any nontrivial fashion.
52 3. AUTOMORPHISMS AND EXTENSIONS
32 Let G be a group of order mn where m and n have no common factor. If G contains
exactly one subgroup M of order m and exactly one subgroup N of order n, prove that G
is the direct product of M and N.
33 Prove that GL
2
(F
2
) ~S
3
.
34 Let G be the quaternion group (1.18). Prove that Aut(G) ~S
4
.
35 Let G be the set of all matrices in GL
3
(1) of the form
¸
a 0 b
0 a c
0 0 J
, aJ =0. Check
that G is a subgroup of GL
3
(1), and prove that it is a semidirect product of 1
2
(additive
group) by 1
×
×1
×
. Is it a direct product of these two groups?
36 Find the automorphism groups of C
o
and S
3
.
37 Let G =N O where N and O are ﬁnite groups, and let g =nq be an element of
G with n ÷ N and q ÷ O. Denote the order of an element . by o(.).
(a) Show that o(g) =k o(q) for some divisor k of [N[.
(b) When O acts trivially on N, show that o(g) =lcm(o(n). o(q)).
(c) Let G =S
5
=·
5
O with O='(1. 2)). Let n =(1. 4. 3. 2. 5) and let q =(1. 2).
Show that o(g) =6, o(n) =5, and o(q) =2.
(d) Suppose that G = (C
;
)
;
O where O is cyclic of order ] and that, for some
generator q of O,
q(a
1
. . . . . a
n
)q
1
=(a
n
. a
1
. . . . . a
n1
).
Show inductively that, for i _],
((1. 0. . . . . 0). q)
i
=((1. . . . . 1. 0. . . . . 0). q
i
)
(i copies of 1). Deduce that ((1. 0. . . . . 0). q) has order ]
2
(hence o(g) =o(n) o(q) in this
case).
(e) Suppose that G =N O where N is commutative, O is cyclic of order 2, and the
generator q of O acts on N by sending each element to its inverse. Show that (n. 1) has
order 2 no matter what n is (in particular, o(g) is independent of o(n)).
38 Let G be the semidirect G =N O of its subgroups N and O, and let
C
1
(O) ={n ÷ N [ nq =qn for all q ÷ O]
(centralizer of O in N). Show that
7(G) ={n q [ n ÷ C
1
(O), q ÷ 7(O), nn
t
n
1
=q
1
n
t
q for all n
t
÷ N].
Let 0 be the homomorphism O÷Aut(N) giving the action of O on N (by conjugation).
Show that if N is commutative, then
7(G) ={n q [ n ÷ C
1
(O), q ÷ 7(O) ¨Ker(0)].
and if N and O are commutative, then
7(G) ={n q [ n ÷ C
1
(O), q ÷ Ker(0)].
CHAPTER 4
Groups Acting on Sets
Deﬁnition and examples
DEFINITION 4.1 Let X be a set and let G be a group. A left action of G on X is a mapping
(g. .) ÷g.: G×X ÷X such that
(a) 1. =., for all . ÷ X:
(b) (g
1
g
2
). =g
1
(g
2
.), all g
1
, g
2
÷ G, . ÷ X.
A set together with a (left) action of G is called a (left) Gset. An action is trivial if g. =.
for all g ÷ G.
The conditions imply that, for each g ÷ G, left translation by g,
g
1
: X ÷X. . ÷g..
has (g
1
)
1
as an inverse, and therefore g
1
is a bijection, i.e., g
1
÷ Sym(X). Axiom (b)
now says that
g ÷g
1
: G ÷Sym(X) (18)
is a homomorphism. Thus, from a left action of G on X, we obtain a homomorphism
G ÷Sym(X): conversely, every such homomorphism deﬁnes an action of G on X. The
action is said to be faithful (or effective) if the homomorphism (18) is injective, i.e., if
g. =. for all . ÷ X == g =1.
EXAMPLE 4.2 (a) Every subgroup of the symmetric group S
n
acts faithfully on {1. 2. .... n].
(b) Every subgroup H of a group G acts faithfully on G by left translation,
H ×G ÷G. (h. .) ÷h..
(c) Let H be a subgroup of G. The group G acts on the set of left cosets of H,
G×G¡H ÷G¡H. (g. C) ÷gC.
The action is faithful if, for example, H =G and G is simple.
(d) Every group G acts on itself by conjugation,
G×G ÷G. (g. .) ÷
v
.
def
=g.g
1
.
53
54 4. GROUPS ACTING ON SETS
For any normal subgroup N, G acts on N and G¡N by conjugation.
(e) For any group G, Aut(G) acts on G.
(f) The group of rigid motions of 1
n
is the group of bijections 1
n
÷1
n
preserving
lengths. It acts on 1
n
on the left.
A right action X ×G ÷G is deﬁned similarly. To turn a right action into a left action,
set g + . = .g
1
. For example, there is a natural right action of G on the set of right
cosets of a subgroup H in G, namely, (C. g) ÷Cg, which can be turned into a left action
(g. C) ÷Cg
1
.
A map of Gsets (alternatively, a Gmap or a Gequivariant map) is a map c: X ÷Y
such that
c(g.) =gc(.). all g ÷ G. . ÷ X.
An isomorphism of Gsets is a bijective Gmap; its inverse is then also a Gmap.
ORBITS
Let G act on X.
A subset S cX is said to be stable under the action of G if
g ÷ G. . ÷ S == g. ÷ S.
The action of G on X then induces an action of G on S.
Write . 
G
. if . = g., some g ÷ G. This relation is reﬂexive because . = 1.,
symmetric because
. =g. == . =g
1
.
(multiply by g
1
on the left and use the axioms), and transitive because
. =g.. : =g
t
. == : =g
t
(g.) =(g
t
g)..
It is therefore an equivalence relation. The equivalence classes are called Gorbits. Thus
the Gorbits partition X. Write G\X for the set of orbits.
By deﬁnition, the Gorbit containing .
0
is
G.
0
={g.
0
[ g ÷ G].
It is the smallest Gstable subset of X containing .
0
.
EXAMPLE 4.3 (a) Suppose G acts on X, and let ˛ ÷ G be an element of order n. Then the
orbits of '˛) are the sets of the form
{.
0
. ˛.
0
. . . . . ˛
n1
.
0
].
(These elements need not be distinct, and so the set may contain fewer than n elements.)
(b) The orbits for a subgroup H of G acting on G by left multiplication are the right
cosets of H in G. We write H\G for the set of right cosets. Similarly, the orbits for H
acting by right multiplication are the left cosets, and we write G¡H for the set of left cosets.
Note that the group law on G will not induce a group law on G¡H unless H is normal.
(c) For a group G acting on itself by conjugation, the orbits are called conjugacy
classes: for . ÷ G, the conjugacy class of . is the set
{g.g
1
[ g ÷ G]
Deﬁnition and examples 55
of conjugates of .. The conjugacy class of .
0
always contains .
0
, and it consists only of .
0
if and only if .
0
is in the centre of G. In linear algebra the conjugacy classes in G =GL
n
(k)
are called similarity classes, and the theory of rational canonical forms provides a set of
representatives for the conjugacy classes: two matrices are similar (conjugate) if and only
if they have the same rational canonical form.
Note that a subset of X is stable if and only if it is a union of orbits. For example, a
subgroup H of G is normal if and only if it is a union of conjugacy classes.
The action of G on X is said to be transitive, and G is said to act transitively on X,
if there is only one orbit, i.e., for any two elements . and . of X, there exists a g ÷ G
such that g. = .. The set X is then called a homogeneous Gset. For example, S
n
acts
transitively on {1. 2. ...n]. For any subgroup H of a group G, G acts transitively on G¡H,
but the action of G on itself is never transitive if G =1 because {1] is always a conjugacy
class.
The action of G on X is doubly transitive if for any two pairs (.
1
. .
2
), (.
1
. .
2
) of
elements of X with .
1
=.
2
and .
1
=.
2
, there exists a (single) g ÷ G such that g.
1
=.
1
and g.
2
=.
2
. Deﬁne kfold transitivity for k _3 similarly.
STABILIZERS
Let G act on X. The stabilizer (or isotropy group) of an element . ÷ X is
Stab(.) ={g ÷ G [ g. =.].
It is a subgroup, but it need not be a normal subgroup (see the next lemma). The action is
free if Stab(.) ={e] for all ..
LEMMA 4.4 For any g ÷ G and . ÷ X,
Stab(g.) =g Stab(.) g
1
.
PROOF. Certainly, if g
t
. =., then
(gg
t
g
1
)g. =gg
t
. =g. =..
and so g Stab(.) g
1
cStab(g.). Conversely, if g
t
(g.) =g., then
(g
1
g
t
g). =g
1
g
t
(g.) =g
1
. =..
and so g
1
g
t
g ÷ Stab(.), i.e., g
t
÷ g Stab(.) g
1
.
2
Clearly
¸
x÷A
Stab(.) =Ker(G ÷Sym(X)).
which is a normal subgroup of G. The action is faithful if and only if
¸
Stab(.) ={1].
EXAMPLE 4.5 (a) Let G act on itself by conjugation. Then
Stab(.) ={g ÷ G [ g. =.g].
56 4. GROUPS ACTING ON SETS
This group is called the centralizer C
G
(.) of . in G. It consists of all elements of G that
commute with, i.e., centralize, .. The intersection
¸
x÷G
C
G
(.) ={g ÷ G [ g. =.g for all . ÷ G]
is the centre of G.
(b) Let G act on G¡H by left multiplication. Then Stab(H) =H, and the stabilizer of
gH is gHg
1
.
(c) Let G be the group of rigid motions of 1
n
(4.2f). The stabilizer of the origin is the
orthogonal group O
n
for the standard positive deﬁnite form on 1
n
(Artin 1991, Chap. 4,
5.16). Let T .(1
n
. ÷) be the subgroup of G of translations of 1
n
, i.e., maps of the form
· ÷· ÷·
0
some ·
0
÷ 1
n
. Then T is a normal subgroup of G and G .T O (cf. Artin
1991, Chap. 5, :2).
For a subset S of X, we deﬁne the stabilizer of S to be
Stab(S) ={g ÷ G [ gS =S].
Then Stab(S) is a subgroup of G, and the same argument as in the proof of (4.4) shows that
Stab(gS) =g Stab(S) g
1
.
EXAMPLE 4.6 Let G act on G by conjugation, and let H be a subgroup of G. The stabi
lizer of H is called the normalizer N
G
(H) of H in G:
N
G
(H) ={g ÷ G [ gHg
1
=H].
Clearly N
G
(H) is the largest subgroup of G containing H as a normal subgroup.
It is possible for gS cS but g ÷ Stab(S) (see 1.33).
TRANSITIVE ACTIONS
PROPOSITION 4.7 If G acts transitively on X, then for any .
0
÷ X, the map
gStab(.
0
) ÷g.
0
: G¡Stab(.
0
) ÷X
is an isomorphism of Gsets.
PROOF. It is welldeﬁned because, if h ÷Stab(.
0
), then gh.
0
=g.
0
. It is injective because
g.
0
=g
t
.
0
== g
1
g
t
.
0
=.
0
== g. g
t
lie in the same left coset of Stab(.
0
).
It is surjective because G acts transitively. Finally, it is obviously Gequivariant.
2
Thus every homogeneous Gset X is isomorphic to G¡H for some subgroup H of G,
but such a realization of X is not canonical: it depends on the choice of .
0
÷ X. To say
this another way, the Gset G¡H has a preferred point, namely, the coset H; to give a
homogeneous Gset X together with a preferred point is essentially the same as to give a
subgroup of G.
Deﬁnition and examples 57
COROLLARY 4.8 Let G act on X, and let O =G.
0
be the orbit containing .
0
. Then the
cardinality of O is
[O[ =(G : Stab(.
0
)). (19)
For example, the number of conjugates gHg
1
of a subgroup H of G is (G: N
G
(H)).
PROOF. The action of G on O is transitive, and so g ÷g.
0
deﬁnes a bijection G¡Stab(.
0
) ÷
G.
0
.
2
The equation (19) is frequently useful for computing [O[.
PROPOSITION 4.9 Let .
0
÷ X. If G acts transitively on X, then
Ker(G ÷Sym(X))
is the largest normal subgroup contained in Stab(.
0
).
PROOF. When
Ker(G ÷Sym(X)) =
¸
x÷A
Stab(.) =
¸
v÷G
Stab(g.
0
)
(4.4)
=
¸
g Stab(.
0
) g
1
.
Hence, the proposition is a consequence of the following lemma.
2
LEMMA 4.10 For any subgroup H of a group G,
¸
v÷G
gHg
1
is the largest normal
subgroup contained in H.
PROOF. Note that N
0
def
=
¸
v÷G
gHg
1
, being an intersection of subgroups, is itself a sub
group. It is normal because
g
1
N
0
g
1
1
=
¸
v÷G
(g
1
g)N
0
(g
1
g)
1
=N
0
— for the second equality, we used that, as g runs over the elements of G, so also does
g
1
g. Thus N
0
is a normal subgroup of G contained in eHe
1
=H. If N is a second such
group, then
N =gNg
1
cgHg
1
for all g ÷ G, and so
N c
¸
v÷G
gHg
1
=N
0
.
2
58 4. GROUPS ACTING ON SETS
THE CLASS EQUATION
When X is ﬁnite, it is a disjoint union of a ﬁnite number of orbits:
X =
n
¸
i=1
O
i
(disjoint union).
Hence:
PROPOSITION 4.11 The number of elements in X is
[X[ =
n
¸
i=1
[O
i
[ =
n
¸
i=1
(G : Stab(.
i
)). .
i
in O
i
. (20)
When G acts on itself by conjugation, this formula becomes:
PROPOSITION 4.12 (CLASS EQUATION)
[G[ =
¸
(G : C
G
(.)) (21)
(. runs over a set of representatives for the conjugacy classes), or
[G[ =[7(G)[ ÷
¸
(G : C
G
(.)) (22)
(. runs over set of representatives for the conjugacy classes containing more than one ele
ment).
THEOREM 4.13 (CAUCHY) If the prime ] divides [G[, then G contains an element of
order ].
PROOF. We use induction on [G[. If for some . not in the centre of G, ] does not divide
(G : C
G
(.)), then ][C
G
(.) and we can apply induction to ﬁnd an element of order ]
in C
G
(.). Thus we may suppose that ] divides all of the terms (G : C
G
(.)) in the class
equation (second form), and so also divides 7(G). But 7(G) is commutative, and it follows
from the structure theorem
1
of such groups that 7(G) will contain an element of order ].
2
COROLLARY 4.14 A ﬁnite group G is a ]group if and only if every element has order a
power of ].
PROOF. If [G[ is a power of ], then Lagrange’s theorem (1.26) shows that the order of
every element is a power of ]. The converse follows from Cauchy’s theorem.
2
COROLLARY 4.15 Every group of order 2], ] an odd prime, is cyclic or dihedral.
1
Here is a direct proof that the theorem holds for an abelian group 7. We use induction on the order of 7.
It sufﬁces to show that 7 contains an element whose order is divisible by ]. because then some power of the
element will have order exactly ]. Let g =1 be an element of 7. If ] doesn’t divide the order of g, then it
divides the order of 7¡'g), in which case there exists (by induction) an element of G whose order in 7¡'g) is
divisible by ]. But the order of such an element must itself be divisible by ].
Deﬁnition and examples 59
PROOF. From Cauchy’s theorem, we know that such a G contains elements s and r of or
ders 2 and ] respectively. Let H ='r). Then H is of index 2, and so is normal. Obviously
s … H, and so G =H LHs :
G ={1. r. . . . . r
;1
. s. rs. . . . . r
;1
s].
As H is normal, srs
1
=r
i
, some i . Because s
2
=1, r =s
2
rs
2
=s(srs
1
)s
1
=r
i
2
,
and so i
2
÷1 mod ]. Because Z¡]Z is a ﬁeld, its only elements with square 1 are ˙1, and
so i ÷1 or ÷1 mod ]. In the ﬁrst case, the group is commutative (any group generated by
a set of commuting elements is obviously commutative); in the second srs
1
= r
1
and
we have the dihedral group (2.9).
2
]GROUPS
THEOREM 4.16 Every nontrivial ﬁnite ]group has nontrivial centre.
PROOF. By assumption, (G : 1) is a power of ], and so (G : C
G
(.)) is power of ] (=]
0
)
for all . not in the centre of G. As ] divides every term in the class equation (22) except
(perhaps) [7(G)[, it must divide [7(G)[ also.
2
COROLLARY 4.17 A group of order ]
n
has normal subgroups of order ]
n
for all m_n.
PROOF. We use induction on n. The centre of G contains an element g of order ], and so
N ='g) is a normal subgroup of G of order ]. Now the induction hypothesis allows us to
assume the result for G¡N. and the correspondence theorem (1.47) then gives it to us for
G.
2
PROPOSITION 4.18 Every group of order ]
2
is commutative, and hence is isomorphic to
C
;
×C
;
or C
;
2.
PROOF. We know that the centre 7 is nontrivial, and that G¡7 therefore has order 1 or ].
In either case it is cyclic, and the next result implies that G is commutative.
2
LEMMA 4.19 Suppose G contains a subgroup H in its centre (hence H is normal) such
that G¡H is cyclic. Then G is commutative.
PROOF. Let a be an element of G whose image in G¡H generates it. Then every element
of G can be written g =a
i
h with h ÷ H, i ÷ Z. Now
a
i
h a
i
0
h
t
=a
i
a
i
0
hh
t
because H c7(G)
=a
i
0
a
i
h
t
h
=a
i
0
h
t
a
i
h.
2
REMARK 4.20 The above proof shows that if H c 7(G) and G contains a set of repre
sentatives for G¡H whose elements commute, then G is commutative.
60 4. GROUPS ACTING ON SETS
For ] odd, it is now not difﬁcult to show that any noncommutative group of order ]
3
is isomorphic to exactly one of the groups constructed in (3.14, 3.15) (Exercise 43). Thus,
up to isomorphism, there are exactly two noncommutative groups of order ]
3
.
EXAMPLE 4.21 Let G be a noncommutative group of order 8. Then G must contain an
element a of order 4 (see Exercise 16). If G contains an element b of order 2 not in
'a), then G . 'a)
0
'b) where 0 is the unique isomorphism Z¡2Z ÷(Z¡4Z)
×
, and so
G ~ D
4
. If not, any element b of G not in 'a) must have order 4, and a
2
= b
2
. Now
bab
1
is an element of order 4 in 'a). It can’t equal a, because otherwise G would be
commutative, and so bab
1
=a
3
. Therefore G is the quaternion group (1.18, 2.7b).
ACTION ON THE LEFT COSETS
The action of G on the set of left cosets G¡H of H in G is a very useful tool in the study
of groups. We illustrate this with some examples.
Let X =G¡H. Recall that, for any g ÷ G,
Stab(gH) =gStab(H)g
1
=gHg
1
and the kernel of
G ÷Sym(X)
is the largest normal subgroup
¸
v÷G
gHg
1
of G contained in H.
REMARK 4.22 (a) Let H be a subgroup of G not containing a normal subgroup of G other
than 1. Then G ÷Sym(G¡H) is injective, and we have realized G as a subgroup of a
symmetric group of order much smaller than (G : 1)Š. For example, if G is simple, then the
Sylow theorems (see Chapter 5) show that G has many proper subgroups H =1 (unless G
is cyclic), but (by deﬁnition) it has no such normal subgroup.
(b) If (G : 1) does not divide (G : H)Š, then
G ÷Sym(G¡H)
can’t be injective (Lagrange’s theorem, 1.26), and we can conclude that H contains a nor
mal subgroup =1 of G. For example, if G has order 99, then it will have a subgroup N of
order 11 (Cauchy’s theorem, 4.13), and the subgroup must be normal. In fact, G =N ×O.
EXAMPLE 4.23 Corollary 4.15 shows that every group G of order 6 is either cyclic or
dihedral. Here we present a slightly different argument. According to Cauchy’s theorem
(4.13), G must contain an element r of order 3 and an element s of order 2. Moreover
N
def
= 'r) must be normal because 6 doesn’t divide 2Š (or simply because it has index 2).
Let H = 's). Either (a) H is normal in G, or (b) H is not normal in G. In the ﬁrst
case, rsr
1
= s, i.e., rs = sr, and so G . 'r) ×'s) ~ C
2
×C
3
. In the second case,
G ÷Sym(G¡H) is injective, hence surjective, and so G ~S
3
~D
3
.
Permutation groups
Consider Sym(X) where X has n elements. Since (up to isomorphism) a symmetry group
Sym(X) depends only on the number of elements in X, we may take X = {1. 2. . . . . n],
Permutation groups 61
and so work with S
n
. The symbol
1 2 3 4 5 6 T
2 5 T 4 3 1 6
denotes the permutation sending 1 ÷2,
2 ÷5, 3 ÷7, etc..
Consider a permutation
o =
1 2 3 . . . n
o(1) o(2) o(3) . . . o(n)
.
The pairs (i. ; ) with i < ; and o(i ) > o(; ) are called the inversions of o, and o is said
to be even or odd according as the number its inversions is even or odd.. The signature,
sign(o), of o is ÷1 or ÷1 according as o is even or odd. For example, sign(o) =÷1 if o is
a transposition.
REMARK 4.24 To compute the signature of o, connect (by a line) each element i in the top
row to the element i in the bottom row, and count the number of times that the lines cross:
o is even or odd according as this number is even or odd. For example,
1 2 3 4 5
3 5 1 4 2
is even (6 intersections). This works, because there is one crossing for each inversion.
For a permutation o, consider the products
V =
¸
1_i~}_n
(; ÷i ) = (2÷1)(3÷1) (n÷1)
(3÷2) (n÷2)
(n÷(n÷1))
oV =
¸
1_i~}_n
(o(; ) ÷o(i )) = (o(2) ÷o(1))(o(3) ÷o(1)) (o(n) ÷o(1))
(o(3) ÷o(2)) (o(n) ÷o(2))
(o(n) ÷o(n÷1)).
The terms in the products are the same except that each inversion introduces a negative
sign.
2
Therefore,
oV =sign(o)V.
Now let 1 be the additive group of maps Z
n
÷Z. For } ÷ 1 and o ÷ S
n
, let o} be
the element of 1 deﬁned by
(o} )(:
1
. . . . . :
n
) =}(:
c(1)
. . . . . :
c(n)
).
For o. t ÷ S
n
, one ﬁnds that
3
o(t} ) =(ot)}. (23)
2
Each is a product over the 2element subsets of {1. 2. . . . . n]; the factor corresponding to the subset {i. ; ]
is ˙(; ÷i ).
3
For . ÷ Z
n
and o ÷ S
n
, let .
c
be the element of Z
n
such that (.
c
)
i
= .
c(i)
. Then (.
c
)
r
= .
cr
. By
deﬁnition, (o} )(.) =}(.
c
). Therefore
(o(t} ))(.) =(t} )(.
c
) =}((.
c
)
r
) =}(.
cr
) =((ot)} )(.).
62 4. GROUPS ACTING ON SETS
Let ] be the element of 1 deﬁned by
](:
1
. . . . . :
n
) =
¸
1_i~}_n
(:
}
÷:
i
).
The same argument as above shows that
o] =sign(o)].
On putting } =] in (23) and using that ] =0, one ﬁnds that
sign(o)sign(t) =sign(ot).
Therefore, “sign” is a homomorphism S
n
÷{˙1]. When n _ 2, it is surjective, and so its
kernel is a normal subgroup of S
n
of order
nŠ
2
, called the alternating group ·
n
.
REMARK 4.25 Clearly sign is the unique homomorphism S
n
÷{˙1] such that sign(o) =
÷1 for every transposition o. Now let G = Sym(X) where X is a set with n elements.
Once we have chosen an ordering of X, we can speak of the inversions of an element o of
G. Deﬁne c(o) to be ÷1 or ÷1 according as o has an even or an odd number of inversions.
The same arguments as above show that c is the unique homomorphism G ÷{˙1] such
that c(o) = ÷1 for every transposition o. In particular, it is independent of the choice of
the ordering. In other words, the parity of the number of inversions of o is independent of
the choice of the ordering on X. Can you prove this directly?
A cycle is a permutation of the following form
i
1
÷i
2
÷i
3
÷ ÷i
i
÷i
1
. remaining i ’s ﬁxed.
The i
}
are required to be distinct. We denote this cycle by (i
1
i
2
...i
i
), and call r its length
— note that r is also its order as an element of S
n
. A cycle of length 2 is a transposition.
A cycle (i ) of length 1 is the identity map. The support of the cycle (i
1
. . . i
i
) is the set
{i
1
. . . . . i
i
], and cycles are said to be disjoint if their supports are disjoint. Note that disjoint
cycles commute. If
o =(i
1
...i
i
)(;
1
...;
x
) (l
1
...l
u
) (disjoint cycles).
then
o
n
=(i
1
...i
i
)
n
(;
1
...;
x
)
n
(l
1
...l
u
)
n
(disjoint cycles).
and it follows that o has order lcm(r. s. .... u).
PROPOSITION 4.26 Every permutation can be written (in essentially one way) as a product
of disjoint cycles.
PROOF. Let o ÷ S
n
, and let O c {1. 2. . . . . n] be an orbit for 'o). If [O[ =r, then for any
i ÷ O.
O ={i. o(i ). . . . . o
i1
(i )].
Therefore o and the cycle (i o(i ) . . . o
i1
(i )) have the same action on any element of O.
Let
{1. 2. . . . . n] =
n
¸
}=1
O
}
Permutation groups 63
be the decomposition of {1. . . . . n] into a disjoint union of orbits for 'o), and let ,
}
be the
cycle associated (as above) with O
}
. Then
o =,
1
,
n
is a decomposition of o into a product of disjoint cycles. For the uniqueness, note that
a decomposition o = ,
1
,
n
into a product of disjoint cycles must correspond to a de
composition of {1. .... n] into orbits (ignoring cycles of length 1 and orbits with only one
element). We can drop cycles of length one, change the order of the cycles, and change how
we write each cycle (by choosing different initial elements), but that’s all because the orbits
are intrinsically attached to o.
2
For example,
1 2 3 4 5 6 7 8
5 7 4 2 1 3 6 8
=(15)(27634)(8). (24)
It has order lcm(2. 5) =10.
COROLLARY 4.27 Each permutation o can be written as a product of transpositions; the
number of transpositions in such a product is even or odd according as o is even or odd.
PROOF. The cycle
(i
1
i
2
...i
i
) =(i
1
i
2
) (i
i2
i
i1
)(i
i1
i
i
).
and so the ﬁrst statement follows from the proposition. Because sign is a homomorphism,
and the signature of a transposition is ÷1, sign(o) =(÷1)
#transpositions
.
2
Note that the formula in the proof shows that the signature of a cycle of length r is
(÷1)
i1
, that is, an rcycle is even or odd according as r is odd or even.
It is possible to deﬁne a permutation to be even or odd according as it is a product of an
even or odd number of transpositions, but then one has to go through an argument as above
to show that this is a welldeﬁned notion.
The corollary says that S
n
is generated by transpositions. For ·
n
there is the following
result.
COROLLARY 4.28 The alternating group ·
n
is generated by cycles of length three.
PROOF. Any o ÷ ·
n
is the product (possibly empty) of an even number of transpositions,
o =t
1
t
t
1
t
n
t
t
n
, but the product of two transpositions can always be written as a product
of 3cycles:
(i; )(kl) =
ˆ
¸
ˆ
¸
(i; )(;l) =(i;l) case ; =k.
(i; )(;k)(;k)(kl) =(i;k)(;kl) case i. ;. k. l distinct,
1 case (i; ) =(kl).
2
Recall that two elements a and b of a group G are said to be conjugate a  b if there
exists an element g ÷G such that b =gag
1
, and that conjugacy is an equivalence relation.
For a group G, it is useful to determine the conjugacy classes in G.
64 4. GROUPS ACTING ON SETS
EXAMPLE 4.29 In S
n
, the conjugate of a cycle is given by:
g(i
1
. . . i
k
)g
1
=(g(i
1
). . . g(i
k
)).
Hence g(i
1
. . . i
i
) (l
1
. . . l
u
)g
1
= (g(i
1
). . . g(i
i
)) (g(l
1
)...g(l
u
)) (even if the cycles
are not disjoint, because conjugation is a homomorphism). In other words, to obtain gog
1
,
replace each element in each cycle of o by its image under g.
We shall now determine the conjugacy classes in S
n
. By a partition of n, we mean a
sequence of integers n
1
. . . . . n
k
such that
1 _n
1
_n
2
_ _n
k
_n and
n
1
÷n
2
÷ ÷n
k
=n.
For example, there are exactly 5 partitions of 4, namely,
4 =1÷1÷1÷1. 4 =1÷1÷2. 4 =1÷3. 4 =2÷2. 4 =4.
and 1. 121. 505 partitions of 61. Note that a partition
{1. 2. .... n] =O
1
L... LO
k
(disjoint union)
of {1. 2. . . . . n] determines a partition of n,
n =n
1
÷n
2
÷... ÷n
k
. n
i
=[O
i
[.
provided the numbering has been chosen so that [O
i
[ _ [O
i÷1
[. Since the orbits of an
element o of S
n
form a partition of {1. . . . . n], we can attach to each such o a partition of
n. For example, the partition of 8 attached to (15)(27634)(8) is 1. 2. 5 and the partition
attached of n attached to
o =(i
1
. . . i
n
1
) (l
1
. . . l
n
k
). (disjoint cycles) 1 < n
i
_n
i÷1
.
is 1. 1. . . . . 1. n
1
. . . . . n
k
(n÷
¸
n
i
ones).
PROPOSITION 4.30 Two elements o and t of S
n
are conjugate if and only if they deﬁne
the same partitions of n.
PROOF. ==: We saw in (4.29) that conjugating an element preserves the type of its dis
joint cycle decomposition.
==: Since o and t deﬁne the same partitions of n, their decompositions into products
of disjoint cycles have the same type:
o =(i
1
. . . i
i
)(;
1
. . . ;
x
). . . (l
1
. . . l
u
).
t =(i
t
1
. . . i
t
i
)(;
t
1
. . . ;
t
x
). . . (l
t
1
. . . l
t
u
).
If we deﬁne g to be
i
1
i
i
;
1
;
x
l
1
l
u
i
t
1
i
t
i
;
t
1
;
t
x
l
t
1
l
t
u
.
then
gog
1
=t.
2
Permutation groups 65
EXAMPLE 4.31 (i;k) =(
1234...
i}k4...
)(123)(
1234...
i}k4...
)
1
.
REMARK 4.32 For 1 < k _ n, there are
n(n1)···(nk÷1)
k
distinct kcycles in S
n
. The
1
k
is
needed so that we don’t count
(i
1
i
2
. . . i
k
) =(i
k
i
1
. . . i
k1
) =. . .
k times. Similarly, it is possible to compute the number of elements in any conjugacy
class in S
n
, but a little care is needed when the partition of n has several terms equal. For
example, the number of permutations in S
4
of type (ab)(cJ) is
1
2
4×3
2
×
2×1
2
=3.
The
1
2
is needed so that we don’t count (ab)(cJ) = (cJ)(ab) twice. For S
4
we have the
following table:
Partition Element No. in Conj. Class Parity
1÷1÷1÷1 1 1 even
1÷1÷2 (ab) 6 odd
1÷3 (abc) 8 even
2÷2 (ab)(cJ) 3 even
4 (abcJ) 6 odd
Note that ·
4
contains exactly 3 elements of order 2, namely those of type 2 ÷2, and that
together with 1 they form a subgroup V . This group is a union of conjugacy classes, and is
therefore a normal subgroup of S
4
.
THEOREM 4.33 (GALOIS) The group ·
n
is simple if n _5
REMARK 4.34 For n =2, ·
n
is trivial, and for n =3, ·
n
is cyclic of order 3, and hence
simple; for n =4 it is nonabelian and nonsimple — it contains the normal, even character
istic, subgroup V (see 4.32).
LEMMA 4.35 Let N be a normal subgroup of ·
n
(n _ 5); if N contains a cycle of length
three, then it contains all cycles of length three, and so equals ·
n
(by 4.28).
PROOF. Let , be the cycle of length three in N, and let o be a second cycle of length three
in ·
n
. We know from (4.30) that o =g,g
1
for some g ÷ S
n
. If g ÷ ·
n
, then this shows
that o is also in N. If not, because n _ 5, there exists a transposition t ÷ S
n
disjoint from
o. Then tg ÷ ·
n
and
o =t ot
1
=tg,g
1
t
1
.
and so again o ÷ N.
2
The next lemma completes the proof of the Theorem.
LEMMA 4.36 Every normal subgroup N of ·
n
, n _ 5, N =1, contains a cycle of length
3.
66 4. GROUPS ACTING ON SETS
PROOF. Let o ÷ N, o =1. If o is not a 3cycle, we shall construct another element o
t
÷ N,
o
t
=1, which ﬁxes more elements of {1. 2. . . . . n] than does o. If o
t
is not a 3cycle, then
we can apply the same construction. After a ﬁnite number of steps, we arrive at a 3cycle.
Suppose o is not a 3cycle. When we express it as a product of disjoint cycles, either it
contains a cycle of length _3 or else it is a product of transpositions, say
(i) o =(i
1
i
2
i
3
...) or
(ii) o =(i
1
i
2
)(i
3
i
4
) .
In the ﬁrst case, o moves two numbers, say i
4
, i
5
, other than i
1
, i
2
, i
3
, because o =
(i
1
i
2
i
3
), (i
1
. . . i
4
). Let , =(i
3
i
4
i
5
). Then o
1
def
=,o,
1
=(i
1
i
2
i
4
. . .) ÷N, and is distinct
from o (because it acts differently on i
2
). Thus o
t
def
=o
1
o
1
=1, but o
t
=,o,
1
o
1
ﬁxes
i
2
and all elements other than i
1
. .... i
5
ﬁxed by o — it therefore ﬁxes more elements than
o.
In the second case, form ,, o
1
, o
t
as in the ﬁrst case with i
4
as in (ii) and i
5
any
element distinct from i
1
. i
2
. i
3
. i
4
. Then o
1
= (i
1
i
2
)(i
4
i
5
) is distinct from o because
it acts differently on i
4
. Thus o
t
= o
1
o
1
= 1, but o
t
ﬁxes i
1
and i
2
, and all elements
=i
1
. .... i
5
not ﬁxed by o — it therefore ﬁxes at least one more element than o.
2
COROLLARY 4.37 For n _5, the only normal subgroups of S
n
are 1. ·
n
, and S
n
.
PROOF. If N is normal in S
n
, then N ¨·
n
is normal in ·
n
. Therefore either N ¨·
n
=·
n
or N ¨·
n
={1]. In the ﬁrst case, N ¯ ·
n
, which has index 2 in S
n
, and so N =·
n
or
S
n
. In the second case, the map . ÷.·
n
: N ÷S
n
¡·
n
is injective, and so N has order 1
or 2, but it can’t have order 2 because no conjugacy class in S
n
(other than {1]) consists of
a single element.
2
ASIDE 4.38 There exists a description of the conjugacy classes in ·
n
, from which it is possible to
deduce its simplicity for n _5 (see Exercise 411).
ASIDE 4.39 A group G is said to be solvable if there exist subgroups
G =G
0
¯G
1
¯ ¯G
i1
¯G
i
¯ ¯G
i
={1]
such that each G
i
is normal in G
i1
and each quotient G
i1
¡G
i
is commutative. Thus ·
n
(also S
n
)
is not solvable if n _5. Let }(X) ÷ QŒX be of degree n.
In Galois theory, one attaches to } a subgroup G
(
of the group of permutations of the roots
of } , and shows that the roots of } can be obtained from the coefﬁcients of } by the algebraic
operations of addition, subtraction, multiplication, division, and the extraction of mth roots if and
only if G
(
is solvable (Galois’s theorem). For every n, there exist lots of polynomials } of degree
n with G
(
~S
n
, and hence (when n _5) lots of polynomials not solvable in radicals.
The ToddCoxeter algorithm.
Let G be a group described by a ﬁnite presentation, and let H be a subgroup described by a
generating set. Then the ToddCoxeter algorithm
4
is a strategy for writing down the set of
4
To solve a problem, an algorithm must always terminate in a ﬁnite time with the correct answer to the
problem. The ToddCoxeter algorithm does not solve the problem of determining whether a ﬁnite presentation
deﬁnes a ﬁnite group (in fact, there is no such algorithm). It does, however, solve the problem of determining
the order of a ﬁnite group from a ﬁnite presentation of the group (use the algorithm with H the trivial subgroup
1.)
The ToddCoxeter algorithm. 67
left cosets of H in G together with the action of G on the set. I illustrate it with an example
(from Artin 1991, 6.9, which provides more details, but note that he composes permutations
in the reverse direction from us).
Let G = 'a. b. c [ a
3
. b
2
. c
2
. cba) and let H be the subgroup generated by c (strictly
speaking, H is the subgroup generated by the element of G represented by the reduced
word c). The operation of G on the set of cosets is described by the action of the generators,
which must satisfy the following rules:
(i) Each generator (a. b. c in our example) acts as a permutation.
(ii) The relations (a
3
. b
2
. c
2
. cba in our example) act trivially.
(iii) The generators of H (c in our example) ﬁx the coset 1H.
(iv) The operation on the cosets is transitive.
The strategy is to introduce cosets, denoted 1. 2. . . . with 1 =1H, as necessary.
Rule (iii) tells us simply that c1 = c. We now apply the ﬁrst two rules. Since we
don’t know what a1 is, let’s denote it 2: a1 = 2. Similarly, let a2 = 3. Now a3 = a
3
1,
which according to (ii) must be 1. Thus, we have introduced three (potential) cosets 1, 2, 3,
permuted by a as follows:
1
o
÷2
o
÷3
o
÷1.
What is b1? We don’t know, and so it is prudent to introduce another coset 4 =b1. Now
b4 =1 because b
2
=1, and so we have
1
b
÷4
b
÷1.
We still have the relation cba. We know a1 =2, but we don’t know what b2 is, and so we
set b2 =5:
1
o
÷2
b
÷5.
By (iii) c1 =1, and by (ii) applied to cba we have c5 =1. Therefore, according to (i) we
must have 5 =1; we drop 5, and so now b2 =1. Since b4 =1 we must have 4 =2, and so
we can drop 4 also. What we know can be summarized by the table:
a a a b b c c a b c
1 2 3 1 2 1 1 1 2 1 1
2 3 1 2 1 2 2 3 2
3 1 2 3 3 3 1 2 3
The bottom right corner, which is forced by (ii), tells us that c2 = 3. Hence also c3 = 2,
and this then determines the rest of the table:
a a a b b c c a b c
1 2 3 1 2 1 1 1 2 1 1
2 3 1 2 1 2 3 2 3 3 2
3 1 2 3 3 3 2 3 1 2 3
We ﬁnd that we have three cosets on which a. b. c act as
a =(123) b =(12) c =(23).
More precisely, we have written down a map G ÷S
3
that is consistent with the above
rules. A theorem (Artin 1991, 9.10) now says that this does in fact describe the action of
68 4. GROUPS ACTING ON SETS
G on G¡H. Since the three elements (123), (12), and (23) generate S
3
, this shows that the
action of G on G¡H induces an isomorphism G ÷S
3
, and that H is a subgroup of order
2.
In Artin 1991, 6.9, it is explained how to make this procedure into an algorithm which,
when it succeeds in producing a consistent table, will in fact produce the correct table.
This algorithm is implemented in Gap — see the worksheet.
Primitive actions.
Let G be a group acting on a set X, and let ¬ be a partition of X. We say that ¬ is stabilized
by G if
· ÷ ¬ == g· ÷ ¬.
It sufﬁces to check the condition for a set of generators for G.
EXAMPLE 4.40 (a) The subgroup G='(1234)) of S
4
stabilizes the partition {{1. 3]. {2. 4]]
of {1. 2. 3. 4].
(b) Identify X = {1. 2. 3. 4] with the set of vertices of the square on which D
4
acts
in the usual way, namely, with r = (1234), s = (2. 4). Then D
4
stabilizes the partition
{{1. 3]. {2. 4]] (opposite vertices stay opposite).
(c) Let X be the set of partitions of {1. 2. 3. 4] into two sets, each with two elements.
Then S
4
acts on X, and Ker(S
4
÷Sym(X)) is the subgroup V deﬁned in (4.32).
The group G always stabilizes the trivial partitions of X, namely, the set of all one
element subsets of X, and {X]. When it stabilizes only those partitions, we say that the
action is primitive; otherwise it is imprimitive. A subgroup of Sym(X) (e.g., of S
n
) is said
to be primitive if it acts primitively on X. Obviously, S
n
itself is primitive, but Example
4.40b shows that D
4
, regarded as a subgroup of S
4
in the obvious way, is not primitive.
EXAMPLE 4.41 A doubly transitive action is primitive: if it stabilized
{{.. .
t
. ...]. {.. ...]...],
then there would be no element sending (.. .
t
) to (.. .).
REMARK 4.42 The Gorbits form a partition of X that is stabilized by G. If the action is
primitive, then the partition into orbits must be one of the trivial ones. Hence
action primitive == action transitive or trivial.
For the remainder of this section, G is a ﬁnite group acting transitively on a set X with at
least two elements.
PROPOSITION 4.43 The group G acts imprimitively if and only if there is a proper subset
· of X with at least 2 elements such that,
for each g ÷ G, either g· =· or g·¨· =0. (25)
PROOF. ==: The partition ¬ stabilized by G contains such an ·.
==: From such an ·, we can form a partition {·. g
1
·. g
2
·. ...] of X, which is stabi
lized by G.
2
Exercises 69
A subset · of X satisfying (25) is called block.
PROPOSITION 4.44 Let · be a block in X with [·[ _2 and · =X. For any . ÷ ·,
Stab(.) ´Stab(·) ´G.
PROOF. We have Stab(·) ¯Stab(.) because
g. =. == g·¨· =0 == g· =·.
Let . ÷ ·, . = .. Because G acts transitively on X, there is a g ÷ G such that g. = ..
Then g ÷ Stab(·), but g … Stab(.).
Let . … ·. There is a g ÷ G such that g. =., and then g … Stab(·).
2
THEOREM 4.45 The group G acts primitively on X if and only if, for one (hence all) . in
X, Stab(.) is a maximal subgroup of G.
PROOF. If G does not act primitively on X, then (see 4.43) there is a block · ´X with at
least two elements, and so (4.44) shows that Stab(.) will not be maximal for any . ÷ ·.
Conversely, suppose that there exists an . in X and a subgroup H such that
Stab(.) ´H ´G.
Then I claim that · =H. is a block =X with at least two elements.
Because H =Stab(.), H. ={.], and so {.] ´· ´X.
If g ÷ H, then g· =·. If g … H, then g· is disjoint from ·: for suppose gh. =h
t
.
some h
t
÷ H; then h
t1
gh ÷ Stab(.) cH, say h
t1
gh =h
tt
, and g =h
t
h
tt
h
1
÷ H.
2
Exercises
41 Let H
1
and H
2
be subgroups of a group G. Show that the maps of Gsets G¡H
1
÷
G¡H
2
are in natural onetoone correspondence with the elements gH
2
of G¡H
2
such that
H
1
cgH
2
g
1
.
42 (a) Show that a ﬁnite group can’t be equal to the union of the conjugates of a proper
subgroup.
(b) Give an example of a proper subset S of a ﬁnite group G such that G =
v÷G
gSg
1
.
43 Prove that any noncommutative group of order ]
3
, ] an odd prime, is isomorphic to
one of the two groups constructed in (3.14, 3.15).
44 Let ] be the smallest prime dividing (G : 1) (assumed ﬁnite). Show that any subgroup
of G of index ] is normal.
45 Show that a group of order 2m, m odd, contains a subgroup of index 2. (Hint: Use
Cayley’s theorem 1.22)
70 4. GROUPS ACTING ON SETS
46 For n _ 5, show that the kcycles in S
n
generate S
n
or ·
n
according as k is even or
odd.
47 Let G =GL
3
(F
2
).
(a) Show that (G : 1) =168.
(b) Let X be the set of lines through the origin in F
3
2
; show that X has 7 elements, and
that there is a natural injective homomorphism G ·÷Sym(X) =S
T
.
(c) Use Jordan canonical forms to show that G has six conjugacy classes, with 1, 21, 42,
56, 24, and 24 elements respectively. [Note that if M is a free F
2
Œ˛module of rank
one, then End
F
2
Œ˛]
(M) =F
2
Œ˛.]
(d) Deduce that G is simple.
48 Let G be a group. If Aut(G) is cyclic, prove that G is commutative; if further, G is
ﬁnite, prove that G is cyclic.
49 Show that S
n
is generated by (12). (13). . . . . (1n); also by (12). (23). . . . . (n÷1n).
410 Let 1 be a conjugacy class of a ﬁnite group G contained in a normal subgroup H
of G. Prove that 1 is a union of k conjugacy classes of equal size in H, where k =(G :
H C
G
(.)) for any . ÷ 1.
411 (a) Let o ÷ ·
n
. From Exercise 410 we know that the conjugacy class of o in S
n
either remains a single conjugacy class in ·
n
or breaks up as a union of two classes of equal
size. Show that the second case occurs ” o does not commute with an odd permutation
” the partition of n deﬁned by o consists of distinct odd integers.
(b) For each conjugacy class 1 in ·
T
, give a member of 1, and determine [1[.
412 Let G be the group with generators a. b and relations a
4
=1 =b
2
, aba =bab.
(a) Use the ToddCoxeter algorithm (with H = 1) to ﬁnd the image of G under the
homomorphism G ÷S
n
, n =(G : 1), given by Cayley’s Theorem 1.11. [No need to
include every step; just an outline will do.]
(b) Use Sage/Gap to check your answer.
413 Show that if the action of G on X is primitive and effective, then the action of any
normal subgroup H =1 of G is transitive.
414 (a) Check that ·
4
has 8 elements of order 3, and 3 elements of order 2. Hence it has
no element of order 6.
(b) Prove that ·
4
has no subgroup of order 6 (cf. 1.30). (Use 4.23.)
(c) Prove that ·
4
is the only subgroup of S
4
of order 12.
415 Let G be a group with a subgroup of index r. Prove:
(a) If G is simple, then (G : 1) divides rŠ.
(b) If r =2. 3. or 4, then G can’t be simple.
(c) There exists a nonabelian simple group with a subgroup of index 5.
Exercises 71
416 Prove that S
n
is isomorphic to a subgroup of ·
n÷2
.
417 Let H and 1 be subgroups of a group G. A double coset of H and 1 in G is a set
of the form
Ha1 ={hak [ h ÷ H, k ÷ 1]
for some a ÷ G.
(a) Show that the double cosets of H and 1 in G partition G.
(b) Let H ¨a1a
1
act on H ×1 by b(h. k) =(hb. a
1
b
1
ak). Show that the orbits
for this action are exactly the ﬁbres of the map (h. k) ÷hak: H ×1 ÷Ha1.
(c) (Double coset counting formula). Use (b) to show that
[Ha1[ =
[H[[1[
[H ¨a1a
1
[
.
CHAPTER 5
The Sylow Theorems; Applications
In this chapter, all groups are ﬁnite.
Let G be a group and let ] be a prime dividing (G: 1). A subgroup of G is called a
Sylow ]subgroup of G if its order is the highest power of ] dividing (G : 1). In other
words, H is a Sylow ]subgroup of G if it is a ]group and its index in G is prime to ].
The Sylow theorems state that there exist Sylow ]subgroups for all primes ] dividing
(G: 1), that the Sylow ]subgroups for a ﬁxed ] are conjugate, and that every ]subgroup
of G is contained in such a subgroup; moreover, the theorems restrict the possible number
of Sylow ]subgroups in G.
The Sylow theorems
In the proofs, we frequently use that if O is an orbit for a group H acting on a set X, and
.
0
÷ O, then the map H ÷X, h ÷h.
0
induces a bijection
H¡Stab(.
0
) ÷O:
see (4.7). Therefore
(H : Stab(.
0
)) =[O[.
In particular, when H is a ]group, [O[ is a power of ]: either O consists of a single
element, or [O[ is divisible by ]. Since X is a disjoint union of the orbits, we can conclude:
LEMMA 5.1 Let H be a ]group acting on a ﬁnite set X, and let X
1
be the set of points
ﬁxed by H; then
[X[ ÷[X
1
[ (mod ]).
When the lemma is applied to a ]group H acting on itself by conjugation, we ﬁnd that
(7(H) : 1) ÷(H : 1) mod ]
and so ][(7(H): 1) (cf. the proof of 4.16).
THEOREM 5.2 (SYLOW I) Let G be a ﬁnite group, and let ] be prime. If ]
i
[(G : 1), then
G has a subgroup of order ]
i
.
73
74 5. THE SYLOW THEOREMS; APPLICATIONS
PROOF. According to (4.17), it sufﬁces to prove this with ]
i
the highest power of ] divid
ing (G : 1), and so from now on we assume that (G : 1) =]
i
m with m not divisible by ].
Let
X ={subsets of G with ]
i
elements].
with the action of G deﬁned by
G×X ÷X. (g. ·) ÷g·
def
={ga [ a ÷ ·].
Let · ÷ X, and let
H =Stab(·)
def
={g ÷ G [ g· =·].
For any a
0
÷·, h ÷ha
0
: H ÷·is injective (cancellation law), and so (H : 1) _[·[ =]
i
.
In the equation
(G : 1) =(G : H)(H : 1)
we know that (G : 1) =]
i
m, (H : 1) _]
i
, and that (G : H) is the number of elements in
the orbit of ·. If we can ﬁnd an · such that ] doesn’t divide the number of elements in its
orbit, then we can conclude that (for such an ·), H =Stab· has order ]
i
.
The number of elements in X is
[X[ =
]
i
m
]
i
=
(]
i
m)(]
i
m÷1) (]
i
m÷i ) (]
i
m÷]
i
÷1)
]
i
(]
i
÷1) (]
i
÷i ) (]
i
÷]
i
÷1)
.
Note that, because i < ]
i
, the power of ] dividing ]
i
m÷i is the power of ] dividing
i . The same is true for ]
i
÷i . Therefore the corresponding terms on top and bottom are
divisible by the same powers of ], and so ] does not divide [X[. Because the orbits form a
partition of X,
[X[ =
¸
[O
i
[. O
i
the distinct orbits.
and so at least one of the [O
i
[ is not divisible by ].
2
EXAMPLE 5.3 Let F
;
= Z¡]Z, the ﬁeld with ] elements, and let G = GL
n
(F
;
). The
n×n matrices in G are precisely those whose columns form a basis for F
n
;
. Thus, the ﬁrst
column can be any nonzero vector in F
n
;
, of which there are ]
n
÷1; the second column
can be any vector not in the span of the ﬁrst column, of which there are ]
n
÷]; and so on.
Therefore, the order of G is
(]
n
÷1)(]
n
÷])(]
n
÷]
2
) (]
n
÷]
n1
).
and so the power of ] dividing (G : 1) is ]
1÷2÷···÷(n1)
. Consider the upper triangular
matrices with 1’s down the diagonal:
¸
¸
¸
¸
¸
¸
1 + + +
0 1 + +
0 0 1 +
.
.
.
.
.
.
.
.
.
.
.
.
0 0 0 1
.
They form a subgroup U of order ]
n1
]
n2
], which is therefore a Sylow ]subgroup
G.
The Sylow theorems 75
REMARK 5.4 The theorem gives another proof of Cauchy’s theorem (4.13). If a prime ]
divides (G: 1), then G will have a subgroup H of order ], and any g ÷ H, g = 1, is an
element of G of order ].
REMARK 5.5 The proof of Theorem 5.2 can be modiﬁed to show directly that for each
power ]
i
of ] dividing (G : 1) there is a subgroup H of G of order ]
i
. One again writes
(G : 1) =]
i
m and considers the set X of all subsets of order ]
i
. In this case, the highest
power ]
i
0
of ] dividing [X[ is the highest power of ] dividing m, and it follows that there
is an orbit in X whose order is not divisible by ]
i
0
÷1
. For an · in such an orbit, the same
counting argument shows that Stab(·) has ]
i
elements. We recommend that the reader
write out the details.
THEOREM 5.6 (SYLOW II) Let G be a ﬁnite group, and let [G[ =]
i
m with m not divisi
ble by ].
(a) Any two Sylow ]subgroups are conjugate.
(b) Let s
;
be the number of Sylow ]subgroups in G; then s
;
÷1 mod ] and s
;
[m.
(c) Every ]subgroup of G is contained in a Sylow ]subgroup.
Let H be a subgroup of G. Recall (4.6, 4.8) that the normalizer of H in G is
N
G
(H) ={g ÷ G [ gHg
1
=H].
and that the number of conjugates of H in G is (G : N
G
(H)).
LEMMA 5.7 Let 1 be a Sylow ]subgroup of G, and let H be a ]subgroup. If H nor
malizes 1, i.e., if H c N
G
(1), then H c 1. In particular, no Sylow ]subgroup of G
other than 1 normalizes 1.
PROOF. Because H and 1 are subgroups of N
G
(1) with 1 normal in N
G
(1), H1 is a
subgroup, and H¡H¨1 .H1¡1 (apply 1.46). Therefore (H1 : 1) is a power of ] (here
is where we use that H is a ]group), but
(H1 : 1) =(H1 : 1)(1 : 1).
and (1 : 1) is the largest power of ] dividing (G : 1), hence also the largest power of ]
dividing (H1 : 1). Thus (H1 : 1) =]
0
=1, and H c1.
2
PROOF (OF SYLOW II) (a) Let X be the set of Sylow ]subgroups in G, and let G act on
X by conjugation,
(g. 1) ÷g1g
1
: G×X ÷X.
Let O be one of the Gorbits: we have to show O is all of X.
Let 1 ÷ O, and let 1 act on O through the action of G. This single Gorbit may break
up into several 1orbits, one of which will be {1]. In fact this is the only onepoint orbit
because
{O] is a 1orbit ”1 normalizes O.
which we know (5.7) happens only for O = 1. Hence the number of elements in every
1orbit other than {1] is divisible by ], and we have that [O[ ÷1 mod ].
76 5. THE SYLOW THEOREMS; APPLICATIONS
Suppose there exists a 1 … O. We again let 1 act on O, but this time the argument
shows that there are no onepoint orbits, and so the number of elements in every 1orbit is
divisible by ]. This implies that #O is divisible by ], which contradicts what we proved in
the last paragraph. There can be no such 1, and so O is all of X.
(b) Since s
;
is now the number of elements in O, we have also shown that s
;
÷1 (mod
]).
Let 1 be a Sylow ]subgroup of G. According to (a), s
;
is the number of conjugates
of 1, which equals
(G : N
G
(1)) =
(G : 1)
(N
G
(1) : 1)
=
(G : 1)
(N
G
(1) : 1) (1 : 1)
=
m
(N
G
(1) : 1)
.
This is a factor of m.
(c) Let H be a ]subgroup of G, and let H act on the set X of Sylow ]subgroups by
conjugation. Because [X[ =s
;
is not divisible by ], X
1
must be nonempty (Lemma 5.1),
i.e., at least one Horbit consists of a single Sylow ]subgroup. But then H normalizes 1
and Lemma 5.7 implies that H c1.
2
COROLLARY 5.8 A Sylow ]subgroup is normal if and only if it is the only Sylow ]
subgroup.
PROOF. Let 1 be a Sylow ]subgroup of G. If 1 is normal, then (a) of Sylow II implies
that it is the only Sylow ]subgroup. The converse statement follows from (3.7c) (which
shows, in fact, that 1 is even characteristic).
2
COROLLARY 5.9 Suppose that a group G has only one Sylow ]subgroup for each prime
] dividing its order. Then G is a direct product of its Sylow ]subgroups.
PROOF. Let 1
1
. . . . . 1
k
be Sylow subgroups of G, and let [1
i
[ = ]
i
i
i
; the ]
i
are distinct
primes. Because each 1
i
is normal in G, the product 1
1
1
k
is a normal subgroup of G.
We shall prove by induction on k that it has order ]
i
1
1
]
i
k
k
. If k =1, there is nothing to
prove, and so we may suppose that k _2 and that 1
1
1
k1
has order ]
i
1
1
]
i
k1
k1
. Then
1
1
1
k1
¨1
k
=1; therefore (1.51) shows that (1
1
1
k1
)1
k
is the direct product of
1
1
1
k1
and 1
k
, and so has order ]
i
1
1
]
i
k
k
. Now (1.52) applied to the full set of Sylow
subgroups of G shows that G is their direct product.
2
EXAMPLE 5.10 Let G =GL(V ) where V is a vector space of dimension n over F
;
. There
is a geometric description of the Sylow subgroups of G. A full ﬂag J in V is a sequence
of subspaces
V =V
n
¯V
n1
¯ ¯V
i
¯ ¯V
1
¯{0]
with dimV
i
=i . Given such a ﬂag J, let U(J) be the set of linear maps ˛: V ÷V such
that
(a) ˛(V
i
) cV
i
for all i , and
(b) the endomorphism of V
i
¡V
i1
induced by ˛ is the identity map.
Alternative approach to the Sylow theorems 77
I claim that U(J) is a Sylow]subgroup of G. Indeed, we can construct a basis {e
1
. . . . . e
n
]
for V such {e
1
] is basis for V
1
, {e
1
. e
2
] is a basis for V
2
, and so on. Relative to this basis,
the matrices of the elements of U(J) are exactly the elements of the group U of (5.3).
Let g ÷ GL
n
(F). Then gJ
def
= {gV
n
. gV
n1
. . . .] is again a full ﬂag, and U(gJ) =
g U(J) g
1
. From (a) of Sylow II, we see that the Sylow ]subgroups of G are precisely
the groups of the form U(J) for some full ﬂag J.
ASIDE 5.11 Some books use different numberings for Sylow’s theorems. I have essentially fol
lowed the original (Sylow 1872).
Alternative approach to the Sylow theorems
We brieﬂy forget that we have proved the Sylow theorems.
THEOREM 5.12 Let G be a group, and let 1 be a Sylow ]subgroup of G. For any sub
group H of G, there exists an a ÷ G such that H ¨a1a
1
is a Sylow ]subgroup of H.
PROOF. Recall (Exercise 417) that G is a disjoint union of the double cosets for H and
1, and so
[G[ =
¸
o
[Ha1[ =
¸
o
[H[[1[
[H ¨a1a
1
[
where the sum is over a set of representatives for the double cosets. On dividing by [1[ we
ﬁnd that
[G[
[1[
=
¸
o
[H[
[H ¨a1a
1
[
.
and so there exists an a such that (H: H ¨a1a
1
) is not divisible by ]. For such an a,
H ¨a1a
1
is a Sylow ]subgroup of H.
2
PROOF (OF SYLOW I) According to Cayley’s theorem (1.22), G embeds into S
n
, and S
n
embeds into GL
n
(F
;
) (see 7.1b below). As GL
n
(F
;
) has a Sylow ]subgroup (see 5.3), so
also does G.
2
PROOF (OF SYLOW II(a,c)) Let 1 be a Sylow]subgroup of G, and let 1
t
be a ]subgroup
of G. Then 1
t
is the unique Sylow ]subgroup of 1
t
, and so the theorem with H = 1
t
shows that a1a
1
¯1
t
for some a. This implies (a) and (c) of Sylow II.
2
Examples
We apply what we have learnt to obtain information about groups of various orders.
5.13 (GROUPS OF ORDER 99) Let G have order 99. The Sylow theorems imply that G
has at least one subgroup H of order 11, and in fact s
11
ˇ
ˇ
99
11
and s
11
÷ 1 mod 11. It
follows that s
11
= 1, and H is normal. Similarly, s
9
[11 and s
9
÷ 1 mod 3, and so the
Sylow 3subgroup is also normal. Hence G is isomorphic to the direct product of its Sylow
subgroups (5.9), which are both commutative (4.18), and so G commutative.
Here is an alternative proof. Verify as before that the Sylow 11subgroup N of G is
normal. The Sylow 3subgroup O maps bijectively onto G¡N, and so G = N O. It
78 5. THE SYLOW THEOREMS; APPLICATIONS
remains to determine the action by conjugation of O on N. But Aut(N) is cyclic of order
10 (see 3.5), and so there is only the trivial homomorphism O ÷Aut(N). It follows that
G is the direct product of N and O.
5.14 (GROUPS OF ORDER ]q, ]. q PRIMES, ] < q) Let G be such a group, and let 1 and
O be Sylow ] and q subgroups. Then (G : O) =], which is the smallest prime dividing
(G : 1), and so (see Exercise 44) O is normal. Because 1 maps bijectively onto G¡O, we
have that
G =O1.
and it remains to determine the action of 1 on O by conjugation.
The group Aut(O) is cyclic of order q ÷1 (see 3.5), and so, unless ][q ÷1, G =O×1.
If ][q ÷1, then Aut(O) (being cyclic) has a unique subgroup 1
t
of order ]. In fact 1
t
consists of the maps
. ÷.
i
. {i ÷ Z¡qZ [ i
;
=1].
Let a and b be generators for 1 and O respectively, and suppose that the action of a on O
by conjugation is . ÷.
i
0
. i
0
=1 (in Z¡qZ). Then G has generators a. b and relations
a
;
. b
q
. aba
1
=b
i
0
.
Choosing a different i
0
amounts to choosing a different generator a for 1, and so gives an
isomorphic group G.
In summary: if ] [ q ÷1, then the only group of order ]q is the cyclic group C
;q
; if
][q ÷1, then there is also a nonabelian group given by the above generators and relations.
5.15 (GROUPS OF ORDER 30) Let G be a group of order 30. Then
s
3
=1. 4. 7. 10. . . . and divides 10:
s
5
=1. 6. 11. . . . and divides 6.
Hence s
3
=1 or 10, and s
5
=1 or 6. In fact, at least one is 1, for otherwise there would be
20 elements of order 3 and 24 elements of order 5, which is impossible. Therefore, a Sylow
3subgroup 1 or a Sylow 5subgroup O is normal, and so H = 1O is a subgroup of G.
Because 3 doesn’t divide 5 ÷1 = 4, (5.14) shows that H is commutative, H ~ C
3
×C
5
.
Hence
G =(C
3
×C
5
)
0
C
2
.
and it remains to determine the possible homomorphisms 0: C
2
÷Aut(C
3
×C
5
). But such
a homomorphism 0 is determined by the image of the nonidentity element of C
2
, which
must be an element of order 2. Let a, b, c generate C
3
, C
5
, C
2
. Then
Aut(C
3
×C
5
) =Aut(C
3
) ×Aut(C
5
).
and the only elements of Aut C
3
and Aut C
5
of order 2 are a ÷a
1
and b ÷b
1
. Thus
there are exactly 4 homomorphisms 0, and 0(c) is one of the following elements:
a ÷a
b ÷b
a ÷a
b ÷b
1
a ÷a
1
b ÷b
a ÷a
1
b ÷b
1
.
The groups corresponding to these homomorphisms have centres of order 30, 3 (generated
by a), 5 (generated by b), and 1 respectively, and hence are nonisomorphic. We have shown
Examples 79
that (up to isomorphism) there are exactly 4 groups of order 30. For example, the third on
our list has generators a. b. c and relations
a
3
. b
5
. c
2
. ab =ba. cac
1
=a
1
. cbc
1
=b.
5.16 (GROUPS OF ORDER 12) Let G be a group of order 12, and let 1 be its Sylow 3
subgroup. If 1 is not normal, then 1 doesn’t contain a nontrivial normal subgroup of G,
and so the map (4.2, action on the left cosets)
c : G ÷Sym(G¡1) ~S
4
is injective, and its image is a subgroup of S
4
of order 12. From Sylow II we see that G
has exactly 4 Sylow 3subgroups, and hence it has exactly 8 elements of order 3. But all
elements of S
4
of order 3 are in ·
4
(see the table in 4.32), and so c(G) intersects ·
4
in a
subgroup with at least 8 elements. By Lagrange’s theorem c(G) =·
4
, and so G ~·
4
.
Now assume that 1 is normal. Then G =1 O where O is the Sylow 4subgroup. If
O is cyclic of order 4, then there is a unique nontrivial map O(=C
4
) ÷Aut(1)(=C
2
),
and hence we obtain a single noncommutative group C
3
C
4
. If O =C
2
×C
2
, there are
exactly 3 nontrivial homomorphism 0: O÷Aut(1), but the three groups resulting are all
isomorphic to S
3
×C
2
with C
2
=Ker 0. (The homomorphisms differ by an automorphism
of O, and so we can also apply Lemma 3.18.)
In total, there are 3 noncommutative groups of order 12 and 2 commutative groups.
5.17 (GROUPS OF ORDER ]
3
) Let G be a group of order ]
3
, with ] an odd prime, and
assume G is not commutative. We know from (4.17) that G has a normal subgroup N of
order ]
2
.
If every element of G has order ] (except 1), then N ~C
;
×C
;
and there is a subgroup
O of G of order ] such that O¨N ={1]. Hence
G =N
0
O
for some homomorphism 0: O÷N. The order of Aut(N) ~GL
2
(F
;
) is (]
2
÷1)(]
2
÷])
(see 5.3), and so its Sylow ]subgroups have order ]. By the Sylow theorems, they are
conjugate, and so Lemma 3.19 shows that there is exactly one nonabelian group in this
case.
Suppose G has elements of order ]
2
, and let N be the subgroup generated by such an
element a. Because (G : N) =] is the smallest (in fact only) prime dividing (G : 1), N is
normal in G (Exercise 44). We next show that G contains an element of order ] not in N.
We know 7(G) = 1, and, because G isn’t commutative, that G¡7(G) is not cyclic
(4.19). Therefore (7(G) : 1) =] and G¡7(G) ~C
;
×C
;
. In particular, we see that for
all . ÷ G, .
;
÷ 7(G). Because G¡7(G) is commutative, the commutator of any pair of
elements of G lies in 7(G), and an easy induction argument shows that
(..)
n
=.
n
.
n
Œ.. .
n.n1/
2
. n _1.
Therefore (..)
;
=.
;
.
;
, and so . ÷.
;
: G ÷G is a homomorphism. Its image is con
tained in 7(G), and so its kernel has order at least ]
2
. Since N contains only ] ÷1 el
ements of order ], we see that there exists an element b of order ] outside N. Hence
80 5. THE SYLOW THEOREMS; APPLICATIONS
G ='a) 'b) ~C
;
2 C
;
, and it remains to observe (3.19) that the nontrivial homomor
phisms C
;
÷Aut(C
;
2) ~C
;
×C
;1
give isomorphic groups.
Thus, up to isomorphism, the only noncommutative groups of order ]
3
are those con
structed in (3.14, 3.15).
5.18 (GROUPS OF ORDER 2]
n
, 4]
n
. AND 8]
n
, ] ODD) Let G be a group of order 2
n
]
n
,
1 _ m_ 3, ] an odd prime, 1 _ n. We shall show that G is not simple. Let 1 be a Sylow
]subgroup and let N =N
G
(1), so that s
;
=(G : N).
From Sylow II, we know that s
;
[2
n
, s
;
=1. ]÷1. 2]÷1. . . .. If s
;
=1, 1 is normal.
If not, there are two cases to consider:
(i) s
;
=4 and ] =3, or
(ii) s
;
=8 and ] =7.
In the ﬁrst case, the action by conjugation of G on the set of Sylow3subgroups
1
deﬁnes
a homomorphism G ÷S
4
, which, if G is simple, must be injective. Therefore (G : 1)[4Š,
and so n =1; we have (G : 1) =2
n
3. Now the Sylow 2subgroup has index 3, and so we
have a homomorphism G ÷S
3
. Its kernel is a nontrivial normal subgroup of G.
In the second case, the same argument shows that (G : 1)[8Š, and so n =1 again. Thus
(G : 1) =56 and s
T
=8. Therefore G has 48 elements of order 7, and so there can be only
one Sylow 2subgroup, which must therefore be normal.
Note that groups of order ]q
i
, ]. q primes, ] <q are not simple, because Exercise 44
shows that the Sylow qsubgroup is normal. An examination of cases now reveals that ·
5
is the smallest noncyclic simple group.
5.19 (GROUPS OF ORDER 60) Let G be a simple group of order 60. We shall show that G
is isomorphic to ·
5
. Let 1 be a Sylow 2subgroup and N =N
G
(1), so that s
2
=(G : N).
According to the Sylow theorems, s
2
=1. 3. 5. or 15.
(a) The case s
2
=1 is impossible, because 1 would be normal (see 5.8).
(b) The case s
2
=3 is impossible, because the kernel of G ÷Sym(G¡N) would be a
nontrivial normal subgroup of G.
(c) In the case s
2
=5, we get an inclusion G ·÷Sym(G¡N) =S
5
, which realizes G as
a subgroup of index 2 in S
5
, but we saw in (4.37) that, for n _ 5, ·
n
is the only subgroup
of index 2 in S
n
.
(d) In the case s
2
=15, a counting argument (using that s
5
=6) shows that there exist
two Sylow 2subgroups 1 and O intersecting in a group of order 2. The normalizer N
of 1 ¨O contains 1 and O, and so it has index 1, 3, or 5 in G. The ﬁrst two cases are
impossible for the same reasons as in (a) and (b). If (G: N) =5, the argument in (c) gives
an isomorphism G ~·
5
; but this is impossible because s
2
(·
5
) =5.
Exercises
51 Show that a ﬁnite group (not necessarily commutative) is cyclic if, for each n > 0, it
contains at most n elements of order dividing n.
1
Equivalently, the usual map G ÷Sym(G¡N).
CHAPTER 6
Subnormal Series; Solvable and
Nilpotent Groups
Subnormal Series.
Let G be a group. A chain of subgroups
G =G
0
¯G
1
¯ ¯G
i
¯G
i÷1
¯ ¯G
n
={1].
is called a subnormal series if G
i
is normal in G
i1
for every i , and it is called a normal
series if G
i
is normal in G for every i .
1
The series is said to be without repetitions if all the
inclusions G
i1
¯G
i
are proper (i.e., G
i1
=G
i
). Then n is called the length of the series.
The quotient groups G
i1
¡G
i
are called the quotient (or factor) groups of the series.
A subnormal series is said to be a composition series if it has no proper reﬁnement that
is also a subnormal series. In other words, it is a composition series if G
i
is maximal among
the proper normal subgroups G
i1
for each i . Thus a subnormal series is a composition
series if and only if each quotient group is simple and nontrivial. Obviously, every ﬁnite
group has a composition series (usually many): choose G
1
to be a maximal proper normal
subgroup of G; then choose G
2
to be a maximal proper normal subgroup of G
1
, etc.. An
inﬁnite group may or may not have a ﬁnite composition series.
Note that from a subnormal series
G =G
0
>G
1
> >G
i
>G
i÷1
> >G
n
={1]
we obtain a sequence of exact sequences
1 ÷G
n1
÷G
n2
÷G
n2
¡G
n1
÷1
1 ÷G
i÷1
÷G
i
÷G
i
¡G
i÷1
÷1
1 ÷G
1
÷G
0
÷G
0
¡G
1
÷1.
Thus G is built up out of the quotients G
0
¡G
1
. G
1
¡G
2
. . . . . G
n1
by forming successive ex
tensions. In particular, since every ﬁnite group has a composition series, it can be regarded
1
Some authors write “normal series” where we write “subnormal series” and “invariant series” where we
write “normal series”.
81
82 6. SUBNORMAL SERIES; SOLVABLE AND NILPOTENT GROUPS
as being built up out of simple groups. The JordanH¨ older theorem, which is the main topic
of this section, says that these simple groups are independent of the composition series (up
to order and isomorphism).
Note that if G has a subnormal series G =G
0
>G
1
> >G
n
={1], then
(G : 1) =
¸
1_i_n
(G
i1
: G
i
) =
¸
1_i_n
(G
i1
¡G
i
: 1).
EXAMPLE 6.1 (a) The symmetric group S
3
has a composition series
S
3
>·
3
>1
with quotients C
2
, C
3
.
(b) The symmetric group S
4
has a composition series
S
4
>·
4
>V >'(13)(24)) >1.
where V ~C
2
×C
2
consists of all elements of order 2 in ·
4
(see 4.32). The quotients are
C
2
, C
3
, C
2
, C
2
.
(c) Any full ﬂag in F
n
;
, ] a prime, is a composition series. Its length is n, and its
quotients are C
;
. C
;
. . . . . C
;
.
(d) Consider the cyclic group C
n
='a). For any factorization m=]
1
]
i
of m into
a product of primes (not necessarily distinct), there is a composition series
C
n
> C m
p
1
> C m
p
1
p
2
>
  
'a) 'a
;
1
) 'a
;
1
;
2
)
The length is r, and the quotients are C
;
1
. C
;
2
. . . . . C
;
r
.
(e) Suppose G is a direct product of simple groups, G =H
1
× ×H
i
. Then G has a
composition series
G>H
2
× ×H
i
>H
3
× ×H
i
>
of length r and with quotients H
1
. H
2
. . . . . H
i
. Note that for any permutation o of {1. 2. . . . r],
there is another composition series with quotients H
c(1)
. H
c(2)
. . . . . H
c(i)
.
(f) We saw in (4.37) that for n _ 5, the only normal subgroups of S
n
are S
n
, ·
n
, {1],
and in (4.33) that ·
n
is simple. Hence S
n
>·
n
>{1] is the only composition series for S
n
.
THEOREM 6.2 (JORDANH¨ OLDER)
2
Let G be a ﬁnite group. If
G =G
0
>G
1
> >G
x
={1]
G =H
0
>H
1
> >H
t
={1]
are two composition series for G, then s = t and there is a permutation o of {1. 2. . . . . s]
such that G
i
¡G
i÷1
~H
c(i)
¡H
c(i)÷1
.
PROOF. We use induction on the order of G.
Case I: H
1
=G
1
. In this case, we have two composition series for G
1
, to which we can
apply the induction hypothesis.
2
Jordan showed that corresponding quotients had the same order, and H¨ older that they were isomorphic.
Solvable groups 83
Case II: H
1
=G
1
. Because G
1
and H
1
are both normal in G, the product G
1
H
1
is a
normal subgroup of G. It properly contains both G
1
and H
1
, which are maximal normal
subgroups of G, and so G
1
H
1
=G. Therefore
G¡G
1
=G
1
H
1
¡G
1
.H
1
¡G
1
¨H
1
(see 1.46).
Similarly G¡H
1
.G
1
¡G
1
¨H
1
. Let 1
2
=G
1
¨H
1
; then 1
2
is a maximal normal sub
group in both G
1
and H
1
, and
G¡G
1
.H
1
¡1
2
. G¡H
1
.G
1
¡1
2
. (26)
Choose a composition series
1
2
>1
3
> >1
u
.
We have the picture:
G
1
> G
2
> > G
x
`
G 1
2
> > 1
u
`
H
1
> H
2
> > H
t
.
On applying the induction hypothesis to G
1
and H
1
and their composition series in the
diagram, we ﬁnd that
Quotients(G>G
1
>G
2
> ) = {G¡G
1
. G
1
¡G
2
. G
2
¡G
3
. . . .] (deﬁnition)
 {G¡G
1
. G
1
¡1
2
. 1
2
¡1
3
. . . .] (induction)
 {H
1
¡1
2
. G¡H
1
. 1
2
¡1
3
. . . .] (apply (26))
 {G¡H
1
. H
1
¡1
2
. 1
2
¡1
3
. . . .] (reorder)
 {G¡H
1
. H
1
¡H
2
. H
2
¡H
3
. . . .] (induction)
= Quotients(G>H
1
>H
2
> ) (deﬁnition).
2
Note that the theorem applied to a cyclic group C
n
implies that the factorization of an
integer into a product of primes is unique.
REMARK 6.3 There are inﬁnite groups having ﬁnite composition series (there are even
inﬁnite simple groups). For such a group, let J(G) be the minimum length of a composition
series. Then the JordanH¨ older theorem extends to show that all composition series have
length J(G) and have isomorphic quotient groups. The same proof works except that you
have to use induction on J(G) instead of [G[ and verify that a normal subgroup of a group
with a ﬁnite composition series also has a ﬁnite composition series (Exercise 61).
The quotients of a composition series are sometimes called composition factors.
Solvable groups
A subnormal series whose quotient groups are all commutative is called a solvable series.
A group is solvable (or soluble) if it has a solvable series. Alternatively, we can say that
a group is solvable if it can be obtained by forming successive extensions of commutative
groups. Since a commutative group is simple if and only if it is cyclic of prime order, we
84 6. SUBNORMAL SERIES; SOLVABLE AND NILPOTENT GROUPS
see that G is solvable if and only if for one (hence every) composition series the quotients
are all cyclic groups of prime order.
Every commutative group is solvable, as is every dihedral group. The results in Chapter
5 show that every group of order < 60 is solvable. By contrast, a noncommutative simple
group, e.g., ·
n
for n _5, will not be solvable.
THEOREM 6.4 (FEITTHOMPSON) Every ﬁnite group of odd order is solvable.
3
PROOF. The proof occupies an entire issue of the Paciﬁc Journal of Mathematics (Feit and
Thompson 1963).
2
In other words, every noncommutative ﬁnite simple group has even order, and therefore
contains an element of order 2. For the role this theorem played in the classiﬁcation of the
ﬁnite simple groups, see p. 51. For a more recent look at the FeitThompson theorem, see
Glauberman 1999.
EXAMPLE 6.5 Consider the subgroups T =
+ +
0 +
¸
and U =
1 +
0 1
¸
of GL
2
(J),
some ﬁeld J. Then U is a normal subgroup of T, and T¡U . J
×
×J
×
, U . (J. ÷).
Hence T is solvable.
PROPOSITION 6.6 (a) Every subgroup and every quotient group of a solvable group is
solvable.
(b) An extension of solvable groups is solvable.
PROOF. (a) Let G>G
1
> >G
n
be a solvable series for G, and let H be a subgroup of G.
The homomorphism
. ÷.G
i÷1
: H ¨G
i
÷G
i
¡G
i÷1
has kernel (H ¨G
i
) ¨G
i÷1
=H ¨G
i÷1
. Therefore, H ¨G
i÷1
is a normal subgroup of
H ¨G
i
and the quotient H ¨G
i
¡H ¨G
i÷1
injects into G
i
¡G
i÷1
, which is commutative.
We have shown that
H >H ¨G
1
> >H ¨G
n
is a solvable series for H.
Let
¨
G be a quotient group of G, and let
¨
G
i
be the image of G
i
in
¨
G. Then
¨
G>
¨
G
1
> >
¨
G
n
={1]
is a solvable series for
¨
G.
3
Burnside (1897, p. 379) wrote:
No simple group of odd order is at present known to exist. An investigation as to the existence
or nonexistence of such groups would undoubtedly lead, whatever the conclusion might be, to
results of importance; it may be recommended to the reader as well worth his attention. Also,
there is no known simple group whose order contains fewer than three different primes. . . .
Signiﬁcant progress in the ﬁrst problem was not made until Suzuki, M., The nonexistence of a certain type of
simple group of ﬁnite order, 1957. However, the second problem was solved by Burnside himself, who proved
using characters that any group whose order contains fewer than three different primes is solvable (see Alperin
and Bell 1995, p. 182).
Solvable groups 85
(b) Let N be a normal subgroup of G, and let
¨
G =G¡N. We have to show that if N
and
¨
G are solvable, then so also is G. Let
¨
G>
¨
G
1
> >
¨
G
n
={1]
N >N
1
> >N
n
={1]
be solvable series for
¨
G and N, and let G
i
be the inverse image of
¨
G
i
in G. Then
G
i
¡G
i÷1
.
¨
G
i
¡
¨
G
i÷1
(see 1.48), and so
G>G
1
> >G
n
(=N) >N
1
> >N
n
is a solvable series for G.
2
COROLLARY 6.7 A ﬁnite ]group is solvable.
PROOF. We use induction on the order the group G. According to (4.16), the centre 7(G)
of G is nontrivial, and so the induction hypothesis implies that G¡7(G) is solvable. Be
cause 7(G) is commutative, (b) of the proposition shows that G is solvable.
2
Let G be a group. Recall that the commutator of .. . ÷ G is
Œ.. . =...
1
.
1
=..(..)
1
Thus
Œ.. . =1 ” .. =...
and G is commutative if and only if every commutator equals 1.
EXAMPLE 6.8 For any ﬁnitedimensional vector space V over a ﬁeld k and any full ﬂag
J ={V
n
. V
n1
. . . .] in V , the group
T(J) ={˛ ÷ Aut(V ) [ ˛(V
}
) cV
}
all ; ]
is solvable. Indeed, let U(J) be the group deﬁned in Example 5.10. Then T(J)¡U(J) is
commutative, and, when k = F
;
, U(J) is a ]group. This proves that T(J) is solvable
when k =F
;
, and in the general case one deﬁnes subgroups T
0
¯T
1
¯ of T(J) with
T
i
={˛ ÷ T(J) [ ˛(V
}
) cV
}i
all ; ]
and notes that the commutator of two elements of T
i
lies in T
i÷1
.
For any homomorphism c: G ÷H
c(Œ.. .) =c(...
1
.
1
) =Œc(.). c(.).
i.e., c maps the commutator of .. . to the commutator of c(.). c(.). In particular, we see
that if H is commutative, then c maps all commutators in G to 1.
The group G
t
=G
(1)
generated by the commutators in G is called the commutator or
ﬁrst derived subgroup of G.
86 6. SUBNORMAL SERIES; SOLVABLE AND NILPOTENT GROUPS
PROPOSITION 6.9 The commutator subgroup G
t
is a characteristic subgroup of G; it is the
smallest normal subgroup of G such that G¡G
t
is commutative.
PROOF. An automorphism ˛ of G maps the generating set for G
t
into G
t
, and hence maps
G
t
into G
t
. Since this is true for all automorphisms of G, G
t
is characteristic.
Write g ÷ ¨ g for the homomorphism g ÷gG
t
: G ÷G¡G
t
. Then Œ ¨ g.
¨
h =Œg. h, which
is 1 because Œg. h ÷ G
t
. Hence Œ ¨ g.
¨
h =1 for all ¨ g,
¨
h ÷ G¡G
t
, which shows that G¡G
t
is
commutative.
Let N be a second normal subgroup of G such that G¡N is commutative. Then Œg. h ÷
1 in G¡N, and so Œg. h ÷ N. Since these elements generate G
t
, N ¯G
t
.
2
For n _ 5. ·
n
is the smallest normal subgroup of S
n
giving a commutative quotient.
Hence (S
n
)
t
=·
n
.
The second derived subgroup of G is (G
t
)
t
; the third is G
(3)
=(G
tt
)
t
; and so on. Since
a characteristic subgroup of a characteristic subgroup is characteristic (3.7a), each derived
group G
(n)
is a characteristic subgroup of G. Hence we obtain a normal series
G ¯G
(1)
¯G
(2)
¯ .
which is called the derived series of G. For example, when n _ 5, the derived series of S
n
is
S
n
¯·
n
¯·
n
¯·
n
¯ .
PROPOSITION 6.10 A group G is solvable if and only if its kth derived subgroup G
(k)
=1
for some k.
PROOF. If G
(k)
=1, then the derived series is a solvable series for G. Conversely, let
G =G
0
>G
1
>G
2
> >G
x
=1
be a solvable series for G. Because G¡G
1
is commutative, G
1
¯ G
t
. Now G
t
G
2
is a
subgroup of G
1
, and from
G
t
¡G
t
¨G
2
:
÷G
t
G
2
¡G
2
cG
1
¡G
2
we see that
G
1
¡G
2
commutative == G
t
¡G
t
¨G
2
commutative == G
tt
cG
t
¨G
2
cG
2
.
Continuing in the fashion, we ﬁnd that G
(i)
cG
i
for all i , and hence G
(x)
=1.
2
Thus, a solvable group G has a canonical solvable series, namely the derived series, in
which all the groups are normal in G. The proof of the proposition shows that the derived
series is the shortest solvable series for G. Its length is called the solvable length of G.
ASIDE 6.11 Not every element of the commutator subgroup of a group is itself a commutator, but
the smallest groups where this occurs have order 94. This was shown by a computer search through
the libraries of small groups.
Nilpotent groups 87
Nilpotent groups
Let G be a group. Recall that we write 7(G) for the centre of G. Let 7
2
(G) c G be the
subgroup of G corresponding to 7(G¡7(G)) cG¡7(G). Thus
g ÷ 7
2
(G) ” Œg. . ÷ 7(G) for all . ÷ G.
Continuing in this fashion, we get a sequence of subgroups (ascending central series)
{1] c7(G) c7
2
(G) c
where
g ÷ 7
i
(G) ” Œg. . ÷ 7
i1
(G) for all . ÷ G.
If 7
n
(G) =G for some m, then G is said to be nilpotent, and the smallest such m is called
the (nilpotency) class of G. For example, all ﬁnite ]groups are nilpotent (apply 4.16).
Only the group {1] has class 0, and the groups of class 1 are exactly the commutative
groups. A group G is of class 2 if and only if G¡7(G) is commutative — such a group is
said to be metabelian.
EXAMPLE 6.12 (a) A nilpotent group is obviously solvable, but the converse is false. For
example, for a ﬁeld J, let
T =
a b
0 c
ˇ
ˇ
ˇ
ˇ
a. b. c ÷ J. ac =0
¸
.
Then 7(T) ={a1 [ a =0], and the centre of T¡7(T) is trivial. Therefore T¡7(T) is not
nilpotent, but we saw in (6.5) that it is solvable.
(b) The group G =
¸
¸
¸
1 + +
0 1 +
0 0 1
¸
¸
is metabelian: its centre is
¸
¸
¸
1 0 +
0 1 0
0 0 1
¸
¸
, and
G¡7(G) is commutative.
(c) Any nonabelian group G of order ]
3
is metabelian. In fact, G
t
=7(G) has order
] (see 5.17), and G¡G
t
is commutative (4.18). In particular, the quaternion and dihedral
groups of order 8, O and D
4
, are metabelian. The dihedral group D
2
n is nilpotent of class
n — this can be proved by induction, using that 7(D
2
n) has order 2, and D
2
n¡7(D
2
n) ~
D
2
n1. If n is not a power of 2, then D
n
is not nilpotent (use Theorem 6.18 below).
PROPOSITION 6.13 (a) A subgroup of a nilpotent group is nilpotent.
(b) · quotient of a nilpotent group is nilpotent.
PROOF. (a) Let H be a subgroup of a nilpotent group G. Clearly, 7(H) ¯ 7(G) ¨H.
Assume (inductively) that 7
i
(H) ¯7
i
(G) ¨H; then 7
i÷1
(H) ¯7
i÷1
(G) ¨H, because
(for h ÷ H)
h ÷ 7
i÷1
(G) == Œh. . ÷ 7
i
(G) all . ÷ G == Œh. . ÷ 7
i
(H) all . ÷ H.
(b) Straightforward.
2
88 6. SUBNORMAL SERIES; SOLVABLE AND NILPOTENT GROUPS
REMARK 6.14 It should be noted that if H is a subgroup of G, then 7(H) may be bigger
than 7(G). For example, the centre of
H =
a 0
0 b
ˇ
ˇ
ˇ
ˇ
ab =0
¸
cGL
2
(J).
is H itself, but the centre of GL
2
(J) consists only of the scalar matrices.
PROPOSITION 6.15 A group G is nilpotent of class _m if and only if
Œ. . . ŒŒg
1
. g
2
. g
3
. . . . . . g
n÷1
 =1
for all g
1
. .... g
n÷1
÷ G.
PROOF. Recall, g ÷ 7
i
(G) ” Œg. . ÷ 7
i1
(G) for all . ÷ G.
Assume G is nilpotent of class _m; then
G =7
n
(G) == Œg
1
. g
2
 ÷ 7
n1
(G) all g
1
. g
2
÷ G
== ŒŒg
1
. g
2
. g
3
 ÷ 7
n2
(G) all g
1
. g
2
. g
3
÷ G
== Œ ŒŒg
1
. g
2
. g
3
. .... g
n
 ÷ 7(G) all g
1
. . . . . g
n
÷ G
== Œ ŒŒg
1
. g
2
. g
3
. . . . . g
n÷1
 =1 all g
1
. . . . . g
n
÷ G.
For the converse, let g
1
÷ G. Then
ŒŒ...ŒŒg
1
. g
2
. g
3
..... g
n
. g
n÷1
 =1 for all g
1
. g
2
. .... g
n÷1
÷ G
== Œ...ŒŒg
1
. g
2
. g
3
. .... g
n
 ÷ 7(G). for all g
1
. .... g
n
÷ G
== Œ...ŒŒg
1
. g
2
. g
3
. .... g
n1
 ÷ 7
2
(G). for all g
1
. .... g
n1
÷ G
== g
1
÷ 7
n
(G) all g
1
÷ G.
2
An extension of nilpotent groups need not be nilpotent, i.e.,
N and G¡N nilpotent =G nilpotent. (27)
For example, the subgroup U of the group T in Examples 6.5 and 6.12 is commutative and
T¡U is commutative, but T is not nilpotent.
However, the implication (27) holds when N is contained in the centre of G. In fact,
we have the following more precise result.
COROLLARY 6.16 For any subgroup N of the centre of G,
G¡N nilpotent of class m == G nilpotent of class _m÷1.
PROOF. Write ¬ for the map G ÷G¡N. Then
¬(Œ...ŒŒg
1
. g
2
. g
3
. .... g
n
. g
n÷1
) =Œ...ŒŒ¬g
1
. ¬g
2
. ¬g
3
. .... ¬g
n
. ¬g
n÷1
 =1
all g
1
. .... g
n÷1
÷ G. Hence Œ...ŒŒg
1
. g
2
. g
3
. .... g
n
. g
n÷1
 ÷ N c7(G), and so
Œ...ŒŒg
1
. g
2
. g
3
. .... g
n÷1
. g
n÷2
 =1 all g
1
. .... g
n÷2
÷ G.
2
Nilpotent groups 89
COROLLARY 6.17 A ﬁnite ]group is nilpotent.
PROOF. We use induction on the order of G. Because 7(G) =1, G¡7(G) nilpotent, which
implies that G is nilpotent.
2
Recall that an extension
1 ÷N
t
÷G
t
÷O÷1
is central if (N) c7(G). Then:
the nilpotent groups are those that can be obtained from commutative groups
by successive central extensions.
Contrast:
the solvable groups are those that can be obtained from commutative groups
by successive extensions (not necessarily central).
THEOREM 6.18 A ﬁnite group is nilpotent if and only if it is equal to a direct product of
its Sylow subgroups.
PROOF. Adirect product of nilpotent groups is obviously nilpotent, and so the “if” direction
follows from the preceding corollary. For the converse, let G be a ﬁnite nilpotent group.
According to (5.9) it sufﬁces to prove that all Sylow subgroups are normal. Let 1 be such a
subgroup of G, and let N =N
G
(1). The ﬁrst lemma below shows that N
G
(N) =N, and
the second then implies that N =G, i.e., that 1 is normal in G.
2
LEMMA 6.19 Let 1 be a Sylow ]subgroup of a ﬁnite group G. For any subgroup H of
G containing N
G
(1), we have N
G
(H) =H.
PROOF. Let g ÷ N
G
(H), so that gHg
1
=H. Then H ¯g1g
1
=1
t
, which is a Sylow
]subgroup of H. By Sylow II, h1
t
h
1
=1 for some h ÷ H, and so hg1g
1
h
1
c 1.
Hence hg ÷ N
G
(1) cH, and so g ÷ H.
2
LEMMA 6.20 Let H be proper subgroup of a ﬁnite nilpotent group G; then H =N
G
(H).
PROOF. The statement is obviously true for commutative groups, and so we can assume
G to be noncommutative. We use induction on the order of G. Because G is nilpotent,
7(G) = 1. Certainly the elements of 7(G) normalize H, and so if 7(G) _ H, we have
H ´ 7(G) H c N
G
(H). Thus we may suppose 7(G) c H. Then the normalizer of H
in G corresponds under (1.47) to the normalizer of H¡7(G) in G¡7(G), and we can apply
the induction hypothesis.
2
REMARK 6.21 For a ﬁnite abelian group G we recover the fact that G is a direct product
of its ]primary subgroups.
PROPOSITION 6.22 (FRATTINI’S ARGUMENT) Let H be a normal subgroup of a ﬁnite
group G, and let 1 be a Sylow ]subgroup of H. Then G =H N
G
(1).
90 6. SUBNORMAL SERIES; SOLVABLE AND NILPOTENT GROUPS
PROOF. Let g ÷ G. Then g1g
1
c gHg
1
=H, and both g1g
1
and 1 are Sylow ]
subgroups of H. According to Sylow II, there is an h ÷ H such that g1g
1
=h1h
1
, and
it follows that h
1
g ÷ N
G
(1) and so g ÷ H N
G
(1).
2
THEOREM 6.23 A ﬁnite group is nilpotent if and only if every maximal proper subgroup
is normal.
PROOF. We saw in Lemma 6.20 that for any proper subgroup H of a nilpotent group G,
H ´N
G
(H). Hence,
H maximal == N
G
(H) =G.
i.e., H is normal in G.
Conversely, suppose every maximal proper subgroup of G is normal. We shall check
the condition of Theorem 6.18. Thus, let 1 be a Sylow ]subgroup of G. If 1 is not
normal in G, then there exists a maximal proper subgroup H of G containing N
G
(1).
Being maximal, H is normal, and so Frattini’s argument shows that G =H N
G
(1) =H
— contradiction.
2
Groups with operators
Recall that the set Aut(G) of automorphisms of a group G is again a group. Let · be
a group. A pair (G. c) consisting of a group G together with a homomorphism c: · ÷
Aut(G) is called an ·group, or G is said to have · as a group of operators.
Let G be an ·group, and write
˛
. for c(˛).. Then
(a)
(˛ˇ)
. =
˛
(
ˇ
.) (c is a homomorphism);
(b)
˛
(..) =
˛
.
˛
. (c(˛) is a homomorphism);
(c)
1
. =. (c is a homomorphism).
Conversely, a map (˛. .) ÷
˛
. : ·×G ÷G satisfying (a), (b), (c) arises from a homomor
phism · ÷Aut(G). Conditions (a) and (c) show that . ÷
˛
. is inverse to . ÷
(˛
1
)
.,
and so . ÷
˛
. is a bijection G ÷G. Condition (b) then shows that it is an automorphism
of G. Finally, (a) shows that the map c(˛) =(. ÷
˛
.) is a homomorphism · ÷Aut(G).
Let G be a group with operators ·. A subgroup H of G is admissible or ·invariant if
. ÷ H ==
˛
. ÷ H, all ˛ ÷ ·.
An intersection of admissible groups is admissible. If H is admissible, so also are its
normalizer N
G
(H) and centralizer C
G
(H).
An ·homomorphism (or admissible homomorphism) of ·groups is a homomor
phism ,: G ÷G
t
such that ,(
˛
g) =
˛
,(g) for all ˛ ÷ ·, g ÷ G.
EXAMPLE 6.24 (a) A group G can be regarded as a group with {1] as group of operators.
In this case all subgroups and homomorphisms are admissible, and so the theory of groups
with operators includes the theory of groups without operators.
(b) Consider G acting on itself by conjugation, i.e., consider G together with the homo
morphism
g ÷i
v
: G ÷Aut(G).
In this case, the admissible subgroups are the normal subgroups.
(c) Consider G with · = Aut(G) as group of operators. In this case, the admissible
subgroups are the characteristic subgroups.
Groups with operators 91
Almost everything we have proved for groups also holds for groups with operators. In
particular, the Theorems 1.45, 1.46, and 1.47 hold for groups with operators. In each case,
the proof is the same as before except that admissibility must be checked.
THEOREM 6.25 For any admissible homomorphism ,: G ÷G
t
of ·groups, N
def
=Ker(,)
is an admissible normal subgroup of G, ,(G) is an admissible subgroup of G
t
, and , factors
in a natural way into the composite of an admissible surjection, an admissible isomorphism,
and an admissible injection:
G G¡N
:
÷,(G) ·÷G
t
.
THEOREM 6.26 Let G be a group with operators ·, and let H and N be admissible
subgroups with N normal. Then H ¨N is a normal admissible subgroup of H, HN
is an admissible subgroup of G, and h(H ¨N) ÷ hH is an admissible isomorphism
H¡H ¨N ÷HN¡N.
THEOREM 6.27 Let c: G ÷
¨
G be a surjective admissible homomorphism of ·groups.
Under the onetoone correspondence H 
¨
H between the set of subgroups of G contain
ing Ker(c) and the set of subgroups of
¨
G (see 1.47), admissible subgroups correspond to
admissible subgroups.
Let c: · ÷Aut(G) be a group with · operating. An admissible subnormal series is a
chain of admissible subgroups of G
G ¯G
1
¯G
2
¯ ¯G
i
with each G
i
normal in G
i1
. Deﬁne similarly an admissible composition series. The quo
tients of an admissible subnormal series are ·groups, and the quotients of an admissible
composition series are simple ·groups, i.e., they have no normal admissible subgroups
apart from the obvious two.
The JordanH¨ older theorem continues to hold for ·groups. In this case the isomor
phisms between the corresponding quotients of two composition series are admissible. The
proof is the same as that of the original theorem, because it uses only the isomorphism
theorems, which we have noted also hold for ·groups.
EXAMPLE 6.28 (a) Consider G with G acting by conjugation. In this case an admissible
subnormal series is a sequence of subgroups
G =G
0
¯G
1
¯G
2
¯ ¯G
x
={1].
with each G
i
normal in G, i.e., a normal series. The action of G on G
i
by conjugation
passes to the quotient, to give an action of G on G
i
¡G
i÷1
. The quotients of two admissible
composition series are isomorphic as Ggroups.
(b) Consider G with ·=Aut(G) as operator group. In this case, an admissible subnor
mal series is a sequence
G =G
0
¯G
1
¯G
2
¯ ¯G
x
={1]
with each G
i
a characteristic subgroup of G, and the quotients of two admissible composi
tion series are isomorphic as Aut(G)groups.
92 6. SUBNORMAL SERIES; SOLVABLE AND NILPOTENT GROUPS
KrullSchmidt theorem
A group G is indecomposable if G =1 and G is not isomorphic to a direct product of two
nontrivial groups, i.e., if
G ~H ×H
t
== H =1 or H
t
=1.
EXAMPLE 6.29 (a) A simple group is indecomposable, but an indecomposable group need
not be simple: it may have a normal subgroup. For example, S
3
is indecomposable but has
C
3
as a normal subgroup.
(b) A ﬁnite commutative group is indecomposable if and only if it is cyclic of prime
power order.
Of course, this is obvious fromthe classiﬁcation, but it is not difﬁcult to prove it directly.
Let G be cyclic of order ]
n
, and suppose that G ~H ×H
t
. Then H and H
t
must be ]
groups, and they can’t both be killed by ]
n
, m < n. It follows that one must be cyclic
of order ]
n
, and that the other is trivial. Conversely, suppose that G is commutative and
indecomposable. Since every ﬁnite commutative group is (obviously) a direct product of
]groups with ] running over the primes, G is a ]group. If g is an element of G of highest
order, one shows that 'g) is a direct factor of G, G ~'g) ×H, which is a contradiction.
(c) Every ﬁnite group can be written as a direct product of indecomposable groups
(obviously).
THEOREM 6.30 (KRULLSCHMIDT) Suppose that G is a direct product of indecompos
able subgroups G
1
. . . . . G
x
and of indecomposable subgroups H
1
. . . . . H
t
:
G .G
1
× ×G
x
. G .H
1
× ×H
t
.
Then s = t , and there is a reindexing such that G
i
~ H
i
. Moreover, given r, we can
arrange the numbering so that
G =G
1
× ×G
i
×H
i÷1
× ×H
t
.
PROOF. See Rotman 1995, 6.36.
2
EXAMPLE 6.31 Let G =F
;
×F
;
, and think of it as a twodimensional vector space over
F
;
. Let
G
1
='(1. 0)). G
2
='(0. 1)): H
1
='(1. 1)). H
2
='(1. ÷1)).
Then G =G
1
×G
2
, G =H
1
×H
2
, G =G
1
×H
2
.
REMARK 6.32 (a) The KrullSchmidt theorem holds also for an inﬁnite group provided it
satisﬁes both chain conditions on subgroups, i.e., ascending and descending sequences of
subgroups of G become stationary.
(b) The KrullSchmidt theorem also holds for groups with operators. For example,
let Aut(G) operate on G; then the subgroups in the statement of the theorem will all be
characteristic.
(c) When applied to a ﬁnite abelian group, the theorem shows that the groups C
n
i
in a
decomposition G =C
n
1
×... ×C
n
r
with each m
i
a prime power are uniquely determined
up to isomorphism (and ordering).
Exercises 93
Exercises
61 Let G be a group (not necessarily ﬁnite) with a ﬁnite composition series
G =G
0
¯G
1
¯ ¯G
n
=1.
and let N be a normal subgroup of G. Show that
N =N ¨G
0
¯N ¨G
1
¯ ¯N ¨G
n
=1
becomes a composition series for N once the repetitions have been omitted.
62 If G
1
and G
2
are groups such that G
t
1
~G
t
2
and G
1
¡G
t
1
~G
2
¡G
t
2
, are G
1
and G
2
necessarily isomorphic? (Here
t
denotes the commutator subgroup.)
CHAPTER 7
Representations of Finite Groups
Throughout this chapter, G is a ﬁnite group and J is a ﬁeld. All vector spaces are ﬁnite
dimensional.
An Jalgebra is a ring · containing J in its centre and ﬁnite dimensional as an J
vector space. We do not assume · to be commutative; for example, · could be the ma
trix algebra M
n
(J). Let e
1
. . . . . e
n
be a basis for · as an Jvector space; then e
i
e
}
=
¸
k
a
k
i}
e
k
for some a
k
i}
÷ J, called the structure constants of · relative to the basis (e
i
)
i
;
once a basis has been chosen, the algebra · is uniquely determined by its structure con
stants.
All ·modules are ﬁnite dimensional when regarded as Jvector spaces. For an ·
module V , mV denotes the direct sum of m copies of V .
The opposite ·
opp
of an Jalgebra · is the same Jalgebra as · but with the multipli
cation reversed, i.e., ·
opp
=(·. ÷.
t
) with a
t
b =ba. In other words, there is a onetoone
correspondence a a
t
: ··
opp
which is an isomorphism of Jvector spaces and has the
property that a
t
b
t
=(ba)
t
.
An ·module M is simple if it is nonzero and contains no submodules except 0 and M,
and it is semisimple if it is isomorphic to a direct sum of simple modules.
Matrix representations
A matrix representation of degree n of G over J is a homomorphism G ÷ GL
n
(J).
The representation is said to be faithful if the homomorphism is injective. Thus a faithful
representation identiﬁes G with group of n×n matrices.
EXAMPLE 7.1 (a) There is a representation O ÷GL
2
(C) of the quaternion group O =
'a. b) sending a to
0
1
1 0
and b to
0 1
1 0
. In fact, that is how we originally deﬁned
O in (1.18).
(b) Let G = S
n
. For each o ÷ S
n
, let 1(o) be the matrix obtained from the identity
matrix by using o to permute the rows. Then, for any n×n matrix ·, 1(o)· is obtained
from·by using o to permute the rows. In particular, 1(o)1(o
t
) =1(oo
t
), and so o ÷1(o)
is a representation of S
n
. Clearly, it is faithful. As every ﬁnite group embeds into S
n
for
some n (Cayley’s theorem, see 1.22), this shows that every ﬁnite group has a faithful matrix
representation.
(c) Let G =C
n
='o). If J contains a nth root of 1, say ¯, then there is representation
o
i
÷¯
i
: C
n
÷GL
1
(J) = J
×
. The representation is faithful if and only if ¯ has order
95
96 7. REPRESENTATIONS OF FINITE GROUPS
exactly n. If n =] is prime and J has characteristic ], then X
;
÷1 =(X ÷1)
;
, and so 1
is the only ]th root of 1 in J. In this case, the representation is trivial, but there is a faithful
representation
o
i
÷
1 i
0 1
: C
;
÷GL
2
(J).
ASIDE 7.2 Recall that the Burnside problem asks whether every ﬁnitely generated group with ﬁnite
exponent is ﬁnite (see p. 37). Burnside proved that the problem has a positive answer for subgroups
of GL
n
(C). Therefore, no inﬁnite ﬁnitely generated group with ﬁnite exponent has a faithful repre
sentation over C.
Roots of 1 in ﬁelds
As the last example indicates, the representations of a group over a ﬁeld J depend on the
roots of 1 in the ﬁeld. The nth roots of 1 in a ﬁeld J form a subgroup u
n
(J) of J
×
, which
is cyclic (see 1.56).
If the characteristic of J divides n, then [u
n
(J)[ < n. Otherwise, X
n
÷1 has distinct
roots (a multiple root would have to be a root of its derivative nX
n1
), and we can always
arrange that [u
n
(J)[ =n by extending J, for example, by replacing a subﬁeld J of C with
JŒ¯ where ¯ =e
2ti¡n
, or by replacing J with JŒX¡(g(X)) where g(X) is an irreducible
factor of X
n
÷1 not dividing X
n
÷1 for any proper divisor m of n.
An element of order n in J
×
is called a primitive nth root of 1. To say that J contains
a primitive nth root of 1, ¯, means that u
n
(J) is a cyclic group of order n and that ¯
generates it (and it implies that either J has characteristic 0 or it has characteristic a prime
not dividing n).
Linear representations
Recall (4.1) that we have deﬁned the notion of a group G acting a set. When the set is an
Jvector space V , we say that the action is linear if the map
g
V
: V ÷V , . ÷g..
is linear for each g ÷ G. Then g
V
has inverse the linear map (g
1
)
V
, and g ÷g
V
: G ÷
GL(V ) is a homomorphism. Thus, from a linear action of G on V , we obtain a homo
morphism of groups G ÷GL(V ); conversely, every such homomorphism deﬁnes a linear
action of G on V . We call a homomorphism G ÷GL(V ) a linear representation of G on
V . Note that a linear representation of G on J
n
is just a matrix representation of degree n.
EXAMPLE 7.3 (a) Let G =C
n
='o), and assume that J contains a primitive nth root of
1, say ¯. Let G ÷GL(V ) be a linear representation of G. Then (o
1
)
n
=(o
n
)
1
=1, and so
the minimum polynomial of o
1
divides X
n
÷1. As X
n
÷1 has n distinct roots ¯
0
. . . . . ¯
n1
in J, the vector space V decomposes into a direct sum of eigenspaces
V =
0_i_n1
V
i
. V
i
def
={· ÷ V [ o· =¯
i
·].
Conversely, every such direct sum decomposition of G arises from a representation of G.
Maschke’s theorem 97
(b) Let G be a commutative group of exponent n, and assume that J contains a primitive
nth root of 1. Let
G
·
=Hom(G. J
×
) =Hom(G. u
n
(J))
To give a representation of G on a vector space V is the same as to give a direct sum
decomposition
V =
;÷G
_
V
;
. V
;
def
={· ÷ V [ o· =,(o)·].
When G is cyclic, this is a restatement of (a), and the general case follows easily (decom
pose V with respect to the action of one cyclic factor of G; then decompose each summand
with respect to the action of a second cyclic factor of G; and so on).
Maschke’s theorem
Let G ÷GL(V ) be a linear representation of G on an Jvector space V . A subspace W
of V is said to be Ginvariant if gW c W for all g ÷ G. An Jlinear map ˛: V ÷V
t
of
vector spaces on which G acts linearly is said to be Ginvariant if
˛(g·) =g(˛·) for all g ÷ G. · ÷ V.
Finally, a bilinear form ç: V ×V ÷J is said to be Ginvariant if
ç(g·. g·
t
) =ç(·. ·
t
) for all g ÷ G, ·. ·
t
÷ V.
THEOREM 7.4 (MASCHKE) Let G ÷GL(V ) be a linear representation of G. If the char
acteristic of J does not divide [G[, then every Ginvariant subspace W of V has a G
invariant complement, i.e., there exists a Ginvariant subspace W
t
such that V =W ˚W
t
.
Note that the theorem always applies when J has characteristic zero.
The condition on the characteristic is certainly necessary: let G = 'o) be the cyclic
group of order ], where ] is the characteristic of J, and let o acts on V =J
2
as the matrix
1 1
0 1
(see 7.3b); the subspace (
+
0
) is Ginvariant, and this complementary subspaces are
those of the form J
o
b
, b =0; none of them is Ginvariant.
Because of the importance of the ideas involved, we present two proofs of Maschke’s
theorem.
PROOF OF MASCHKE’S THEOREM (CASE J =1 OR C)
LEMMA 7.5 Let ç be a symmetric bilinear form on V , and let W be a subspace of V . If ç
and W are Ginvariant, then so also is W
J
def
={· ÷ V [ ç(n. ·) =0 for all n ÷ W].
PROOF. Let · ÷ W
J
and let g ÷ G. For any n ÷ W , ç(n. g·) =ç(g
1
n. ·) because ç is
Ginvariant, and ç(g
1
n. ·) =0 because W is Ginvariant. This shows that g· ÷ W
J
.
2
Recall from linear algebra that if ç is nondegenerate, then V =W ˚W
J
. Therefore,
in order to prove Maschke’s theorem, it sufﬁces to show that there exists a Ginvariant
nondegenerate symmetric bilinear from ç: V ×V ÷J.
98 7. REPRESENTATIONS OF FINITE GROUPS
LEMMA 7.6 For any symmetric bilinear form ç on V ,
¨
ç(·. n)
def
=
¸
v÷G
ç(g·. gn)
is a Ginvariant symmetric bilinear form on V .
PROOF. The form ç is obviously bilinear and symmetric, and for g
0
÷ G,
¨
ç(g
0
·. g
0
n)
def
=
¸
v÷G
ç(gg
0
·. gg
0
n).
which equals
¸
v÷G
ç(g·. gn) because, as g runs over G, so also does gg
0
.
2
Unfortunately, we can’t conclude that
¨
ç is nondegenerate when ç is (otherwise we
could prove that all JŒGmodules are semisimple, with no restriction on J or G).
LEMMA 7.7 Let J =1. If ç is a positive deﬁnite symmetric bilinear form on V , then so
also is
¨
ç.
PROOF. If
¨
ç is positive deﬁnite, then for any nonzero · in V ,
¨
ç(·. ·) =
¸
v÷G
ç(g·. g·) > 0.
2
This completes the proof of Maschke’s theorem when J =1, because there certainly
exist positive deﬁnite symmetric bilinear forms ç on V . A similar argument using hermitian
forms applies when J =C (or, indeed, when J is any subﬁeld of C).
ASIDE 7.8 A representation of a group G on a real vector space V is unitary if there exists a
Ginvariant positive deﬁnite symmetric bilinear form on V . Lemma 7.6 shows that every unitary
representation is semisimple, and Lemma 7.7 shows that every real representation of a ﬁnite group
is unitary.
PROOF OF MASCHKE’S THEOREM (GENERAL CASE)
An endomorphism¬ of an Jvector space V is called a projector if ¬
2
=¬. The minimum
polynomial of a projector ¬ divides X
2
÷X =X(X÷1), and so V decomposes into a direct
sum of eigenspaces,
V =V
0
(¬) ˚V
1
(¬), where
V
0
(¬) ={· ÷ V [ ¬· =0] =Ker(¬)
V
1
(¬) ={· ÷ V [ ¬· =·] =Im(¬).
Conversely, a decomposition V =V
0
˚V
1
arises from a projector (·
0
. ·
1
) ÷(0. ·
1
).
Now suppose that G acts linearly on V . If a projector ¬ is Ginvariant, then V
1
(¬) and
V
0
(¬) are obviously Ginvariant. Thus, to prove the theorem it sufﬁces to show that W is
the image of a G–invariant projector ¬.
We begin by choosing an Jlinear projector ¬ with image W , which certainly exists,
and we modify it to obtain a Ginvariant projector ¨ ¬ with the same image. For · ÷ V , let
¨ ¬(·) =
1
[G[
¸
v÷G
g
¬(g
1
·)
.
The group algebra; semisimplicity 99
This makes sense because [G[ 1 ÷ J
×
, and it deﬁnes an Jlinear map ¨ ¬: V ÷V . Let
n ÷ W ; then g
1
n ÷ W , and so
¨ ¬(n) =
1
[G[
¸
v÷G
g(g
1
n) =
1
[G[
¸
v÷G
n =n. (28)
The image of ¨ ¬ is contained in W , because Im(¬) cW and W is Ginvariant, and so
¨ ¬
2
(·)
def
= ¨ ¬( ¨ ¬ (·))
(2S)
= ¨ ¬(·)
for any · ÷ V . Thus, ¨ ¬ is a projector, and (28) shows that Im( ¨ ¬) ¯W , and hence Im( ¨ ¬) =
W . It remains to show that ¨ ¬ is Ginvariant. For g
0
÷ V
¨ ¬(g
0
·) =
1
[G[
¸
v÷G
g
¬(g
1
g
0
·)
=g
0
1
[G[
¸
v÷G
(g
1
0
g)
¬(g
1
g
0
·)
.
which equals g
0
¨ ¬ (·) because, as g runs over G, so also does g
1
0
g.
The group algebra; semisimplicity
The group algebra JŒG of G is deﬁned to be the Jvector space with basis the elements
of G endowed with the multiplication extending that on G. Thus,
˘ an element of JŒG is a sum
¸
v÷G
c
v
g, c
v
÷ J,
˘ two elements
¸
v÷G
c
v
g and
¸
v÷G
c
t
v
g of JŒG are equal if and only if c
v
= c
t
v
for all g, and
˘
¸
v÷G
c
v
g
¸
v÷G
c
t
v
g
=
¸
v÷G
c
tt
v
g. c
tt
v
=
¸
v
1
v
2
=v
c
v
1
c
t
v
2
.
A linear action
g. · ÷g·: G×V ÷V
of G on an Jvector space extends uniquely to an action of JŒG on V ,
¸
v÷G
c
v
g. · ÷
¸
v÷G
c
v
g·: JŒG ×V ÷V.
which makes V into an JŒGmodule. The submodules for this action are exactly the G
invariant subspaces.
Let G ÷GL(V ) be a linear representation of G. When V is simple (resp. semisimple)
as an JŒGmodule, the representation is usually said to be irreducible (resp. completely
reducible). However, I will call them simple (resp. semisimple) representations.
PROPOSITION 7.9 If the characteristic of J does not divide [G[, then every JŒGmodule
is a direct sum of simple submodules.
PROOF. Let V be a JŒGmodule. If V is simple, then there is nothing to prove. Otherwise,
it contains a nonzero proper submodule W . According to Maschke’s theorem, V =W ˚
W
t
with W
t
an JŒGsubmodule. If W and W
t
are simple, then the proof is complete;
otherwise, we can continue the argument, which terminates in a ﬁnite number of steps
because V has ﬁnite dimension as an Jvector space.
2
As we have observed, the linear representations of G can be regarded as JŒGmodules.
Thus, to understand the linear representations of G, we need to understand the JŒG
modules, and for this we need to understand the structure of the Jalgebra JŒG. In the next
three sections we study Jalgebras and their modules; in particular, we prove the famous
Wedderburn theorems concerning Jalgebras whose modules are all semisimple.
100 7. REPRESENTATIONS OF FINITE GROUPS
Semisimple modules
In this section, · is an Jalgebra.
PROPOSITION 7.10 Every ·module V admits a ﬁltration
V =V
0
¯V
1
¯ ¯V
x
={0]
such that the quotients V
i
¡V
i÷1
are simple ·modules. If
V =W
0
¯W
1
¯ ¯W
t
={0]
is a second such ﬁltration, then s = t and there is a permutation o of {1. . . . . s] such that
V
i
¡V
i÷1
~W
c(i)
¡W
c(i)÷1
for all i .
PROOF. This is a variant of the JordanH¨ older theorem (6.2), which can be proved by the
same argument.
2
COROLLARY 7.11 Suppose
V ~V
1
˚ ˚V
x
~W
1
˚ ˚W
t
with all the ·modules V
i
and W
}
simple. Then s = t and there is a permutation o of
{1. . . . . s] such that V
i
~W
c(i)
.
PROOF. Each decomposition deﬁnes a ﬁltration, to which the proposition can be applied.
2
PROPOSITION 7.12 Let V be an ·module. If V is a sum of simple submodules, say
V =
¸
i÷J
S
i
(the sum need not be direct), then for any submodule W of V , there is a
subset J of 1 such that
V =W ˚
i÷J
S
i
.
PROOF. Let J be maximal among the subsets of 1 such the sum S
J
def
=
¸
}÷J
S
}
is direct
and W ¨S
J
=0. I claim that W ÷S
J
=V (hence V is the direct sum of W and the S
}
with ; ÷ J). For this, it sufﬁces to show that each S
i
is contained in W ÷S
J
. Because S
i
is simple, S
i
¨(W ÷S
J
) equals S
i
or 0. In the ﬁrst case, S
i
cW ÷S
J
, and in the second
S
J
¨S
i
=0 and W ¨(S
J
÷S
i
) =0, contradicting the deﬁnition of 1.
2
COROLLARY 7.13 The following conditions on an ·module V are equivalent:
(a) V is semisimple;
(b) V is a sum of simple submodules;
(c) every submodule of V has a complement.
PROOF. The proposition shows that (b) implies (c), and the argument in the proof of (7.9)
shows that (c) implies (a). It is obvious that (a) implies (b).
2
COROLLARY 7.14 Sums, submodules, and quotient modules of semisimple modules are
semisimple.
PROOF. Each is a sum of simple modules.
2
Simple Jalgebras and their modules 101
Simple Jalgebras and their modules
An Jalgebra · is said to be simple if it contains no twosided ideals except 0 and ·. We
shall make frequent use of the following observation:
The kernel of a homomorphism } : · ÷T of Jalgebras is an ideal in · not
containing 1; therefore, if · is simple, then } is injective.
EXAMPLE 7.15 We consider the matrix algebra M
n
(J). Let e
i}
be the matrix with 1 in
the (i. ; )th position and zeros elsewhere.
(a) Let 1 be a twosided ideal in M
n
(J), and suppose that 1 contains a nonzero matrix
M =(m
i}
) with, say, m
i
0
}
0
=0. As
e
i i
0
M e
}
0
}
=m
i
0
}
0
e
i}
and e
i i
0
M e
}
0
}
÷ 1, we see that 1 contains all the matrices e
i}
and so equals
M
n
(J). We have shown that M
n
(J) is simple.
(b) For M. N ÷ M
n
(J), the ; th column of M N is M N
}
where N
}
is the ; th column
of N. Therefore, for a given matrix N,
N
}
=0 = (M N)
}
=0
N
}
=0 = (M N)
}
can be arbitrary.
(29)
For 1 _ i _ n, let 1(i ) be the set of matrices whose ; th columns are zero for ; =i
and whose i th column is arbitrary. For example, when n =4,
1(3) =
ˆ
ˆ
¸
ˆ
ˆ
¸
¸
¸
¸
0 0 + 0
0 0 + 0
0 0 + 0
0 0 + 0
¸
¸
¸
¸
¸
¸
cM
4
(J).
It follows from (29) that 1(i ) is a minimal left ideal in M
n
(J). Note that M
n
(J) is
a direct sum
M
n
(J) =1(1) ˚ ˚1(n)
of minimal left ideals.
EXAMPLE 7.16 An Jalgebra is said to be a division algebra if every nonzero element
a has an inverse, i.e., there exists a b such that ab = 1 = ba. Thus a division algebra
satisﬁes all the axioms to be a ﬁeld except commutativity (and for this reason is sometimes
called a skew ﬁeld). Clearly, a division algebra has no nonzero proper ideals, left, right, or
twosided, and so is simple.
If D is a division algebra, then the argument in (7.15a) shows that the algebra M
n
(D)
is simple.
EXAMPLE 7.17 For a. b ÷ J
×
, let H(a. b) be the Jalgebra with basis 1. i. ;. k (as an
Jvector space) and with the multiplication determined by
i
2
=a. ;
2
=b. i; =k =÷;i
(so i k =i i; =a; etc.). Then H(a. b) is an Jalgebra, called a quaternion algebra over
J. For example, if J =1, then H(÷1. ÷1) is the usual quaternion algebra. One can show
that H(a. b) is either a division algebra or it is isomorphic to M
2
(J). In particular, it is
simple.
102 7. REPRESENTATIONS OF FINITE GROUPS
7.18 Much of linear algebra does not require that the ﬁeld be commutative. For example,
the usual arguments show that a ﬁnitely generated module V over a division algebra D has
a basis, and that all bases have the same number n of elements — n is called the dimension
of V . In particular, all ﬁnitely generated Dmodules are free.
7.19 Let · be an Jalgebra, and let
¹
· denote · regarded as a left ·module. Right
multiplication . ÷.a on
¹
· by an element a of · is an ·linear endomorphism of
¹
·.
Moreover, every ·linear map c:
¹
· ÷
¹
· is of this form with a =c(1). Thus,
End
¹
(
¹
·) .· (as Jvector spaces).
Let c
o
be the map . ÷.a. Then
(c
o
ıc
o
0 )(1)
def
=c
o
(c
o
0 (1)) =c
o
(a
t
) =a
t
a =c
o
0
o
(1).
and so
End
¹
(
¹
·) .·
opp
(as Jalgebras).
More generally,
End
¹
(V ) .·
opp
for any ·module V that is free of rank 1, and
End
¹
(V ) .M
n
(·
opp
)
for any free ·module V of rank n (cf. 7.32 below).
CENTRALIZERS
Let · be an Jsubalgebra of an Jalgebra T. The centralizer of · in T is
C
B
(·) ={b ÷ T [ ba =ab for all a ÷ ·].
It is again an Jsubalgebra of T.
EXAMPLE 7.20 In the following examples, the centralizers are taken in M
n
(J).
(a) Let · be the set of scalar matrices in M
n
(J), i.e., · = J 1
n
. Clearly, C(·) =
M
n
(J).
(b) Let · =M
n
(J). Then C(·) is the centre of M
n
(J), which we now compute. Let
e
i}
be the matrix with 1 in the (i. ; )th position and zeros elsewhere, so that
e
i}
e
/n
=
e
in
if ; =l
0 if ; =l.
Let ˛ = (a
i}
) ÷ M
n
(J). Then ˛ =
¸
i,}
a
i}
e
i}
, and so ˛e
/n
=
¸
i
a
i/
e
in
and
e
/n
˛ =
¸
}
a
n}
e
/}
. If ˛ is in the centre of M
n
(J), then ˛e
/n
= e
/n
˛, and so
a
i/
=0 for i =l, a
n}
=0 for ; =m, and a
/ /
=a
nn
. It follows that the centre of
M
n
(J) is set of scalar matrices J 1
n
. Thus C(·) =J 1
n
.
(c) Let · be the set of diagonal matrices in M
n
(J). In this case, C(·) =·.
Notice that in all three cases, C(C(·)) =·.
Simple Jalgebras and their modules 103
THEOREM 7.21 (DOUBLE CENTRALIZER THEOREM) Let · be an Jalgebra, and let V
be a faithful semisimple ·module. Then C(C(·)) =· (centralizers taken in End
T
(V )).
PROOF. Let D = C(·) and let T = C(D). Clearly · c T, and the reverse inclusion
follows from the next lemma when we take ·
1
. . . . . ·
n
to generate V as a Jvector space.
2
LEMMA 7.22 For any ·
1
. . . . . ·
n
÷ V and b ÷ T, there exists an a ÷ · such that
a·
1
=b·
1
. a·
2
=b·
2
. . . . . a·
n
=b·
n
.
PROOF. We ﬁrst prove this for n =1. Note that ··
1
is an ·submodule of V , and so (see
7.13) there exists an ·submodule W of V such that V =··
1
˚W . Let ¬: V ÷V be the
map (a·
1
. n) ÷(a·
1
. 0) (projection onto ··
1
). It is ·linear, hence lies in D, and has the
property that ¬(·) =· if and only if · ÷ ··
1
. Now
¬(b·
1
) =b(¬·
1
) =b·
1
.
and so b·
1
÷ ··
1
, as required.
We now prove the general case. Let W be the direct sum of n copies of V with · acting
diagonally, i.e.,
a(·
1
. . . . . ·
n
) =(a·
1
. . . . . a·
n
). a ÷ ·. ·
i
÷ V.
Then W is again a semisimple ·module (7.14). The centralizer of · in End
T
(W ) consists
of the matrices (,
i}
)
1_i,}_n
, ,
i}
÷ End
T
(V ), such that (,
i}
a) =(a,
i}
) for all a ÷ ·, i.e.,
such that ,
i}
÷ D (cf. 7.32). In other words, the centralizer of · in End
T
(·) is M
n
(D). An
argument as in Example 7.20(b), using the matrices e
i}
(ı) with ı in the i; th position and
zeros elsewhere, shows that the centralizer of M
n
(D) in End
T
(W ) consists of the diagonal
matrices
¸
¸
¸
¸
ˇ 0 0
0 ˇ 0
.
.
.
.
.
.
.
.
.
.
.
.
0 0 ˇ
with ˇ ÷ T. We now apply the case n = 1 of the lemma to ·, W , b, and the vector
(·
1
. . . . . ·
n
) to complete the proof.
2
THEOREM 7.23 Every simple Jalgebra is isomorphic to M
n
(D) for some n and some
division Jalgebra D.
PROOF. Choose a simple ·module S, for example, any minimal left ideal of ·. Then ·
acts faithfully on S, because the kernel of ·÷End
T
(S) will be a twosided ideal of · not
containing 1, and hence is 0.
Let D be the centralizer of · in the Jalgebra End
T
(S) of Jlinear maps S ÷S.
According to the double centralizer theorem (7.21), the centralizer of D in End
T
(S) is
·, i.e., · = End
T
(S). Schur’s lemma (7.24 below) implies that D is a division algebra.
Therefore S is a free Dmodule (7.18), say, S ~D
n
, and so End
T
(S) ~M
n
(D
opp
) (see
7.19).
2
104 7. REPRESENTATIONS OF FINITE GROUPS
LEMMA 7.24 (SCHUR’S LEMMA) For every Jalgebra ·and simple ·module S, End
¹
(S)
is a division algebra.
PROOF. Let , be an ·linear map S ÷S. Then Ker(,) is an ·submodule of S, and so
it is either S or 0. In the ﬁrst case, , is zero, and in the second it is an isomorphism, i.e., it
has an inverse that is also ·linear.
2
MODULES OVER SIMPLE JALGEBRAS
For any Jalgebra ·, the submodules of
¹
· are the left ideals in ·, and the simple sub
modules of
¹
· are the minimal left ideals.
PROPOSITION 7.25 Any two minimal left ideals of a simple Jalgebra are isomorphic as
left ·modules, and
¹
· is a direct sum of its minimal left ideals.
PROOF. After Theorem 7.23, we may assume that · =M
n
(D) for some division algebra
D. We saw in (7.16) that the minimal left ideals in M
n
(D) are those of the form 1({; ]).
Clearly · =
¸
1_}_n
1({; ]) and each 1({; ]) is isomorphic to D
n
with its natural action
of M
n
(D).
2
THEOREM 7.26 Let ·be a simple Jalgebra, and let S be a simple ·module. Then every
·module is isomorphic to a direct sum of copies of S.
PROOF. Let S
0
be a minimal left ideal of ·. The proposition shows that
¹
·~S
n
0
for some
n. Let e
1
. . . . . e
i
be a set of generators for V as an ·module. The map
(a
1
. . . . . a
i
) ÷
¸
a
i
e
i
realizes V as a quotient of a direct sum of r copies of
¹
·, and hence as a quotient of nrS
0
.
Thus, V is a sum of simple submodules each isomorphic to S
0
, and so Proposition 7.12
shows that V ~mS
0
for some m.
2
COROLLARY 7.27 Let · be a simple Jalgebra. Then any two simple ·modules are
isomorphic, and any two ·modules having the same dimension over J are isomorphic.
PROOF. Obvious from the Theorem.
2
COROLLARY 7.28 The integer n in Theorem 7.23 is uniquely determined by ·, and D is
uniquely determined up to isomorphism.
PROOF. If · ~M
n
(D), then D ~End
¹
(S) for any simple ·module S. Thus, the state
ment follows from Theorem 7.26.
2
Semisimple Jalgebras and their modules 105
CLASSIFICATION OF THE DIVISION ALGEBRAS OVER J
After Theorem 7.23, to classify the simple algebras over J, it remains to classify the divi
sion algebras over J.
PROPOSITION 7.29 When J is algebraically closed, the only division algebra over J is J
itself.
PROOF. Let D be division algebra over J. For any ˛ ÷ D, the Jsubalgebra JŒ˛ of D
generated by ˛ is a ﬁeld (because it is an integral domain of ﬁnite degree over J). Therefore
˛ ÷ J.
2
ASIDE 7.30 The classiﬁcation of the isomorphism classes of division algebras over a ﬁeld J is
one the most difﬁcult and interesting problems in algebra and number theory. For J =1, the only
division algebra is the usual quaternion algebra. For J ﬁnite, the only division algebra with centre
J is J itself (theorem of Wedderburn).
A division algebra over J whose centre is J is said to be central (formerly normal). Brauer
showed that the set of isomorphism classes of central division algebras over a ﬁeld form a group,
called (by Hasse and Noether) the Brauer group
1
of the ﬁeld. The statements in the last paragraph
show that the Brauer groups of algebraically closed ﬁelds and ﬁnite ﬁelds are zero, and the Brauer
group of 1 has order 2. The Brauer groups of Q and its ﬁnite extensions were computed by Albert,
Brauer, Hasse, and Noether in the 1930s as a consequence of class ﬁeld theory.
Semisimple Jalgebras and their modules
An Jalgebra · is said to be semisimple if every ·module is semisimple. Theorem 7.26
shows that simple Jalgebras are semisimple, and Maschke’s theorem shows that the group
algebra JŒG is semisimple when the order of G is not divisible by the characteristic of J
(see 7.9).
EXAMPLE 7.31 Let · be a ﬁnite product of simple Jalgebras. Then every minimal left
ideal of a simple factor of · is a simple ·submodule of
¹
·. Therefore,
¹
· is a direct sum
of simple ·modules, and so is semisimple. Since every ·module is a quotient of a direct
sum of copies of
¹
·, this shows that · is semisimple.
Before stating the main result of this section, we recall some elementary module theory.
7.32 Let · be an Jalgebra, and consider modules
M =M
1
˚ ˚M
n
N =N
1
˚ ˚N
n
.
Let ˛ be an ·linear map M ÷N. For .
}
÷ M
}
, let
˛(0. . . . . 0. .
}
. 0. . . . . 0) =(.
1
. . . . . .
n
).
1
The tensor product D˝
T
D
t
of two central simple algebras over J is again a central simple algebra, and
hence is isomorphic to M
i
(D
tt
) for some central simple algebra D
tt
. Deﬁne
ŒDŒD
t
 =ŒD
tt
.
This product is associative because of the associativity of tensor products, the isomorphism class of J is an
identity element, and ŒD
opp
 is an inverse for ŒD.
106 7. REPRESENTATIONS OF FINITE GROUPS
Then .
}
÷.
i
is an ·linear map M
}
÷N
i
, which we denote ˛
i}
. Thus, ˛ deﬁnes an
m×n matrix whose i; th coefﬁcient is an ·linear map M
}
÷N
i
. Conversely, every such
matrix (˛
i}
) deﬁnes an ·linear map M ÷N, namely,
¸
¸
¸
¸
¸
¸
¸
.
1
.
.
.
.
}
.
.
.
.
n
÷
¸
¸
¸
¸
¸
¸
¸
˛
11
˛
1}
˛
1n
.
.
.
.
.
.
.
.
.
˛
i1
˛
i}
˛
}n
.
.
.
.
.
.
.
.
.
˛
n1
˛
n}
˛
nn
¸
¸
¸
¸
¸
¸
¸
.
1
.
.
.
.
}
.
.
.
.
n
def
=
¸
¸
¸
¸
¸
¸
¸
˛
11
(.
1
) ÷ ÷˛
1n
(.
n
)
.
.
.
˛
i1
(.
1
) ÷ ÷˛
i n
(.
n
)
.
.
.
˛
n1
(.
1
) ÷ ÷˛
nn
(.
n
)
.
Thus, we see
Hom
¹
(M. N) .
Hom
¹
(M
}
. N
i
)
1_}_n, 1_i_n
(30)
(isomorphism of Jvector spaces). When M = N, this becomes an isomorphism of J
algebras. For example, if M is a direct sum of m copies of M
0
, then
End
¹
(M) .M
n
(End
¹
(M
0
)) (31)
(m×m matrices with coefﬁcients in the ring End
¹
(M
0
)).
THEOREM 7.33 Let V be a ﬁnite dimensional Jvector space, and let ·be an Jsubalgebra
of End
T
(V ). If V is semisimple as an ·module, then the centralizer of · in End
T
(V ) is
a product of simple Jalgebras (hence it is a semisimple Jalgebra).
PROOF. By assumption, we can write V ~
¸
i
r
i
S
i
where the S
i
are simple ·modules,
no two of which are isomorphic. The centralizer of · in End
T
(V ) is End
¹
(V ), and
End
¹
(V ) ~End
¹
(
¸
i
r
i
S
i
). Because Hom
¹
(S
}
. S
i
) =0 for i =; ,
End
¹
(
r
i
S
i
) .
¸
i
End
¹
(r
i
S
i
) by (30)
.
¸
i
M
i
i
(D
i
) by (31)
where D
i
=End
¹
(S
i
). According to Schur’s lemma (7.24), D
i
is a division algebra, and
therefore M
i
i
(D
i
) is a simple Jalgebra (see 7.16).
2
THEOREM 7.34 Every semisimple Jalgebra is isomorphic to a product of simple J
algebras.
PROOF. Choose an ·module V on which · acts faithfully, for example, V =
¹
·. Then ·
is equal to its double centralizer C(C(·)) in End
T
(V ) (see 7.21). According to Theorem
7.33, C(·) is semisimple, and so C(C(·)) is a product of simple algebras.
2
Modules over a semisimple Jalgebra
Let · =T ×C be a product of Jalgebras. A Tmodule M becomes an ·module with
the action
(b. c)m=bm.
THEOREM 7.35 Let · be a semisimple Jalgebra, say, · = ·
1
× ×·
t
with the ·
i
simple. For each ·
i
, let S
i
be a simple ·
i
module (cf. 7.27).
The representations of G 107
(a) Each S
i
is a simple ·module, and every simple ·module is isomorphic to exactly
of of the S
i
.
(b) Every ·module is isomorphic to
¸
r
i
S
i
for some r
i
÷ N, and two modules
¸
r
i
S
i
and
¸
r
t
i
S
i
are isomorphic if and only if r
i
=r
t
i
for all i .
PROOF. (a) It is obvious that each S
i
is simple when regarded as an ·module, and that no
two of themare isomorphic. It follows from(7.25) that
¹
·~
¸
r
i
S
i
for some r
i
÷N. Let S
be a simple ·module, and let . be a nonzero element of S. Then the map a ÷a.:
¹
·÷S
is surjective, and so its restriction to some S
i
in
¹
· is nonzero, and hence an isomorphism.
(b) The ﬁrst part follows from (a) and the deﬁnition of a semisimple ring, and the second
part follows from (7.11).
2
The representations of G
PROPOSITION 7.36 The dimension of the centre of JŒG as an Jvector space is the num
ber of conjugacy classes in G.
PROOF. Let C
1
. . . . . C
t
be the conjugacy classes in G, and, for each i , let c
i
be the element
¸
o÷C
i
a in JŒG. We shall prove the stronger statement:
centre of JŒG =Jc
1
˚ ˚Jc
t
(32)
As c
1
. . . . . c
t
are obviously linearly independent, it sufﬁces to showthat they span the centre.
For any g ÷ G and
¸
o÷G
m
o
a ÷ JŒG,
g
¸
o÷G
m
o
a
g
1
=
¸
o÷G
m
o
gag
1
.
The coefﬁcient of a in the right hand sum is m
v
1
ov
, and so
g
¸
o÷G
m
o
a
g
1
=
¸
o÷G
m
v
1
ov
a.
This shows that
¸
o÷G
m
o
a lies in the centre of JŒG if and only if the function a ÷m
o
is
constant on conjugacy classes, i.e., if and only if
¸
o÷G
m
o
a ÷
¸
i
Jc
i
.
2
REMARK 7.37 An element
¸
o÷G
m
o
a of JŒG can be regarded as a map a ÷m
o
: G ÷
J. In this way, JŒG . Map(G. J). The action of G on JŒG corresponds to the action
(g} )(a) = }(g
1
a) of g ÷ G on } : G ÷ J. In the above proof, we showed that the
elements of the centre of JŒG correspond exactly to the functions } : G ÷ J that are
constant on each conjugacy class. Such functions are called class functions.
In the remainder of this chapter, we assume that J is an algebraically closed ﬁeld of
characteristic zero (e.g., C)
PROPOSITION 7.38 The group algebra JŒG is isomorphic to a product of matrix algebras
over J.
PROOF. Recall that, when J has characteristic zero, Maschke’s theorem (7.9) implies that
JŒG is semisimple, and so is a product of simple algebras (7.35). Each of these is a matrix
algebra over a division algebra (7.23), but the only division algebra over an algebraically
closed ﬁeld is the ﬁeld itself (7.29).
2
108 7. REPRESENTATIONS OF FINITE GROUPS
The representation G ÷GL(
TŒG]
JŒG) is called the regular representation.
THEOREM 7.39 (a) The number of isomorphism classes of simple JŒGmodules is equal
to the number of conjugacy classes in G.
(b) The multiplicity of any simple representation S in the regular representation is equal
to its degree dim
T
S.
(c) Let S
1
. . . . . S
t
be a set of representatives for the isomorphism classes of simple JG
modules, and let }
i
=dim
T
S
i
. Then
¸
1_i_t
}
2
i
=[G[.
PROOF. (a) Under our hypothesis, JŒG ~M
(
1
(J)× ×M
(
t
(J) for some integers }
1
. . . . . }
t
.
According to Theorem 7.35, the number of isomorphism classes of simple JŒGmodules
is the number of factors t . The centre of a product of Jalgebras is the product of their
centres, and so the centre of JŒG is isomorphic to tJ. Therefore t is the dimension of the
centre of J, which we know equals the number of conjugacy classes of G.
(b) With the notations of (7.15), M
(
(J) .1(1) ˚ ˚1(} ).
(c) The equality is simply the statement
¸
1_i_t
dim
T
M
(
i
(J) =dim
T
JŒG.
2
The characters of G
Recall that the trace Tr
V
(˛) of an endomorphism ˛: V ÷V of a vector space V is
¸
a
i i
where (a
i}
) is the matrix of ˛ with respect to some basis for V . It is independent of the
choice of the basis (the traces of conjugate matrices are equal).
From each representation of g ÷g
V
: G ÷GL(V ), we obtain a function ,
V
on G.
,
V
(g) =Tr
V
(g
V
).
called the character of j. Note that ,
V
depends only on the isomorphism class of the
JŒGmodule V , and that ,
V
is a class function. The character , is said to be simple (or
irreducible) if it is deﬁned by a simple JGmodule. The principal character ,
1
is that
deﬁned by the trivial representation of G (so ,
1
(g) = 1 for all g ÷ G), and the regular
character ,
reg
is that deﬁned by the regular representation. On computing ,
reg
(g) by using
the elements of G as a basis for JŒG, one see that ,
reg
(g) is the number of elements a of
G such that ga =a, and so
,
reg
(g) =
[G[ if g =e
0 otherwise.
When V has dimension 1, the character ,
V
of G is said to be linear. In this case, GL(V ) .
J
×
, and so ,
V
(g) = j(g). Therefore, ,
V
is a homomorphism G ÷ J
×
, and so this
deﬁnition of “linear character” essentially agrees with the earlier one.
LEMMA 7.40 For any Gmodules V and V
t
,
,
V ˚V
0 =,
V
÷,
V
0 .
The characters of G 109
PROOF. Compute the matrix of g
1
with respect to a basis of V ˚V
t
that is made by com
bining a basis for V with a basis for V
t
.
2
Let S
1
. . . . . S
t
be a set of representatives for the isomorphism classes of simple JG
modules with S
1
chosen to be the trivial representation, and let ,
1
. . . . . ,
t
be the corre
sponding characters.
PROPOSITION 7.41 The functions ,
1
. . . . . ,
t
are linearly independent over J, i.e., if c
1
. . . . . c
t
÷
J are such that
¸
i
c
i
,
i
(g) =0 for all g ÷ G, then the c
i
are all zero.
PROOF. Write JŒG ~ M
(
1
(J) × ×M
(
t
(J), and let e
}
= (0. . . . . 0. 1. 0. . . . . 0). Then
e
}
acts as 1 on S
}
and as 0 on S
}
for i =; , and so
,
i
(e
}
) =
}
}
=dim
T
S
}
if i =;
0 otherwise.
(33)
Therefore,
¸
i
c
i
,
i
(e
}
) =c
}
}
}
,
from which the claim follows.
2
PROPOSITION 7.42 Two JŒGmodules are isomorphic if and only if their characters are
equal.
PROOF. We have already observed that the character of a representation depends only
on its isomorphism class. Conversely, if V =
¸
1_i_t
c
i
S
i
, c
i
÷ N, then its character is
,
V
=
¸
1_i_t
c
i
,
i
, and (33) shows that c
i
= ,
V
(e
i
)¡}
i
. Therefore ,
V
determines the
multiplicity with which each S
i
occurs in V , and hence it determines the isomorphism
class of V .
2
ASIDE 7.43 The proposition is false if J is allowed to have characteristic ] =0. For example, the
representation o
i
÷
1 i
0 1
: C
;
÷GL
2
(J) of (7.1c) is not trivial, but it has the same character
as the trivial representation. The proposition is false even when the characteristic of J doesn’t
divide the order of the group, because, for any representation G ÷GL(V ), the character of the
representation of G on ]V is identically zero.
Any function G ÷J that can be expressed as a Zlinear combination of characters is
called a virtual character.
2
PROPOSITION 7.44 The simple characters of G form a Zbasis for the virtual characters
of G.
PROOF. Let ,
1
. . . . . ,
t
be the simple characters of G. Then the characters of G are exactly
the class functions that can be expressed in the form
¸
m
i
,
i
, m
i
÷ N, and so the virtual
characters are exactly the class functions that can be expressed
¸
m
i
,
i
, m
i
÷ Z. Therefore
the simple characters certainly generate the Zmodule of virtual characters, and Proposition
7.41 shows that they are linearly independent over Z (even over J).
2
2
Some authors call it a generalized character, but this is to be avoided: there is more than one way to
generalize the notion of a character.
110 7. REPRESENTATIONS OF FINITE GROUPS
PROPOSITION 7.45 The simple characters of G form an Jbasis for the class functions on
G.
PROOF. The class functions are the functions from the set of conjugacy classes in G to
J. As this set has t elements, they form an Jvector space of dimension t . As the simple
characters are a set of t linearly independent elements of this vector space, they must form
a basis.
2
We now assume that J is a subﬁeld of C stable under complex conjugation c ÷ ¨ c.
For class functions }
1
and }
2
on G, deﬁne
(}
1
[}
2
) =
1
[G[
¸
o÷G
}
1
(a)}
2
(a).
LEMMA 7.46 The pairing ( [ ) is an inner product on the Jspace of class functions on G.
PROOF. We have to check:
˘ (}
1
÷}
2
[} ) =(}
1
[} ) ÷(}
2
[} ) for all class functions }
1
. }
2
. } ;
˘ (c}
1
[}
2
) =c(}
1
. }
2
) for c ÷ J and class functions }
1
. }
2
;
˘ (}
2
[}
1
) =(}
1
[}
2
) for all class functions }
1
. }
2
;
˘ (} [} ) > 0 for all nonzero class functions } .
All of these are obvious from the deﬁnition.
2
For an JŒGmodule V , V
G
denotes the submodule of elements ﬁxed by G:
V
G
={· ÷ V [ g· =· for all g ÷ G]
LEMMA 7.47 Let ¬ be the element
1
G
¸
o÷G
a of JŒG. For any JŒGmodule V , ¬
V
is
a projector with image V
G
.
PROOF. For any g ÷ G,
g¬ =
1
[G[
¸
o÷G
ga =
1
[G[
¸
o÷G
a =¬. (34)
from which it follows that ¬¬ = ¬ (in the Jalgebra JŒG). Therefore, for any JŒG
module V , ¬
2
V
=¬
V
and so ¬
V
is a projector. If · is in its image, say · =¬·
0
, then
g· =g¬·
0
(34)
= ¬·
0
=·
and so · lies in V
G
. Conversely, if · ÷ V
G
, the obviously ¬· =
1
G
¸
o÷G
a· =·, and so
· is in the image of ¬.
2
PROPOSITION 7.48 For any JŒGmodule V ,
dim
T
V
G
=
1
[G[
¸
o÷G
,
V
(a).
The character table of a group 111
PROOF. Let ¬ be as in Lemma 7.47. Because ¬
V
is a projector, V is the direct sum of
its 0eigenspace and its 1eigenspace, and we showed that the latter is V
G
. Therefore,
Tr
V
(¬
V
) =dim
T
V
G
. On the other hand, because the trace is a linear function,
Tr
V
(¬
V
) =
1
[G[
¸
o÷G
Tr
V
(a
V
) =
1
[G[
¸
o÷G
,
V
(a).
2
THEOREM 7.49 For any JŒGmodules V and W.
dim
T
Hom
TŒG]
(V. W ) =(,
V
[,
W
).
PROOF. The group G acts on the space Hom
T
(V. W ) of Jlinear maps V ÷W by the
rule,
(gc)(·) =g(c(g·)). g ÷ G. c ÷ Hom
T
(V. W ). · ÷ V.
and Hom
T
(V. W )
G
=Hom
TG
(V. W ).
2
COROLLARY 7.50 If , and ,
t
are simple characters, then
(,[,
t
) =
1 if , =,
t
0 otherwise.
Therefore the simple characters form an orthonormal basis for the space of class functions
on G.
The character table of a group
To be written.
Examples
To be written.
Exercises
71 Let C be an n×r matrix with coefﬁcients in a ﬁeld J. Show that
{M ÷ M
n
(J) [ MC =0]
is a left ideal in M
n
(J), and that every left ideal is of this form for some C.
72 This exercise shows how to recover a ﬁnite group G from its category of representa
tions over a ﬁeld k. Let S be a ﬁnite set, and let · be the set of maps S ÷k.
(a) Show that · becomes a commutative ring with the product
(}
1
}
2
)(g) =}
1
(g)}
2
(g). }
1
, }
2
÷ ·. g ÷ G.
Moreover, when we identify c ÷k with the constant function, ·becomes a kalgebra.
112 7. REPRESENTATIONS OF FINITE GROUPS
(b) Show that
· .
¸
x÷S
k
x
(product of copies of k indexed by the elements of S).
and that the k
x
are exactly the minimal ksubalgebras of ·. Deduce that End
kalg
(·) .
Sym(S).
(c) Let (}
1
. }
2
) ÷ ·×· act on S ×S by (}
1
. }
2
)(s
1
. s
2
) =}
1
(s
1
)}
2
(s
2
); show that this
deﬁnes a bijection ·˝· .Map(S ×S. k). Now take S =G.
(d) Show that the map r
¹
: G ÷End
klinear
(·),
(r
¹
(g)} )(g
t
) =}(gg
t
). } ÷ ·. g. g
t
÷ G
is a representation of G (this is the regular representation).
(e) Deﬁne z: · ÷·˝· by z(} )(g
1
. g
2
) = }(g
1
g
2
). Show that, for any homomor
phism˛: ·÷·of kalgebras such (1˝˛)ız=zı˛, there exists a unique element
g ÷ G such that ˛(} ) =g} for all } ÷ ·. [Hint: Deduce from (b) that there exists a
bijection ç: G ÷G such that (˛} )(g) =}(çg) for all g ÷ G. From the hypothesis
on ˛, deduce that ç(g
1
g
2
) =g
1
ç(g
2
) for all g
1
. g
2
÷ G(1). Hence ç(g) =g ç(e)
for all g ÷ G. Deduce that ˛(} ) =ç(e)} for all } ÷ ·.]
(f) Show that the following maps are Gequivariant
e:k ÷· (trivial representation on k; r
¹
on ·)
m:·˝· ÷· (r
¹
˝r
¹
on ·˝·; r
¹
on ·)
z:· ÷·˝· (r
¹
on ·; 1˝r
¹
on ·˝·).
(g) Suppose that we are given, for each ﬁnite dimensional representation (V. r
V
), a k
linear map z
V
. If the family (z
V
) satisﬁes the conditions
i) for all representations V , W , z
V ˝W
=z
V
˝z
W
:
ii) for k with its trivial representation, z
k
=id
k
;
iii) for all Gequivariant maps ˛: V ÷W , z
W
ı˛ =˛ ız
V
:
then there exists a unique g ÷ G(1) such that z
V
=r
V
(g) for all V . [Hint: show that
z
¹
satisﬁes the conditions of (d).]
NOTES For a historical account of the representation theory of ﬁnite groups, emphasizing the work
of “the four principal contributors to the theory in its formative stages: Ferdinand Georg Frobenius,
William Burnside, Issai Schur, and Richard Brauer”, see Curtis 1999.
APPENDIX A
Additional Exercises
34. Prove that a ﬁnite group G having just one maximal subgroup must be a cyclic ]group,
] prime.
35. Let a and b be two elements of S
T6
. If a and b both have order 146 and ab =ba, what
are the possible orders of the product ab?
37. Suppose that the group G is generated by a set X.
(a) Show that if g.g
1
÷ X for all . ÷ X. g ÷ G, then the commutator subgroup of G is
generated by the set of all elements ...
1
.
1
for .. . ÷ X.
(b) Show that if .
2
=1 for all . ÷ X, then the subgroup H of G generated by the set of
all elements .. for .. . ÷ X has index 1 or 2.
38. Suppose ] _3 and 2]÷1 are both prime numbers (e.g., ] =3. 7. 19. 31. . . .). Prove, or
disprove by example, that every group of order ](2]÷1) is commutative.
39. Let H be a subgroup of a group G. Prove or disprove the following:
(a) If G is ﬁnite and 1 is a Sylow ]subgroup, then H ¨1 is a Sylow ]subgroup of
H.
(b) If G is ﬁnite, 1 is a Sylow ]subgroup, and H ¯N
G
(1), then N
G
(H) =H.
(c) If g is an element of G such that gHg
1
cH, then g ÷ N
G
(H).
40. Prove that there is no simple group of order 616.
41. Let n and k be integers 1 _ k _ n. Let H be the subgroup of S
n
generated by the
cycle (a
1
. . . a
k
). Find the order of the centralizer of H in S
n
. Then ﬁnd the order of the
normalizer of H in S
n
. [The centralizer of H is the set of g ÷ G such ghg
1
=h for all
h ÷ H. It is again a subgroup of G.]
42. Prove or disprove the following statement: if H is a subgroup of an inﬁnite group G,
then for all . ÷ G, .H.
1
cH == .
1
H. cH.
43. Let H be a ﬁnite normal subgroup of a group G, and let g be an element of G. Suppose
that g has order n and that the only element of H that commutes with g is 1. Show that:
(a) the mapping h ÷g
1
h
1
gh is a bijection from H to H;
(b) the coset gH consists of elements of G of order n.
113
114 A. ADDITIONAL EXERCISES
44. Show that if a permutation in a subgroup G of S
n
maps . to ., then the normalizers of
the stabilizers Stab(.) and Stab(.) of . and . have the same order.
45. Prove that if all Sylow subgroups of a ﬁnite group G are normal and abelian, then the
group is abelian.
46. A group is generated by two elements a and b satisfying the relations: a
3
=b
2
, a
n
=1,
b
n
=1 where m and n are positive integers. For what values of m and n can G be inﬁnite.
47. Show that the group G generated by elements . and . with deﬁning relations .
2
=
.
3
=(..)
4
=1 is a ﬁnite solvable group, and ﬁnd the order of G and its successive derived
subgroups G
t
, G
tt
, G
ttt
.
48. A group G is generated by a normal set X of elements of order 2. Show that the com
mutator subgroup G
t
of G is generated by all squares of products .. of pairs of elements
of X.
49. Determine the normalizer N in GL
n
(J) of the subgroup H of diagonal matrices, and
prove that N¡H is isomorphic to the symmetric group S
n
.
50. Let G be a group with generators . and . and deﬁning relations .
2
, .
5
, (..)
4
. What
is the index in G of the commutator group G
t
of G.
51. Let G be a ﬁnite group, and H the subgroup generated by the elements of odd order.
Show that H is normal, and that the order of G¡H is a power of 2.
52. Let G be a ﬁnite group, and 1 a Sylow ]subgroup. Show that if H is a subgroup of
G such that N
G
(1) cH cG, then
(a) the normalizer of H in G is H;
(b) (G : H) ÷1 (mod ]).
53. Let G be a group of order 33 25. Show that G is solvable. (Hint: A ﬁrst step is to ﬁnd
a normal subgroup of order 11 using the Sylow theorems.)
54. Suppose that ˛ is an endomorphism of the group G that maps G onto G and commutes
with all inner automorphisms of G. Show that if G is its own commutator subgroup, then
˛. =. for all . in G.
55. Let G be a ﬁnite group with generators s and t each of order 2. Let n =(G : 1)¡2.
(a) Show that G has a cyclic subgroup of order n. Now assume n odd.
(b) Describe all conjugacy classes of G.
(c) Describe all subgroups of G of the form C(.) ={. ÷ G[.. =..], . ÷ G.
(d) Describe all cyclic subgroups of G.
(e) Describe all subgroups of G in terms of (b) and (d).
(f) Verify that any two ]subgroups of G are conjugate (] prime).
56. Let G act transitively on a set X. Let N be a normal subgroup of G, and let Y be the
set of orbits of N in X. Prove that:
(a) There is a natural action of G on Y which is transitive and shows that every orbit of
N on X has the same cardinality.
(b) Show by example that if N is not normal then its orbits need not have the same
cardinality.
115
57. Prove that every maximal subgroup of a ﬁnite ]group is normal of prime index (] is
prime).
58. A group G is metacyclic if it has a cyclic normal subgroup N with cyclic quotient G¡N.
Prove that subgroups and quotient groups of metacyclic groups are metacyclic. Prove or
disprove that direct products of metacyclic groups are metacylic.
59. Let G be a group acting doubly transitively on X, and let . ÷ X. Prove that:
(a) The stabilizer G
x
of . is a maximal subgroup of G.
(b) If N is a normal subgroup of G, then either N is contained in G
x
or it acts transitively
on X.
60. Let .. . be elements of a group G such that ...
1
=.
5
, . has order 3, and . =1 has
odd order. Find (with proof) the order of ..
61. Let H be a maximal subgroup of G, and let · be a normal subgroup of H and such
that the conjugates of · in G generate it.
(a) Prove that if N is a normal subgroup of G, then either N cH or G =N·.
(b) Let M be the intersection of the conjugates of H in G. Prove that if G is equal to its
commutator subgroup and · is abelian, then G¡M is a simple group.
62. (a) Prove that the centre of a nonabelian group of order ]
3
, ] prime, has order ].
(b) Exhibit a nonabelian group of order 16 whose centre is not cyclic.
63. Show that the group with generators ˛ and ˇ and deﬁning relations
˛
2
=ˇ
2
=(˛ˇ)
3
=1
is isomorphic with the symmetric group S
3
of degree 3 by giving, with proof, an explicit
isomorphism.
64. Prove or give a counterexample:
(a) Every group of order 30 has a normal subgroup of order 15.
(b) Every group of order 30 is nilpotent.
65. Let t ÷ Z, and let G be the group with generators .. . and relations ...
1
= .
t
,
.
3
=1.
(a) Find necessary and sufﬁcient conditions on t for G to be ﬁnite.
(b) In case G is ﬁnite, determine its order.
66. Let G be a group of order ]q, ] =q primes.
(a) Prove G is solvable.
(b) Prove that G is nilpotent ” G is abelian ” G is cyclic.
(c) Is G always nilpotent? (Prove or ﬁnd a counterexample.)
67. Let X be a set with ]
n
elements, ] prime, and let G be a ﬁnite group acting transitively
on X. Prove that every Sylow ]subgroup of G acts transitively on X.
68. Let G ='a. b. c [ bc =cb, a
4
=b
2
=c
2
=1. aca
1
=c, aba
1
=bc). Determine
the order of G and ﬁnd the derived series of G.
69. Let N be a nontrivial normal subgroup of a nilpotent group G. Prove that N ¨7(G) =
1.
70. Do not assume Sylow’s theorems in this problem.
116 A. ADDITIONAL EXERCISES
(a) Let H be a subgroup of a ﬁnite group G, and 1 a Sylow ]subgroup of G. Prove
that there exists an . ÷ G such that .1.
1
¨H is a Sylow ]subgroup of H.
(b) Prove that the group of n ×n matrices
¸
¸
¸
1 + . . .
0 1
. . .
0 1
is a Sylow ]subgroup of
GL
n
(F
;
).
(c) Indicate how (a) and (b) can be used to prove that any ﬁnite group has a Sylow ]
subgroup.
71. Suppose H is a normal subgroup of a ﬁnite group G such that G¡H is cyclic of order
n, where n is relatively prime to (G : 1). Prove that G is equal to the semidirect product
H S with S a cyclic subgroup of G of order n.
72. Let H be a minimal normal subgroup of a ﬁnite solvable group G. Prove that H is
isomorphic to a direct sum of cyclic groups of order ] for some prime ].
73. (a) Prove that subgroups · and T of a group G are of ﬁnite index in G if and only if
·¨T is of ﬁnite index in G.
(b) An element . of a group G is said to be an FCelement if its centralizer C
G
(.) has ﬁnite
index in G. Prove that the set of all JC elements in G is a normal.
74. Let G be a group of order ]
2
q
2
for primes ] > q. Prove that G has a normal subgroup
of order ]
n
for some n _1.
75. (a) Let 1 be a ﬁnite nilpotent group, and let 1be a subgroup of 1 such that 1 ı1 =1,
where ı1 is the derived subgroup. Prove that 1 =1. [You may assume that a ﬁnite group
is nilpotent if and only if every maximal subgroup is normal.]
(b) Let G be a ﬁnite group. If G has a subgroup H such that both G¡ıH and H are
nilpotent, prove that G is nilpotent.
76. Let G be a ﬁnite noncyclic ]group. Prove that the following are equivalent:
(a) (G : 7(G)) _]
2
.
(b) Every maximal subgroup of G is abelian.
(c) There exist at least two maximal subgroups that are abelian.
77. Prove that every group G of order 56 can be written (nontrivially) as a semidirect
product. Find (with proofs) two nonisomorphic nonabelian groups of order 56.
78. Let G be a ﬁnite group and c : G ÷G a homomorphism.
(a) Prove that there is an integer n _0 such that c
n
(G) =c
n
(G) for all integers m_n.
Let ˛ =c
n
.
(b) Prove that G is the semidirect product of the subgroups Ker ˛ and Im˛.
(c) Prove that Im˛ is normal in G or give a counterexample.
79. Let S be a set of representatives for the conjugacy classes in a ﬁnite group G and let H
be a subgroup of G. Show that S cH == H =G.
80. Let G be a ﬁnite group.
(a) Prove that there is a unique normal subgroup 1 of G such that (i) G¡1 is solvable
and (ii) if N is a normal subgroup and G¡N is solvable, then N ¯1.
(b) Show that 1 is characteristic.
(c) Prove that 1 =Œ1. 1 and that 1 =1 or 1 is nonsolvable.
APPENDIX B
Solutions to the Exercises
These solutions fall somewhere between hints and complete solutions. Students were ex
pected to write out complete solutions.
11 By inspection, the only element of order 2 is c =a
2
=b
2
. Since gcg
1
also has order
2, it must equal c, i.e., gcg
1
=c for all g ÷ O. Thus c commutes with all elements of O,
and {1. c] is a normal subgroup of O. The remaining subgroups have orders 1, 4, or 8, and
are automatically normal (see 1.36a).
12 The product ab =
1 1
0 1
, and
1 1
0 1
n
=
1 n
0 1
.
13 Consider the subsets {g. g
1
] of G. Each set has exactly 2 elements unless g has order
1 or 2, in which case it has 1 element. Since G is a disjoint union of these sets, there must
be a (nonzero) even number of sets with 1 element, and hence at least one element of order
2.
14 The symmetric group S
n
contains a subgroup that is a direct product of subgroups S
n
1
,
. . . , S
n
r
.
15 Because the group G¡N has order n, (gN)
n
= 1 for every g ÷ G (see 1.27). But
(gN)
n
=g
n
N, and so g
n
÷ N. For the second statement, consider the subgroup {1. s] of
D
3
. It has index 3 in D
3
, but the element t has order 2, and so t
3
=t … {1. s].
16 (a) Let a. b ÷ G. We are given that a
2
=b
2
=(ab)
2
=e. In particular, abab =e. On
multiplying this on right by ba, we ﬁnd that ab =ba. (b) Show by induction that
¸
1 a b
0 1 c
0 0 1
n
=
¸
1 na nb ÷
n(n1)
2
ac
0 1 nc
0 0 1
.
17 Let H be a subgroup of a ﬁnite group G, and assume that H contains at least one
element from each conjugacy class of G. Then G is the union of the conjugates of H, but
there are at most (G: H) of these, and they overlap (at least) in {1], and so this can happen
only if H =G. (According to Serre 2003, this result goes back to Jordan in the 1870s.)
117
118 B. SOLUTIONS TO THE EXERCISES
18 Commensurability is obviously reﬂexive and symmetric, and so it sufﬁces to prove
transitivity. We shall use that if a subgroup H of a group G has ﬁnite index in G, then
H ¨G
t
has ﬁnite index in G
t
for any subgroup G
t
of G (because the natural map G
t
¡H ¨
G
t
÷G¡H is injective). Using this, it follows that if H
1
and H
3
are both commensurable
with H
2
, then H
1
¨H
2
¨H
3
is of ﬁnite index in H
1
¨H
2
and in H
2
¨H
3
(and therefore
also in H
1
and H
3
). As H
1
¨H
3
¯ H
1
¨H
2
¨H
3
, it also has ﬁnite index in each of H
1
and H
3
.
19 By assumption, the set G is nonempty, so let a ÷ G. Because G satisﬁes the can
cellation law, the map . ÷a.: G ÷G is a permutuation of G, and some power of this
permutation is the identity permutation. Therefore, for some n _1, a
n
. =. for all . ÷ G,
and so a
n
is a left neutral element. By counting, one sees that every element has a left
inverse, and so we can apply (1.10a).
21 The key point is that 'a) ='a
2
) ×'a
n
). Apply (1.50) to see that D
2n
breaks up as a
product.
22 Note ﬁrst that any group generated by a commuting set of elements must be commu
tative, and so the group G in the problem is commutative. According to (2.8), any map
{a
1
. . . . . a
n
] ÷· with · commutative extends uniquely to homomorphism G ÷·, and so
G has the universal property that characterizes the free abelian group on the generators a
i
.
23 (a) If a =b, then the word a ab
1
b
1
is reduced and =1. Therefore, if a
n
b
n
=
1, then a =b. (b) is similar. (c) The reduced form of .
n
, . =1, has length at least n.
24 (a) Universality. (b) C
o
×C
o
is commutative, and the only commutative free groups
are 1 and C
o
. (c) Suppose a is a nonempty reduced word in .
1
. . . . . .
n
, say a =.
i
(or
.
1
i
). For ; =i , the reduced form of Œ.
}
. a
def
=.
}
a.
1
}
a
1
can’t be empty, and so a and
.
}
don’t commute.
25 The unique element of order 2 is b
2
. Since gb
2
g
1
also has order 2 for any g ÷ O
n
,
we see that gb
2
g
1
= b
2
, and so b
2
lies in the centre. [Check that it is the full centre.]
The quotient group O
n
¡'b
2
) has generators a and b, and relations a
2
n2
= 1, b
2
= 1,
bab
1
=a
1
, which is a presentation for D
2
n2 (see 2.9).
26 (a) A comparison of the presentation D
n
= 'r. s [ r
n
. s
2
. srsr = 1) with that for G
suggests putting r =ab and s =a. Check (using 2.8) that there are homomorphisms:
D
n
÷G. r ÷ab. s ÷a. G ÷D
n
. a ÷s. b ÷s
1
r.
The composites D
n
÷G ÷D
n
and G ÷D
n
÷G are the both the identity map on gen
erating elements, and therefore (2.8 again) are identity maps. (b) Omit.
27 The hint gives ab
3
a
1
=bc
3
b
1
. But b
3
=1. So c
3
=1. Since c
4
=1, this forces
c =1. From acac
1
=1 this gives a
2
=1. But a
3
=1. So a =1. The ﬁnal relation then
gives b =1.
119
28 The elements .
2
, .., .
2
lie in the kernel, and it is easy to see that '.. .[.
2
. ... .
2
)
has order (at most) 2, and so they must generate the kernel (at least as a normal group —
the problem is unclear). One can prove directly that these elements are free, or else apply
the NielsenSchreier theorem (2.6). Note that the formula on p. 34 (correctly) predicts that
the kernel is free of rank 2 2÷2÷1 =3
29 We have to show that if s and t are elements of a ﬁnite group satisfying t
1
s
3
t =s
5
,
then the given element g is equal to 1. Because the group is ﬁnite, s
n
=1 for some n. If
3[n, the proof is easy, and so we suppose that gcd(3. n) =1. But then
3r ÷nr
t
=1, some r. r
t
÷ Z.
and so s
3i
=s. Hence
t
1
st =t
1
s
3i
t =(t
1
s
3
t )
i
=s
5i
.
Now,
g =s
1
(t
1
s
1
t )s(t
1
st ) =s
1
s
5i
ss
5i
=1.
as required. [In such a question, look for a pattern. Note that g has two conjugates in it, as
does the relation for G, and so it is natural to try to relate them.]
31 Let N be the unique subgroup of order 2 in G. Then G¡N has order 4, but there is no
subgroup O c G of order 4 with O¨N = 1 (because every group of order 4 contains a
group of order 2), and so G =N O for any O. A similar argument applies to subgroups
N of order 4.
32 For any g ÷ G, gMg
1
is a subgroup of order m, and therefore equals M. Thus M
(similarly N) is normal in G, and MN is a subgroup of G. The order of any element of
M ¨N divides gcd(m. n) = 1, and so equals 1. Now (1.51) shows that M ×N ~ MN,
which therefore has order mn, and so equals G.
33 Show that GL
2
(F
2
) permutes the 3 nonzero vectors in F
2
×F
2
(2dimensional vector
space over F
2
).
34 The following solutions were suggested by readers. We write the quaternion group as
O={˙1. ˙i. ˙;. ˙k].
(A) Take a cube. Write the six elements of O of order 4 on the six faces with i opposite ÷i ,
etc.. Each rotation of the cube induces an automorphism of O, and Aut(O) is the symmetry
group of the cube, S
4
. (B) The group O has one element of order 2, namely ÷1, and six
elements of order 4, namely, ˙i , ˙; , ˙k. Any automorphism ˛ of O must map ÷1 to
itself and permute the elements of order 4. Note that i; = k, ;k = i , ki = ; , so ˛ must
send the circularly ordered set i. ;. k to a similar set, i.e., to one of the eight sets in the
following table:
i ; k ÷i ÷; k
i ÷; ÷k ÷i ; ÷k
i k ÷; ÷i ÷k ÷;
i ÷k ; ÷i k ;
Because ˛(÷1) =÷1, ˛ must permute the rows of the table, and it is not difﬁcult to see that
all permutations are possible.
120 B. SOLUTIONS TO THE EXERCISES
35 The pair
N =
¸
¸
¸
1 0 b
0 1 c
0 0 1
¸
¸
and O=
¸
¸
¸
a 0 0
0 a 0
0 0 J
¸
¸
satisﬁes the conditions (i), (ii), (iii) of (3.8). For example, for (i) (Maple says that)
¸
a 0 b
0 a c
0 0 J
¸
1 0 b
0 1 c
0 0 1
¸
a 0 b
0 a c
0 0 J
1
=
¸
1 0 ÷
b
d
÷
1
d
(b ÷ab)
0 1 ÷
c
d
÷
1
d
(c ÷ac)
0 0 1
It is not a direct product of the two groups because it is not commutative.
36 Let g generate C
o
. Then the only other generator is g
1
, and the only nontrivial
automorphism is g ÷g
1
. Hence Aut(C
o
) ={˙1]. The homomorphism S
3
÷Aut(S
3
)
is injective because 7(S
3
) = 1, but S
3
has exactly 3 elements a
1
. a
2
. a
3
of order 2 and 2
elements b. b
2
of order 3. The elements a
1
. b generate S
3
, and there are only 6 possibilities
for ˛(a
1
), ˛(b), and so S
3
÷Aut(S
3
) is also onto.
37 (a) The element g
o(q)
÷ N, and so has order dividing [N[. (c) The element g =
(1. 4. 3)(2. 5), and so this is obvious. (d) By the ﬁrst part, ((1. 0. . . . . 0). q)
;
=((1. . . . . 1). 1),
and (1. . . . . 1) has order ] in (C
;
)
;
. (e) We have (n. q)(n. q) =(nn
1
. qq) =(1. 1).
38 Let n q ÷ 7(G). Then
(n q)(1 q
t
) = n qq
t
(1 q
t
)(n q) = q
t
nq
t1
q
t
q
all q
t
÷ O
¸
==
n ÷ C
1
(O)
q ÷ 7(O)
and
(n q)(n
t
1) = nqn
t
q
1
q
(n
t
1)(n q) = n
t
n q
n
t
÷ N
¸
== n
1
n
t
n =qn
t
q
1
.
The converse and the remaining statements are easy.
41 Let c: G¡H
1
÷G¡H
2
be a Gmap, and let c(H
1
) = gH
2
. For a ÷ G, c(aH
1
) =
ac(H
1
) = agH
2
. When a ÷ H
1
, c(aH
1
) = gH
2
, and so agH
2
= gH
2
; hence g
1
ag ÷
H
2
, and so a ÷ gH
2
g
1
. We have shown H
1
c gH
2
g
1
. Conversely, if g satisﬁes this
condition, the aH
1
÷agH
2
is a welldeﬁned map of Gsets.
42 Let H be a proper subgroup of G, and let N =N
G
(H). The number of conjugates of
H is (G : N) _(G : H) (see 4.8). Since each conjugate of H has (H : 1) elements and the
conjugates overlap (at least) in {1], we see that
ˇ
ˇ
ˇ
¸
gHg
1
ˇ
ˇ
ˇ < (G : H)(H : 1) =(G : 1).
For the second part, choose S to be a set of representatives for the conjugacy classes.
43 According to 4.17, 4.18, there is a normal subgroup N of order ]
2
, which is commuta
tive. Now show that G has an element c of order ] not in N, and deduce that G =N 'c),
etc..
121
44 Let H be a subgroup of index ], and let N be the kernel of G ÷Sym(G¡H) — it is
the largest normal subgroup of G contained in H (see 4.22). If N =H, then (H : N) is
divisible by a prime q _ ], and (G : N) is divisible by ]q. But ]q doesn’t divide ]Š —
contradiction.
45 Embed G into S
2n
, and let N = ·
2n
¨G. Then G¡N ·÷S
2n
¡·
2n
= C
2
, and so
(G : N) _ 2. Let a be an element of order 2 in G, and let b
1
. . . . . b
n
be a set of right
coset representatives for 'a) in G, so that G ={b
1
. ab
1
. . . . . b
n
. ab
n
]. The image of a in
S
2n
is the product of the m transpositions (b
1
. ab
1
). . . . . (b
n
. ab
n
), and since m is odd, this
implies that a … N.
46 The set X of kcycles in S
n
is normal, and so the group it generates is normal (1.38).
But, when n _ 5, the only nontrivial normal subgroups of S
n
are ·
n
and S
n
itself. If k is
odd, then X is contained in ·
n
, and if k is even, then it isn’t.
47 (a) The number of possible ﬁrst rows is 2
3
÷1; of second rows 2
3
÷2; of third rows
2
3
÷2
2
; whence (G : 1) =7×6×4 =168. (b) Let V =F
3
2
. Then [V [ =2
3
=8. Each line
through the origin contains exactly one point = origin, and so [X[ =7. (c) We make a list
of possible characteristic and minimal polynomials:
Characteristic poly. Min’l poly. Size Order of element in class
1 X
3
÷X
2
÷X ÷1 X ÷1 1 1
2 X
3
÷X
2
÷X ÷1 (X ÷1)
2
21 2
3 X
3
÷X
2
÷X ÷1 (X ÷1)
3
42 4
4 X
3
÷1 =(X ÷1)(X
2
÷X ÷1) Same 56 3
5 X
3
÷X ÷1 (irreducible) Same 24 7
6 X
3
÷X
2
÷1 (irreducible) Same 24 7
Here size denotes the number of elements in the conjugacy class. Case 5: Let ˛ be an
endomorphism with characteristic polynomial X
3
÷X ÷1. Check from its minimal poly
nomial that ˛
T
=1, and so ˛ has order 7. Note that V is a free F
2
Œ˛module of rank one,
and so the centralizer of ˛ in G is F
2
Œ˛ ¨G ='˛). Thus [C
G
(˛)[ =7, and the number of
elements in the conjugacy class of ˛ is 168¡7 =24. Case 6: Exactly the same as Case 5.
Case 4: Here V =V
1
˚V
2
as an F
2
Œ˛module, and
End
F
2
Œ˛]
(V ) =End
F
2
Œ˛]
(V
1
) ˚End
F
2
Œ˛]
(V
2
).
Deduce that [C
G
(˛)[ =3, and so the number of conjugates of ˛ is
16S
3
=56. Case 3: Here
C
G
(˛) = F
2
Œ˛ ¨G = '˛), which has order 4. Case 1: Here ˛ is the identity element.
Case 2: Here V =V
1
˚V
2
as an F
2
Œ˛module, where ˛ acts as 1 on V
1
and has minimal
polynomial X
2
÷1 on V
2
. Either analyse, or simply note that this conjugacy class contains
all the remaining elements. (d) Since 168 =2
3
×3×7, a proper nontrivial subgroup H of
G will have order
2. 4. 8. 3. 6. 12. 24. 7. 14. 28. 56. 21. 24, or 84.
If H is normal, it will be a disjoint union of {1] and some other conjugacy classes, and so
(N : 1) =1÷
¸
c
i
with c
i
equal to 21, 24, 42, or 56, but this doesn’t happen.
48 Since G¡7(G) ·÷Aut(G), we see that G¡7(G) is cyclic, and so by (4.19) that G is
commutative. If G is ﬁnite and not cyclic, it has a factor C
;
r ×C
;
s etc..
122 B. SOLUTIONS TO THE EXERCISES
49 Clearly (i; ) =(1; )(1i )(1; ). Hence any subgroup containing (12). (13). . . . contains all
transpositions, and we know S
n
is generated by transpositions.
410 Note that C
G
(.)¨H =C
1
(.), and so H¡C
1
(.) ~H C
G
(.)¡C
G
(.)). Prove each
class has the same number c of elements. Then
[1[ =(G : C
G
(.)) =(G : H C
G
(.))(H C
G
(.) : C
G
(.)) =kc.
411 (a) The ﬁrst equivalence follows fromthe preceding problem. For the second, note that
o commutes with all cycles in its decomposition, and so they must be even (i.e., have odd
length); if two cycles have the same odd length k, one can ﬁnd a product of k transpositions
which interchanges them, and commutes with o; conversely, show that if the partition of n
deﬁned by o consists of distinct integers, then o commutes only with the group generated
by the cycles in its cycle decomposition. (b) List of conjugacy classes in S
T
, their size,
parity, and (when the parity is even) whether it splits in ·
T
.
Cycle Size Parity Splits in ·
T
‹ C
T
(o) contains
1 (1) 1 1 N
2 (12) 21 O
3 (123) 70 1 N (67)
4 (1234) 210 O
5 (12345) 504 1 N (67)
6 (123456) 840 O
7 (1234567) 720 1 Y 720 doesn’t divide 2520
8 (12)(34) 105 1 N (67)
9 (12)(345) 420 O
10 (12)(3456) 630 1 N (12)
11 (12)(3456) 504 O
12 (123)(456) 280 1 N (14)(25)(36)
13 (123)(4567) 420 O
14 (12)(34)(56) 105 O
15 (12)(34)(567) 210 1 N (12)
412 According to Gap, n =6, a ÷(13)(26)(45), b ÷(12)(34)(56).
413 Since Stab(g.
0
) =gStab(.
0
)g
1
, if H cStab(.
0
) then H cStab(.) for all ., and
so H = 1, contrary to hypothesis. Now Stab(.
0
) is maximal, and so H Stab(.
0
) = G,
which shows that H acts transitively.
51 Let ] be a prime dividing [G[ and let 1 be a Sylow ]subgroup, of order ]
n
say. The
elements of 1 all have order dividing ]
n
, and it has at most
1÷]÷ ÷]
n1
=
]
n
÷1
]÷1
< ]
n
elements of order dividing ]
n1
; therefore 1 must have an element of order ]
n
, and so it
is cyclic. Each Sylow ]subgroup has exactly ]
n
elements of order dividing ]
n
, and so
there can be only one. Now (5.9) shows that G is a product of its Sylow subgroups.
123
62 No, D
4
and the quaternion group have isomorphic commutator subgroups and quotient
groups but are not isomorphic. Similarly, S
n
and ·
n
×C
2
are not isomorphic when n _5.
APPENDIX C
TwoHour Examination
1. Which of the following statements are true (give brief justiﬁcations for each of (a), (b),
(c), (d); give a correct set of implications for (e)).
(a) If a and b are elements of a group, then a
2
=1. b
3
=1 == (ab)
6
=1.
(b) The following two elements are conjugate in S
T
:
1 2 3 4 5 6 7
3 4 5 6 7 2 1
.
1 2 3 4 5 6 7
2 3 1 5 6 7 4
.
(c) If G and H are ﬁnite groups and G×·
594
~H ×·
594
. then G ~H.
(d) The only subgroup of ·
5
containing (123) is ·
5
itself.
(e) Nilpotent == cyclic == commutative == solvable (for a ﬁnite group).
2. How many Sylow 11subgroups can a group of order 110 = 2 5 11 have? Classify
the groups of order 110 containing a subgroup of order 10. Must every group of order 110
contain a subgroup of order 10?
3. Let G be a ﬁnite nilpotent group. Show that if every commutative quotient of G is cyclic,
then G itself is cyclic. Is the statement true for nonnilpotent groups?
4. (a) Let G be a subgroup of Sym(X), where X is a set with n elements. If G is commu
tative and acts transitively on X, show that each element g =1 of G moves every element
of X. Deduce that (G : 1) _n.
(b) For each m_1, ﬁnd a commutative subgroup of S
3n
of order 3
n
.
(c) Show that a commutative subgroup of S
n
has order _3
n
3
.
5. Let H be a normal subgroup of a group G, and let 1 be a subgroup of H. Assume that
every automorphism of H is inner. Prove that G =H N
G
(1).
6. (a) Describe the group with generators . and . and deﬁning relation ...
1
=.
1
.
(b) Describe the group with generators . and . and deﬁning relations ...
1
= .
1
,
...
1
=.
1
.
You may use results proved in class or in the notes, but you should indicate clearly what
you are using.
125
126 C. TWOHOUR EXAMINATION
SOLUTIONS
1. (a) False: in 'a. b[a
2
. b
3
), ab has inﬁnite order.
(b) True, the cycle decompositions are (1357)(246), (123)(4567).
(c) True, use the KrullSchmidt theorem.
(d) False, the group it generates is proper.
(e) Cyclic == commutative == nilpotent == solvable.
2. The number of Sylow 11subgroups s
11
=1. 12. . . . and divides 10. Hence there is only
one Sylow 11subgroup 1. Have
G =1
0
H. 1 =C
11
. H =C
10
or D
5
.
Now have to look at the maps 0 : H ÷Aut(C
11
) = C
10
. Yes, by the SchurZassenhaus
lemma.
3. Suppose G has class > 1. Then G has quotient H of class 2. Consider
1 ÷7(H) ÷H ÷H¡7(H) ÷1.
Then H is commutative by (4.17), which is a contradiction. Therefore G is commutative,
and hence cyclic.
Alternatively, by induction, which shows that G¡7(G) is cyclic.
No! In fact, it’s not even true for solvable groups (e.g., S
3
).
4. (a) If g. =., then gh. =hg. =h.. Hence g ﬁxes every element of X, and so g =1.
Fix an . ÷ X; then g ÷g. : G ÷X is injective. [Note that Cayley’s theorem gives an
embedding G ·÷S
n
, n =(G : 1).]
(b) Partition the set into subsets of order 3, and let G =G
1
× ×G
n
.
(c) Let O
1
. . . . . O
i
be the orbits of G, and let G
i
be the image of G in Sym(O
i
). Then
G ·÷G
1
× ×G
i
, and so (by induction),
(G : 1) _(G
1
: 1) (G
i
: 1) _3
n
1
3
3
nr
3
=3
n
3
.
5. Let g ÷ G, and let h ÷ H be such that conjugation by h on H agrees with conjugation
by g. Then g1g
1
=h1h
1
, and so h
1
g ÷ N
G
(1).
6. (a) It’s the group .
G ='.) '.) =C
o
0
C
o
with 0: C
o
÷Aut(C
o
) =˙1. Alternatively, the elements can be written uniquely in the
form .
i
.
}
, i. ; ÷ Z, and .. =.
1
..
(b) It’s the quaternion group. From the two relations get
.. =.
1
.. .. =..
1
and so .
2
=.
2
. The second relation implies
..
2
.
1
=.
2
. =.
2
.
and so .
4
=1.
Alternatively, the ToddCoxeter algorithm shows that it is the subgroup of S
S
generated
by (1287)(3465) and (1584)(2673).
Bibliography
ALPERIN, J. L. AND BELL, R. B. 1995. Groups and representations, volume 162 of
Graduate Texts in Mathematics. SpringerVerlag, New York.
ARTIN, M. 1991. Algebra. Prentice Hall Inc., Englewood Cliffs, NJ.
ASCHBACHER, M. AND SMITH, S. D. 2004. The classiﬁcation of quasithin groups. I, II,
volume 111, 112 of Mathematical Surveys and Monographs. American Mathematical
Society, Providence, RI. Structure of strongly quasithin 1groups.
BESCHE, H. U., EICK, B., AND O’BRIEN, E. A. 2001. The groups of order at most 2000.
Electron. Res. Announc. Amer. Math. Soc. 7:1–4 (electronic).
BRAUER, R. AND FOWLER, K. A. 1955. On groups of even order. Ann. of Math. (2)
62:565–583.
BURNSIDE, W. 1897. Theory of groups of ﬁnite order. Cambridge: at the University Press,
Cambridge.
CURTIS, C. W. 1999. Pioneers of representation theory: Frobenius, Burnside, Schur, and
Brauer, volume 15 of History of Mathematics. American Mathematical Society, Provi
dence, RI.
FEIT, W. 1995. On the work of Eﬁm Zelmanov. In Proceedings of the International
Congress of Mathematicians, Vol. 1, 2 (Z¨ urich, 1994), pp. 17–24, Basel. Birkh¨ auser.
FEIT, W. AND THOMPSON, J. G. 1963. Solvability of groups of odd order. Paciﬁc J. Math.
13:775–1029.
GLAUBERMAN, G. 1999. A new look at the FeitThompson odd order theorem. Mat.
Contemp. 16:73–92. 15th School of Algebra (Portuguese) (Canela, 1998). Available at
www.mat.unb.br/
~
matcont/16_5.ps.
HALL, JR., M. 1959. The theory of groups. The Macmillan Co., New York, N.Y.
HUMPHREYS, J. E. 1990. Reﬂection groups and Coxeter groups, volume 29 of Cambridge
Studies in Advanced Mathematics. Cambridge University Press, Cambridge.
MASSEY, W. S. 1967. Algebraic topology: An introduction. Harcourt, Brace & World,
Inc., New York.
PYBER, L. 1993. Enumerating ﬁnite groups of given order. Ann. of Math. (2) 137:203–220.
127
128 BIBLIOGRAPHY
RONAN, M. 2006. Symmetry and the monster. One of the greatest quests of mathematics.
Oxford University Press, Oxford.
ROTMAN, J. J. 1995. An introduction to the theory of groups, volume 148 of Graduate
Texts in Mathematics. SpringerVerlag, New York, fourth edition.
SCHUPP, P. E. 1987. A characterization of inner automorphisms. Proc. Amer. Math. Soc.
101:226–228.
SERRE, J.P. 1980. Trees. SpringerVerlag, Berlin. Translated from the French by John
Stillwell.
SERRE, J.P. 2003. On a theorem of Jordan. Bull. Amer. Math. Soc. (N.S.) 40:429–440.
SOLOMON, R. 2001. A brief history of the classiﬁcation of the ﬁnite simple groups. Bull.
Amer. Math. Soc. (N.S.) 38:315–352.
SYLOW, M. L. 1872. Th´ eor` emes sur les groupes de substitutions. Math. Ann. 5:584–594.
WILD, M. 2005. The groups of order sixteen made easy. Amer. Math. Monthly 112:20–31.
Index
action
doubly transitive, 55
effective, 53
faithful, 53
free, 55
imprimitive, 68
kfold transitive, 55
left, 53
linear, 96
primitive, 68
right, 54
transitive, 55
trivial, 53
algebra
division, 101
group, 99
quaternion, 101
semisimple, 105
simple, 101
algorithm
ToddCoxeter, 38, 66
automorphism
inner, 41
of a bilinear form, 10
of a group, 41
outer, 41
basis
for a commutative group, 25
block, 69
centralizer
of element, 56
of subgroup, 113
centralizer, of a subalgebra, 102
centre
of a group, 12
class
nilpotency, 87
class equation
class, 58
commutator, 24, 35
composition factors, 83
conjugacy class, 54
coset
left, 16
right, 16
Coxeter system, 38
cycle, 62
dimension, 102
disjoint cycles
disjoint, 62
element
neutral, 8
elementary divisors, 26
equivariant map, 54
exact sequence, 48
exponent
of a group, 30, 37
extension
central, 48
isomorphic, 48
split, 49
extension, of groups, 48
faithful representation, 95
ﬂag
full, 76
Frattini’s argument, 89
Gmap, 54
Gset, 53
generates, 13
generators
of a group, 35
group, 7
4, 14
·, 90
abelian, 9
additive, 7
alternating, 62
Burnside, 37
commutative, 9
complete, 42
Coxeter, 38
129
130 INDEX
cyclic, 13
dihedral, 13
ﬁnite reﬂection, 39
ﬁnitely presented, 37
free, 33
free abelian, 35
general linear, 10
indecomposable, 92
isotropy, 55
metabelian, 87
metacyclic, 115
multiplicative, 7
nilpotent, 87
of rigid motions, 54
of symmetries, 9
orthogonal, 10
], 8
permutation, 9
primitive, 68
quaternion, 14
generalized, 35
quotient, 21
reﬂection, 39
simple, 19
soluble, 83
solvable, 66, 83
special linear, 20
symplectic, 10
with operators, 90
group.
factor, 21
groups
of order 12, 79
of order 2
n
]
n
, m_3., 80
of order 2], 58
of order 30, 78
of order 60, 80
of order 99, 77
of order ], 17
of order ]
2
, 59
of order ]
3
, 79
of order ]q, 78
of small order, 14
homogeneous, 55
homomorphism
admissible, 90
of groups, 15
index
of a subgroup, 17
invariant factors, 26
inverse, 8
inversion, of a permutation, 61
isomorphism
of Gsets, 54
of groups, 8, 15
kernel, 20
Klein Viergruppe, 14
length
of a subnormal series, 81
solvable, 86
length of a cycle, 62
map, of Gsets, 54
module
semisimple, 95
simple, 95
monoid
free, 31
negative, 8
normalizer
of a subgroup, 56
opposite algebra, 95
orbit, 54
order
of a group, 8
of an element, 9
partition
of a natural number, 64
stabilized, 68
permutation
even, 61
odd, 61
presentation
of a group, 35
problem
Burnside, 37
restricted Burnside, 37
word, 37
product
direct, 9, 23
knit, 48
semidirect, 44
ZappaSz´ ep, 48
projector, 98
quotients, of a normal series
of a normal series, 81
rank
of a commutative group, 26
of a Coxeter system, 38
of a free group, 34
Index 131
reduced form
of a word, 32
reﬂection, 39
relations, 35
deﬁning, 35
representation
linear, 96
matrix, 95
series
admissible subnormal, 91
ascending central, 87
composition, 81
derived, 86
normal, 81
solvable, 83
subnormal, 81
without repetitions, 81
signature, 61
skew ﬁeld, 101
stabilizer
of a subset, 56
of an element, 55
stable subset
stable, 54
subgroup, 12
·invariant, 90
admissible, 90
characteristic, 43
commutator, 85
ﬁrst derived, 85
generated by a set, 13
normal, 18
normal generated by a set, 20
second derived, 86
Sylow ], 73
torsion, 10
subset
normal, 20
support
of a cycle, 62
table
multiplication, 11
theorem
Cauchy, 58
Cayley, 16
centre of a ]group, 59
commutative groups have bases, 25
correspondence, 22
double centralizer, 103
factorization of homomorphisms, 22
FeitThompson, 84
Galois, 65
isomorphism, 22
JordanH¨ older, 82
KrullSchmidt, 92
Lagrange, 17
Maschke, 97
NielsenSchreier, 34
nilpotency condition, 89
primitivity condition, 69
SchurZassenhaus, 49
structure of commutative groups, 26
structure of Coxeter groups, 39
Sylow I, 73
Sylow II, 75
Sylow subgroups of subgroups, 77
transposition, 14
word, 31
reduced, 32
words
equivalent, 33
The ﬁrst version of these notes was written for a ﬁrstyear graduate algebra course. As in most such courses, the notes concentrated on abstract groups and, in particular, on ﬁnite groups. However, it is not as abstract groups that most mathematicians encounter groups, but rather as algebraic groups, topological groups, or Lie groups, and it is not just the groups themselves that are of interest, but also their linear representations. It is my intention (one day) to expand the notes to take account of this, and to produce a volume that, while still modest in size (c200 pages), will provide a more comprehensive introduction to group theory for beginning graduate students in mathematics, physics, and related ﬁelds.
BibTeX information @misc{milneGT, author={Milne, James S.}, title={Group Theory (v3.10)}, year={2010}, note={Available at www.jmilne.org/math/}, pages={131} }
Please send comments and corrections to me at the address on my website www.jmilne. org/math/. v2.01 (August 21, 1996). First version on the web; 57 pages. v2.11 (August 29,2003). Fixed many minor errors; numbering unchanged; 85 pages. v3.00 (September 1, 2007). Revised and expanded; 121 pages. v3.01 (May 17, 2008). Minor ﬁxes and changes; 124 pages. v3.02 (September 21, 2009). Minor ﬁxes; changed TeX styles; 127 pages. v3.10 (September 24, 2010). Many minor improvements; 131 pages.
The multiplication table of S3 on the front page was produced by Group Explorer.
Copyright c 1996, 2002, 2003, 2007, 2008, 2010 J.S. Milne. Single paper copies for noncommercial personal use may be made without explicit permission from the copyright holder.
Contents
Contents 1 Basic Deﬁnitions and Results Deﬁnitions and examples . . . . . . . Multiplication tables . . . . . . . . . Subgroups . . . . . . . . . . . . . . . Groups of small order . . . . . . . . . Homomorphisms . . . . . . . . . . . Cosets . . . . . . . . . . . . . . . . . Normal subgroups . . . . . . . . . . . Kernels and quotients . . . . . . . . . Theorems concerning homomorphisms Direct products . . . . . . . . . . . . Commutative groups . . . . . . . . . The order of ab . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . 3 7 7 11 12 14 15 16 18 20 21 23 25 29 29 31 31 32 35 37 38 40 41 41 43 44 48 50 51 53 53 60 66 68 69 73 3
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
2 Free Groups and Presentations; Coxeter Groups Free monoids . . . . . . . . . . . . . . . . . . . . Free groups . . . . . . . . . . . . . . . . . . . . . Generators and relations . . . . . . . . . . . . . . Finitely presented groups . . . . . . . . . . . . . . Coxeter groups . . . . . . . . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . 3 Automorphisms and Extensions Automorphisms of groups . . . . Characteristic subgroups . . . . . Semidirect products . . . . . . . . Extensions of groups . . . . . . . The H¨ lder program. . . . . . . . o Exercises . . . . . . . . . . . . . 4 Groups Acting on Sets Deﬁnition and examples . . Permutation groups . . . . . The ToddCoxeter algorithm. Primitive actions. . . . . . . Exercises . . . . . . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
5 The Sylow Theorems; Applications
. . . . . . . . . . . . . . . . . . . . . . The character table of a group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The group algebra. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Representations of Finite Groups Matrix representations . . . . . . . . . . . . . . . 73 77 77 80 81 81 83 87 90 92 93 95 95 96 96 97 99 100 101 105 107 108 111 111 111 113 117 125 127 129 6 Subnormal Series. . . . . . . . . . . . . . . . . . . Groups with operators . . . . . Semisimple F algebras and their modules The representations of G . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Linear representations . . Maschke’s theorem . . . . . . . . . . . . . . . . . . . . . . 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .The Sylow theorems . . . . . . . . . . . . . . . . . Solvable and Nilpotent Groups Subnormal Series. . . . KrullSchmidt theorem . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nilpotent groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Examples . . . . . . . . . . . . . . Solvable groups . . . . . . . Simple F algebras and their modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . Roots of 1 in ﬁelds . . . . A Additional Exercises B Solutions to the Exercises C TwoHour Examination Bibliography Index . . . . . . . . . . . . . . . . semisimplicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The characters of G . . . . . . . . . . . . . . . . . . . . . . . . . Alternative approach to the Sylow theorems Examples . . . . . . . . . . . . . . . . Semisimple modules . . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Given an equivalence relation. Q is the ﬁeld of rational numbers. emphasizing computational group theory. denoted . I have beneﬁted from the posts to mathoverﬂow by Richard Borcherds.i/ j i 2 Zg is a set with three elements whereas .e. then ff . Richard Stanley. a family of elements of A indexed by I . Also.net/questions/9990/). C is the ﬁeld of complex numbers. X Y X is a subset of Y (not necessarily proper). I also recommend N. Demetres Christoﬁdes. Noah Snyder. 2. p D 2. Œ denotes the equivalence class containing . n 2 mZ. The Sage page http://www.f . Benoˆt Claudon.. The empty set is denoted by .colostate. X and Y are canonically isomorphic (or there is a given or unique isomorphism). if f is the function Z ! Z=3Z sending an integer to its equivalence class.edu/~hulpke/CGT/education. The identity element of a ring is required to act as 1 on a module over the ring. 7. 3. Dave Simpson.math. For integers m and n. i. Robin Chapman.sagemath.org/ provides a front end for GAP and other programs.02) described how to use Maple for computations in group theory.ai /i 2I .1 Rings are required to have an identity element 1. We use the standard (Bourbaki) notations: N D f0. Qiaochu Yuan. Fp D Z=pZ for p a prime number.N OTATIONS . Keith Conrad. P REREQUISITES An undergraduate “abstract algebra” course. Mark Meckes. 5 . 1. Fq is a ﬁnite ﬁeld with q elements where q is a power of a prime number. Throughout the notes. : : :. C OMPUTER ALGEBRA PROGRAMS GAP is an open source computer algebra program. In particular. For example. : : :g. 11. R is the ﬁeld of real numbers. Leonid Positselski. I recommend going to Alexander Hulpke’s page http:// www. Z is the ring of integers. or equals Y by deﬁnition. Earlier versions of these notes (v3. 1000000007. Michiel Vermeulen.. 5. The cardinality of a set S is denoted by jSj (so jS j is the number of elements in S when S is ﬁnite). p is a prime number.html where you will ﬁnd versions of GAP for both Windows and Macs and a guide “Abstract Algebra in GAP”. X DY X Y X 'Y def X is deﬁned to be Y . To get started with GAP. i. Adam Glesser. Efthymios Sofos.i//i 2Z is family with an inﬁnite index set. : : : . Victor Petrov. 1A family should be distinguished from a set. Let I and A be sets. ı Sylvan Jacques. X is isomorphic to Y .e.sourceforge. An element a of a ring is a unit if it has an inverse (element b such that ab D 1 D ba). is a function i 7! ai W I ! A. Carter’s “Group Explorer” http://groupexplorer. and others (a reference mo9990 means http://mathoverflow.. mjn means that m divides n. and homomorphisms of rings are required to take 1 to 1. net for exploring the structure of groups of small order. ACKNOWLEDGEMENTS I thank the following for providing corrections and comments for earlier versions of these notes: Dustin Clausen. Robert Thompson. Steve Dalton. Martin Klazar.
To him are due the ﬁrst attempts at classiﬁcation with a view to forming a theory from a number of isolated facts. . Galois indicated how far reaching the applications of the theory might be. the theory has advanced continuously. Burnside. W. and Camille Jordan in the preface to his Trait´ . R. This was followed in 1870 by M. and thereby contributed greatly. in the Trait´ . it was unclear until quite late in the classiﬁcation whether there would be a ﬁnite or inﬁnite number of sporadic groups. in 1870 ﬂagged Galois’ distinction between groupes simples e and groupes compos´ es as the most important dichotomy in the theory of permutation e groups. as apart from its applications. it had a double cover. which turned out to be the monster. Why are the ﬁnite simple groups classiﬁable? It is unlikely that there is any easy reason why a classiﬁcation is possible.The theory of groups of ﬁnite order may be said to date from the time of Cauchy. mainly by French mathematicians. by shewing that to every equation of ﬁnite degree there corresponds a group of ﬁnite order on which all the properties of the equation depend. during the middle part of the [nineteenth] century. One problem. Jordan’s treatise is devoted to a devele ´ opement of the ideas of Galois and to their application to the theory of equations. Ludwig Sylow published his famous theorems on subgroups of prime power order. 1897. each of which has a double cover that is a centralizer of an involution in the next one. The monster happens to have no double cover so the process stopped there. Because of this problem (among others). is that every simple group has to be tested to see if it leads to new simple groups containing it in the centralizer of an involution. if indirectly. Many additions were made.” which was published e e in 1866. 2001. Bull. Finally. Moreover. which was a potential centralizer of an involution in a larger simple group. but more especially in recent years. and the corresponding division of groups into simple and composite. Solomon. when the baby monster was discovered. . mo38161. No considerable progress in the theory.. to its subsequent developement. Galois introduced the concept of a normal subgroup in 1832. Galois introduced into the theory the exceedingly important idea of a [normal] subgroup. was made till the appearance in 1872 of Herr Sylow’s memoir “Th´ or` mes sur les groupes de e e substitutions” in the ﬁfth volume of the Mathematische Annalen. in 1872. Soc. unless someone comes up with a completely new way to classify groups. . Jordan’s “Trait´ des substitutions et des e equations alg´ briques. Richard Borcherds.” The greater part of M. Math. Amer. Since the date of this memoir. Serret’s “Cours d’Alg` bre Sup´ rieure. Theory of Groups of Finite Order. Moreover. The ﬁrst connected exposition of the theory was given in the third edition of M. but without checking every ﬁnite simple group there seems no obvious reason why one cannot have an inﬁnite chain of larger and larger sporadic groups. Jordan began building a database of ﬁnite simple e groups — the alternating groups of degree at least 5 and most of the classical projective linear groups over ﬁelds of prime cardinality. For example. at least with the current methods of classiﬁcation via centralizers of involutions.
Richard Borcherds. a huge and extraordinary mathematical object. b. the group is said to be multiplicative. (1) G!G 7 . 1. Deﬁnitions and examples D EFINITION 1. . Yet somehow hidden behind these axioms is the monster simple group. / to G.2 In the following. b/ 7! a bW G satisfying the following conditions: G1: (associativity) for all a. which appears to rely on numerous bizarre coincidences to exist. The axioms for groups give no obvious hint that anything like this exists. b. . it is said to be additive. there exists an a0 2 G such that a a0 D e D a0 a: We usually abbreviate .b c/I G2: (existence of a neutral element) there exists an element e 2 G such that a eDaDe a for all a 2 G. G3: (existence of inverses) for each a 2 G. c 2 G.C HAPTER 1 Basic Deﬁnitions and Results The axioms for a group are short and natural. we write a C b for a b and 0 for e. In the ﬁrst case.a. . : : : are elements of a group G.a b/ c D a . Also. in Mathematicians 2009. alternatively. .G. and in the second. a. we usually write ab for a b and 1 for e.1 A group is a set G together with a binary operation .
the inverse of a product is the product of the inverses in the reverse order.e. then b D b e D b . in particular. For an element a of a group G. We call it the inverse of a.aj C1 an / D .ai C1 ai /. deﬁne 8 n > 0 . it has a left inverse. an of elements of G is unambiguously deﬁned. Two groups . we might end up with . . the result is the same. a2 . G $ G 0 . Conversely.n copies of a/ < aa a n nD0 a D e : a 1 a 1 a 1 n < 0 (jnj copies of a 1 ) . In fact. If i ¤ j . a3 of elements of G is unambiguously deﬁned: whether we form a1 a2 ﬁrst and then ..a2 a3 /. (G1) implies that the product of any ordered ntuple a1 . if G is ﬁnite. . a2 . (c) Note that (G1) shows that the product of any ordered triple a1 .ai C1 aj /.ai C1 an / (2) as the ﬁnal product.aj C1 aj / .aj C1 an /: (3) Note that the expression within each pair of parentheses is well deﬁned because of the induction hypotheses.a1 . whereas in another we might end up with . e is the unique element of G such that e e D e (apply G3). ab D ac H) b D c. (e) (G3) implies that the cancellation laws hold in groups. (2) equals (3).a1 ai /. . 0 / are isomorphic if there exists a onetoone correspondence a $ a0 .a1 ai / . and denote it a 1 (or the negative of a. if i D j . / and . ba D ca H) b D c (multiply on left or right by a 1 ).a c/ D . If e 0 is a second such element.G. and hence (by counting) bijective. such that . similarly. (d) The inverse of a1 a2 an is an 1 an 11 a1 1 .b a/ c D e c D c: Hence the element a0 in (G3) is uniquely determined by a. or a2 a3 ﬁrst and then a1 . (b) If b a D e and a c D e. then e 0 D e e 0 D e. BASIC D EFINITIONS AND R ESULTS (a) An element e satisfying (1) is called a neutral element. and so a has a right inverse. Then . then the cancellation laws imply (G3): the map x 7! axW G ! G is injective. In fact. We prove this by induction on n. i. The order jGj of a group G is its cardinality. A ﬁnite group whose order is a power of a prime p is called a pgroup. Thus. e is in the image. and the argument in (b) above shows that the two inverses are equal.ai C1 aj /. and denote it a). b 2 G.8 1.a1 aj /.G 0 ..a1 an / D .aj C1 an / an / and the expressions on the right are equal because of (G1).a b/0 D a0 0 b 0 for all a.a1 a2 /a3 .a1 ai /. In one multiplication. we may suppose i < j .
m C n/a.gg 0 . When m D 0. Usually. let Cm be the group .˛ ı . Therefore Sym. 1. C/. Let S be a set and let Sym. (5) and so associativity holds. With this notation.s/// D . but I prefer to use descriptive names where possible. C/. n 2 Z: 9 (4) is an ideal in Z. called the group of symmetries of S. The identity map s 7! s is an identity element for Sym.na/ D mna: group” is more common than “commutative group”.S / be the set of bijections ˛W S ! S . and multiplication is deﬁned by . In this case. an ¤ e unless n D 0.S / to be bijections. the permutation group on n letters Sn is deﬁned to be the group of symmetries of the set f1.g. and an D e ” mjn: E XAMPLES 1. 2 Sym.6 A group G is commutative (or abelian)1 if ab D ba. all a.g 0 . :::. ng — it has order nŠ. for example. 1. .3 Let C1 be the group .S / is a group.5 When G and H are groups. 1 “Abelian m. for an integer m 1. it is the smallest integer m > 0 such that am D e. h0 / D . and a is said to have inﬁnite order. it is the cartesian product of G and H .ˇ ı // . .s/ D . It follows from (4) that the set fn 2 Z j an D eg .Z.ˇ.˛ ı ˇ/ ı / .˛ ı ˇ/.s// D ˛. b 2 G: In a commutative group. .S /. and so equals mZ for some integer m 0. m. and inverses exist because we required the elements of Sym. called the (direct) product of G and H . For example. When G is commutative. a 1 D am 1 . We deﬁne the product of two elements of Sym. Equation (4) becomes: ma C na D . all m. As a set.S / and s 2 S. we can construct a new group G H .Deﬁnitions and examples The usual rules hold: am an D amCn . When m ¤ 0.. the empty product is e.S / to be their composite: ˛ˇ D ˛ ı ˇ: For any ˛. hh0 /: 1. h/. .a C b/ D ma C mb for m 2 Z and a. we write commutative groups additively.am /n D amn . and.Z=mZ. and a is said to have ﬁnite order m. the product of any ﬁnite (not necessarily ordered) family S of elements is well deﬁned. ˇ.s/.4 Permutation groups. b 2 G.
e. / is called the orthogonal group of . : : : .j Än be the matrix of .w. : : : .e. i. Aut. . Note that if V has dimension n. en be a basis for V .v.V / called the general linear group of V . w 2 V: The automorphisms of form a group Aut. Aut. an / AT P A @ : A D . An automorphism of such a is an isomorphism ˛W V ! V such that . v/ all v. the F linear automorphisms of V form a group GL.w. an / P @ : A : : : : 1 : : : an bn bn a1 ! a1 ! If A is the matrix of ˛ with respect to the basis. The n n matrices with coefﬁcients in F and nonzero determinant form a group GLn .10 and so the map 1. : : : . In this case. the elements of ﬁnite order form a subgroup Gtors of G.v. called the torsion subgroup.ei . there exists a basis for V for which the matrix of is Â Ã 0 Im J2m D . 1. w 2 V.8 Let V be a ﬁnite dimensional vector space over a ﬁeld F . then the choice of a basis determines an isomorphism GL. w 2 V. Im 0 2 When we use the basis to identify V with F n . . a/ 7! maW Z G ! G makes A into a Zmodule. and let P D .m.V / ! GLn . . / with the group of invertible matrices A such that2 AT P A D P . BASIC D EFINITIONS AND R ESULTS . : : : .. ej //1Äi. In a commutative group G. / is called the symplectic group of . : : : . @ : A 7! . Let e1 . then ˛ corresponds to the map (6) becomes the statement that 0 b1 bn an : : : 7! A an : : : :Therefore. .7 Let F be a ﬁeld. ˛w/ D . 1 0 b1 bn 1 a1 ! an 0 b1 bn 1 . w/ D . w/ D . an / P @ : A for all : : : : On examining this statement on the standard basis vectors for F n.a1 . (7) When is symmetric. v/ all v.a .v.˛v. 1. i.a1 .F / sending an automorphism to its matrix with respect to the basis. The choice of the basis identiﬁes Aut. 2m D n. A bilinear form on V is a mapping W V V ! F that is linear in each variable. When is skewsymmetric.F / called the general linear group of degree n. (6) and nondegenerate. @ : A 2 F n: : : we see that it is equivalent to (7). /. the pairing becomes 0 1 0 1 a1 ! b1 b1 : . and nondegenerate.. w/ for all v. For a ﬁnite dimensional F vector space V .
and apply (G30 ) to ﬁnd a0 and a00 such that a0 a D e and a00 a0 D e. these conditions obviously imply those in (a).a a0 / D a00 whence (G3). They are minimal in the sense that there exist sets with a binary operation satisfying any two of them but not the third.a a0 / D . / is called a semigroup. For example.2(c)).9 A set S together with a binary operation . t A/ D A D . Q Q Q . be? Clearly. (G30 ) for each a 2 G. As there is at most one neutral element in a set with an associative binary operation. Multiplication tables A binary operation on a ﬁnite set can be described by its multiplication table: e a b c : : : e ee ae be ce : : : a ea a2 ba ca : : : b eb ab b2 cb : : : c ec ac bc c2 : : : ::: ::: ::: ::: ::: . 11 R EMARK 1.S. . Q What should . B/ D . To see that these imply (G2) and (G3). In a monoid. and for any pair A and B of such sequences. (b) A group can be deﬁned to be a set G with a binary operation satisfying the following conditions: (g1) is associative. (g2) G is nonempty. The product Q def A D a1 an of any sequence A D . there exists an a0 2 G such that a0 a is neutral./ . be the empty sequence. . .. C/ satisﬁes (g1) and (g2) but not (g3). let a 2 G. and (8) holds. A/ D . (8) Let .a0 a/ a0 D a00 a0 D e. we should have Q Q Q Q Q Q Q . When the binary operation is associative. . the set of 2 2 matrices with coefﬁcents in a ﬁeld and with A B D AB BA satisﬁes (g2) and (g3) but not (g1).ai /1Äi Än of elements in a semigroup S is welldeﬁned (see 1. the sequence of elements in S indexed by the empty set.. (g3) for each a 2 G.N.e. should be a neutral element. . .G3) can be replaced by the following weaker conditions (existence of a left neutral element and left inverses): (G20 ) there exists an e such that e a D a for all a. the empty set satisﬁes (g1) and (g3) but not (g2). whence (G2). A SIDE 1.A t B/ . A semigroup with a neutral element is called a monoid.a0 a/ D a e.a./ D .A t .a00 a0 / . there exists an a0 2 G such that a0 a D e. A/ ./ : Q In other words.10 (a) The group conditions (G2.Multiplication tables and the group of invertible matrices A such that AT J2m A D J2m is called the symplectic group Sp2m . b/ 7! a bW S S ! S is called a magma. the product of any ﬁnite (possibly empty) sequence of elements is welldeﬁned. Then a a0 D e .a a0 / a D a . i. A/ . and a D e a D .
C/ .Z. an intersection of subrings of a ring is a subring. (S1) implies that the binary operation on G deﬁnes a binary operation S S ! S on S .N. When S is ﬁnite. E XAMPLE 1. there are 864 D 6277 101 735 386 680 763 835 789 423 207 666 416 102 355 444 464 034 512 896 binary operations on a set with 8 elements.14 It is generally true that an intersection of subobjects of an algebraic object is a subobject. For the multiplication table of S3 .11 Let S be a nonempty subset of a group G. By assumption S contains at least one element a. and if we allow each position to hold any of the n 2 elements. this is not practical! The table has n2 positions.12 1. : : :g S. Inverses exist if and only if each element occurs exactly once in each row and in each column (see 1.15 For any subset X of a group G. namely.and S2: a 2 S H) a 1 2 S.G/ D fg 2 G j gx D xg for all x 2 Gg: It is a subgroup of G. For example. Note that each colour occurs exactly once in each row and and each column. then verifying the associativity law requires checking n3 equalities. now a 1 D an 1 2 S . If S1: a. a2 . then fa. its inverse a 1 . there is a smallest subgroup of G containing X. It consists of all ﬁnite products of elements of X and their inverses (repetitions allowed).12 The centre of a group G is the subset Z. P ROPOSITION 1. see the front page. and the product e D aa 1 . and (S1) and (S2) obviously hold. The example . list all possible multiplication tables and check the axioms. P ROOF. an intersection of submodules of a module is a submodule. then the binary operation on G makes S into a group. This suggests an algorithm for ﬁnding all groups of a given ﬁnite order n. If there are n elements. Subgroups P ROPOSITION 1. and so on. 2 A nonempty subset S satisfying (S1) and (S2) is called a subgroup of G. 2 R EMARK 1. b 2 S H) ab 2 S. Except for very small n. . BASIC D EFINITIONS AND R ESULTS The element e is an identity element if and only if the ﬁrst row and column of the table simply repeat the elements. It is nonempty because it contains e. say an D e.13 An intersection of subgroups of G is a subgroup of G: P ROOF. For example.2e). Finally (S2) shows that the inverses of elements in S lie in S. and so a has ﬁnite order. which is automatically associative. but only ﬁve isomorphism classes of groups of order 8 (see 4. condition (S1) implies (S2): let a 2 S . P ROPOSITION 1. C/ shows that (S1) does not imply (S2) when S is inﬁnite.21). then that gives a total of nn possible tables very few of which deﬁne groups.
then G D f: : : .Subgroups 13 P ROOF.i D feg. : : : .e. and it is evidently the smallest such group. For example. If r has inﬁnite order.. e. ri $ i mod n.e.16 The cyclic groups. 1. 2 The subgroup S given by the proposition is denoted hX i. A group is said to be cyclic if it is generated by a single element. But the set of such products satisﬁes (S1) and (S2) and hence is a subgroup containing X. Dn is the group of symmetries of a regular polygon with nsides. r i $ i: Thus. 4 More formally. : : :g C1 .4 Number the vertices 1. Let r be the rotation through 2 =n about the centre of polygon (so i 7! i C 1 mod n/. r 2 . if G D hri for some r 2 G. Then all the statements concerning Dn can proved without appealing to geometry (or illuminating the reader). n in the counterclockwise direction. Note that the order of an element a of a group is the order of the subgroup hai it generates. then the set of all ﬁnite products of elements of X is already a group and so equals hX i. s 2 4 r D1!2!3!1 . up to isomorphism. we shall loosely use Cn to denote any cyclic group of order n (not necessarily Z=nZ or Z). For example. for example. If every element of X has ﬁnite order. all ﬁnite products of elements of X and their inverses. r. In future. the pictures s 1 r sD 2 3 1$1 2$3 r 1 8 < 1$1 sD 2$4 : 3$3 r D1!2!3!4!1 3 3 This group is denoted D or D n 2n depending on whether the author is viewing it concretely (as the symmetries of an npolygon) or abstractly. It therefore equals S . if every element of G can be written as a ﬁnite product of elements from X and their inverses. r i . r. i. then G D fe. E XAMPLES 1. h. and is called the subgroup generated by X.:::. If r has ﬁnite order n. and G can be thought of as the group of rotational symmetries about the centre of a regular polygon with nsides. if G is ﬁnite. r i .17 The dihedral groups Dn . The intersection S of all subgroups of G containing X is again a subgroup containing X . We say that X generates G if G D hXi. : : : . there is exactly one cyclic group of order n for each n Ä 1. D can be deﬁned to be the subgroup of S generated by rW i 7! i C 1 (mod n/ and n n sW i 7! n C 2 i (mod n). r n 1 g Cn .3 For n 3. and let s be the reﬂection in the line (= rotation about the line) through the vertex 1 and the centre of the polygon (so i 7! n C 2 i mod n). :::. Clearly S contains with X .. i.r 1 .
rs.14 1. In the following table. Hence Dn D hs. a group C6 . 3g. j 7! b extends uniquely to a homomorphism H ! M2 . 51.26) below for a more precise statement). a2 b. a2 . It is not difﬁcult to see that Sn is generated by transpositions (see (4. ij D k D j i: The map i 7! a. which maps the group hi.28 below). Let t be the reﬂection in the line through the midpoint of the side joining the vertices 1 and 2 and the centre of the polygon (so i 7! n C 3 i mod n/. p0 p 1 1 0 1 Á and b D 01 10 . Groups of small order [For] n D 6. Recall that H D R1 ˚ Ri ˚ Rj ˚ Rk with the multiplication determined by i 2 D 1 D j 2. . s. r n 1 sg.t s/n D e D . t 2 D e. BASIC D EFINITIONS AND R ESULTS illustrate the groups D3 and D4 . In the general case r n D eI These equalites imply that Dn D fe. :::. rg and D2 to be C2 C2 D f1. 1. c C n D t means that there are c commutative groups and n noncommutative groups (up to isomorphism. namely Cp (see 1.C/ of Ralgebras. Cayley. a2 D b 2 . Math. and two groups C2 C3 and S3 . r n 1 s 2 D eI srs D r 1 (so sr D r n 1 s/: . ˙i. there are three groups. and it is clear from the geometry that the elements of the set are distinct. :::. of course). 1 (1878). The subgroup of GL2 . ng.18 The quaternion group Q: Let a D a4 D e. Note that D3 is the full group of permutations of f1. bi.C/ generated by a and b is Q D fe. more simply. The group D2 is also called the Klein Viergruppe or. b. there is only one group of order p. t i and s 2 D e. the 4group. p. j i isomorphically onto ha. . and so jDn j D 2n. a3 . a. 2. 1. American J. r. ˙j. ˙kg of the quaternion algebra H. For each prime p.19 Recall that Sn is the permutation group on f1. s. ab. r.st /n : We deﬁne D1 to be C2 D f1. Then r D t s. Then bab D a3 (so ba D a3 b). It is the smallest noncommutative group. A transposition is a permutation that interchanges two elements and leaves all other elements unchanged. a3 bg: The group Q can also be described as the subset f˙1. :::. 2. rsg.
529.a2 D ˛. there are exactly 49.n/ Ä n. D9 .C3 C4 C3 / C2 . b j a4 D b 5 D 1. 27 Co.23 4. . 1.a1 .21 4. A4 . 2001).1/ ! 0 as e.am /. C3 C10 . the more high powers of primes divide n. Homomorphisms D EFINITION 1. 4. D4 C2 15 Ref. D11 See opensourcemath. b j a3 D b 7 D 1. An isomorphism is a bijective homomorphism.14 5.18 5.n/ where e. D7 C15 See Wild 2005 C18 .n/ is the number of isomorphism classes of groups of order n.ha. C2 C14 . Q.20 A homomorphism from a group G to a second G 0 is a map ˛W G ! G 0 such that ˛. a complete irredundant list of groups of order Ä 2000 had been found — up to isomorphism.910.1//e.a/˛.14 Here ha. ba D ab 2 i 5. ba D ab 2 i C6 . if f . am of G. C2 S3 .a1 am / D ˛. C2 C9 .ab/ D ˛.C5 C3 . In fact. ba D ab 2 i is the group with generators a and b and relations a4 D b 5 D 1 and ba D ab 2 (see Chapter 2).14 5. D5 C12 .21 Let ˛ be a homomorphism.n/ is the largest exponent of a prime dividing n and o.html C6 .org/gap/small groups. For any elements a1 . b j a4 D b 5 D 1.484 (Besche et al.F / ! F is a homomorphism. C2 C6 . then 2 2 f . S3 C8 .18 4.C2 C22 .n/ ! 1 (see Pyber 1993). ha. For example. the determinant map detW GLn . C3 C20 .a2 D ˛. . Roughly speaking.D10 . b 2 G.14 C21 . the more groups of order n there should be.Homomorphisms jGj 4 6 8 9 10 12 14 15 16 18 20 21 22 24 c Cn D t 2C0 D 2 1C1 D 2 3C2 D 5 2C0 D 2 1C1 D 2 2C3 D 5 1C1 D 2 1C0 D 1 5 C 9 D 14 2C3 D 5 2C3 D 5 1C1 D 2 1C1 D 2 3 C 12 D 15 Groups C4 .a1 /˛.16 5. S3 C10 . : : : . ˛. C2 C3 C2 C2 . C4 C3 C4 .b/ for all a.a1 / am // am / ˛. By 2001.
e/ D ˛. 2 Unfortunately.e.bx/ D abx D .x/ D aL . . we let aS D fas j s 2 S g Sa D fsa j s 2 S g: Because of the associativity law.x/. For x 2 G. Also aa 1 DeDa 1 a H) ˛.e/. List the elements of the group as a1 . and so a homomorphism of commutative groups is also a homomorphism of Zmodules.a 1 /L D id D .a 1 / D e D ˛. and so ˛.16 1. this implies that aL ı . and the sets of the form Ha are called the right cosets of H in G. unless n is small. As eL D id. In particular. this allows one to realize the group as a group of permutations. and so we can denote this set unambiguously by abS: When H is a subgroup of G. deﬁne aL W G ! G to be the map x 7! ax (left multiplication by a).bS / D . 2 C OROLLARY 1.G/.e/ D e (apply 1.aL ı bL /. Cosets For a subset S of a group G and an element a of G.22 (C AYLEY ) There is a canonical injective homomorphism ˛W G ! Sym.ab/S . We shall see later (4.a/ 1 .x// D aL .G/. Hence a 7! aL is a homomorphism G ! Sym.ab/L . BASIC D EFINITIONS AND R ESULTS 1. : : : . .a 1 /˛.a/m : Moreover ˛. a.bL .ee/ D ˛. each row of the multiplication table of a group is a permutation of the elements of the group.a 1 /L ı aL .G/: P ROOF. the sets of the form aH are called the left cosets of H in G. P ROOF.e/˛. Sn is too large to be manageable. (9) and so homomorphisms preserve all products.2a). an . and so aL is a bijection. As Cayley pointed out. i.ab/L D aL ı bL . and so . aH D H if and only if a 2 H .23 A ﬁnite group of order n can be realized as a subgroup of Sn . T HEOREM 1. As we noted above.a 1 / D ˛. and so ˛. It follows that (9) holds for all m 2 Z. and it is injective because of the cancellation law. aL 2 Sym. Because e 2 H .a/˛.am / D ˛. for m ˛..22) that G can often be embedded in a permutation group of much smaller order than nŠ. For a 2 G.a/.
G W H / of them. for example.24 Let G D .g/.G W 1/ is the order of G. up to isomorphism. a prime. . (c) aH D bH if and only if a 1 b 2 H: (d) Any two left cosets have the same number of elements (possibly inﬁnite). and each left coset has . G is cyclic and so G Cp . and so a 1 b 2 H . In particular. there is only one group of order 1. and C D aH and D aH by (a). 000.26 (L AGRANGE ) If G is ﬁnite. 000. then H D a 1 bH .H W 1/: In particular. 5 More formally.27 The order of each element of a ﬁnite group divides the order of the group. 2 C OROLLARY 1. and let H be a subspace of dimension 1 (line through the origin). (a) An element a of G lies in a left coset C of H if and only if C D aH: (b) Two left cosets are either disjoint or equal.28 If G has order p. In fact there are only two groups of order 1. 000. In other words.G W H / is the cardinality of the set faH j a 2 Gg. if aH D bH . As the left cosets of H in G cover G. Apply Lagrange’s theorem to H D hgi. T HEOREM 1.G W 1/ D . But only e has order 1.H W 1/ D order.ba 1 /L W ah 7! bh is a bijection aH ! bH: 2 The index . that. and so aH D aa 1 bH D bH . then every element of G has order 1 or p. and so aH D bhH D bH: C0 (b) If C and C 0 are not disjoint. recalling that . (c) If a 1 b 2 H .5 For example.18). (1. . 2 E XAMPLE 1. (a) Certainly a 2 aH . Conversely.R2 . 007 (because this number is prime).H W 1/ elements. and so G is generated by any element a ¤ e.25 Let H be a subgroup of a group G. then .G W H /. 000. (d) The map . P ROPOSITION 1. 014. 000.Cosets 17 E XAMPLE 1. 000. the order of every subgroup of a ﬁnite group divides the order of the group.G W H / of H in G is deﬁned to be the number of left cosets of H in G. This shows. P ROOF. then a D bh for some h. then they have a common element a. P ROOF.25b) shows that they form a partition G. then H D a 1 bH . 049 (see 4. P ROOF. The left cosets of H in G form a partition of G. the condition “a and b lie in the same left coset” is an equivalence relation on G. C/. . Then the cosets (left or right) of H are the lines a C H parallel to H . there are . Conversely. if a lies in the left coset bH .
G W H / is also the number of right cosets of H in G: But.S T / D . in fact H ' Z. On j multiplying F second equality by gi .aH / 1 is the right coset Ha 1 .2).Q/. Hence . and so we can denote this set unambiguously as RS T .31 For any subgroups H K of G. but has no element of order 4. then G has a subgroup of order p n (Sylow’s theorem 5. note that the 4group C2 C2 has order 4.Ha/ 1 D a 1 H . . More generally.13). we have the following result. BASIC D EFINITIONS AND R ESULTS 1.34 below).30 Lagrange’s theorem has a partial converse: if a prime p divides m D . and let H G. if gNg 1 D N for all g 2 G.29 For a subset S of G.32 To show that N is normal. Write G D i 2I gi H (disjoint union). the next example shows that there can exist a subgroup N of a group G and an element g of G such that gNg 1 N but gNg 1 ¤ N .G W K/ D . . denoted N G G.RS /T . A subgroup N of G is normal. R.18 1. Let g D 5 0 . Then . This shows that . Then 01 Â Ã Â ÃÂ 1 n 5 0 1 1 g g D 0 1 0 1 0 Hence gHg 1 D n 1 ˚ 1n 0 1 ˇ « ˇ n 2 Z . we ﬁnd that gi H D j 2J gi hj K (disjoint union).G W H /. P ROPOSITION 1. but has no subgroup of order 6 (see Exercise 414). the and so G D i.G W H /. we let S T D fst j s 2 S . let S 1 D fg 1 j g 2 S g.j 2I J gi hj K (disjoint union). E XAMPLE 1. if a prime power p n divides m. and rewriting this with g 1 for g gives that N gNg 1 for all g. R EMARK 1. then G has an element of order p (Cauchy’s theorem 4.33 Let G D GL2 .G W 1/. t 2 T g: Because of the associativity law.H W K/ (meaning either both are inﬁnite or both are ﬁnite and equal). However. Then H is a subgroup of Ã Â Ã 0 1 5n D : 1 0 1 ÃÂ 5 0 1 H (and g 1 Hg 6 H ). and H D F2J hj K (disjoint union). a left coset will not be a right coset (see 1. F F P ROOF. 1. because multiplying this inclusion on the left and right with g 1 and g respectively gives the inclusion N g 1 Ng. and A4 has order 12.G W K/ D jI jjJ j D . and . Therefore S 7! S 1 deﬁnes a onetoone correspondence between the set of left cosets and the set of right cosets under which aH $ Ha 1 .H W K/: 2 Normal subgroups When S and T are two subsets of a group G. However. in general. it sufﬁces to check that gNg 1 N for all g.
and so gNg 1 D N . Hence gN D Ng. if the left coset gN is also a right coset.hn/ 1 1.25a). but the converse group Q is not commutative. : : : . E XAMPLE 1. and . if N is normal. r n is not normal because r (c) Every subgroup is false: the quaternion Exercise 11). gN D Ng for all g 2 G: P ROOF. in order for N to be normal. : : : . we must have that for all g 2 G and n 2 N . and so gH D Hg: (b) Consider the dihedral group Dn D fe. 2 1. then it must be the right coset Ng by (1.34 A subgroup N of G is normal if and only if every left coset of N in G is also a right coset. sg D r n 2 s … fe. to say that N is normal amounts to saying that an element of G can be moved past an element of N at the cost of replacing the element of N by another element of N . Clearly. Then G D H t gH (disjoint union).35 The proposition says that. 1g 1 sr 1 .Normal subgroups 19 P ROPOSITION 1. If both H and N are normal. The set HN is nonempty. there exists an n0 such that ng D gn0 ). r. Since . and so HN is a subgroup. but every subgroup is normal (see A group G is said to be simple if it has no normal subgroups other than G and feg. gN D Ng). If H is also normal. s.h1 n1 /. Indeed. P ROOF. For n 3 the subgroup fe.35 Dn 1 h 1 1. Hence gH is the complement of H in G. r n Then Cn D fe. then every left coset is a right coset (in fact.36 (a) Every subgroup of index two is normal. of a commutative group is normal (obviously). Conversely. then HN is a subgroup of G. Such a group can still have lots of nonnormal subgroups — in fact. Hg is the complement of H in G.h2 n2 / D h1 h2 n0 n2 2 HN. there exists an n0 2 N such that gn D n0 g (equivalently.35 D h 1 0 n 2 HN it is also closed under the formation of inverses. then HN is a normal subgroup of G. r. and hence is normal. gNg 1 D N ” gN D Ng: Thus. : : : . let g 2 G H .37 If H and N are subgroups of G and N is normal. sg. In other words. then gHNg 1 D gHg 1 gNg 1 D HN for all g 2 G. the Sylow theorems (Chapter 5) imply that every ﬁnite group has nontrivial subgroups unless it is cyclic of prime order. 2 . Similarly. for all g 2 G and n 2 N . P ROPOSITION 1. in which case. r n 1 sg: has index 2. 1 and so it is closed under multiplication.
P ROOF. Conversely. then gag 1 D .g/ D eg: If ˛ is injective. BASIC D EFINITIONS AND R ESULTS An intersection of normal subgroups of a group is again a normal subgroup (cf. is a homomorphism G ! G. P ROPOSITION 1.a/˛. then ˛ is injective. 1. P ROOF. then ˛. a D x1 xm with each xi or its inverse in X. the kernel of the homomorphism detW GLn . if Ker.gxm g 1 /.14). because ˛.˛/ D feg. P ROPOSITION 1. if N is a normal subgroup of G. It is obviously a subgroup. Kernels and quotients The kernel of a homomorphism ˛W G ! G 0 is Ker. and g 2 G.a/ D e. L EMMA 1. a 7! gag 1 . P ROOF.20 1. 1 2 As X is closed under conjugation.g 0 / H) ˛.F / ! F is the group of n n matrices with determinant 1 — this group SLn . . so that ˛. 1 or its inverse lies in X. each gxi g hX i.g 1 0 g / D e H) g 1 0 g D e H) g D g 0 . Obvious. we obtain the following proposition.g/ 1 D ˛.g/˛. If a 2 hXi.g/ D ˛. The map “conjugation by g”.˛/ D fg 2 Gj ˛.g/˛. then Ker.˛/. and so ghX ig L EMMA 1. then the subgroup hXi generated by it is normal. we can deﬁne the normal subgroup generated by a subset X of a group G to be the intersection of the normal subgroups containing X.42 Every normal subgroup occurs as the kernel of a homomorphism. say. Therefore.˛/.41 The kernel of a homomorphism is a normal subgroup.F / is called the special linear group of degree n.˛/ D feg.39 For any subset X of G.gag Hence gag 1 1 / D ˛. For example. We say that a subset X of a group G is normal or closed under conjugation if gXg 1 X for all g 2 G. S g2G gXg 1 is normal.40 The normal subgroup generated by a subset X of G is h S g2G gXg 1 i. P ROPOSITION 1. the subset smallest normal set containing X. and if a 2 Ker. and it is the 2 On combining these lemmas. then there is a unique group structure on the set G=N of cosets of N in G for which the natural map a 7! ŒaW G ! G=N is a homomorphism.38 If X is normal.gx1 g 1 / . More precisely.g/ 1 D e: 2 2 Ker . Its description in terms of X is a little complicated.
.g/. . the coset N is an identity element.g 0 N // D ˛.n/ D ˛. N E XAMPLE 1.43 The map a 7! aN W G ! G=N has the following universal property: for any homomorphism ˛W G ! G 0 of groups such that ˛. The product is certainly associative. or ﬁrst. P ROPOSITION 1.bN / D a. s g (cyclic group of order 2).g 0 N /.gg 0 N / D ˛. but this can be confused with “direct factor”.gN /˛. . .g/˛. We have to check (a) that this is welldeﬁned. . The quotient group Z=mZ is a cyclic group of order m. .44 (a) Consider the subgroup mZ of Z.gn/ D ˛.s/ j s 2 Sg. (a). N and ˛ is a homomorphism because N ˛. we have to show that abN D a0 b 0 N . second. and (b) that it gives a group structure on the set of cosets.gN / .S / D f˛. (c) For n 2. the quotient Dn =hri D fe. N N Theorems concerning homomorphisms The theorems in this subsection are sometimes called the isomorphism theorems (ﬁrst.N / D feg. ).g/.34 The group G=N is called the6 quotient of G by N .g/˛.b 0 N / D aN b 0 D a0 N b 0 D a0 b 0 N: (b).gN / D ˛. Note that for n 2 N . 1N 2 1. and so ˛ is constant on each left coset gN of N in G. (b) Let L be a line through the origin in R2 . there exists a unique homomorphism G=N ! G 0 making the diagram at right commute: G a 7! aN G=N ˛ G 0: P ROOF.. .34 1. . or . 6 Some authors say “factor” instead of “quotient”. FACTORIZATION OF HOMOMORPHISMS Recall that the image of a map ˛W S ! T is ˛.bN / D . It will then be obvious that the map g 7! gN is a homomorphism with kernel N . N N N N The uniqueness of ˛ follows from the surjectivity of G ! G=N . Propositions 1. Then R2 =L is isomorphic to R (because it is a onedimensional vector space over R).41 and 1. N 2 ˛. Let aN D a0 N and bN D b 0 N . Write the cosets as left cosets.ab/N . third.g 0 / D ˛. But abN D a. . . and deﬁne . .aN /.gg 0 / D ˛. and a is an inverse for aN .42 show that the normal subgroups are exactly the kernels of homomorphisms. It therefore deﬁnes a map ˛W G=N ! G 0 .Theorems concerning homomorphisms 21 P ROOF. ˛.
g/ isomorphism ! I: P ROOF.e.a 1 /. The universal property of quotients (1. then N N N N H . 20). P ROOF. h 7! hN: This is a homomorphism.22 1. and let N D Ker. but it need not be a normal subgroup of G. According to Theorem 1.37) that HN is a subgroup.˛/.G/ is a subgroup of G 0 . H \ N is a normal subgroup of H . then the lattice of subgroups in N captures the structure of the lattice of subgroups of G lying over the kernel of G ! G.g/.a0 /. Consider the map H ! G=N. We have already seen (1. an isomorphism. which is therefore normal in H . and ˛ factors in a natural way into the composite of a surjection. Moreover.H / and a subgroup N of G corresponds to H D ˛ 1 . and the map h.a/ def and b 0 D ˛..45.aa0 / and b 1 D ˛. and its kernel is H \ N .e. and so ˛ has trivial kernel. T HE CORRESPONDENCE THEOREM N The next theorem shows that if G is a quotient group of G. The homomorphism ˛ is certainly N N N surjective. 2 T HE ISOMORPHISM THEOREM T HEOREM 1.47 (C ORRESPONDENCE T HEOREM ) Let ˛W G G be a surjective homomorphism. i. N G N T HEOREM 1.gN / D e. the map induces an isomorphism H=H \ N ! I where I is its image. if H $ H and H 0 $ H 0 .41) that the kernel is a normal subgroup of G. I D HN=N .x/W G ! I deﬁnes a homomorphism ˛W G=N ! I with ˛. This N N implies that it is injective (p.. and if ˛. then g 2 Ker. the image I of ˛ is a subgroup of G 0 . H is contained in the normalizer of N — see later). We have already seen (1.46 (I SOMORPHISM T HEOREM ) Let H be a subgroup of G and N a normal subgroup of G.gN / D ˛.45 (H OMOMORPHISM T HEOREM ) For any homomorphism ˛W G ! G 0 of groups.43) shows that the map x 7! ˛. Then H \ N is still normal in H . then bb 0 D ˛.H /.˛/ D N .H \ N / 7! hN W H=H \ N ! HN=N is an isomorphism. BASIC D EFINITIONS AND R ESULTS T HEOREM 1. the kernel N of ˛ is a normal subgroup of G. Then HN is a subgroup of G. If b D ˛. 2 It is not necessary to assume that N be normal in G as long as hN h 1 D N for all h 2 H (i. But I is the set of cosets of the form hN with h 2 H . Then there is a onetoone correspondence 1W1 N fsubgroups of G containing N g $ fsubgroups of Gg N under which a subgroup H of G containing N corresponds to H D ˛. and an injection: ˛ G ! G0 x ? ?injective ? surjectiveyg7!gN ? G=N gN 7!˛. and so I D ˛.
rsi hN i s hri N hr s i NN hsi hr 2 si hr 2 i hrsi hr 3 si 1 1 Direct products Let G be a group. in which case. 2 ' E XAMPLE 1. Therefore N is normal. P ROOF.H 0 W H /. and ˛˛ the two operations give the required bijection. then ˛. This means that each element g of G can be written uniquely in the form g D h1 h2 hk .H / D H . 2 C OROLLARY 1. H $ H=N . 1 ˛.hk h0 /: k .H / is a subgroup of G (see 1. Recall (1. ˛ induces an N (b) H isomorphism ' N N G=H ! G=H : N N N P ROOF. Clearly.h1 h0 /. and let H1 . The remaining statements are easily veriﬁed. hk / 7! h1 h2 hk W H1 H2 Hk ! G is an isomorphism of groups.G=N /=. : : : .H=N /. N N ˛ N .H / D HN .H 0 W H / D . F For example. The groups G and G=N have the following lattices of subgroups: D4 D4 =hr 2 i hr 2 . in which case the homomorphism g 7! gN W G ! G=N induces an isomorphism G=H ! .17) that srs 1 D r 3 .Direct products 23 N N N N (a) H H 0 ” H H 0 . hi 2 Hi .48 Let N be a normal subgroup of G. h2 .h1 . then 1 2 k gg 0 D . Hk be subgroups of G. then ˛ 1 . We say that G is a direct product of the subgroups Hi if the map . 2 and so sr 2 s 1 D r 3 D r 2 . in which case . and that if g D h1 h2 hk and g 0 D h0 h0 h0 .h2 h0 / 1 2 . which equals H if and only if H 1 . and if H is a subgroup of G.H / is easily seen to be a subgroup of G conN taining N . Therefore. a decomposition H 0 D iF ai H of H 0 into a disjoint union of left cosets of 2I N N N H gives a similar decomposition H 0 D i 2I ai H of H 0 . then there is a onetoone correspondence between the set of subgroups of G containing N and the set of subgroups of G=N . Moreover H is normal in G if and only if H=N is normal in G=N . N is normal in G if and only if H is normal in G. si hri hr 2 . This is the special case of the theorem in which ˛ is g 7! gN W G ! G=N . If H is a subgroup of G.49 Let G D D4 and let N be its subgroup hr 2 i. : : : .45).
and is in H1 because H1 is normal. h2 / 7! h1 h2 is a homomorphism. BASIC D EFINITIONS AND R ESULTS The following propositions give criteria for a group to be a direct product of subgroups. e/. : : : . Conversely.24 1. P ROOF. hk 1/ 7! h1 h2 hk 1 W H1 H2 Hk 1 ! H1 H2 Hk 1 .h1 . (b) H1 \ H2 D feg.h2 h1 / is e. and so we argue by induction on k.b. we shall prove only the sufﬁciency.50 A group G is a direct product of subgroups H1 .H1 Hj 1 Hj C1 Hk / D feg. we have just done this.h1 h2 / . (b) for each j . H2 . and (b) implies that it is injective: h1 h2 D e H) h1 D h2 1 2 H1 \ H2 D feg: Finally. Certainly. (c) implies that . and so the induction hypothesis shows that .e. The necessity of the conditions being obvious. g 1 / maps to e under . : : : .h1 h2 / .h1 . H2 if and only if (a) G D H1 H2 . P ROOF. If G is the direct product of H1 and H2 . : : : . The conditions (a. P ROPOSITION 1.g. H2 . and (c) H1 and H2 are both normal in G. and (c) each of H1 . and (b) holds because. Hj \ . h2 of a group commute if and only if their commutator Œh1 . : : : .h2 h1 / 1 def 2 1 D h1 h2 h1 1 h2 1 D . and so it remains to show that they imply that each element h1 of H1 commutes with each element h2 of H2 . for any g 2 H1 \ H2 . then certainly (a) and (c) hold. Therefore (b) implies h1 and h2 commute.c) hold for the subgroups H1 . h1 h2 h1 1 h2 1 which is in H2 because H2 is normal. But . (b) H1 \ H2 D feg. h2 . An induction argument using (1. and (c) every element of H1 commutes with every element of H2 . P ROOF. Hk is normal in G. 2 P ROPOSITION 1. Hk 1 of H1 Hk 1 . (a) implies that it is surjective. Hk if and only if (a) G D H1 H2 Hk . P ROPOSITION 1. Two elements h1 . the element . H2 if and only if (a) G D H1 H2 . h2 D .h1 h2 h1 1 / h2 1 .51 A group G is a direct product of subgroups H1 . these conditions are implied by those in the previous proposition.h1 .37) shows that H1 Hk 1 is a normal subgroup of G. h2 / 7! h1 h2 and so equals . For k D 2.52 A group G is a direct product of subgroups H1 .
xk g and let c1 . c2 . A subset fx1 .h. The composite of these isomorphisms H1 Hk 1 Hk . Then there exist generators y1 .hk /7!. : : : . : : : .hk /7!hhk !G 2 sends . We argue by induction on s D c1 C C ck . 25 Hk satisﬁes the hypotheses of (1. : : : . we change the signs of both ci and xi .c1 c2 . but. c1 c2 > 0.c1 c2 /x1 C c2 . x3 . and ˘ . : : : . written additively.H1 Hk ! G is also an isomorphism.Commutative groups is an isomorphism. : : : . ck / D 1. for ﬁnite commutative groups. xk consists of the sums mi xi . : : : . This allows us to assume that all ci 2 N. this is similar to the ﬁrst proof of the theorem. Then. h2 . ck be integers such that gcd. c3 . and so we may assume that it requires at least k > 1 Stillwell tells me that. The lemma certainly holds if s D 1.c1 c2 / C c2 C C ck < s.h1 .h1 hk 1 .c1 .hk / ! H1 Hk 1 Hk . yk for M such that y1 D . the statement is trivial. xk i of M generated by the elements x1 .h. Let M be a commutative group. 7 We argue by induction on the number of generators of M . for the sake of completeness. ck / D 1. hk / to h1 h2 hk : Commutative groups The classiﬁcation of ﬁnitely generated commutative groups is most naturally studied as part of the theory of modules over a principal ideal domain. : : : . P ROOF. say. by induction. hence it is a ﬁnite direct sum of cyclic groups. yk for M such that y1 D c1 x1 C C ck xk . xk g of M is a basis for M if it generates M and m1 x1 C then M D hx1 i ˚ ˚ hxk i: L EMMA 1. and so. and so we assume s > 1. there exist generators y1 .51).h1 . P ROOF. and so Hk 1/ . x2 C x1 . If ci < 0. : : : . xk g generates M . : : : . mi 2 Z. mi 2 Z H) mi xi D 0 for every iI D c1 x1 C T HEOREM 1. The P subgroup hx1 . : : : . given by Kronecker in 1870. 7 John . Now ˘ fx1 . If M can be generated by one element.x1 C x2 / C c3 x3 C C ck xk . : : : . at least two ci are nonzero. I include an elementary exposition here. hk / 7! hhk W .54 Every ﬁnitely generated commutative group M has a basis.53 Suppose that M is generated by fx1 . C ck xk 2 C mk xk D 0. ˘ gcd. The pair H1 Hk 1.:::.
it contains at most n elements of order dividing n.x1 /. : : : . and then they are uniquely determined by M . 2 E XAMPLE 1.x1 /. . ns 1 jns . Statement (c) says that M can be expressed M Cpe1 1 Cpet t r C1 . they are called the elementary divisors of M . we may suppose that G D Cn1 Cnr with ni 2 N. there are at most n of these. then G has more than n elements of order dividing n. : : : . : : : . xk i. (c) the ni can be chosen to be powers of prime numbers.55 A ﬁnite commutative group is cyclic if. : : : . The number r is called the rank of M .a1 . xk g for M with k elements there is one for which the order of x1 is the smallest possible. T HEOREM 1.54. P ROOF. But dy1 D m1 x1 C m2 x2 C C mk xk D 0 2 and d Ä m1 < order. (12) e e e for certain prime powers pi i (repetitions of primes allowed). and so the corollary shows that every ﬁnite subgroup of F is cyclic. then there exists a relation m1 x1 C m2 x2 C C mk xk D 0 (10) with m1 x1 ¤ 0. : : : . According to the lemma. By r being uniquely determined by M . Therefore. : : : .56 Let F be a ﬁeld. Among the generating sets fx1 . xk i. and then they are uniquely determined by M . Because unique factorization holds in F ŒX . The integers n1 . the number of copies of C1 will be the same (and similarly for the ni in (b) and (c)). Let ai generate the i th factor. and so generates G. The elements of order dividing n in F are the roots of the polynomial X n 1. ar / has order n1 nr . and so this is a contradiction. there exists a generating set y1 . If n divides ni and nj with i ¤ j . After the Theorem 1. mk / > 0. If M is not the direct sum of hx1 i and hx2 .26 1. : : : . which together with x1 forms a basis for M . (a) r is uniquely determined by M .m1 . BASIC D EFINITIONS AND R ESULTS generators. ei 1. This will complete the proof. Let d D gcd. : : : . (b) the ni can be chosen so that n1 2 and n1 jn2 . ns Cn1 Cns r C1 (11) 2 and r 0. We shall show that M is then the direct sum of hx1 i and hx2 . ns in (b) are called the invariant factors of M . and that the integers p11 .57 A nonzero ﬁnitely generated commutative group M can be expressed M for certain integers n1 . : : : . : : : . Then . we mean that in any two decompositions of M of the form (11). yk such that y1 D c1 x1 C C ck xk . we may suppose that m1 2 N and m1 < order. because the induction hypothesis provides us with a basis for the second group. for each n > 0. C OROLLARY 1. the hypothesis implies that the ni are relatively prime. pt t are uniquely determined by M . and let ci D mi =d . Moreover.
Z=pZ/r . n 1 g where D e 2 i=n is a primitive nth root of 1. we can replace M with its torsion subgroup (and so assume r D 0). Continuing in this fashion. (a) For a prime p not dividing any of the di .C/ n .c) If gcd.C/ of D fe 2 i m=n j0ÄmÄn 1g D f1. each commutative group of order 90 is isomorphic to exactly one of C90 or C3 C30 — to see this. Cns D Cpei where the product is over the distinct primes among the i pi and ei is the highest exponent for the prime pi . The homomorphism a 7! 1 is called the trivial (or principal) character. In proving the uniqueness statements in (b) and (c). the set elements of order dividing n is cyclic of order n. E XAMPLE 1. p 2 will divide some pi i in (12) 2 .54. ns 1 is the smallest integer > 0 such that ns 1 M is cyclic. 2 S UMMARY 1. n1 jn2 . : : : .C1 =pC1 /r . Once this has been achieved. Similarly. or it can be proved directly: ns is the smallest integer > 0 such that ns M D 0. : : : . . we ﬁnd that the elementary divisors of M can be read off from knowing the numbers of elements of M of each prime power order.C/ D fz 2 C j jzj D 1g. and so on. The uniqueness of the invariant factors can be derived from that of the elementary divisors. This is an inﬁnite group. T HE LINEAR CHARACTERS OF A COMMUTATIVE GROUP Let .59 The quadratic residue modulo p of an integer a not divisible by p is deﬁned by Â Ã a 1 if a is a square in Z=pZ D 1 otherwise. for example.m. A prime p will occur as one of the primes pi in (12) if and only M has an element of order p. note that the largest invariant factor must be a factor of 90 divisible by all the prime factors of 90. For example. then Cm Cn contains an element of order mn. A linear character (or just character) of a group G is a homomorphism G ! . in fact. For any integer n. in which case p will occur exact a times where p a e is the number of elements of order dividing p. M=pM .58 Each ﬁnite commutative group is isomorphic to exactly one of the groups Cn1 Cnr . and so Cm Cn Cmn : (13) Use (13) to decompose the Cni into products of cyclic groups of prime power order. n/ D 1.Commutative groups P ROOF. nr 1 jnr : The order of this group is n1 nr . 27 and so r is the dimension of M=pM as an Fp vector space. in which case it will divide exactly b of the if and only if M has an element of order p e pi i where p a b p 2b is the number of elements in M of order dividing p 2 . n . (b.C/. ns 2 is the smallest integer such that ns 2 can be expressed as a product of two cyclic groups. The ﬁrst assertion is a restatement of Theorem 1. p . (13) can be used to combine factors to achieve a decomposition as Q in (b).
a/ D jGj if is trivial 0 otherwise.a/ .b/ ¤ .C/ is a character of .ab/ .g 0 /.a/ .b/ 1 ¤ 1. and so X X X . The set of characters of a group G becomes a group G _ under the addition.C/ ! . Apply the theorem to G _ .g/ . The statements are obvious for cyclic groups.b/ . 2 7! .C/ not of this form. 2 P ROOF.G _ /_ ' G. For any characters and of G.C/. In other words. let G be a commutative group endowed with a locally compact topology for which the group operations are continuous.62 (O RTHOGONALITY R ELATIONS ) Let G be a ﬁnite commutative group.C/ is larger than Z.C/ has a natural topology for which it is locally compact.C/ ! . Therefore Œa 7! p W . In particular.a 1 / D 0.b/ 1 .G A SIDE 1.a/ .a/ . it is necessary to consider groups together with a topology.28 1.a/ . this implies that a2G . then the group G _ of continuous characters G ! . Each m 2 Z does deﬁne a character 7! m W . (a) The dual of G _ is isomorphic to G. noting that .a/ of G _ is P ROOF. G G _ and G ' G __ . X a2G .a 1 / D . this depends only on a modulo p. (b) The map G ! G __ sending an element a of G to the character an isomorphism. 2 C OROLLARY 1. then . .60 Let G be a ﬁnite commutative group. BASIC D EFINITIONS AND R ESULTS Clearly.b/. as we observed above.C/. X a2G . the dual group Z_ of Z is isomorphic to . H /_ ' G _ H _.63 For any a 2 G. these are the only continuous homomorphisms.a 1 / D 1. Otherwise there exists a b 2 G such that . For example. and . Z_ ' . In general. T HEOREM 1.g/ D .ab/ 1 / D .b/ .1/. As a runs over G.Z=pZ/ . and the Pontryagin duality theorem says that the natural map G ! G __ is an isomorphism. so also does ab.a/ D jGj if a D e 0 otherwise. P ROOF. and so the sum is jGj. then is Á Á Á a b a ab p D p p (because . X 2G _ . For inﬁnite groups.Z=pZ/ is cyclic).a 1 /D jGj if D 0 otherwise. T HEOREM 1. If D . and so the dual of . called the dual group of G. .Z=pZ/ ! f˙1g D 2 . For example.C/ by the map 7! .61 The statement that the natural map G ! G __ is an isomorphism is a special case of the Pontryagin theorem. but there are many homomorphisms . and if neither a nor b Á divisible by p. C 0 /. However..a 1 /: a2G a2G a2G P Because .
P ROOF.X u 1 /. T HEOREM 1. If a has order m and b has order n. u 1t u 1v 1 has characteristic polynomial X2 . and 2r respectively.X w/. Exercises 11 Show that the quaternion group has only one element of order 2. v. r > 1. 2n. As the group Fq has order q 1 and is cyclic (see 1. .X w 1 /. Deduce that Q is not isomorphic to D4 . there exist elements a and b of SL2 . 2n.Fq / such that a. bi is inﬁnite. As I is the unique element of order 2 in SL2 . and that it commutes with all elements of Q. b. and ab have orders 2m. there exist elements u. The characteristic polynomial of a is . 2 and so ab is similar to diag.u. Let p be a prime number not dividing 2mnr.uv C t C u 1 v 1 /X C 1 D .X u/. 0 u 1 t v 1 where t has been chosen so that uv C t C u 1 v 1 D wCw 1 : 1 /.Z/. Similarly b has order 2n. Let Â Ã Â Ã u 1 v 0 aD and b D (elements of SL2 .The order of ab 29 The order of ab Let a and b be elements of a group G.56). and so some power of it. ab in SL2 .w. n. Then p is a unit in the ﬁnite ring Z=2mnrZ.64 For any integers m. 12 Consider the elements Â 0 aD 1 Ã 1 0 Â bD 0 1 Ã 1 1 in GL2 . there exists a ﬁnite group G with elements a and b such that a has order m. and that every subgroup of Q is normal. q say. and r as required. and w of Fq having orders 2m. 13 Show that every ﬁnite group of even order contains an element of order 2. b. is 1 in the ring. and 2r respectively. Therefore ab has order 2r. The matrix Â Ã uv C t v 1 ab D . n. We shall show that. but that ab has inﬁnite order. for a suitable prime power q.Fq /.Fq /). This means that 2mnr divides q 1. Show that a4 D 1 and b 3 D 1. the images of a. and ab has order r. what can we say about the order of ab? The next theorem shows that we can say nothing at all. u Therefore a has order 2m. and hence that the group ha.Fq /=f˙I g will then have orders m. b has order n. w 1 /. and so a is similar to diag.
the smallest such m is then called the exponent of G. (a) Show that every group of exponent 2 is commutative. 17 Show that any set of representatives for the conjugacy classes in a ﬁnite group generates the group. BASIC D EFINITIONS AND R ESULTS 14 Let n D n1 C C nr be Qpartition of the positive integer n. c 2 Fp ˇ . Show that commensurability is an equivalence relation on the subgroups of G. b. . (b) Show that. 18 Two subgroups H and H 0 of a group G are said to be commensurable if H \ H 0 is of ﬁnite index in both H and H 0 . for an odd prime p.30 1. but is not commutative. Use Lagrange’s theorem a to show that nŠ is divisible by rD1 ni Š. i 15 Let N be a normal subgroup of G of index n. 19 Show that a nonempty ﬁnite set with an associative binary operation satisfying the cancellation laws is a group. : 0 0 1 ˇ has exponent p. the group of matrices 9 80 1ˇ = < 1 a b ˇ ˇ @0 1 c A ˇ a. Give an example to show that this may be false when the subgroup is not normal. then g n 2 N . Show that if g 2 G. 16 A group G is said to have ﬁnite exponent if there exists an m > 0 such that am D e for every a in G.
) Then 1 serves as an identity element.C HAPTER 2 Free Groups and Presentations. Free monoids Let X D fa. Recall that a monoid is a set S with an associative binary operation having an identity element e. for example. we make precise what this means. For example. c. The empty sequence is allowed. Two words can be multiplied by juxtaposition. For example. b 2 S and ˛. aa. Write SX for the set of words together with this binary operation. A word is a ﬁnite sequence of symbols from X in which repetition is allowed. aabac. Dn can be described as the group with generators r. srsr D e: In this chapter. the second condition is not automatic. A homomorphism of monoids preserves all ﬁnite products. b. Because inverses cause problems.b/ for all a. Coxeter Groups It is frequently useful to describe a group by giving a set of generators for the group and a set of relations for the generators from which every other relation in the group can be deduced. 31 .ab/ D ˛. First we need to deﬁne the free group on a set X of generators — this is a group generated by X and with no relations except for those implied by the group axioms. Then SX is a monoid. : : :g be a (possibly inﬁnite) set of symbols.e/ D e — unlike the case of groups. we denote it by a different symbol. b are distinct words. aaaa aabac D aaaaaabac: This deﬁnes on the set of all words an associative binary operation. (In the unfortunate case that the symbol 1 is already an element of X . called the free monoid on X. and we denote it by 1.a/˛. A homomorphism ˛W S ! S 0 of monoids is a map such that ˛. s 2 D e. we ﬁrst do this for monoids. s and relations r n D e.
Deﬁne X 0 to be the set consisting of the symbols in X and also one additional symbol. because the result of such a sequence of cancellations will not be affected if a0 a0 1 is cancelled ﬁrst. Moreover. since the argument is the same in both cases. at least one of a0 or a0 1 must . b 1 . a 1 . Starting with a word w. This becomes a monoid under juxtaposition. b. the map X ! SX has the following universal property: for any map of sets ˛W X ! S from X to a monoid S. we can perform a ﬁnite sequence of cancellations to arrive at a reduced word (possibly empty). 1 a c ca ! cabb a a ! cabb ! ca: We have underlined the pair we are cancelling. for example. because the induction hypothesis can be applied to the (shorter) word obtained by cancelling a0 a0 1 . there exists a unique homomorphism SX ! S making the diagram at right commute: X a 7! a SX ˛ S: Free groups We want to construct a group FX containing X and having the same universal property as SX with “monoid” replaced by “group”. consider a reduced form w0 obtained by a sequence in which no cancellation cancels a0 a0 1 directly. which will be called the reduced form w0 of w. F REE G ROUPS AND P RESENTATIONS . P ROOF. C OXETER G ROUPS When we identify an element a of X with the word a. Otherwise a pair of the form a0 a0 1 or a0 1 a0 occurs — assume the ﬁrst. and that different terms survive in the two cases (the ca at the right in the ﬁrst cancellation. There may be many different ways of performing the cancellations. Observe that any two reduced forms of w obtained by a sequence of cancellations in which a0 a0 1 is cancelled ﬁrst are equal. cabb cabb 1 1 a 1 1 c 1 1 ca ! caa 1 c 1 1 ca ! cc 1 1 ca ! ca. and the next result says that this always happens. Next observe that any two reduced forms of w obtained by a sequence of cancellations in which a0 a0 1 is cancelled at some point are equal. and we can’t cancel out the obvious terms in words of the following form: aa 1 or a 1 a A word is said to be reduced if it contains no pairs of the form aa 1 or a 1 a. X becomes a subset of SX and generates it (i. If w is reduced.32 2. for each a 2 X. Note that the middle a 1 is cancelled with different a’s.. and the ca at left in the second). there is nothing to prove. Nevertheless we ended up with the same answer.e. no proper submonoid of SX contains X). denoted a 1 . Since a0 a0 1 does not remain in w0 . thus X 0 D fa. P ROPOSITION 2. : : :g: Let W 0 be the set of words using symbols from X 0 .1 There is only one reduced form of a word. We use induction on the length of the word w. Finally. but it is not a group because a 1 is not yet the inverse of a.
w 0 are equivalent. because . Alternatively. i. inverses are obtained in the obvious way. then X becomes identiﬁed with a subset of FX — clearly it generates FX .. there exists a unique homomorphism FX ! G making the following diagram commute: X a 7! a FX ˛ G: . and so we may cancel the original pair instead.e. 2 Let FX be the set of equivalence classes of words. It also has inverses. two words represent the same element of FX if and only if they have the same reduced forms. P ROPOSITION 2. if they have the same reduced P ROPOSITION 2. Thus we are back in the case just proved.3 For any map of sets ˛W X ! G from X to a group G. The next proposition is a precise statement of the fact that there are no relations among the elements of X when regarded as elements of FX except those imposed by the group axioms.ab gh/ h 1 g 1 b 1 a 1 1: Thus FX is a group. denoted w form. we can ﬁrst cancel as much as possible in w and v separately. If the pair itself is not cancelled. and so we obtain the same result in the two cases. to obtain w0 v0 and then continue cancelling. w w0. then the ﬁrst cancellation involving the pair must look like 6 a0 1 6 a0 a0 1 or a0 6 a0 1 6 a0 where our original pair is underlined. Thus the reduced form of wv is the reduced form of w0 v0 .2 shows that the binary operation on W 0 deﬁnes a binary operation on FX . called the free group on X. each element of FX is represented by a unique reduced word. but (by assumption) the reduced forms of w and v equal the reduced forms of w 0 and v 0 . the empty word represents 1. w 0 . which obviously makes it into a monoid. Proposition 2.Free groups 33 be cancelled at some point. To obtain the reduced form of wv. When we identify a 2 X with the equivalence class of the (reduced) word a. But the word obtained after this cancellation is the same as if our original pair were cancelled.2 Products of equivalent words are equivalent. A similar statement holds for w 0 v 0 . Let w0 and v0 be the reduced forms of w and of v. This is an equivalence relation (obviously). 2 We say two words w. multiplication is deﬁned by juxtaposition. To summarize: the elements of FX are represented by words in X 0 . multiplication is deﬁned by juxtaposition and passage to the reduced form. v v 0 H) wv w0v0: P ROOF.
3). and so ˇ ı ˛ D idF X . in particular. there exists a unique homomorphism ˛W FX ! F 0 such that ˛ ı Ã D Ã0 . ˛ extends to a homomorphism of monoids SX 0 ! G. (1921) proved this for ﬁnitely generated subgroups.a/ 1 . then there is a unique isomorphism ˛W FX ! F 0 such that ˛ ı Ã D Ã0 .44. P ROOF.5 Every group is a quotient of a free group. According to (2. by the universality of Ã0 . The best proof uses topology. now .4 The universal property of the map ÃW X ! FX . some important results on free groups. a monoid.34 2. Then there is an algorithm for constructing from any ﬁnite set of generators for H a free ﬁnite set of generators. Theorem 11.a 1 / D ˛. similarly. but the free group on a set consisting of two elements is already very complicated.ˇ ı ˛/ ı Ã D Ã. We extend it to a map of sets X 0 ! G by setting ˛. and so ˛ and ˇ are inverse isomorphisms. x 7! x. and Hall 1959. there exists a unique homomorphism ˇW F 0 ! FX such that ˇ ı Ã0 D Ã. but by the universality of Ã. C OXETER G ROUPS P ROOF. and so will factor through FX D SX 0 = . and in fact gave an algorithm for deciding whether a word lies in the subgroup. 2 For example. 1 Nielsen . and the image.G W H / D i < 1. Two free groups FX and F Y are isomorphic if and only if X and Y have the same cardinality. If G has ﬁnite rank n and . Thus we can deﬁne the rank of a free group G to be the cardinality of any free generating set (subset X of G for which the homomorphism FX ! G given by (2. ˛ ı ˇ D idF 0 . and in particular covering spaces—see Serre 1980 or Rotman 1995. Schreier (1927) proved the general case. and let F be the free group generated by X .3) is an isomorphism). Consider a map ˛W X ! G. see Rotman 1995. then H is free of rank ni i C 1: In particular. I now discuss.g. Choose a set X of generators for G (e. the commutator subgroup of the free group on two generators has inﬁnite rank. C OROLLARY 2. must equal G: 2 The free group on the set X D fag is simply the inﬁnite cyclic group C1 generated by a. being a subgroup containing X. T HEOREM 2.. Let H be a ﬁnitely generated subgroup of a free group G. For proofs. Chapter 7. The resulting map FX ! G is a group homomorphism. Chapter 11. H may have rank greater than that of F (or even inﬁnite rank2 ). F REE G ROUPS AND P RESENTATIONS .6 (N IELSEN S CHREIER ) 1 Subgroups of free groups are free. Because G is. X D G). It is unique because we know it on a set of generators for FX . This map will send equivalent words to the same element of G. characterizes it: if Ã0 W X ! F 0 is a second map with the same universal property. 2 R EMARK 2. idF X is the unique homomorphism FX ! FX such that idF X ıÃ D Ã. the map a 7! aW X ! G extends to a homomorphism F ! G. without proof. FX Ã X Ã0 ˛ F 0: We recall the proof: by the universality of Ã.
see Exercise 25). Then G D f1g because s D ss 3 t D s 4 t D t 1 D s3t t 3 D s 3 ss 3 D s: For the remaining examples. n a2 n 1 3. r ( i is a loop around the i th point) and a single relation 1 r D 1: (g) The fundamental group of a compact Riemann surface of genus g has 2g generators u1 . s 2 . :::.1).X. In general. R/ is a presentation for G.7 (a) The dihedral group Dn has generators r. t j s 3 t. a2 n 2 D b 2 . The free abelian group on generators a1 . . I should say the relations a2 n 1 . bab Da 1 : For n D 3 this is the group Q of (1.7). One also says that .40) is said to have X as generators and R as relations (or as a set of deﬁning relations).9 below for a proof. E XAMPLE 2. s 4 i. has generators a. and denotes G by hX j Ri.Generators and relations 35 Generators and relations Consider a set X and a set R of words made up of symbols in X 0 . t 3 . a2 n 2 b 2. an and relations Œai . vg and a single relation u1 v1 u1 1 v1 1 (ibid. it has order 2n (for more on it. : : : . and the quotient G of FX by the normal subgroup generated by these elements (1. which contains a good account of the interplay between group theory and topology. b and relations3 1 D 1. i ¤ j: (d) Let G D hs. IV Exercise 5. s and deﬁning relations r n . srsr: (See 2. (e) The fundamental group of the open disk with one point removed is the free group on where is any loop around the point (ibid. a2 . II 5. bab 1 a. def (c) Two elements a and b in a group commute if and only if their commutator Œa. Each element of R represents an element of the free group FX . for many types of topological spaces. ug . aj . (f) The fundamental group of the sphere with r points removed has generators 1 . : : : . b D aba 1 b 1 is 1. 3 Strictly ug vg ug 1 vg 1 D 1 speaking. an has generators a1 .) (b) The generalized quaternion group Qn . there is an algorithm for obtaining a presentation for the fundamental group.18). :::. see Massey 1967. v1 . For example.
b 2 D 1. ba D an 1 b imply that each element of G is represented by one of the following elements.xy/ D ab. Then x has order m.8 Let G be the group deﬁned by the presentation . y j x m . we know that ˛ extends to a homomorphism FX ! H .43). there exists a group H with elements a and b such that a. n. By the universal property of quotients (see 1. ˛ factors through FX=N D G. n > 1. n. s 2 Dn satisfy these relations. 4 Each . As ˛. there exists a homomorphism ˛W G ! H such that ˛. To see this. b j a2 . and ab have orders m.x/ D a and ˛. As this is true for all r > 1. bg ! Dn .64). E XAMPLE 2. b j an .17).X. which we again denote ˛.8). y n i where m. F REE G ROUPS AND P RESENTATIONS . an 1 b. ab. The equalities an D 1. For any group H and map of sets ˛W X ! H sending each element of R to 1 (in the obvious sense4 ). Similarly.10 Let G D hx. babai. By assumption ÃR Ker. there exists a unique homomorphism G ! H making the following diagram commute: X a 7! a G ˛ H: P ROOF. b. b 7! s extends uniquely to a homomorphism G ! Dn . and so jGj Ä 2n D jDn j. b. R/. r > 1. Because the elements r. Similarly. : : : . and xy has inﬁnite order in G.ab/n i ' Dn by a 7! s. This proves the existence.˛/. b 2 . This homomorphism is surjective because r and s generate Dn .y/ D b. b 2 . : : : . and the fact that ˛. y has order n. Let ÃR be the image of R in FX . According to (2. . The order of x certainly divides m. a 7! r.9 Let G D ha. b 7! t. recall that for any integers m. Therefore the homomorphism is bijective (and these symbols represent distinct elements of G). From the universal property of free groups (2. the element xy must have order at least r.˛/. 1.36 2. and r respectively (Theorem 1. element of R represents an element of FX . and the condition requires that the unique extension of ˛ to FX sends each of these elements to 1. the element xy has inﬁnite order. We prove that G is isomorphic to the dihedral group Dn (see 1. and therefore the normal subgroup N generated by ÃR is contained in Ker. 2 E XAMPLE 2.3). ha.x/ has order m shows that x has order exactly m. and the uniqueness follows from the fact that we know the map on a set of generators for X. the map fa. y has order n. C OXETER G ROUPS P ROPOSITION 2. an 1 .
simple. T HE WORD PROBLEM Let G be the group deﬁned by a ﬁnite presentation . Therefore the homomorphism is bijective. e/ has only ﬁnitely many ﬁnite quotients. it asks whether there is one ﬁnite quotient of B. The classiﬁcation of the ﬁnite simple groups (see p. Adjan and Novikov showed the answer is yes: there do exist inﬁnite ﬁnitely generated groups of ﬁnite exponent. e/ having all other ﬁnite quotients as quotients. solvable. The word problem for G asks whether there exists an algorithm (decision procedure) for deciding whether a word on X 0 represents 1 in G. E XAMPLE 2.r.11 Consider a ﬁnite group G. The answer is negative: Novikov and Boone showed that there exist ﬁnitely presented groups G for which no such algorithm exists.X. equivalently. The extension of a 7! aW X ! G to FX sends each element of R to 1. e/ is ﬁnite. R/. and so G is ﬁnitely presented. Let G 0 D hX j Ri. The Burnside problem asked whether B. Although it is easy to deﬁne a group by a ﬁnite presentation.r.X. nilpotent. but does there exist such a group with ﬁnite exponent? (Burnside problem). The restricted Burnside problem asks whether B.Finitely presented groups 37 Finitely presented groups A group is said to be ﬁnitely presented if it admits a presentation . Of course. free. T HE B URNSIDE PROBLEM Recall that a group is said to have exponent e if g e D 1 for all g 2 G and e is the smallest natural number with this property. e/ is the quotient of the free group on r generators by the subgroup generated by all eth powers. as the quotient of a huge free group by a huge subgroup. Let X D G.X. and so jG 0 j Ä jGj. But every element of G 0 is represented by an element of X.r. In 1968. R/ with both X and R ﬁnite. there do exist other groups for which there is an algorithm. 50) showed that in order prove that B. it sufﬁces to . which may be quite small. R/ is a presentation of G. calculating the properties of the group can be very difﬁcult — note that we are deﬁning the group. e/ always has only ﬁnitely many ﬁnite quotients. and let R be the set of words fabc 1 j ab D c in Gg: I claim that . It is easy to write down examples of inﬁnite groups generated by a ﬁnite number of elements of ﬁnite order (see Exercise 12 or Example 2. torsion.10). which is obviously surjective. and it is known to be inﬁnite except some small values of r and e. torsionfree. or has a solvable word problem. and therefore deﬁnes a homomorphism G 0 ! G. ﬁnite. Chapter 12. for proofs of these statements. T HE RESTRICTED B URNSIDE PROBLEM The Burnside group of exponent e on r generators B. The same ideas lead to the following result: there does not exist an algorithm that will determine for an arbitrary ﬁnite presentation whether or not the corresponding group is trivial.r. See Rotman 1995.r. abelian. I list some negative results.
G. . According to (2. See Feit 1995. (a) If m. then G D hS j Ri where R D f. 2. We shall develop various methods for recognizing groups from their presentations (see also the exercises).10). s and t each have order 2.s. S ! N [ f1g E XAMPLES 2. According to (2. but for which this is very difﬁcult to prove. Thus a Coxeter system is deﬁned by a set S and a mapping mW S satisfying (14).s.s. S / consisting of a group G and a set of generators S for G subject only to relations of the form . t / 2.st /m.12 Up to isomorphism. and st has inﬁnite order.t. s/: When no relation occurs between s and t. t / D m. t / D 1.st /m. This was shown by Eﬁm Zelmanov in 1989 after earlier work of Kostrikin. s/ D 1 all s. fs.G.s. F REE G ROUPS AND P RESENTATIONS . fs..G.xi /1Äi Än . Coxeter groups A Coxeter system is a pair . t g/ where G D hs. t j s 2 . The cardinality of S is called the rank of the Coxeter system. C OXETER G ROUPS prove it for e equal to a prime power.13 The Coxeter systems of rank 2 are indexed by m.s. t 2 . In particular. we set m. T ODD C OXETER ALGORITHM There are some quite innocuous looking ﬁnite presentations that are known to deﬁne quite small groups. t g/ where G D hs.t / j m.C2 . then the Coxeter system is .s. the only Coxeter system of rank 1 is . t j s 2 . where 8 < m.9). t / 2 (14) : m. then the Coxeter system is . The Coxeter groups are those that arise as part of a Coxeter system.s.38 2.s. The standard approach to these questions is to use the ToddCoxeter algorithm (see Chapter 4 below). m. . 2. s ¤ t and st has order n. fsg/. G ' Dn . (b) If m. t / D 1. t / is an integer n.t / D 1.yi /1Äi Än / D xi yi : .s. t / < 1g.s.st /n i.14 Let V D Rn endowed with the standard positive deﬁnite symmetric bilinear form X . t 2 i.
: : : . We write s˛ for the reﬂection deﬁned by ˛. t / in G 0 (see 2. : : : . : : : .12). and so the theorem says that the elements of S remain distinct in G and that each s and each st has the largest possible order. sends the vector i j ˛ D . or v3. t to t. (c) For each s ¤ t in S .G. We know that s and t are distinct elements of order 2 in G 0 and that st has order m. 0. 1.s. t j s 2 . : : : . See Humphreys 1990.t / i.16 Let . t / in G. .s. st has order m. P ROOF. As Sn is generated by the transpositions.˛.8). and all other elements of S to 1 extends to a homomorphism G ! G 0 . If S has only a single element. a .a . they are precisely the ﬁnite Coxeter groups. and it realizes G as a group of automorphisms of a “geometry”. t /. According to (2. an / D .16 deﬁnes a “geometry” on V for which there exist “reﬂections” s . The standard proof of Theorem 2.V /. and so the statements are obvious.s.n/ /: The transposition .s. : : : . this shows that it is a ﬁnite reﬂection group (hence also a Coxeter group). Note that the order of s is 1 or 2. . Therefore. For such a group G. Otherwise. (a) The natural map S ! G is injective. ai . : : :/ to its negative.02 of these notes. . such that s t has order m. 1.a1 .13). : : : . S / be the the Coxeter system deﬁned by a map mW S f1g satisfying (14). the map s 7! s extends to homomorphism of group G ! GL. s 2 S .Coxeter groups 39 A reﬂection is a linear map sW V ! V sending some nonzero vector ˛ to ˛ and ﬁxing the points of the hyperplane H˛ orthogonal to ˛.15 Let Sn act on Rn by permuting the coordinates. . 0. ai .a1 . 2 R EMARK 2.4). and leaves the points of the hyperplane H˛ D . ibid.0.9). i j T HE STRUCTURE OF C OXETER GROUPS T HEOREM 2. : : : . 0. and the order of st divides m. The map S ! G 0 sending s to s. A ﬁnite reﬂection group is a ﬁnite group generated by reﬂections.. t 2 .17 Let V be the Rvector space with basis a family . (b) Each s 2 S has order 2 in G. t /.st /m. 1. ˛/ because this is certainly correct for v D ˛ and for v 2 H˛ . the ﬁnite reﬂection groups are all Coxeter groups (in fact. 2. Thus. Chapter 5. S/ is a Coxeter system (Humphreys 1990.s. and hence (by linearity) on the whole of V D h˛i ˚ H˛ . it is given by the formula 2.ij / is a reﬂection. and it follows that the same is true in G. then G ' C2 (see 2.ij / interchanging i and j . let s and t be distinct elements of S . an / ﬁxed. and let G 0 D hs.es /s2S indexed by S .1/ . 0.G. ˛/ s˛ v D v ˛. S ! N[ . it is possible to choose a set S of generating reﬂections for which . 6.v. This proves the theorem.
F REE G ROUPS AND P RESENTATIONS . b 2 . (b) Prove that F1 F1 is not a free group. (This is usually known as the inﬁnite dihedral group. b 3 . Prove: (a) If an D b n with n > 1. b.y/. . aba 1 bc group f1g. bi. then a D b.40 2. and he asked for a uniﬁed proof. bjan . 24 Let Fn denote the free group on n generators. (c) Prove that the centre Z. but it hit me while I was in the middle of the road. prove that D2n han i ha2 . Prove that G is the trivial 28 Let F be the free group on the set fx. . 1. [Hint: Use universal properties. Find a minimal generating set for the kernel of ˛. As I was walking through Cambridge.9). 25 Prove that Qn (see 2.] 23 Let a and b be elements of an arbitrary free group F . Prove that the element gDs 1 t 1 s 1 t st 1 st is in the kernel of every map from G to a ﬁnite group. C OXETER G ROUPS Exercises 21 Let Dn D ha. then ab D ba.abc/ then c 610 D 1. Coxeter came to Cambridge and gave a lecture [in which he stated a] problem for which he gave proofs for selected examples. then Fn is isomorphic to both a subgroup and a quotient group of Fm . . Math. suddenly the idea hit me. t j t 1s3t D s 5 i. ababi is an inﬁnite group. pp. I left the lecture room thinking. So I pretended that Coxeter had calculated the difﬁculty of this problem so precisely that he knew that I would get the solution just in the middle of the road. yg and let G D C2 . then a D 1. (b) If am b n D b n am with mn 6D 0. I’ve called that theorem “the murder weapon”.) 27 Let G D ha. . 22 Prove that the group with generators a1 . Prove that Qn =Z. .ab/n i ' Dn (cf. (b) Prove that G D ha. (c) If the equation x n D a has a solution x for every n. Let ˛ be the homomorphism F ! G such that ˛. Intelligencer 23 (2001). an and relations Œai . c j a3 . c 4 . Prove: (a) If n < m. John Conway.] 1 b 1 i. no.Qn / is isomorphic to D2n 1 .x/ D a D ˛. with generator a ¤ 1.bcb 1 /3 . i ¤ j . . an . 8–9.Qn /. Is the kernel a free group? 29 Let G D hs. b j a2 . .7b) has a unique subgroup of order 2.Fn / D 1 provided n > 1. : : : . Ever since.aba 1 /3 D . When the idea hit me I stopped and a large truck ran into me. : : : . b 2 . [Hint: Expand . If n is odd. 2. One consequence of it is that in a group if a2 D b 3 D c 5 D . which is Z. aj D 1. is the free abelian group on a1 . acac 1 . b j a2 . 26 (a) Prove that ha. 2. . and hence that D2n C2 Dn . ababi be the nth dihedral group.
and the remaining automorphisms are said to be outer.G/: In fact. the identity map g 7! g is an identity element.1 (a) Let G D Fn .F2 /: 41 . an automorphism is a bijection. and so we obtain from (1.x/ D ˛. x 7! gxg 1 WG!G is an automorphism of G.G/.G/ ! Inn. Note that .g/ 1 D i˛. which is again an automorphism.ig ı ih /. composition of maps is always associative (see (5). and therefore has an inverse. the map ig “conjugation by g”.˛ ı ig ı ˛ 1 /.G/ is a normal subgroup of Aut. The automorphisms of G as a commutative group are just p the automorphisms of G as a vector space over Fp .G/. (b) As a particular case of (a). Because G is commutative.C HAPTER 3 Automorphisms and Extensions Automorphisms of groups An automorphism of a group G is an isomorphism of the group with itself. we see that Aut.G/ D fg 2 G j gx D xg all x 2 Gg.x/: E XAMPLE 3. p.g/ .G/ is a homomorphism. igh . For g 2 G.x/ g 1 / D ˛. Z.x/.G/ of automorphisms of G becomes a group under composition: the composite of two automorphisms is again an automorphism. An automorphism of this form is called an inner automorphism. all nontrivial automorphisms of G are outer. Its kernel is the centre of G.g/ x ˛. Inn.g ˛ 1 . The set Aut.Fp /. thus Aut.G/: for g 2 G and ˛ 2 Aut.hxh 1 /g 1 .45) an isomorphism G=Z. Its image is denoted by Inn.e. .G/ D GLn .gh/ 1 D g. i. 9).gh/x.x/ D . and so the map g 7! ig W G ! Aut..C2 C2 / D GL2 .
Z=nZ/ ' . 7.b/ D ˛. An automorphism ˛ of G generators of G are exactly the elements a must send a to another generator of G. and so ˛.m. Therefore. say G D hai. n/ D 1. ˛.14.F2 / S3 .ai / D ˛.n/ . The map ˛ 7! m deﬁnes an isomorphism Aut.Z=p11 Z/ r .Cn / ! . The group S2 is commutative and hence fails (a). and (b) every automorphism of G is inner.Q/ S4 .m. Aut. The smallest n n multiple of m divisible by n is m gcd. See Exercise 34. (b) If G is a simple noncommutative group. C2 AUTOMORPHISMS OF CYCLIC GROUPS Let G be a cyclic group of order n. n/ D 1g: This isomorphism is independent of the choice of a generator a for G: if ˛. and so the nonisomorphic groups C2 and S3 have isomorphic automorphism groups. and so the m with gcd. Sn is complete. then r Z=nZ ' Z=p11 Z r Z=ps s Z.Q/ ' Q=ha2 i C2 C2 : In fact.S6 /=Inn.m. According to Exercise 33.n/ . Aut. r ps s is the factorization of n into a m mod n $ . C OMPLETE GROUPS D EFINITION 3. Thus. See Rotman 1995.2 The inner automorphisms of a group are the only automorphisms that extend to every overgroup (Schupp 1987).G/ is complete.10. we have that Inn.ai /m D .a/ D am . am has order gcd.G/ of G is trivial.Z=nZ/ .4 (a) For n ¤ 2.42 3.m.Z=nZ/ where . 6.3 A group G is complete if the map g 7! ig W G ! Aut.5.m mod p r1 . Theorem 7. Let m be an integer 1.a/ D am for some m relatively prime to n. GL2 . A SIDE 3.a/i D ami D . and so r . If n D p11 product of powers of distinct primes. a group G is complete if and only if (a) the centre Z.Z=nZ/ D funits in the ring Z=nZg D fm C nZ j gcd.Z=ps s Z/ : . AUTOMORPHISMS AND E XTENSIONS (c) Since the centre of the quaternion group Q is ha2 i. then for any other element b D ai of G. See Rotman 1995. E XAMPLE 3. This is an isomorphism of rings.G/ is an isomorphism. Theorems 7. then Aut.S6 / C2 and hence S6 fails (b).b/m : r It remains to determine . : : :/ by the Chinese remainder theorem.
Z=p r Z/ ! . R EMARK 3.G/.Z=p Z/ r 2 p odd. p. a characteristic subgroup is normal. Therefore . 5i is not cyclic. (13). m mod n $ . p prime.m mod p r1 .Characteristic subgroups 43 It remains to consider the case n D p r . Also a characteristic subgroup of a characteristic subgroup is a characteristic subgroup. The homomorphism .1 C p/ (cf. : : : . r > 2: Characteristic subgroups D EFINITION 3. S UMMARY 3. and it is characteristic if it stable under all automorphisms. Suppose ﬁrst that p is odd.Z=nZ/ sends an element a of G to am . Thus a normal subgroup of N need not be a normal subgroup of G. Hence . 1. The same argument as in (1.G/ ' . p r 1g is a complete set of representatives 1 for Z=p r Z.Z=nZ/ : The automorphism of G corresponding to Œm 2 . a characteristic subgroup of N will be a normal subgroup of G.Z=p r Z/ is cyclic with generator . Therefore .p 1/pr < C ˆ 2 : C2 C2r 1 .Z=p r Z/ .Z=pZ/ is cyclic.1 C p/j . one ﬁnds that 1 C p has order p r 1 in .7 (a) Consider a group G and a normal subgroup N . N N N N N N . a subgroup H of G is normal if it is stable under all inner automorphisms of G.Z=pZ/ . Using the binomial theorem. Aut. 5. see 3.H / H for all ˛ 2 Aut.p 1/ D 1 and ap again maps to a r def generator of . and p of these elements are divisible by p. r r (b) If n D p11 ps s is the factorization of n into a product of powers of distinct primes pi .Z=p r Z/ map to a generator of . then r . Let a 2 r r .p 1/. An inner automorphism of G restricts to an automorphism of N . However. and every element can be written uniquely in the form i .Z=pZ/ is surjective with kernel of order p r 1 . Thus.6 A characteristic subgroup of a group G is a subgroup H such that ˛. which may be outer (for an example. Then ap . p r D 22 p D 2.Z=nZ/ ' . .5 (a) For a cyclic group of G of order n. 3.Z=p r Z/ contains an element D ap of order p 1. and we know that . The set f0.Z=pZ/ .16 below). 0 Ä j < pr C2 C2 1 : On the other hand. 0Äi <p 1.Z=p r Z/ has order p r pr p D pr 1 . 8 ˆC.Z=ps s Z/ .Z=8Z/ D f1.H / D H for all automorphisms ˛ of G.Z=p11 Z/ r . 7g D h3. 27). In particular.32) shows that it sufﬁces to check that ˛. : : :/: (c) For a prime p.
44 3. p Semidirect products Let N be a normal subgroup of G.z/˛. an element g of G can be written uniquely in the form g D nq. Expect subgroups with a general grouptheoretic deﬁnition to be characteristic.G/ of G is a characteristic subgroup. because zg D gz all g 2 G H) ˛.r i / D r i Â hsi D Cn Â C2 (see 1. then I claim that we can reconstruct G from N . then Dn D hri where Â. we write G D N Q (or N Â Q where ÂW Q ! Aut. Each element g of G deﬁnes an automorphism of N . (c) If H is the only subgroup of G of order m.8 A group G is a semidirect product of its subgroups N and Q if N is normal and the homomorphism G ! G=N induces an isomorphism Q ! G=N .s/. and as g runs over G. . and n must be gq we have a onetoone correspondence of sets G If g D nq and g 0 D n0 q 0 . (d) Every subgroup of a commutative group is normal but not necessarily characteristic. E XAMPLE 3. NQ D G. Indeed. and this deﬁnes a homomorphism ÂW G ! Aut. q 2 Q. For example. n 2. then gg 0 D .q n0 q 1 11 Thus. N \ Q D f1g: (15) Note that Q need not be a normal subgroup of G.g/ also runs over G. because ˛. but it is not p p characteristic because it is not stable under Aut. then it must be characteristic. n 2 N.17). ˛. !N Q: /qq 0 D n Â. let Cn D hri and C2 D hsi. and the restriction of Â to Q. G is a semidirect product of subgroup N and Q if N G G.z/ all g 2 G. Q. n 7! gng 1 . g 7! ig jN: If there exists a subgroup Q of G such that G ! G=N maps Q isomorphically onto G=N . When G is the semidirect product of subgroups N and Q.H / is again a subgroup of G of order m.n0 / qq 0 : D EFINITION 3.N /.N / gives the action of Q on N by inner automorphisms). every subspace of dimension 1 in F2 is subgroup of F2 .nq/ n0 q 0 D n.g/ D ˛. — q must be the unique element of Q mapping to gN 2 G=N .Fp /. Equivalently.9 (a) In Dn . 1.F2 / D GL2 . AUTOMORPHISMS AND E XTENSIONS (b) The centre Z.q/.g/˛.
0 b 1 0 b 0 1 0 1 and so B is not the direct product of T and U .q/. q/.n0 .N / arises from a semidirect product.q 1 /. the action of Q on N is that given by Â .1/ D 1.N.n. q/. We have seen that.1. the action of T on U is not trivial.. (c) The quaternion group can not be written as a semidirect product in any nontrivial fashion (see Exercise 31).n00 . let G D N Q. q/.n00 . U is a normal subgroup of B. Thus. Â / consisting of two groups N and Q and a homomorphism Â W Q ! Aut. Therefore Sn D An C2 . qq 0 /. q/.n. q/. 1/ D . the semidirect product of N and Q: P ROOF.1/ D 1 and Â. T D . for example.n. and deﬁne .n0 . NQ D G. Let B be the subgroup of upper triangular matrices in G. Because Â. B DU T.n. and so G D N Q. qq 0 /: P ROPOSITION 3.n Â. Write q n for Â.n0 . so that the composition law becomes . Therefore. Â ÃÂ ÃÂ Ã Â Ã 0 a 0 1 c a 1 1 ac=b D .n.n Then .q/. n n . Moreover. Â W Q ! Aut. q 0 /.q/.q 1 / D . (e) Let G D GLn . We now prove that every triple .n. q/ D . q/. in fact. As a set. and N \ Q D f1g. q/. q 00 // and so the associative law holds. UD . 1/. Next .10 The composition law above makes G into a group.n0 . Q.q 1 n 1 . 1/. and so . 2 .n. q 0 //. (d) A cyclic group of order p 2 .1.n q 0 qq 0 00 q 0 Q. 0 0 0 1 Then. from a semidirect product G D N . and it is obvious that N G G. q 0 / D . . when n 2. and that the triple determines G.n0 /.q n 1 . T the subgroup of diagonal matrices in G.1.Semidirect products 45 (b) The alternating subgroup An is a normal subgroup of Sn (because it has index 2). Note that. and so .n. q 1 / is an inverse for . qq 0 q 00 / D . when N and Q are regarded as subgroups of G. when n D 2..n.N. we obtain a triple n . U T D B. q/ D . 1/ is an identity element.12/g maps isomorphically onto Sn =An .N //. is not a semidirect product (because it has only one subgroup of order p).n/. Thus G is a group. and U \ T D f1g. Â Ã Â Ã Â Ã 0 1 BD . q 0 / D . p prime.n.F /.q 1 1 n 1 . q 00 / D . . q/. Q. and C2 D f. and U the subgroup of upper triangular matrices with all their diagonal coefﬁcients equal to 1.1.
N / to be the homomorphism such that Â. Let N D hai be cyclic of order p 2 . If we denote the generators of C3 and C4 by a and b. all elements except 1 have order p. Let N D ha. The group G D N ap D 1. with generators a. Deﬁne ÂW Q ! Aut. then p Â Ã def i / is the automorphism of N deﬁned by the matrix 1 0 . and Cp is generated by ˛W a 7! a1Cp (note that ˛ 2 . i. and have the deﬁning relations a3 D 1. Both S3 and C6 are semidirect products of C3 by C2 — they correspond to the two distinct homomorphisms C2 ! C2 ' Aut.9a) C4 C2 . Œb. which does have an element of order 22 : Note that this shows that a group can have quite different representations as a semidirect product: D4 . AUTOMORPHISMS AND E XTENSIONS E XAMPLES 3. 3. (If we regard N as the additive group N D F2 with a and b the standard basis elements. the middle equality shows that the group is not commutative. Let Â be the (unique) nontrivial homomorphism C4 ! Aut. D a1Cp : It is a noncommutative group of order p 3 .) The group G D N Â.5).c ÂQ i 1 is a group of order p 3 . ab D cac 1 .46 3. a D 1 D Œb. that sending a generator of C4 to the map a 7! a2 .C2 C2 / C2 : .. G D4 . : : :). c and deﬁning relations ap D b p D c p D 1.a/ D ab i . 3.11 A group of order 12. Deﬁne Q ! Aut N by b 7! ˛.n.15 Groups of order p 3 (no element of order p 2 ). When p D 2.c i /.13 Groups of order 6. b 4 D 1. not isomorphic to A4 .3.C3 / ' C2 . c: Because b ¤ 1.c i /. When p is odd.N /. q/W N Q!N Â Q is an isomorphism of groups if and only if Â is the trivial homomorphism Q ! Aut.b/ D b. 3. namely. Â. q/ 7! . and let Q D hci be a cyclic group of order p. and let Q D hbi be cyclic of order p. bi be the product of two cyclic groups hai and hbi of order p.a/ D a1C2p . where p is an odd prime.q/.12 Direct products.C3 /.e.n. Then Aut N Cp 1 Cp (see 3. The bijection of sets . Â. then a and b generate G. Then G D C3 Â C4 is a noncommutative group of order 12. b and deﬁning relations bab 1 b p D 1. n 2 N . b.n/ D n for all q 2 Q. bab 1 def D a2 : 3. 2 def Â Q has generators a. and possesses an element of order p 2 .14 Groups of order p 3 (element of order p 2 ).
n/.n0 /.17 If there exists an ˛ 2 Aut. In particular.Q/. and . q/ 7! .n Â. To see this. then the map .n.˛.n/ ˛. a/ of G has the property that g.n.˛ 1 . 1/g 1 D . 3.n.n0 .n0 //.n0 . q/W N is an isomorphism. Â 0 / determine isomorphic groups.˛. q/ 7! . q/ . q/W N Â0 Q!N Â Q 2 is also a homomorphism. q/ . let .N / be the homomorphism sending a generator a of C1 to ˛ 2 Aut. The map .q/ ı ˛ 1 /.q/. a noncommutative group of order p 3 is isomorphic to the group in (3. q 0 / D . qq 0 / D .˛.n0 / .n.˛. q 0 // D .n/.n. and let G D N Â C1 .˛. q/. and so both are isomorphisms. q/ D .n/ .q/ ı ˛ then the map . L EMMA 3.15) if it doesn’t (see Exercise 43). qq 0 / D .˛.n/.n0 //. Q!N Q Â0 Q. of a group N . q/ 7! . q 0 / D . and it is inverse to .q//W N is an isomorphism.q/.˛.N /.q/. qq 0 /. qq 0 / D . let ÂW C1 ! Aut. ˛. Then . Q. q/ 2 N . Â Q N Â0 Q .˛ ı Â.˛. Â / and .n/.n.n. Let ˛ be an automorphism.˛. Q. C RITERIA FOR SEMIDIRECT PRODUCTS TO BE ISOMORPHIC It will be useful to have criteria for when two triples . possibly outer. qq 0 /: Therefore is a homomorphism.N.16 Making outer automorphisms inner.n0 //.18 If Â D Â 0 ı ˛ with ˛ 2 Aut. For .n/. 1/ for all n 2 N.n0 /.q/ D ˛ ı Â.N. there are exactly two noncommutative groups of order p 3 .n. L EMMA 3. We can realize N as a normal subgroup of a group G in such a way that ˛ becomes the restriction to N of an inner automorphism of G.N / such that Â 0 .14) if it has an element of order p 2 and to the group in (3.n. q/ Â Â 1 . up to isomorphism.Â.˛. all q 2 Q.q/.1..Semidirect products 47 For an odd prime p. The element g D . P ROOF.n/ ˛ Â.n/ Â 0 .˛.
Extensions of groups A sequence of groups and homomorphisms 1!N !G!Q!1 is exact if Ã is injective. AUTOMORPHISMS AND E XTENSIONS 2 P ROOF.N / such that Â 0 . Now the map . q i / is an isomorphism N Â Q ! N Â 0 Q. Two extensions of Q by N are said to be isomorphic if there exists a commutative diagram 1 ! N ! G ! Q ! 1 ? ? y 1 1 This ! N ! G0 ! Q ! 1: is Bourbaki’s terminology (Alg` bre. say a0 D ai with i relatively prime to the order of Q.1 An extension is central if Ã. 44). then G is the ZappaSz´ p (or knit) product of H1 e and H2 (see http://en.19 If Q is cyclic and the subgroup Â. then N Â Q N Â0 Q: P ROOF. I 6). 2 S UMMARY 3.˛. and Ker.Q/. G D H1 H2 ((15).G/.N / ! Q. some authors call (16) an extension of N by Q.N / Z. Let a generate Q.n/.n. (b) If H1 is normal in G. Routine veriﬁcation. p. which is central if and only if Â is the trivial homomorphism.48 3.a/ ˛ 1 : The element Â 0 .N / is conjugate to Â 0 .N / is a normal sub group of G (isomorphic by Ã to N / and G=Ã. ' Ã (16) is surjective. there exists an a0 2 Q and an ˛ 2 Aut. G D H1 H2 (1. By assumption.a0 / D ˛ Â.20 Let G be a group with subgroups H1 and H2 such that G D H1 H2 and H1 \ H2 D feg. 1 ! N ! N Â Q ! Q ! 1. Thus Ã. (c) If neither H1 nor H2 is normal.51).Q/. We often identify N with the subgroup Ã. q/ 7! . (a) If H1 and H2 are both normal. e . and so we can choose a0 to generate Q.a0 / generates Â 0 . L EMMA 3.org/wiki/ZappaSzep_product). then G is the semidirect product of H1 and H2 . so that each element g of G can be written uniquely as g D h1 h2 with h1 2 H1 and h2 2 H2 .Ã/. / D Im.wikipedia. a semidirect product N Â Q gives rise to an extension of Q by N . then G is the direct product of H1 and H2 . For example.Q/ of Aut.N / of G and Q with the quotient G=N: An exact sequence (16) is also called an extension of Q by N .
thus gng 1 D n 1 all n 2 N .21 (S CHUR Z ASSENHAUS ) An extension of ﬁnite groups of relatively prime order is split.51. We give two criteria for an extension to split. Rotman 1995. 7.N /. 1 g/ 2 NQ. Let Q D CG . hence N \ Q D 1.N /.41. Finally. Next note that every element of N \ Q is in the centre of N . CG . n 7! gng 1 W N ! N is an automorphism of N . an extension will not split. the extensions 1 ! N ! Q ! Q=N ! 1 with N any subgroup of order 4 in the quaternion group Q. Since g was arbitrary. and 1 ! Cp ! Cp2 ! Cp ! 1 do not split. 2 An extension 1!N !G!Q!1 gives rise to a homomorphism Â 0 W G ! Aut. We shall check that N and Q satisfy the conditions of Proposition 1. for any g 2 G. In fact. or (b) there exists a homomorphism sW Q ! G such that ı s D id : In general.g/. Â 0 .22 An extension (17) splits if N is complete. namely.qQq 1 /n 1 D nQn 1 DQ (recall that every element of N commutes with every element of Q). gQg 1 def D n. For example. (a) there exists a subgroup Q0 G such that induces an isomorphism Q0 ! Q. which (because N is complete) is trivial. This equation shows that 1 g 2 Q. G is then direct product of N with the centralizer of N in G. 1 ! N ! G ! Q ! 1. Observe ﬁrst that. Therefore Q is normal in G.n/ D gng 1 : .Extensions of groups An extension of Q by N . T HEOREM 3. is said to be split if it is isomorphic to the extension deﬁned by a semidirect product N Equivalent conditions: Ã 49 Â Q. 2 (17) P ROPOSITION 3. for any element g D nq 2 G. it must be the inner automorphism deﬁned by an element of N . and (because N is complete). P ROOF.N / D fg 2 G j gn D ng all n 2 N g. and hence g D . we have shown that G D NQ. P ROOF.
endomorphisms of Q and N act as endomorphisms on Ext1 .q/ in Aut. N /. Math. o It would be of the greatest interest if it were possible to give an overview of the entire collection of ﬁnite simple groups. By far the largest class is (c). The problem of classifying all simple groups falls into two parts: A. When Q and N are commutative and Â is trivial. then Ext1 . if Q and N are killed by m and n respectively. the FischerGriess monster. a group is simple if it can’t be realized as an extension of smaller groups. N /Â for the set of isomorphism classes of extensions with a given Â: These sets have been extensively studied.. Some have even speculated that the largest of them. Classify all extensions of ﬁnite groups. In other words.Q.Q. AUTOMORPHISMS AND E XTENSIONS Let q 2 G map to q in Q. N /. certain inﬁnite families of matrix groups. B.N /=Inn. N /. Every ﬁnite group can be obtained by taking repeated extensions of simple groups. Classify all ﬁnite simple groups.m. then the image of Â 0 .N /=Inn. n/. Thus. multiplication by m on Q or N induces multiplication by m on Ext1 . Thus the simple ﬁnite groups can be regarded as the basic building blocks for all ﬁnite groups.N /: This map Â depends only on the isomorphism class of the extension.Q. In particular. the group G is also commutative. T HE CLASSIFICATION OF FINITE SIMPLE GROUPS There is a complete list of ﬁnite simple groups. is built into the fabric of the universe. Otto H¨ lder.Fq / D fm m matrices A with entries in Fq such that det A D 1g: def .N / D Aut. As an example of a matrix group. consider SLm . and hence by gcd. This proves the SchurZassenhaus theorem in this case. N / is killed by m and by n. and the 26 “sporadic groups”. and there is a commutative group structure on the set Ext1 . and we write Ext1 .50 3.N / depends only on q. Moreover. but the 26 sporadic groups are of more interest than their small number might suggest. They are (a) (b) (c) (d) the cyclic groups of prime order. Q Q therefore we get a homomorphism Â W Q ! Out. A. def The H¨ lder program.Q.Q. the alternating groups An for n 5 (see the next chapter). 1892 o Recall that a group G is simple if it contains no normal subgroup except 1 and G. Ann.
However. are “close” either to simple groups or to groups such as dihedral groups. As we noted earlier.Fq /=fcentreg are simple when m 3 (Rotman 1995. but an old conjecture said that. 347): . except for the cyclic groups of prime order. these have been closed. and the second smallest is PSL3 . the determination of the ﬁnite simple groups with an involution centralizer isomorphic to H is a ﬁnite problem. in reworking the proof. when work of Brauer and others suggested that the key was to study the centralizers of elements of order 2 (the involution centralizers). Exercises 31 Let G be the quaternion group (1. .13). . This group is not simple if q 0 0 0 B0 0C B C m q ¤ 2. Of course.Fq / D SLn .Exercises 51 Here q D p n . . the effort to complete the classiﬁcation of the ﬁnite simple groups began in earnest. def B T HE CLASSIFICATION OF ALL EXTENSIONS OF FINITE GROUPS Much is known about the extensions of ﬁnite groups.. are in the centre for any m C. but these are the only matrices in the centre. Later work showed that the problem is even tractable. p prime. for any ﬁnite group H . 8.F2 /. . in fact.13). 8. With the proof of this conjecture by Feit and Thompson (1963). Howo ever a clear strategy for accomplishing this did not begin to emerge until the 1950s. and so the strategy became: (a) list the groups H that are candidates for being an involution centralizer in some ﬁnite simple group. this approach applies only to the ﬁnite simple groups containing an element of order 2. However. the classiﬁcation of all ﬁnite groups is completely infeasible. but there remained sceptics. as Solomon writes (2001. because the proof depended on thousands of pages of rarely read journal articles. see the book Ronan 2006. a complete irredundant list of ﬁnite groups was available only for those up to an order of about 2000. and. :: @ : A 0 0 dividing q 1. Brauer and Fowler (1955) showed that.23) and when m D 2 and q > 3 (ibid. The smallest noncommutative group is A5 . about the extensions of one simple group by another. and Fq is “the” ﬁeld with 1 elements. For example.8). gaps were discovered. because the scalar matricesB D 1.18). . and the number of groups on the list is overwhelming. Heisenberg groups. and with the publication of Aschbacher and Smith 2004 it has become generally accepted that the proof of the classiﬁcation is indeed complete. Other ﬁnite simple groups can be obtained from the groups in (1. A complete classiﬁcation was announced in 1982. N OTES The dream of classifying the ﬁnite simple groups goes back at least to H¨ lder 1892. and the groups PSLm . for example. which as order 168 (see Exercise 47). etc. and for a more technical account. p. every ﬁnite simple group has even order and hence contains an element of order 2 by Cauchy’s theorem (4. see the expository article Solomon 2001. which arise naturally in the study of simple groups. by the year 2001. Nevertheless experience shows that most of the ﬁnite groups which occur in “nature” . For a popular account of the history of the classiﬁcation. Prove that G can’t be written as a semidirect product in any nontrivial fashion. and (b) for each H in (a) list the ﬁnite simple groups for which H occurs as an involution centralizer.
Prove that Aut. : : : .n/).g/ is independent of o. Q is cyclic of order 2.q//: (c) Let G D S5 D A5 Q with Q D h.R/.q/ D 2.1. : : : . Let n D . : : : . Show that Z. and the generator q of Q acts on N by sending each element to its inverse. an /q 1 D . prove that G is the direct product of M and N .Q/. (d) Suppose that G D . Deduce that . o.g/ D 6. 0/. Show that if N is commutative.1.Q/ \ Ker. nn0 n 1 Q of its subgroups N and Q. Is it a direct product of these two groups? 36 Find the automorphism groups of C1 and S3 . then Z. 1/ has order 2 no matter what n is (in particular.g/ D lcm.N / giving the action of Q on N (by conjugation).q/ in this case).q/ for some divisor k of jN j.n.x/: (a) Show that o. 0.. : : : .Q/.1.g/ D k o. 37 Let G D N Q where N and Q are ﬁnite groups. 34 Let G be the quaternion group (1.g/ D o. : : : .. an 1 /: Show inductively that. o. : : : .F2 / S3 . and if N and Q are commutative.n/ D 5.G/ D fn q j n 2 CN . AUTOMORPHISMS AND E XTENSIONS 32 Let G be a group of order mn where m and n have no common factor. 5/ and let q D .1.R/ of the form @0 a c A.Cp /p Q where Q is cyclic of order p and that. show that o.Q/ D fn 2 N j nq D q n for all q 2 Qg Dq 1 0 n q for all n0 2 N g: Let Â be the homomorphism Q ! Aut.52 3.o. (b) When Q acts trivially on N ..18). q.n/. . 38 Let G be the semidirect G D N (centralizer of Q in N ). and o. and let g D nq be an element of G with n 2 N and q 2 Q. for some generator q of Q. q i / (i copies of 1).Q/. for i Ä p. 0.Â /g.1. (e) Suppose that G D N Q where N is commutative. ad ¤ 0.n/ o.G/ D fn q j n 2 CN . o. q/i D . S4 .1. a1 . q 2 Ker. Show that o.G/ 0 1 a 0 b 35 Let G be the set of all matrices in GL3 . 2/. q 2 Z. Check 0 0 d that G is a subgroup of GL3 . 33 Prove that GL2 .Q/.an . If G contains exactly one subgroup M of order m and exactly one subgroup N of order n.G/ D fn q j n 2 CN . 3. 2. Denote the order of an element x by o. then Z. 0/ .Â /g: . Show that . 0/. 4. q/ has order p 2 (hence o. 0. 1. q 2 Z. 2/i. and prove that it is a semidirect product of R2 (additive group) by R R . and let CN .a1 .
x 7! gx. all g1 .X /.g1 g2 /x D g1 . left translation by g. from a left action of G on X. i.X / (18) is a homomorphism. Thus.e. gL 2 Sym.g. 2. . x/ 7! g x D gxg 53 def 1 : . . for all x 2 XI (b) . every such homomorphism deﬁnes an action of G on X. (d) Every group G acts on itself by conjugation. An action is trivial if gx D x for all g 2 G. for example. x/ 7! hx: (c) Let H be a subgroup of G. has .C HAPTER 4 Groups Acting on Sets Deﬁnition and examples D EFINITION 4. The conditions imply that. and therefore gL is a bijection. g2 2 G.h. x/ 7! gxW G X ! X such that (a) 1x D x. G G=H ! G=H. The action is said to be faithful (or effective) if the homomorphism (18) is injective. G G ! G. H G ! G. gL W X ! X.. x 2 X: A set together with a (left) action of G is called a (left) Gset.g.g. :::. .e. for each g 2 G. Axiom (b) now says that g 7! gL W G ! Sym.1 Let X be a set and let G be a group. C / 7! gC: The action is faithful if.. i.g2 x/.2 (a) Every subgroup of the symmetric group Sn acts faithfully on f1. we obtain a homomorphism G ! Sym.g 1 /L as an inverse. ng. The group G acts on the set of left cosets of H . if gx D x for all x 2 X H) g D 1: E XAMPLE 4. A left action of G on X is a mapping . (b) Every subgroup H of a group G acts faithfully on G by left translation.X /I conversely. H ¤ G and G is simple.
E XAMPLE 4. This relation is reﬂexive because x D 1x. x 2 S H) gx 2 S: The action of G on X then induces an action of G on S . there is a natural right action of G on the set of right cosets of a subgroup H in G. A map of Gsets (alternatively.G/ acts on G: (f) The group of rigid motions of Rn is the group of bijections Rn ! Rn preserving lengths. C / 7! Cg 1 . Thus the Gorbits partition X . namely. a Gmap or a Gequivariant map) is a map 'W X ! Y such that '. It acts on Rn on the left. Note that the group law on G will not induce a group law on G=H unless H is normal. A right action X G ! G is deﬁned similarly. G ROUPS ACTING ON S ETS For any normal subgroup N .) (b) The orbits for a subgroup H of G acting on G by left multiplication are the right cosets of H in G. ˛x0 .g. G acts on N and G=N by conjugation.54 4. its inverse is then also a Gmap. z D g 0 y H) z D g 0 . x 2 X: An isomorphism of Gsets is a bijective Gmap. set g x D xg 1 . : : : . and let ˛ 2 G be an element of order n. the conjugacy class of x is the set fgxg 1 j g 2 Gg . The equivalence classes are called Gorbits. Aut. and we write G=H for the set of left cosets. the orbits are called conjugacy classes: for x 2 G.g 0 g/x: It is therefore an equivalence relation. . O RBITS Let G act on X.C. ˛ n 1 x0 g: (These elements need not be distinct. and transitive because y D gx. We write H nG for the set of right cosets. the Gorbit containing x0 is Gx0 D fgx0 j g 2 Gg: It is the smallest Gstable subset of X containing x0 . For example. the orbits for H acting by right multiplication are the left cosets. By deﬁnition. To turn a right action into a left action. which can be turned into a left action . Write GnX for the set of orbits. some g 2 G. Write x G y if y D gx. all g 2 G. (c) For a group G acting on itself by conjugation.x/. symmetric because y D gx H) x D g 1 y (multiply by g 1 on the left and use the axioms). Similarly.gx/ D . g/ 7! Cg.gx/ D g'. and so the set may contain fewer than n elements. A subset S X is said to be stable under the action of G if g 2 G. Then the orbits of h˛i are the sets of the form fx0 . (e) For any group G.3 (a) Suppose G acts on X .
gx/ D g 1.x/ D fg 2 G j gx D xgg: .X //. The action is faithful if and only if E XAMPLE 4. Sn acts transitively on f1. In linear algebra the conjugacy classes in G D GLn . and it consists only of x0 if and only if x0 is in the centre of G. The set X is then called a homogeneous Gset. 2 and so g 1 g0 g 2 Stab. but the action of G on itself is never transitive if G ¤ 1 because f1g is always a conjugacy class.e. For example. Then Stab. Note that a subset of X is stable if and only if it is a union of orbits.gg 0 g and so g Stab. there exists a (single) g 2 G such that gx1 D y1 and gx2 D y2 . for any two elements x and y of X. y2 / of elements of X with x1 ¤ x2 and y1 ¤ y2 . but it need not be a normal subgroup (see the next lemma). L EMMA 4. Certainly. x2 /. i. For example. then . then .gx/.x/ D fg 2 G j gx D xg: It is a subgroup. . The action is free if Stab.x/. if there is only one orbit. 1 y D x. The stabilizer (or isotropy group) of an element x 2 X is Stab.x/ g P ROOF. S TABILIZERS Let G act on X.y1 . and G is said to act transitively on X .x/ g Clearly \ x2X Stab. Deﬁne kfold transitivity for k 3 similarly.. Stab. Conversely. if g 0 . For any subgroup H of a group G.G ! Sym..gx/ D g Stab. a subgroup H of G is normal if and only if it is a union of conjugacy classes.4 For any g 2 G and x 2 X.x/ D feg for all x. Stab. if g 0 x D x. g 0 2 g Stab. T Stab. and the theory of rational canonical forms provides a set of representatives for the conjugacy classes: two matrices are similar (conjugate) if and only if they have the same rational canonical form. The action of G on X is doubly transitive if for any two pairs . 2. which is a normal subgroup of G. there exists a g 2 G such that gx D y.x/ D f1g.Deﬁnition and examples 55 of conjugates of x. The action of G on X is said to be transitive. The conjugacy class of x0 always contains x0 .x1 . :::ng.x/ D Ker.g 1 0 g g/x D g 1 0 g .5 (a) Let G act on itself by conjugation. i.x/ g 1 1 1 : /gx D gg 0 x D gx D y.k/ are called similarity classes.e.gx/ D gx. G acts transitively on G=H .
The intersection \ CG .H / of H in G: NG . the Gset G=H has a preferred point. i. Let T ' .4) shows that Stab. the map g Stab.16).H / is the largest subgroup of G containing H as a normal subgroup. It is possible for gS S but g 2 Stab. Then T is a normal subgroup of G and G ' T O (cf.S / is a subgroup of G. if h 2 Stab.S / D fg 2 G j gS D S g: Then Stab.2f). the coset H . i.x/ D fg 2 G j gx D xg for all x 2 Gg x2G is the centre of G. centralize.33). 4. 5. g 0 lie in the same left coset of Stab.x0 /. then ghx0 D gx0 . x. It is welldeﬁned because. P ROOF.x0 / 7! gx0 W G= Stab. Chap. T RANSITIVE ACTIONS P ROPOSITION 4.H / D H .e.7 If G acts transitively on X .S / g 1 : E XAMPLE 4. C/ be the subgroup of G of translations of Rn . . and the same argument as in the proof of (4. Chap.x0 / ! X is an isomorphism of Gsets. (b) Let G act on G=H by left multiplication.56 4.e. For a subset S of X . It is injective because gx0 D g 0 x0 H) g 1 0 g x0 D x0 H) g. Finally. maps of the form v 7! v C v0 some v0 2 Rn . namely. we deﬁne the stabilizer of S to be Stab. 5. G ROUPS ACTING ON S ETS This group is called the centralizer CG . Thus every homogeneous Gset X is isomorphic to G=H for some subgroup H of G.Rn . but such a realization of X is not canonical: it depends on the choice of x0 2 X: To say this another way. Then Stab.S / (see 1.6 Let G act on G by conjugation..x/ of x in G. and the stabilizer of gH is gHg 1 : (c) Let G be the group of rigid motions of Rn (4. The stabilizer of the origin is the orthogonal group On for the standard positive deﬁnite form on Rn (Artin 1991. to give a homogeneous Gset X together with a preferred point is essentially the same as to give a subgroup of G. and let H be a subgroup of G. then for any x0 2 X.x0 /: 2 It is surjective because G acts transitively. It consists of all elements of G that commute with. Artin 1991.H / D fg 2 G j gHg 1 D H g: Clearly NG . 2).gS / D g Stab.. The stabilizer of H is called the normalizer NG . it is obviously Gequivariant.
gx0 / D (4. Thus N0 is a normal subgroup of G contained in eHe 1 D H .x/ D \ g2G Stab. then Ker. P ROOF. and let O D Gx0 be the orbit containing x0 .x0 / g 1 : Hence.10 For any subgroup H of a group G. so also does g1 g. 2 The equation (19) is frequently useful for computing jOj. subgroup contained in H . When Ker.Deﬁnition and examples 57 C OROLLARY 4. Then the cardinality of O is jOj D .8 Let G act on X .9 Let x0 2 X. P ROOF. the number of conjugates gHg 1 of a subgroup H of G is . It is normal because 2 1 T g2G gHg is the largest normal 1.GW NG . If G acts transitively on X . and so N \ g2G gHg 1 D N0 : 2 . being an intersection of subgroups. def T P ROOF.g1 g/ D N0 — for the second equality. as g runs over the elements of G.x0 /.H //. then N D gNg 1 gHg 1 for all g 2 G.G ! Sym.G ! Sym.g1 g/N0 . L EMMA 4. we used that. The action of G on O is transitive.4) \ g Stab.x0 / ! Gx0 .X // D \ x2X Stab. and so g 7! gx0 deﬁnes a bijection G= Stab. Note that N0 D g2G gHg group.G W Stab. If N is a second such group. is itself a sub1 g1 N0 g1 1 D \ g2G .X // is the largest normal subgroup contained in Stab.x0 //: (19) For example. the proposition is a consequence of the following lemma. P ROPOSITION 4.
y runs over set of representatives for the conjugacy classes containing more than one element).y//.G W CG . Thus we may suppose that p divides all of the terms . 1 Here . P ROOF. is a direct proof that the theorem holds for an abelian group Z.G/. P ROOF. xi in Oi : (20) When G acts on itself by conjugation. is cyclic or dihedral. it is a disjoint union of a ﬁnite number of orbits: XD Hence: P ROPOSITION 4. If for some y not in the centre of G. 2 C OROLLARY 4. It sufﬁces to show that Z contains an element whose order is divisible by p. p an odd prime.y// in the class equation (second form).G/j C .26) shows that the order of every element is a power of p.xi //.13 (C AUCHY ) If the prime p divides jGj. and it follows from the structure theorem1 of such groups that Z. then Lagrange’s theorem (1.G W Stab. T HEOREM 4.G W CG . because then some power of the element will have order exactly p.15 Every group of order 2p. or X jGj D jZ. p does not divide .y/.x// jGj D . in which case there exists (by induction) an element of G whose order in Z=hgi is divisible by p. If p doesn’t divide the order of g.58 4. G ROUPS ACTING ON S ETS T HE CLASS EQUATION When X is ﬁnite.x runs over a set of representatives for the conjugacy classes). We use induction on the order of Z. and so also divides Z.G/ will contain an element of order p. then it divides the order of Z=hgi. If jGj is a power of p.G/ is commutative. We use induction on jGj.14 A ﬁnite group G is a pgroup if and only if every element has order a power of p.2 C OROLLARY 4.12 (C LASS EQUATION ) X . this formula becomes: P ROPOSITION 4. then pjCG . then G contains an element of order p.11 The number of elements in X is jXj D m X i D1 m [ i D1 Oi (disjoint union): jOi j D m X i D1 . The converse follows from Cauchy’s theorem. Let g ¤ 1 be an element of Z.G W CG . But the order of such an element must itself be divisible by p.y/ and we can apply induction to ﬁnd an element of order p in CG .y// (21) (22) .G W CG . But Z.
Then G is commutative. s.G/ 2 R EMARK 4.17 A group of order p n has normal subgroups of order p m for all m Ä n. Because Z=pZ is a ﬁeld. 2 p. From Cauchy’s theorem. r. By assumption. and so N D hgi is a normal subgroup of G of order p. We use induction on n.G/j also.19 Suppose G contains a subgroup H in its centre (hence H is normal) such that G=H is cyclic. and that G=Z therefore has order 1 or p.G W 1/ is a power of p. . the group is commutative (any group generated by a set of commuting elements is obviously commutative). rs. and the next result implies that G is commutative.20 The above proof shows that if H Z. then G is commutative.Deﬁnition and examples 59 P ROOF. 2 L EMMA 4. We know that the centre Z is nontrivial. Now a i h a i h0 0 D ai ai hh0 0 D ai ai h0 h 0 0 D ai h ai h: 0 because H Z. . As p divides every term in the class equation (22) except (perhaps) jZ. Then H is of index 2. its only elements with square 1 are ˙1.GROUPS T HEOREM 4. Now the induction hypothesis allows us to assume the result for G=N. The centre of G contains an element g of order p. i 2 Z. and hence is isomorphic to Cp Cp or Cp2 . and so . Obviously s … H . r p 1 sg: 2 As H is normal.y// is power of p (¤ p 0 ) for all y not in the centre of G. P ROOF. : : : .47) then gives it to us for G: 2 P ROPOSITION 4. P ROOF.9). Then every element of G can be written g D ai h with h 2 H .18 Every group of order p 2 is commutative. r D s 2 rs 2 D s. In either case it is cyclic.G/ and G contains a set of representatives for G=H whose elements commute. we know that such a G contains elements s and r of orders 2 and p respectively. and so i Á 1 or 1 mod p. : : : . in the second srs 1 D r 1 and we have the dihedral group (2. Because s 2 D 1. and the correspondence theorem (1. it must divide jZ.G W CG . and so is normal. 2 C OROLLARY 4. Let a be an element of G whose image in G=H generates it. and so i 2 Á 1 mod p. Let H D hri. srs 1 D r i . some i . r p 1 . P ROOF.G/j. P ROOF.16 Every nontrivial ﬁnite pgroup has nontrivial centre. In the ﬁrst case.srs 1 /s 1 D r i . and so G D H [ H s W G D f1.
i.7b). : : : . Then G must contain an element a of order 4 (see Exercise 16). According to Cauchy’s theorem (4.Z=4Z/ . E XAMPLE 4. In fact. ng. In the second case. Thus. G ROUPS ACTING ON S ETS For p odd. but (by deﬁnition) it has no such normal subgroup. E XAMPLE 4. Moreover def N D hri must be normal because 6 doesn’t divide 2Š (or simply because it has index 2).15) (Exercise 43).21 Let G be a noncommutative group of order 8.14.13). Let X D G=H . 2. G D N Q. (b) If . then G ' hai Â hbi where Â is the unique isomorphism Z=2Z ! . R EMARK 4. we may take X D f1. In the ﬁrst case.G W 1/ does not divide . for any g 2 G. and so bab 1 D a3 .X / depends only on the number of elements in X. Either (a) H is normal in G. G ! Sym. Now bab 1 is an element of order 4 in hai. hence surjective. 4.13). 2.G=H / is injective. and a2 D b 2 . Permutation groups Consider Sym. Since (up to isomorphism) a symmetry group Sym. Therefore G is the quaternion group (1.e. If G contains an element b of order 2 not in hai. Stab.. rsr 1 D s. 1. then it will have a subgroup N of order 11 (Cauchy’s theorem.22 (a) Let H be a subgroup of G not containing a normal subgroup of G other than 1. Here we present a slightly different argument. then G ! Sym. For example.X / is the largest normal subgroup T g2G gHg 1 1 D gHg 1 of G contained in H . rs D sr. because otherwise G would be commutative. any element b of G not in hai must have order 4. ACTION ON THE LEFT COSETS The action of G on the set of left cosets G=H of H in G is a very useful tool in the study of groups. For example.H /g and the kernel of G ! Sym. there are exactly two noncommutative groups of order p 3 . and so G S3 D3 .60 4. and we have realized G as a subgroup of a symmetric group of order much smaller than .X / where X has n elements.26). and we can conclude that H contains a normal subgroup ¤ 1 of G. . Let H D hsi.gH / D g Stab. then the Sylow theorems (see Chapter 5) show that G has many proper subgroups H ¤ 1 (unless G is cyclic).G=H / is injective. if G has order 99.G W H /Š. Recall that. Then G ! Sym.18. and so G ' hri hsi C2 C3 . up to isomorphism. it is now not difﬁcult to show that any noncommutative group of order p 3 is isomorphic to exactly one of the groups constructed in (3. 3. and the subgroup must be normal. If not.23 Corollary 4. if G is simple.G W 1/Š.15 shows that every group G of order 6 is either cyclic or dihedral. It can’t equal a. or (b) H is not normal in G. G must contain an element r of order 3 and an element s of order 2. We illustrate this with some examples. and so G D4 .G=H / can’t be injective (Lagrange’s theorem.
24 To compute the signature of . 1 3 2 5 3 1 4 4 5 2 is even (6 intersections). : : : . 2 Each 2 Sn .2/ .n/ The pairs .3/ . and count the number of times that the lines cross: is even or odd according as this number is even or odd.z1 .j / .n 1/ 2/ 1// . sign.i / . z .3/ .n . and is said to be even or odd according as the number its inversions is even or odd.Permutation groups 61 and so work with Sn .i. the factor corresponding to the subset fi.1/ . let f be . j / with i < j and . For example. let x be the element of Zn such that .n . etc. / D 1 if is a transposition. consider the products Y V D . Then . Consider a permutation Â Ã 1 2 3 ::: n D : . By deﬁnition. 2.1// . 2574316 2 7! 5.2 1/. connect (by a line) each element i in the top row to the element i in the bottom row.z For . This works.n .x / D x .2// 1//: V D Y .x/ D . The symbol 1 2 3 4 5 6 7 denotes the permutation sending 1 7! 2.x / D f .1// . .2// .n/ The terms in the products are the same except that each inversion introduces a negative sign. 2 Sn .i / > . of is C1 or 1 according as is even or odd.x / / D f .j i / D .x /..n/ . one ﬁnds that3 .1/ .. /.x /i D x . : : : . because there is one crossing for each inversion.x / D . For a permutation . .2/ 1Äi <j Än .j / are called the inversions of . . f /.j i/.3/ : : : . . 3 7! 7. /V: Now let P be the additive group of maps Zn ! Z. /f / .x/ D f .n/ . ng.x/: . f //. 3 For x 2 Zn and 2 Sn . For example. j g is ˙. The signature. .2 Therefore.n . ..3 1Äi <j Än .n/ /: /f: (23) is a product over the 2element subsets of f1.3 1/ 2/ . .i // D . R EMARK 4. f /.1//. V D sign. : : : . f /. f/D. For f 2 P and the element of P deﬁned by . . sign. zn / D f . . Therefore ..
P ROOF. the parity of the number of inversions of is independent of the choice of the ordering on X . A cycle of length 2 is a transposition. When n 2.l1 :::lu / . zn / D The same argument as above shows that p D sign. / D 1 for every transposition . (disjoint cycles). In other words. /: Y . Can you prove this directly? A cycle is a permutation of the following form i1 7! i2 7! i3 7! 7! ir 7! i1 . 2. and let O f1. / D sign. remaining i ’s ﬁxed.i /g: have the same action on any element of O. The support of the cycle .i . The ij are required to be distinct.zj zi /: 1Äi <j Än Therefore. it is independent of the choice of the ordering. then for any O D fi.l1 :::lu /m (disjoint cycles).z1 . In particular.j1 :::js / then m .X / where X is a set with n elements. and cycles are said to be disjoint if their supports are disjoint.i1 :::ir /m . Now let G D Sym. one ﬁnds that sign.26 Every permutation can be written (in essentially one way) as a product of disjoint cycles. and so its kernel is a normal subgroup of Sn of order nŠ .25 Clearly sign is the unique homomorphism Sn ! f˙1g such that sign. called the alternating group An . “sign” is a homomorphism Sn ! f˙1g. Note that disjoint cycles commute. / sign. and the cycle . m [ j D1 f1.62 4.r. : : : . : : : . it is surjective. and call r its length — note that r is also its order as an element of Sn . ng D Oj .i1 :::ir /.i / : : : r 1 . /p: On putting f D p in (23) and using that p ¤ 0. ir g. 2 R EMARK 4.i // r 1 . We denote this cycle by .i1 : : : ir / is the set fi1 . D . If jOj D r. : : : .i /. u/: P ROPOSITION 4. Deﬁne ". ng be an orbit for h i. / D 1 for every transposition . Once we have chosen an ordering of X. s. Therefore Let 2 Sn . / to be C1 or 1 according as has an even or an odd number of inversions. .i / of length 1 is the identity map. A cycle .j1 :::js /m and it follows that has order lcm. The same arguments as above show that " is the unique homomorphism G ! f˙1g such that ". : : : . 2. If D . Let i 2 O. : : : . G ROUPS ACTING ON S ETS Let p be the element of P deﬁned by p. we can speak of the inversions of an element of G. :::.i1 i2 :::ir /.
change the order of the cycles.8/: 5 7 4 2 1 3 6 8 (24) It has order lcm. but the product of two transpositions can always be written as a product of 3cycles: 8 ˆ. but that’s all because the orbits are intrinsically attached to : 2 For example. : : : . / D . C OROLLARY 4. it is useful to determine the conjugacy classes in G. C OROLLARY 4. . 2 Note that the formula in the proof shows that the signature of a cycle of length r is .ir 2 ir 1 /.ij k/. 1/r 1 . < .27634/. 5/ D 10. 1/#transpositions . ng into a disjoint union of orbits for h i. :::. k. note that a decomposition D 1 m into a product of disjoint cycles must correspond to a decomposition of f1.ir 1 ir /.j kl/ case i. Because sign is a homomorphism. and the signature of a transposition is 1.ij /.i1 i2 / . The cycle . The corollary says that Sn is generated by transpositions.15/. We can drop cycles of length one. and change how we write each cycle (by choosing different initial elements). For the uniqueness.kl/: 2 Recall that two elements a and b of a group G are said to be conjugate a b if there exists an element g 2 G such that b D gag 1 .j k/.i1 i2 :::ir / D . P ROOF. the number of transpositions in such a product is even or odd according as is even or odd. It is possible to deﬁne a permutation to be even or odd according as it is a product of an even or odd number of transpositions.j k/. ng into orbits (ignoring cycles of length 1 and orbits with only one element). j. Any 2 An is the product (possibly empty) of an even number of transpositions.2. an rcycle is even or odd according as r is odd or even. and let cycle associated (as above) with Oj .27 Each permutation can be written as a product of transpositions.kl/ D . Â Ã 1 2 3 4 5 6 7 8 D . but then one has to go through an argument as above to show that this is a welldeﬁned notion.kl/ D . For An there is the following result. 0 0 D t1 t1 tm tm . P ROOF. and so the ﬁrst statement follows from the proposition. ˆ : 1 case .ij l/ case j D k.ij /. Then D 1 m 63 be the j is a decomposition of into a product of disjoint cycles.Permutation groups be the decomposition of f1.28 The alternating group An is generated by cycles of length three.ij / D . sign. For a group G.j l/ D .ij /. and that conjugacy is an equivalence relation. that is. l distinct.
4 D 1 C 3. 4 D 1 C 1 C 2.30 Two elements the same partitions of n. because conjugation is a homomorphism). the partition of 8 attached to .29 In Sn . 2. : : : . 1.l1 : : : lu /: If we deﬁne g to be Â i1 0 i1 ir 0 ir j1 0 j1 g g 1 js js0 D : l1 0 l1 Ã lu 0 .j1 : : : js0 / : : : .i1 / : : : g. 4 D 2 C 2.g. : : : .g. 4 D 4. : : : . 2. By a partition of n. and 1. For example.15/. H) W We saw in (4.ik //: Hence g. : : : .i1 / : : : g. the conjugate of a cycle is given by: g. nk . their decompositions into products of disjoint cycles have the same type: D .i1 : : : in1 / is 1.l1 : : : lnk /.i1 : : : ik /g 1 D .n ni ones/: and of Sn are conjugate if and only if they deﬁne P ROPOSITION 4.64 4.i1 : : : ir /. 505 partitions of 61. we can attach to each such a partition of n.j1 : : : js / : : : . we mean a sequence of integers n1 . ng D O1 [ ::: [ Ok of f1. 2. In other words.i1 : : : ir /.i1 : : : ir / .ir // . G ROUPS ACTING ON S ETS E XAMPLE 4. n D n1 C n2 C ::: C nk . namely.29) that conjugating an element preserves the type of its disjoint cycle decomposition. there are exactly 5 partitions of 4.l1 : : : lu /. 4 D 1 C 1 C 1 C 1.27634/.l1 : : : lu /g 1 D . to obtain g g 1 . : : : . (H W Since and deﬁne the same partitions of n. ni D jOi j. nk such that 1 Ä n1 Ä n2 Ä n1 C n2 C Ä nk Ä n and C nk D n: For example.g. ng. Note that a partition f1.l1 /:::g. P ROOF. lu 2 then .8/ is 1. :::. 5 and the partition attached of n attached to D . 121. n1 .lu // (even if the cycles are not disjoint. Since the orbits of an element of Sn form a partition of f1. 1. (disjoint union) provided the numbering has been chosen so that jOi j Ä jOi C1 j. (disjoint cycles) 1 < ni Ä ni C1 . 0 0 0 0 0 D . ng determines a partition of n. replace each element in each cycle of by its image under g: We shall now determine the conjugacy classes in Sn . P .
ab/. For S4 we have the 2 following table: Partition Element No. Class Parity 1C1C1C1 1 1 even 1C1C2 . the number of permutations in S4 of type . and that together with 1 they form a subgroup V . T HEOREM 4. and let be a second cycle of length three in An . and for n D 3.28).31 .ij k/ D .36 Every normal subgroup N of An . The 1 k is . subgroup V (see 4. n 3. 5.cd /. and so equals An (by 4. An is trivial.33 (G ALOIS ) The group An is simple if n 5 R EMARK 4. If g 2 An .Permutation groups E XAMPLE 4. and hence simple. P ROOF. For example.n kC1/ k distinct kcycles in Sn .abc/ 8 even 2C2 .30) that D g g 1 for some g 2 Sn . there exists a transposition t 2 Sn disjoint from . 2 The next lemma completes the proof of the Theorem. it is possible to compute the number of elements in any conjugacy class in Sn . Then tg 2 An and D t t 1 D tg g 1 t 1 .ab/.ab/ twice. and is therefore a normal subgroup of S4 . because n 5. there are needed so that we don’t count 1: 65 n. N ¤ 1.1234::: / ij k4::: ij k4::: R EMARK 4.cd / 3 even 4 .i1 i2 : : : ik / D .ik i1 : : : ik 1/ D ::: k times. and so again 2 N.32). if N contains a cycle of length three. then it contains all cycles of length three.1234::: /. This group is a union of conjugacy classes.35 Let N be a normal subgroup of An (n 5/.ab/. but a little care is needed when the partition of n has several terms equal. then this shows that is also in N .cd / is Â Ã 1 4 3 2 1 D 3: 2 2 2 The 1 is needed so that we don’t count . If not. namely those of type 2 C 2. An is cyclic of order 3.n 1/ .123/. Similarly. in Conj.abcd / 6 odd Note that A4 contains exactly 3 elements of order 2. Let be the cycle of length three in N . L EMMA 4. for n D 4 it is nonabelian and nonsimple — it contains the normal. contains a cycle of length . even characteristic.32 For 1 < k Ä n.34 For n D 2. L EMMA 4. We know from (4.cd / D .ab/ 6 odd 1C3 .
i1 i2 i3 /. solve the problem of determining the order of a ﬁnite group from a ﬁnite presentation of the group (use the algorithm with H the trivial subgroup 1. i5 ﬁxed by — it therefore ﬁxes more elements than . multiplication. then we can apply the same construction. Therefore either N \ An D An or N \ An D f1g. ¤ 1. In the second case. form . and Sn . and so N D An or Sn . G ROUPS ACTING ON S ETS P ROOF. In Galois theory. :::. and all elements it acts differently on i4 . Suppose is not a 3cycle. The ToddCoxeter algorithm does not solve the problem of determining whether a ﬁnite presentation deﬁnes a ﬁnite group (in fact.i1 : : : i4 /. and the extraction of mth roots if and only if Gf is solvable (Galois’s theorem). Let D . i5 not ﬁxed by — it therefore ﬁxes at least one more element than . and hence (when n 5) lots of polynomials not solvable in radicals.i i i : : :/ . The ToddCoxeter algorithm. 0 ¤ 1. When we express it as a product of disjoint cycles. In the ﬁrst case. N An . i2 . Thus 1 1 2 ¤ i1 .i1 i2 /. and is distinct 1 2 4 def 0D 1 ¤ 1. 2. there exist lots of polynomials f of degree n with Gf Sn . we arrive at a 3cycle. 2 C OROLLARY 4. say (i) (ii) D . If 0 is not a 3cycle. P ROOF. i5 . the map x 7! xAn W N ! Sn =An is injective. In the second case. Thus An (also Sn / is not solvable if n 5. An . then N \ An is normal in An . It does. and let H be a subgroup described by a generating set. there is no such algorithm). Let G be a group described by a ﬁnite presentation.66 4. Thus 1 i2 and all elements other than i1 . For every n. i3 . ng than does . but it can’t have order 2 because no conjugacy class in Sn (other than f1g) consists of a single element. however. A SIDE 4. say i4 . If N is normal in Sn . subtraction. Then 1 D . other than i1 . i2 .i3 i4 i5 /. Then the ToddCoxeter algorithm4 is a strategy for writing down the set of solve a problem. the only normal subgroups of Sn are 1. moves two numbers. either it contains a cycle of length 3 or else it is a product of transpositions. which ﬁxes more elements of f1. : : : . division. Let f .i4 i5 / is distinct from because 0 D 1 ¤ 1. one attaches to f a subgroup Gf of the group of permutations of the roots of f . . but 0 ﬁxes i and i . Let 2 N . :::.) 4 To . which has index 2 in Sn . and so N has order 1 or 2. but 0 D 1 1 ﬁxes from (because it acts differently on i2 ). an algorithm must always terminate in a ﬁnite time with the correct answer to the problem.i1 i2 /. i3 . If is not a 3cycle.38 There exists a description of the conjugacy classes in An .i3 i4 / . from which it is possible to deduce its simplicity for n 5 (see Exercise 411).37 For n 5. and shows that the roots of f can be obtained from the coefﬁcients of f by the algebraic operations of addition. 2 A SIDE 4.i1 i2 i3 :::/ or D . After a ﬁnite number of steps. because ¤ def 1 D .39 A group G is said to be solvable if there exist subgroups G D G0 G1 Gi 1 Gi Gr D f1g such that each Gi is normal in Gi 1 and each quotient Gi 1 =Gi is commutative. i4 . In the ﬁrst case. we shall construct another element 0 2 N . Then 1 D 2 N . 0 as in the ﬁrst case with i4 as in (ii) and i5 any element distinct from i1 . 1 .X / 2 QŒX be of degree n.
cba in our example) act trivially. 2. Similarly. What we know can be summarized by the table: a 1 2 3 2 3 1 a 3 1 2 a 1 2 3 b 2 1 b 1 2 3 c 1 c 1 2 3 a 2 3 1 b 1 2 c 1 2 3 b b The bottom right corner. and this then determines the rest of the table: a 1 2 3 2 3 1 a 3 1 2 a 1 2 3 b 2 1 3 b 1 2 3 c 1 3 2 c 1 2 3 a 2 3 1 b 1 3 2 c 1 2 3 We ﬁnd that we have three cosets on which a.123/ b D . Since we don’t know what a1 is. and so it is prudent to introduce another coset 4 D b1. but note that he composes permutations in the reverse direction from us).The ToddCoxeter algorithm. let a2 D 3. Therefore. b. denoted 1. Now b4 D 1 because b 2 D 1. 6. A theorem (Artin 1991. Rule (iii) tells us simply that c1 D c. b 2 . 9. we drop 5.23/: More precisely. I illustrate it with an example (from Artin 1991. : : : with 1 D 1H . according to (i) we must have 5 D 1. which provides more details. c act as a D . c 2 . cbai and let H be the subgroup generated by c (strictly speaking. which is forced by (ii). Since b4 D 1 we must have 4 D 2. but we don’t know what b2 is. and so we have 1 7! 4 7! 1: We still have the relation cba. b. we have introduced three (potential) cosets 1. Let G D ha. 3. We know a1 D 2. H is the subgroup generated by the element of G represented by the reduced word c). c in our example) acts as a permutation. and by (ii) applied to cba we have c5 D 1. The operation of G on the set of cosets is described by the action of the generators. as necessary. The strategy is to introduce cosets. The operation on the cosets is transitive. let’s denote it 2: a1 D 2. 67 left cosets of H in G together with the action of G on the set.10) now says that this does in fact describe the action of . We now apply the ﬁrst two rules. c j a3 .9. which must satisfy the following rules: (i) (ii) (iii) (iv) Each generator (a. Now a3 D a3 1. and so we can drop 4 also. permuted by a as follows: a a a 1 7! 2 7! 3 7! 1: What is b1? We don’t know.12/ c D . c 2 . b 2 . b. The generators of H (c in our example) ﬁx the coset 1H . The relations (a3 . tells us that c2 D 3. which according to (ii) must be 1. 2. Thus. we have written down a map G ! S3 that is consistent with the above rules. and so we set b2 D 5: a b 1 7! 2 7! 5: By (iii) c1 D 1. and so now b2 D 1. Hence also c3 D 2.
2.X / (e. g2 A. and . In Artin 1991.32).12/. Primitive actions. The group G always stabilizes the trivial partitions of X . 4gg (opposite vertices stay opposite). for each g 2 G.42 The Gorbits form a partition of X that is stabilized by G. and Ker. g1 A. the set of all oneelement subsets of X . R EMARK 4. x 0 / to .g. with r D . namely. :::g of X. of Sn ) is said to be primitive if it acts primitively on X.: (25) is stabilized P ROOF. it is explained how to make this procedure into an algorithm which. will in fact produce the correct table. fy. 4g.S4 ! Sym. which is stabilized by G. E XAMPLE 4. (H: From such an A. when it succeeds in producing a consistent table. E XAMPLE 4. but Example 4. 2. 2 . 4gg of f1. 3. 3g. 4/. 3g. G is a ﬁnite group acting transitively on a set X with at least two elements. y/.23/ generate S3 . This algorithm is implemented in Gap — see the worksheet.x.1234/i of S4 stabilizes the partition ff1. regarded as a subgroup of S4 in the obvious way. we say that the action is primitive. each with two elements. Obviously. this shows that the action of G on G=H induces an isomorphism G ! S3 .40b shows that D4 . (b) Identify X D f1. Hence action primitive H) action transitive or trivial. Then S4 acts on X. .41 A doubly transitive action is primitive: if it stabilized ffx. G ROUPS ACTING ON S ETS G on G=H . (c) Let X be the set of partitions of f1. we can form a partition fA. :::g:::g. f2. For the remainder of this section.9. 3. namely.X // is the subgroup V deﬁned in (4.2. A subgroup of Sym. and fXg. Let G be a group acting on a set X. Then D4 stabilizes the partition ff1. If the action is primitive. x 0 . :::g.x.40 (a) The subgroup G D h. and let be a partition of X. 3. then there would be no element sending . H): The partition stabilized by G contains such an A. otherwise it is imprimitive. 6. Sn itself is primitive..1234/. is not primitive.68 4. f2.43 The group G acts imprimitively if and only if there is a proper subset A of X with at least 2 elements such that. 2.123/. Since the three elements . When it stabilizes only those partitions. We say that by G if A 2 H) gA 2 : It sufﬁces to check the condition for a set of generators for G. 4g into two sets. s D . either gA D A or gA \ A D . 4g with the set of vertices of the square on which D4 acts in the usual way. and that H is a subgroup of order 2. then the partition into orbits must be one of the trivial ones. P ROPOSITION 4.
G: 69 Stab. Then g 2 Stab.14. Because G acts transitively on X. S (b) Give an example of a proper subset S of a ﬁnite group G such that G D g2G gSg 1 . is isomorphic to one of the two groups constructed in (3.x/ H . We have Stab. contains a subgroup of index 2. Show that the maps of Gsets G=H1 ! G=H2 are in natural onetoone correspondence with the elements gH2 of G=H2 such that H1 gH2 g 1 . Show that any subgroup of G of index p is normal. 45 Show that a group of order 2m. There is a g 2 G such that gx D y. m odd. If g … H .45 The group G acts primitively on X if and only if. H) gA D A: Let y 2 A.A/: 2 T HEOREM 4.A/ Stab.G W 1/ (assumed ﬁnite). Stab.A/. and then g … Stab.x/: Let y … A. If g 2 H . suppose that there exists an x in X and a subgroup H such that Stab.x/ because gx D x H) gA \ A ¤ . 43 Prove that any noncommutative group of order p 3 . 3. P ROPOSITION 4.Exercises A subset A of X satisfying (25) is called block. then gA is disjoint from A: for suppose ghx D h0 x some h0 2 H . and g D h0 h00 h 1 2 H . then (see 4. y ¤ x. Because H ¤ Stab. Then I claim that A D H x is a block ¤ X with at least two elements. 44 Let p be the smallest prime dividing . For any x 2 A. then gA D A. p an odd prime. for one (hence all) x in X. but g … Stab.x/ H G. then h0 1 gh 2 Stab.x/.43) there is a block A X with at least two elements. there is a g 2 G such that gx D y.15).x/ P ROOF. H x ¤ fxg. If G does not act primitively on X . P ROOF.44 Let A be a block in X with jAj Stab. 2 Exercises 41 Let H1 and H2 be subgroups of a group G. 42 (a) Show that a ﬁnite group can’t be equal to the union of the conjugates of a proper subgroup.x/ is a maximal subgroup of G. and so fxg A X . Conversely. say h0 1 gh D h00 .A/ 2 and A ¤ X .44) shows that Stab.x/ will not be maximal for any x 2 A. and so (4. (Hint: Use Cayley’s theorem 1.22) .
If Aut. and determine jKj. . (b) If r D 2. or 4. (b) For each conjugacy class K in A7 . Prove: (a) If G is simple.! Sym. also by . prove that G is cyclic. where k D . 412 Let G be the group with generators a. .n 1 n/. Prove that K is a union of k conjugacy classes of equal size in H .G W 1/ divides rŠ. b and relations a4 D 1 D b 2 . [Note that if M is a free F2 Œ˛module of rank one. (a) Use the ToddCoxeter algorithm (with H D 1) to ﬁnd the image of G under the homomorphism G ! Sn . 1. 411 (a) Let 2 An . .] (b) Use Sage/Gap to check your answer.1 2/. [No need to include every step. (a) Show that . (Use 4. 413 Show that if the action of G on X is primitive and effective. and 3 elements of order 2.G W 1/ D 168. 414 (a) Check that A4 has 8 elements of order 3.) (c) Prove that A4 is the only subgroup of S4 of order 12.x// for any x 2 K. give a member of K. Hence it has no element of order 6. 56.G/ is cyclic.70 46 For n odd. prove that G is commutative.F2 /. G is ﬁnite. if further. then G can’t be simple. (b) Prove that A4 has no subgroup of order 6 (cf. : : : . 3. then EndF2 Œ˛ . 415 Let G be a group with a subgroup of index r.G W 1/. with 1. 4. n D .30). and 2 that there is a natural injective homomorphism G .M / D F2 Œ˛. show that X has 7 elements.1 3/. (c) There exists a nonabelian simple group with a subgroup of index 5. and 24 elements respectively.1 2/. 49 Show that Sn is generated by . 21. . then .2 3/. then the action of any normal subgroup H ¤ 1 of G is transitive. G ROUPS ACTING ON S ETS 5.1 n/. : : : . From Exercise 410 we know that the conjugacy class of in Sn either remains a single conjugacy class in An or breaks up as a union of two classes of equal size. (c) Use Jordan canonical forms to show that G has six conjugacy classes.11. Show that the second case occurs ” does not commute with an odd permutation ” the partition of n deﬁned by consists of distinct odd integers. aba D bab. 410 Let K be a conjugacy class of a ﬁnite group G contained in a normal subgroup H of G.X / D S7 . 24. . 48 Let G be a group. (b) Let X be the set of lines through the origin in F3 . just an outline will do. given by Cayley’s Theorem 1.23. show that the kcycles in Sn generate Sn or An according as k is even or 47 Let G D GL3 .G W H CG . 42.] (d) Deduce that G is simple.
h. Show that the orbits for this action are exactly the ﬁbres of the map . (a) Show that the double cosets of H and K in G partition G. 71 417 Let H and K be subgroups of a group G.hb. k/ D . (c) (Double coset counting formula). a 1 b 1 ak/. k 2 Kg for some a 2 G.Exercises 416 Prove that Sn is isomorphic to a subgroup of AnC2 . Use (b) to show that jHaKj D jH jjKj jH \ aKa 1j : . A double coset of H and K in G is a set of the form HaK D fhak j h 2 H .h. (b) Let H \ aKa 1 act on H K by b. k/ 7! hakW H K ! HaK.
.
C HAPTER 5 The Sylow Theorems. Applications In this chapter.mod p/: When the lemma is applied to a pgroup H acting on itself by conjugation. we ﬁnd that . and x0 2 O. The Sylow theorems In the proofs.H W Stab. the theorems restrict the possible number of Sylow psubgroups in G.G W 1/. Let G be a group and let p be a prime dividing . and that every psubgroup of G is contained in such a subgroup.Z.x0 / ! OI see (4. If p r j. Since X is a disjoint union of the orbits. The Sylow theorems state that there exist Sylow psubgroups for all primes p dividing . In other words.16). Therefore . or jOj is divisible by p.H / W 1/ Á . and let p be prime. all groups are ﬁnite. moreover.H /W 1/ (cf.2 (S YLOW I) Let G be a ﬁnite group. when H is a pgroup. h 7! hx0 induces a bijection H= Stab. then jXj Á jX H j . then G has a subgroup of order p r : mod p 73 .x0 // D jOj: In particular. jOj is a power of p: either O consists of a single element.GW 1/.GW 1/. that the Sylow psubgroups for a ﬁxed p are conjugate. H is a Sylow psubgroup of G if it is a pgroup and its index in G is prime to p.1 Let H be a pgroup acting on a ﬁnite set X. T HEOREM 5. and let X H be the set of points ﬁxed by H .7). A subgroup of G is called a Sylow psubgroup of G if its order is the highest power of p dividing . then the map H ! X. the proof of 4.G W 1/. we can conclude: L EMMA 5. we frequently use that if O is an orbit for a group H acting on a set X .Z.H W 1/ and so pj.
with the action of G deﬁned by G Let A 2 X . The number of elements in X is Â r Ã . A/ 7! gA D fga j a 2 Ag: def def H D Stab.G W H /. which is therefore a Sylow psubgroup 0 0 0 1pn 2 . and so on.p n C. X jX j D jOi j. Oi the distinct orbits.G W 1/ D . and that . p 1C2C Consider the upper triangular 1 1 : : : C C C C: :C :A : 1 p.G W 1/ D p r m with m not divisible by p. Thus.p r p r C 1/ Note that. h 7! ha0 W H ! A is injective (cancellation law).p n 1/. and so the power of p dividing . the order of G is .Fp /. the power of p dividing p r m i is the power of p dividing i . 2 E XAMPLE 5.G W 1/.17).p r i / .p r m 1/ .G W 1/ is matrices with 1’s down the diagonal: 0 1 B0 1 B B0 0 B B: : @: : : : They form a subgroup U of order p n G.G W H / is the number of elements in the orbit of A. of which there are p n p.p r 1/ .H W 1/ Ä jAj D p r . Because the orbits form a partition of X. . Therefore. According to (4. and so from now on we assume that .A/ D fg 2 G j gA D Ag: For any a0 2 A. and let X ! X. and so at least one of the jOi j is not divisible by p.74 5. the ﬁrst p column can be any nonzero vector in Fn . the ﬁeld with p elements.H W 1/ Ä p r .p r m i / . the second column p can be any vector not in the span of the ﬁrst column.3 Let Fp D Z=pZ. A PPLICATIONS P ROOF.p r m p r C 1/ p m jX j D D : pr p r . of which there are p n 1. The n n matrices in G are precisely those whose columns form a basis for Fn . Therefore the corresponding terms on top and bottom are divisible by the same powers of p. T HE S YLOW T HEOREMS . and so . H D Stab A has order p r .p n p/.n 1/ . If we can ﬁnd an A such that p doesn’t divide the number of elements in its orbit. pn 1 /.g. and so p does not divide jX j. Let X D fsubsets of G with p r elementsg. . and let G D GLn .H W 1/ we know that . In the equation .G W 1/ D p r m.p r m/. The same is true for p r i . it sufﬁces to prove this with p r the highest power of p dividing .p n p2/ . because i < p r . then we can conclude that (for such an A).
and we have that jOj Á 1 mod p. and H P .P /.P W 1/. then G will have a subgroup H of order p. L EMMA 5.G W 1/.G W NG . (b) Let sp be the number of Sylow psubgroups in G. For an A in such an orbit.4 The theorem gives another proof of Cauchy’s theorem (4. .P / with P normal in NG . (c) Every psubgroup of G is contained in a Sylow psubgroup. and let jGj D p r m with m not divisible by p. Therefore . and any g 2 H . and that the number of conjugates of H in G is . i.The Sylow theorems 75 R EMARK 5. P ROOF. and let G act on X by conjugation.46). R EMARK 5.13). One again writes .6 (S YLOW II) Let G be a ﬁnite group. 4. In fact this is the only onepoint orbit because fQg is a P orbit ” P normalizes Q. Recall (4.7) happens only for Q D P . no Sylow psubgroup of G other than P normalizes P .HP W P /.HP W P / D p 0 D 1. (a) Any two Sylow psubgroups are conjugate. Let H be a subgroup of G. the same counting argument shows that Stab. which we know (5. HP is a subgroup.2 can be modiﬁed to show directly that for each power p r of p dividing . the highest power p r0 of p dividing jXj is the highest power of p dividing m.7 Let P be a Sylow psubgroup of G. and it follows that there is an orbit in X whose order is not divisible by p r0 C1 . then sp Á 1 mod p and sp jm. g ¤ 1. and let P act on O through the action of G. T HEOREM 5. Hence the number of elements in every P orbit other than fP g is divisible by p.6.HP W P / is a power of p (here is where we use that H is a pgroup). If H normalizes P . 2 P ROOF ( OF S YLOW II) (a) Let X be the set of Sylow psubgroups in G.G W 1/ there is a subgroup H of G of order p r ..GW 1/.A/ has p r elements.HP W 1/. one of which will be fP g. but . and let H be a psubgroup. Because H and P are subgroups of NG . and H=H \ P ' HP =P (apply 1. In particular.H //. In this case.P /. Let P 2 O.HP W 1/ D .8) that the normalizer of H in G is NG . and . then H P .e. If a prime p divides . This single Gorbit may break up into several P orbits. is an element of G of order p. if H NG . We recommend that the reader write out the details.G W 1/ D p r m and considers the set X of all subsets of order p r . . P / 7! gP g 1 W G X ! X: Let O be one of the Gorbits: we have to show O is all of X.H / D fg 2 G j gHg 1 D H g. Thus .P W 1/ is the largest power of p dividing .5 The proof of Theorem 5. hence also the largest power of p dividing .g.
NG .F / be the set of linear maps ˛W V ! V such that (a) ˛.e.76 5.51) shows that . r r We shall prove by induction on k that it has order p11 pkk . (b) Since sp is now the number of elements in O.1).G W 1/ m D D : . and so the number of elements in every P orbit is divisible by p.NG . we have also shown that sp Á 1 (mod p/. Then G is a direct product of its Sylow psubgroups. We again let P act on O. There is a geometric description of the Sylow subgroups of G. Then P1 Pk 1 \ Pk D 1. and let jPi j D pi i . at least one H orbit consists of a single Sylow psubgroup. X H must be nonempty (Lemma 5. A PPLICATIONS Suppose there exists a P … O.P1 Pk 1 /Pk is the direct product of r r P1 Pk 1 and Pk .P / W 1/ . and let H act on the set X of Sylow psubgroups by conjugation. Now (1. 2 C OROLLARY 5. 2 C OROLLARY 5. i.P / W P / This is a factor of m. A full ﬂag F in V is a sequence of subspaces V D Vn Vn 1 Vi V1 f0g with dim Vi D i . Let P be a Sylow psubgroup of G. which contradicts what we proved in the last paragraph.Vi / Vi for all i . This implies that #O is divisible by p. : : : . and so has order p11 pkk .52) applied to the full set of Sylow subgroups of G shows that G is their direct product. which equals .NG . let U. there is nothing to r r prove. r P ROOF. Let P be a Sylow psubgroup of G.7 implies that H P . There can be no such P . Because jXj D sp is not divisible by p.8 A Sylow psubgroup is normal if and only if it is the only Sylow psubgroup. P ROOF. If k D 1.P / W P / . and (b) the endomorphism of Vi =Vi 1 induced by ˛ is the identity map. . and so we may suppose that k 2 and that P1 Pk 1 has order p11 pkk 11 . but this time the argument shows that there are no onepoint orbits..P W 1/ . (c) Let H be a psubgroup of G.10 Let G D GL.G W 1/ . the pi are distinct primes. 2 E XAMPLE 5.7c) (which shows. Given such a ﬂag F . the product P1 Pk is a normal subgroup of G.9 Suppose that a group G has only one Sylow psubgroup for each prime p dividing its order.P // D . The converse statement follows from (3. then (a) of Sylow II implies that it is the only Sylow psubgroup.V / where V is a vector space of dimension n over Fp .G W NG . Because each Pi is normal in G. and so O is all of X. sp is the number of conjugates of P . According to (a). T HE S YLOW T HEOREMS . that P is even characteristic). in fact. Pk be Sylow subgroups of G. But then H normalizes P and Lemma 5. If P is normal. therefore (1. Let P1 .
Recall (Exercise 417) that G is a disjoint union of the double cosets for H and P . Hence G is isomorphic to the direct product of its Sylow subgroups (5. It .12 Let G be a group.F / are exactly the elements of the group U of (5. 2 P ROOF ( OF S YLOW I) According to Cayley’s theorem (1. e2 g is a basis for V2 . def Let g 2 GLn .c)) Let P be a Sylow psubgroup of G. the matrices of the elements of U. and so the theorem with H D P 0 shows that aP a 1 P 0 for some a.11 Some books use different numberings for Sylow’s theorems. Alternative approach to the Sylow theorems We brieﬂy forget that we have proved the Sylow theorems. Relative to this basis. G embeds into Sn . en g for V such fe1 g is basis for V1 .F/. : : :g is again a full ﬂag. and so X X jH jjP j jGj D jHaP j D a a jH \ aP a 1 j where the sum is over a set of representatives for the double cosets.22). The ˇ Sylow theorems imply that G ˇ 99 and s11 Á 1 mod 11.F / is a Sylow psubgroup of G.18). On dividing by jP j we ﬁnd that jGj X jH j D . fe1 . s9 j11 and s9 Á 1 mod 3. For any subgroup H of G. From (a) of Sylow II. For such an a. which are both commutative (4. The Sylow 3subgroup Q maps bijectively onto G=N .H W H \ aP a H \ aP a 1 is a Sylow psubgroup of H .9). we see that the Sylow psubgroups of G are precisely the groups of the form U. Indeed. It has at least one subgroup H of order 11. Then P 0 is the unique Sylow psubgroup of P 0 . T HEOREM 5. A SIDE 5. and so on. and so the Sylow 3subgroup is also normal. and Sn embeds into GLn .Fp / (see 7.3).Fp / has a Sylow psubgroup (see 5.3).F / g 1 . gVn 1 . and so G commutative.F / for some full ﬂag F . Then gF D fgVn . 2 Examples We apply what we have learnt to obtain information about groups of various orders. a jH \ aP a 1 j jP j and so there exists an a such that . and let P 0 be a psubgroup of G. This implies (a) and (c) of Sylow II. Verify as before that the Sylow 11subgroup N of G is normal. and U. 5. As GLn . : : : . I have essentially followed the original (Sylow 1872).1b below). and H is normal. we can construct a basis fe1 . 2 P ROOF ( OF S YLOW II(a. Similarly.gF / D g U. Here is an alternative proof.Alternative approach to the Sylow theorems 77 I claim that U. P ROOF. 1/ is not divisible by p. there exists an a 2 G such that H \ aP a 1 is a Sylow psubgroup of H . and in fact s11 11 follows that s11 D 1. and so G D N Q. so also does G. and let P be a Sylow psubgroup of G.13 (G ROUPS OF ORDER 99) Let G have order 99.
Then Aut.C5 /. In fact. Then G has generators a. b. and s5 D 1 or 6. c generate C3 .15 (G ROUPS OF ORDER 30) Let G be a group of order 30.5). and Â. (5. which is the smallest prime dividing . In fact P 0 consists of the maps x 7! x i . i0 ¤ 1 (in Z=qZ). If pjq 1. 5. then Aut. Then . then the only group of order pq is the cyclic group Cpq . Because 3 doesn’t divide 5 1 D 4. C5 . and 1 respectively. if pjq 1. Therefore. : : : and divides 6: Hence s3 D 1 or 10. fi 2 Z=qZ j i p D 1g: Let a and b be generators for P and Q respectively. 5. then there is also a nonabelian group given by the above generators and relations. at least one is 1.C3 C5 / D Aut. : : : and divides 10I s5 D 1. 6.14 (G ROUPS OF ORDER pq. for otherwise there would be 20 elements of order 3 and 24 elements of order 5. p. which must be an element of order 2. We have shown .N /. A PPLICATIONS remains to determine the action by conjugation of Q on N .14) shows that H is commutative.5). q PRIMES . In summary: if p q 1. and so H D PQ is a subgroup of G.C3 C5 /. 1.G W 1/. and so gives an isomorphic group G. and it remains to determine the action of P on Q by conjugation. unless pjq 1. 5 (generated by b).Q/ (being cyclic) has a unique subgroup P 0 of order p.N / is cyclic of order 10 (see 3. G D Q P . 10. and so there is only the trivial homomorphism Q ! Aut. 7. and it remains to determine the possible homomorphisms ÂW C2 ! Aut. Hence G D .G W Q/ D p. and the only elements of Aut C3 and Aut C5 of order 2 are a 7! a 1 and b 7! b there are exactly 4 homomorphisms Â. But Aut. aba 1 D b i0 : Choosing a different i0 amounts to choosing a different generator a for P . 11. Then s3 D 1. The group Aut.78 5. we have that G D Q P. 4. and so. and let P and Q be Sylow p and q subgroups. Because P maps bijectively onto G=Q. It follows that G is the direct product of N and Q. which is impossible. p < q) Let G be such a group. 3 (generated by a). b and relations ap . and so (see Exercise 44) Q is normal. and hence are nonisomorphic. bq .Q/ is cyclic of order q 1 (see 3. a Sylow 3subgroup P or a Sylow 5subgroup Q is normal. C2 . and suppose that the action of a on Q by conjugation is x 7! x i0 .C3 / Aut. But such a homomorphism Â is determined by the image of the nonidentity element of C2 .C3 C5 / Â C2 .c/ is one of the following elements: a 7! a b 7! b a 7! a b 7! b 1 Thus a 7! a b 7! b 1 a 7! a b 7! b 1 1 : The groups corresponding to these homomorphisms have centres of order 30. H C3 C5 . Let a. T HE S YLOW T HEOREMS .
b. there are 3 noncommutative groups of order 12 and 2 commutative groups. Then G D P Q where Q is the Sylow 4subgroup. Hence .G/. we see that there exists an element b of order p outside N .G/ is commutative. We next show that G contains an element of order p not in N .32).17 (G ROUPS OF ORDER p 3 ) Let G be a group of order p 3 .G/. and so its kernel has order at least p 2 .18.G/ W 1/ D p and G=Z. and so we can also apply Lemma 3.19).G/ is not cyclic (4.p 2 1/. ab D ba. then N Cp Cp and there is a subgroup Q of G of order p such that Q \ N D f1g.n 1/ 2 . Its image is contained in Z. (The homomorphisms differ by an automorphism of Q. that G=Z. and assume G is not commutative.3). and its image is a subgroup of S4 of order 12. c and relations a3 . Since N contains only p 1 elements of order p. they are conjugate. cbc 1 D b: 5. action on the left cosets) ' W G ! Sym. If Q is cyclic of order 4.xy/p D x p y p .G W N / D p is the smallest (in fact only) prime dividing .G=P / S4 is injective. then P doesn’t contain a nontrivial normal subgroup of G.N / GL2 . From Sylow II we see that G has exactly 4 Sylow 3subgroups. and so x 7! x p W G ! G is a homomorphism.17) that G has a normal subgroup N of order p 2 . and hence we obtain a single noncommutative group C3 C4 .G/ intersects A4 in a subgroup with at least 8 elements. For example. We know from (4.Z. and let N be the subgroup generated by such an element a. x n. and so its Sylow psubgroups have order p.D C2 /. there are exactly 3 nontrivial homomorphism ÂW Q ! Aut. and so Lemma 3.P /. Suppose G has elements of order p 2 . In particular. and an easy induction argument shows that . c2.2. x p 2 Z. n 1: Therefore . If Q D C2 C2 . b5.Fp / is . and so G A4 . and so '. and. then there is a unique nontrivial map Q.G/ ¤ 1. The order of Aut.G/. If P is not normal. We know Z.G/ D A4 . and let P be its Sylow 3subgroup.G W 1/.16 (G ROUPS OF ORDER 12) Let G be a group of order 12. Hence GDN Â Q for some homomorphism ÂW Q ! N .Examples 79 that (up to isomorphism) there are exactly 4 groups of order 30. 5.D C4 / ! Aut. By the Sylow theorems. and hence it has exactly 8 elements of order 3.P /. cac 1 Da 1 . because G isn’t commutative. N is normal in G (Exercise 44). Because G=Z. Because .) In total. Now assume that P is normal.19 shows that there is exactly one nonabelian group in this case. and so the map (4. If every element of G has order p (except 1). but the three groups resulting are all isomorphic to S3 C2 with C2 D Ker Â. with p an odd prime.G/ Cp Cp . Therefore .xy/n D x n y n Œy. the commutator of any pair of elements of G lies in Z. By Lagrange’s theorem '. we see that for all x 2 G. But all elements of S4 of order 3 are in A4 (see the table in 4.p 2 p/ (see 5. the third on our list has generators a.
80
5. T HE S YLOW T HEOREMS ; A PPLICATIONS
G D hai hbi Cp2 Cp , and it remains to observe (3.19) that the nontrivial homomorphisms Cp ! Aut.Cp2 / Cp Cp 1 give isomorphic groups. Thus, up to isomorphism, the only noncommutative groups of order p 3 are those constructed in (3.14, 3.15). 5.18 (G ROUPS OF ORDER 2p n , 4p n ; AND 8p n , p ODD ) Let G be a group of order 2m p n , 1 Ä m Ä 3, p an odd prime, 1 Ä n. We shall show that G is not simple. Let P be a Sylow psubgroup and let N D NG .P /, so that sp D .G W N /. From Sylow II, we know that sp j2m , sp D 1; p C 1; 2p C 1; : : :. If sp D 1, P is normal. If not, there are two cases to consider: (i) sp D 4 and p D 3, or (ii) sp D 8 and p D 7: In the ﬁrst case, the action by conjugation of G on the set of Sylow 3subgroups1 deﬁnes a homomorphism G ! S4 , which, if G is simple, must be injective. Therefore .G W 1/j4Š, and so n D 1; we have .G W 1/ D 2m 3. Now the Sylow 2subgroup has index 3, and so we have a homomorphism G ! S3 . Its kernel is a nontrivial normal subgroup of G. In the second case, the same argument shows that .G W 1/j8Š, and so n D 1 again. Thus .G W 1/ D 56 and s7 D 8. Therefore G has 48 elements of order 7, and so there can be only one Sylow 2subgroup, which must therefore be normal. Note that groups of order pq r , p; q primes, p < q are not simple, because Exercise 44 shows that the Sylow qsubgroup is normal. An examination of cases now reveals that A5 is the smallest noncyclic simple group. 5.19 (G ROUPS OF ORDER 60) Let G be a simple group of order 60. We shall show that G is isomorphic to A5 . Let P be a Sylow 2subgroup and N D NG .P /, so that s2 D .G W N /. According to the Sylow theorems, s2 D 1; 3; 5; or 15: (a) The case s2 D 1 is impossible, because P would be normal (see 5.8). (b) The case s2 D 3 is impossible, because the kernel of G ! Sym.G=N / would be a nontrivial normal subgroup of G. (c) In the case s2 D 5, we get an inclusion G ,! Sym.G=N / D S5 , which realizes G as a subgroup of index 2 in S5 , but we saw in (4.37) that, for n 5, An is the only subgroup of index 2 in Sn . (d) In the case s2 D 15, a counting argument (using that s5 D 6) shows that there exist two Sylow 2subgroups P and Q intersecting in a group of order 2. The normalizer N of P \ Q contains P and Q, and so it has index 1, 3, or 5 in G. The ﬁrst two cases are impossible for the same reasons as in (a) and (b). If .GW N / D 5, the argument in (c) gives an isomorphism G A5 ; but this is impossible because s2 .A5 / D 5.
Exercises
51 Show that a ﬁnite group (not necessarily commutative) is cyclic if, for each n > 0, it contains at most n elements of order dividing n.
1 Equivalently,
the usual map G ! Sym.G=N /.
C HAPTER
6
Subnormal Series; Solvable and Nilpotent Groups
Subnormal Series.
Let G be a group. A chain of subgroups G D G0 G1 Gi Gi C1 Gn D f1g:
is called a subnormal series if Gi is normal in Gi 1 for every i , and it is called a normal series if Gi is normal in G for every i .1 The series is said to be without repetitions if all the inclusions Gi 1 Gi are proper (i.e., Gi 1 ¤ Gi ). Then n is called the length of the series. The quotient groups Gi 1 =Gi are called the quotient (or factor) groups of the series. A subnormal series is said to be a composition series if it has no proper reﬁnement that is also a subnormal series. In other words, it is a composition series if Gi is maximal among the proper normal subgroups Gi 1 for each i. Thus a subnormal series is a composition series if and only if each quotient group is simple and nontrivial. Obviously, every ﬁnite group has a composition series (usually many): choose G1 to be a maximal proper normal subgroup of G; then choose G2 to be a maximal proper normal subgroup of G1 , etc.. An inﬁnite group may or may not have a ﬁnite composition series. Note that from a subnormal series G D G0 F G1 F F Gi F Gi C1 F F Gn D f1g
we obtain a sequence of exact sequences 1 ! Gn
1
! Gn
2
! Gn
2 =Gn 1
!1
1 ! Gi C1 ! Gi ! Gi =Gi C1 ! 1 1 ! G1 ! G0 ! G0 =G1 ! 1: Thus G is built up out of the quotients G0 =G1 ; G1 =G2 ; : : : ; Gn 1 by forming successive extensions. In particular, since every ﬁnite group has a composition series, it can be regarded
authors write “normal series” where we write “subnormal series” and “invariant series” where we write “normal series”.
1 Some
81
82
6. S UBNORMAL S ERIES ; S OLVABLE AND N ILPOTENT G ROUPS
as being built up out of simple groups. The JordanH¨ lder theorem, which is the main topic o of this section, says that these simple groups are independent of the composition series (up to order and isomorphism). Note that if G has a subnormal series G D G0 F G1 F F Gn D f1g, then Y Y .G W 1/ D .Gi 1 W Gi / D .Gi 1 =Gi W 1/:
1Äi Än 1Äi Än
E XAMPLE 6.1 (a) The symmetric group S3 has a composition series S3 F A3 F 1 with quotients C2 , C3 : (b) The symmetric group S4 has a composition series S4 F A4 F V F h.13/.24/i F 1; where V C2 C2 consists of all elements of order 2 in A4 (see 4.32). The quotients are C2 , C3 , C2 , C2 . (c) Any full ﬂag in Fn , p a prime, is a composition series. Its length is n, and its p quotients are Cp ; Cp ; : : : ; Cp : (d) Consider the cyclic group Cm D hai. For any factorization m D p1 pr of m into a product of primes (not necessarily distinct), there is a composition series Cm F k hai
m Cp 1
F
Cpm p
1 2
F
k hap1 i
k hap1 p2 i Hr . Then G has a
The length is r, and the quotients are Cp1 ; Cp2 ; : : : ; Cpr . (e) Suppose G is a direct product of simple groups, G D H1 composition series G F H2 Hr F H3 Hr F
of length r and with quotients H1 ; H2 ; : : : ; Hr . Note that for any permutation of f1; 2; : : : rg, there is another composition series with quotients H .1/ ; H .2/ ; : : : ; H .r/ . (f) We saw in (4.37) that for n 5, the only normal subgroups of Sn are Sn , An , f1g, and in (4.33) that An is simple. Hence Sn F An F f1g is the only composition series for Sn . ¨ T HEOREM 6.2 (J ORDAN H OLDER ) 2 Let G be a ﬁnite group. If G D G0 F G1 F G D H0 F H1 F F Gs D f1g F Ht D f1g of f1; 2; : : : ; sg
are two composition series for G, then s D t and there is a permutation such that Gi =Gi C1 H .i / =H .i /C1 .
P ROOF. We use induction on the order of G. Case I: H1 D G1 . In this case, we have two composition series for G1 , to which we can apply the induction hypothesis.
2 Jordan
showed that corresponding quotients had the same order, and H¨ lder that they were isomorphic. o
: : :g fG=G1 . Choose a composition series K2 F K3 F We have the picture: G1 G H1 F F G2 F F Gs F Ku : F Ht F Ku : G=H1 ' G1 =K2 : (26) K2 F H2 F On applying the induction hypothesis to G1 and H1 and their composition series in the diagram. Since a commutative group is simple if and only if it is cyclic of prime order. Because G1 and H1 are both normal in G. K2 =K3 . K2 =K3 . G=H1 .Solvable groups 83 Case II: H1 ¤ G1 . Let K2 D G1 \ H1 .46).G/ be the minimum length of a composition series. G1 =K2 . : : :g D Quotients. For such a group. R EMARK 6.G F G1 F G2 F / D fG=G1 . Then the JordanH¨ lder theorem extends to show that all composition series have o length d. Solvable groups A subnormal series whose quotient groups are all commutative is called a solvable series. Therefore G=G1 D G1 H1 =G1 ' H1 =G1 \ H1 (see 1. It properly contains both G1 and H1 . 2 Note that the theorem applied to a cyclic group Cm implies that the factorization of an integer into a product of primes is unique.G F H1 F H2 F / (deﬁnition) (induction) (apply (26)) (reorder) (induction) (deﬁnition). we can say that a group is solvable if it can be obtained by forming successive extensions of commutative groups. the product G1 H1 is a normal subgroup of G. Similarly G=H1 ' G1 =G1 \ H1 . : : :g fG=H1 . then K2 is a maximal normal subgroup in both G1 and H1 . : : :g fG=H1 . we . and G=G1 ' H1 =K2 . H1 =H2 . The same proof works except that you have to use induction on d. K2 =K3 . G2 =G3 . and so G1 H1 D G. which are maximal normal subgroups of G. The quotients of a composition series are sometimes called composition factors. H1 =K2 . we ﬁnd that Quotients. A group is solvable (or soluble) if it has a solvable series. Alternatively.G/ and have isomorphic quotient groups. let d.3 There are inﬁnite groups having ﬁnite composition series (there are even inﬁnite simple groups). : : :g fH1 =K2 .G/ instead of jGj and verify that a normal subgroup of a group with a ﬁnite composition series also has a ﬁnite composition series (Exercise 61). H2 =H3 . G1 =G2 .
g. T HEOREM 6. We have shown that H F H \ G1 F F H \ Gn is a solvable series for H ..H \ Gi / \ Gi C1 D H \ GiC1 . (b) An extension of solvable groups is solvable. Every commutative group is solvable. whatever the conclusion might be.6 (a) Every subgroup and every quotient group of a solvable group is solvable. N N N Let G be a quotient group of G.. For a more recent look at the FeitThompson theorem. . and B=U ' F F . and let Gi be the image of Gi in G. and let H be a subgroup of G. it may be recommended to the reader as well worth his attention. 379) wrote: No simple group of odd order is at present known to exist. there is no known simple group whose order contains fewer than three different primes. However. S OLVABLE AND N ILPOTENT G ROUPS see that G is solvable if and only if for one (hence every) composition series the quotients are all cyclic groups of prime order. P ROOF. .3 P ROOF. 1957. Therefore. see p. which is commutative. Â Ã Â Ã 1 E XAMPLE 6. C/. (a) Let G F G1 F The homomorphism F Gn be a solvable series for G. P ROPOSITION 6. every noncommutative ﬁnite simple group has even order. An investigation as to the existence or nonexistence of such groups would undoubtedly lead. M. p. The results in Chapter 5 show that every group of order < 60 is solvable. The proof occupies an entire issue of the Paciﬁc Journal of Mathematics (Feit and Thompson 1963).5 Consider the subgroups B D and U D of GL2 . e. will not be solvable. and therefore contains an element of order 2. a noncommutative simple group. 2 In other words. Then U is a normal subgroup of B. p.84 6.4 (F EITT HOMPSON ) Every ﬁnite group of odd order is solvable. see Glauberman 1999. . For the role this theorem played in the classiﬁcation of the ﬁnite simple groups. 182). Hence B is solvable. 51. Also.F. to results of importance. U ' . The nonexistence of a certain type of simple group of ﬁnite order. . An for n 5. 0 0 1 some ﬁeld F . x 7! xGi C1 W H \ Gi ! Gi =Gi C1 has kernel . S UBNORMAL S ERIES . H \ Gi C1 is a normal subgroup of H \ Gi and the quotient H \ Gi =H \ Gi C1 injects into Gi =Gi C1 . 3 Burnside N F Gn D f1g (1897. Signiﬁcant progress in the ﬁrst problem was not made until Suzuki. the second problem was solved by Burnside himself. By contrast. who proved using characters that any group whose order contains fewer than three different primes is solvable (see Alperin and Bell 1995. Then N N G F G1 F N is a solvable series for G. as is every dihedral group.F /.
the group B. when k D Fp .8 For any ﬁnitedimensional vector space V over a ﬁeld k and any full ﬂag F D fVn . We have to show that if N N are solvable.G/ is commutative.x/. i.F / D f˛ 2 Aut. Then B. Recall that the commutator of x.Vj / Vj i all j g and notes that the commutator of two elements of Bi lies in Bi C1 .48).F / with Bi D f˛ 2 B. and G is commutative if and only if every commutator equals 1.D N / F N1 F F Nm 2 y 1 D xy. According to (4. y 2 G is Œx.xyx 1 y 1 / D Œ'. and let G D G=N .Œx. '. : : :g in V . and so the induction hypothesis implies that G=Z.e. C OROLLARY 6. ' maps the commutator of x. Let and G N N G F G1 F N F N1 F N F Gn D f1g F Nm D f1g N N be solvable series for G and N .16).1/ generated by the commutators in G is called the commutator or ﬁrst derived subgroup of G. y D 1 ” xy D yx. U.yx/ 1 is solvable.F / is a pgroup. Vn 1 . In particular.F /=U. the centre Z.7 A ﬁnite pgroup is solvable.. Indeed. and. and in the general case one deﬁnes subgroups B0 B1 of B. and let Gi be the inverse image of Gi in G. We use induction on the order the group G. .y/. The group G 0 D G . 2 Let G be a group.F / is solvable when k D Fp . y D xyx Thus Œx.F / be the group deﬁned in Example 5.Vj / Vj all j g 1 F Gn . and so N Gi =Gi C1 ' G G F G1 F is a solvable series for G. This proves that B. then so also is G. Because Z. '.F / j ˛. E XAMPLE 6. P ROOF.F / is commutative.G/ of G is nontrivial.10.y/.Solvable groups 85 N (b) Let N be a normal subgroup of G.x/. y/ D '. let U. For any homomorphism 'W G ! H '. Then N i =Gi C1 (see 1. y to the commutator of '.V / j ˛. (b) of the proposition shows that G is solvable. we see that if H is commutative. then ' maps all commutators in G to 1.G/ is solvable.
S OLVABLE AND N ILPOTENT G ROUPS P ROPOSITION 6.k/ D 1 for some k. Then Œg. . h. but the smallest groups where this occurs have order 94. Write g 7! g for the homomorphism g 7! gG 0 W G ! G=G 0 .2/ . and from G 0 =G 0 \ G2 ! G 0 G2 =G2 we see that G1 =G2 commutative H) G 0 =G 0 \ G2 commutative H) G 00 Continuing in the fashion. the derived series of Sn which is called the derived series of G. h 2 G=G 0 . If G . Then Œg.s/ D 1. 5. Its length is called the solvable length of G.G 00 /0 .1/ G . The proof of the proposition shows that the derived series is the shortest solvable series for G. An is the smallest normal subgroup of Sn giving a commutative quotient. h D 1 for all g. N G 0 . Conversely. h D Œg. which N N N 0 .n/ is a characteristic subgroup of G. and hence maps G 0 into G 0 . and hence G . Because G=G1 is commutative. then the derived series is a solvable series for G.i / ' G1 =G2 G 0 \ G2 G2 : 2 Gi for all i . A SIDE 6. The second derived subgroup of G is .9 The commutator subgroup G 0 is a characteristic subgroup of G. when n is Sn An An An : P ROPOSITION 6.86 6. An automorphism ˛ of G maps the generating set for G 0 into G 0 .Sn /0 D An . Since these elements generate G 0 .3/ D . the third is G . S UBNORMAL S ERIES . Hence we obtain a normal series G G . Hence Œg.10 A group G is solvable if and only if its kth derived subgroup G . we ﬁnd that G . in which all the groups are normal in G. each derived group G . h 7! 1 in G=N . P ROOF. This was shown by a computer search through the libraries of small groups. G 0 is characteristic. it is the smallest normal subgroup of G such that G=G 0 is commutative. h 2 N . and so on. Since a characteristic subgroup of a characteristic subgroup is characteristic (3. Hence . which shows that G=G 0 is is 1 because Œg. Thus. G1 subgroup of G1 . let G D G0 F G1 F G2 F F Gs D 1 G 0 .11 Not every element of the commutator subgroup of a group is itself a commutator.k/ D 1. a solvable group G has a canonical solvable series. 2 For n 5.G 0 /0 . Since this is true for all automorphisms of G. For example. namely the derived series. h 2 G N N N N commutative. Now G 0 G2 is a be a solvable series for G. and so Œg. Let N be a second normal subgroup of G such that G=N is commutative. P ROOF.7a).
x 2 Z i 1 G be the Z. but the converse is false. Therefore B=Z.H / Z. because (for h 2 H ) h 2 Z i C1 .G/ \ H .B/ is not nilpotent. b. A group G is of class 2 if and only if G=Z.G=Z. let Â Ãˇ a b ˇ ˇ a. Assume (inductively) that Z i .G/ \ H . x 2 Z i . then Dn is not nilpotent (use Theorem 6. using that Z. If n is not a power of 2. Recall that we write Z. In fact.G/ ” Œg. are metabelian. and the centre of B=Z.H / Z i C1 .G/ is commutative — such a group is said to be metabelian.H / all x 2 H: (b) Straightforward. G 0 D Z. Thus g 2 Z 2 .B/ D faI j a ¤ 0g. and G=G 0 is commutative (4.18 below). For example.G/ H) Œh.Nilpotent groups 87 Nilpotent groups Let G be a group. then Z i C1 .G/ for all x 2 G: If Z m .G/ Z 2 . and the smallest such m is called the (nilpotency) class of G. (6. The dihedral group D2n is nilpotent of class n — this can be proved by induction.G/ is commutative.16). and D2n =Z.D2n / D2n 1 . c 2 F. we get a sequence of subgroups (ascending central series) f1g where g 2 Z i .D2n / has order 2. x 2 Z. Let Z 2 . In particular.G/ D G for some m. For example. but we saw in 80 that it is solvable.5) 80 19 19 < 1 = < 1 0 = A is metabelian: its centre is @0 1 0 A . x 2 Z i . Only the group f1g has class 0.G/ \ H . 0 0 1 0 0 1 G=Z.12 (a) A nilpotent group is obviously solvable.13 (a) A subgroup of a nilpotent group is nilpotent. P ROPOSITION 6. and (b) The group G D @0 1 : . ac ¤ 0 : BD 0 c ˇ Then Z. the quaternion and dihedral groups of order 8.G/. for a ﬁeld F . (a) Let H be a subgroup of a nilpotent group G.G/ has order p (see 5. Clearly.G/ subgroup of G corresponding to Z. Z.18). (c) Any nonabelian group G of order p 3 is metabelian. P ROOF. Q and D4 .G/ for all x 2 G: Continuing in this fashion.G/ for the centre of G.G// G=Z.17). and the groups of class 1 are exactly the commutative groups.B/ is trivial.H / Z i .G/ all x 2 G H) Œh. E XAMPLE 6. then G is said to be nilpotent.G/ . 2 . : . (b) A quotient of a nilpotent group is nilpotent. all ﬁnite pgroups are nilpotent (apply 4.G/ ” Œg.
let g1 2 G.G/. g3 .88 6. g2 . For example. :::. gmC1 2 G H) Œ:::ŒŒg1 . G=N nilpotent of class m H) G nilpotent of class Ä m C 1: P ROOF. g3 . but B is not nilpotent. for all g1 . S UBNORMAL S ERIES . :::. the centre of Â Ãˇ a 0 ˇ ˇ ab ¤ 0 GL2 . g3 2 Z m H) Œ H) Œ . g2 . gmC2 D 1 all g1 . the subgroup U of the group B in Examples 6.G/ all g1 2 G: An extension of nilpotent groups need not be nilpotent. gm 2 G H) Œ:::ŒŒg1 . gmC1 D 1 for all g1 . g2 . g2 . Write for the map G ! G=N . :::. gm 2 Z. Then ŒŒ:::ŒŒg1 . g3 .G/ all g1 .5 and 6. gm 1 2G 2 H) g1 2 Z m . :::.F / consists only of the scalar matrices. then Z. C OROLLARY 6. g3 . :::. :::. g3 . gmC1 D 1 for all g1 . S OLVABLE AND N ILPOTENT G ROUPS R EMARK 6. :::. g3 2 G ŒŒg1 .14 It should be noted that if H is a subgroup of G. :::. we have the following more precise result. gm . N and G=N nilpotent G nilpotent. for all g1 .Œ:::ŒŒg1 .. gmC1 2 N Z. but the centre of GL2 . g3 .e.15 A group G is nilpotent of class Ä m if and only if Œ: : : ŒŒg1 . g2 .G/ for all x 2 G: . g2 . gmC1 / D Œ:::ŒŒ g1 .F /: HD 0 b ˇ is H itself.G/. and so 2 Œ:::ŒŒg1 . P ROPOSITION 6. :::. g2 2 Z m 1 1 .G/ all g1 . gmC1 2 G: P ROOF.16 For any subgroup N of the centre of G. g3 . : : : . In fact. (27) For example. i.G/. gm 2 G ŒŒg1 . g2 . : : : . gmC2 2 G: .G/ H) Œg1 . g3 . then G D Z m . gm . However.G/ all g1 . x 2 Z i Assume G is nilpotent of class Ä m. gmC1 D 1 all g1 . gmC1 2 G. g2 .:::. gmC1 D 1 all g1 . : : : .H / may be bigger than Z. : : : . gm . gm 2 Z. g2 . the implication (27) holds when N is contained in the centre of G. g2 . g2 . :::. Recall. :::. g2 2 G 2 H) ŒŒg1 . Then . :::. gm 1 2 Z 2 . g 2 Z i . g3 .G/. gm 2 G: For the converse. .12 is commutative and B=U is commutative.G/ ” Œg. g2 . g3 . g2 . :::. gm . Hence Œ:::ŒŒg1 . gmC1 .
which implies that G is nilpotent. let G be a ﬁnite nilpotent group. hP 0 h 1 D P for some h 2 H .G/ nilpotent.N / Z.G/. and so we can assume G to be noncommutative. and we can apply the induction hypothesis. For the converse. P ROOF. P ROOF. 2 Recall that an extension is central if Ã. The statement is obviously true for commutative groups. and let P be a Sylow psubgroup of H . then H ¤ NG . Certainly the elements of Z. The ﬁrst lemma below shows that NG . i. A direct product of nilpotent groups is obviously nilpotent. and let N D NG .P / H . and so if Z. which is a Sylow psubgroup of H .H /. We use induction on the order of G.G/ in G=Z. and the second then implies that N D G. Then: Ã 1!N !G!Q!1 the nilpotent groups are those that can be obtained from commutative groups by successive central extensions.G/ ¤ 1. and so g 2 H: 2 L EMMA 6. Hence hg 2 NG .G/ H .21 For a ﬁnite abelian group G we recover the fact that G is a direct product of its pprimary subgroups. According to (5.H /.H / D H . We use induction on the order of G. Then H gP g 1 D P 0 . G=Z. T HEOREM 6. we have NG . Because G is nilpotent. so that gHg 1 D H . and so hgP g 1 h 1 P .N / D N . Because Z. that P is normal in G. By Sylow II.G/ H NG .P /.20 Let H be proper subgroup of a ﬁnite nilpotent group G. Contrast: the solvable groups are those that can be obtained from commutative groups by successive extensions (not necessarily central). Thus we may suppose Z.P /.22 (F RATTINI ’ S A RGUMENT ) Let H be a normal subgroup of a ﬁnite group G. Let P be such a subgroup of G. 2 L EMMA 6.G/ ¤ 1.9) it sufﬁces to prove that all Sylow subgroups are normal. 2 R EMARK 6..G/ normalize H .G/.P /. 89 P ROOF.18 A ﬁnite group is nilpotent if and only if it is equal to a direct product of its Sylow subgroups.e.Nilpotent groups C OROLLARY 6. Then G D H NG .47) to the normalizer of H=Z.17 A ﬁnite pgroup is nilpotent.19 Let P be a Sylow psubgroup of a ﬁnite group G. and so the “if” direction follows from the preceding corollary. Z. P ROPOSITION 6.G/ H . Let g 2 NG . . Then the normalizer of H in G corresponds under (1. For any subgroup H of G containing NG . P ROOF.H /. we have H Z.
g/ for all ˛ 2 A.˛ / x. (c) Consider G with A D Aut.23 A ﬁnite group is nilpotent if and only if every maximal proper subgroup is normal.G.˛/ is a homomorphism). In this case all subgroups and homomorphisms are admissible. (' is a homomorphism). i. Conversely. Let g 2 G. H maximal H) NG . Conversely.e.G/ is called an Agroup.H / and centralizer CG . so also are its normalizer NG . Hence. Finally. there is an h 2 H such that gP g 1 D hP h 1 .˛ g/ D ˛ . i. Let G be an Agroup.H / D G.G/ of automorphisms of a group G is again a group. x/ 7! ˛ x W A G ! G satisfying (a). . A pair . the admissible subgroups are the normal subgroups.˛/ D . We saw in Lemma 6. If P is not normal in G. then there exists a maximal proper subgroup H of G containing NG .xy/ D ˛ x ˛ y (c) 1 x D x (' is a homomorphism).P / and so g 2 H NG .P /. and it follows that h 1 g 2 NG .G/: In this case.90 6. P ROOF.G/. We shall check the condition of Theorem 6. According to Sylow II. or G is said to have A as a group of operators. Being maximal. H NG . If H is admissible. 2 T HEOREM 6. Then gP g 1 gHg 1 D H . '/ consisting of a group G together with a homomorphism 'W A ! Aut.18. Let A be a group. (b) Consider G acting on itself by conjugation. and so Frattini’s argument shows that G D H NG .˛ˇ / x D ˛ .P /. (a) shows that the map '. (c) arises from a homomor1 phism A ! Aut. g 2 G: E XAMPLE 6..G/ as group of operators. all ˛ 2 A: An intersection of admissible groups is admissible. Thus.ˇ x/ (b) ˛ .H /. Conditions (a) and (c) show that x 7! ˛ x is inverse to x 7! . 2 Groups with operators Recall that the set Aut. ('.˛. Then (a) . S OLVABLE AND N ILPOTENT G ROUPS P ROOF. (b). a map . In this case.x 7! ˛ x/ is a homomorphism A ! Aut.24 (a) A group G can be regarded as a group with f1g as group of operators.P / D H — contradiction. and so x 7! ˛ x is a bijection G ! G. H is normal in G. Let G be a group with operators A. and write ˛ x for '. and both gP g 1 and P are Sylow psubgroups of H .20 that for any proper subgroup H of a nilpotent group G. H is normal. the admissible subgroups are the characteristic subgroups.e. Condition (b) then shows that it is an automorphism of G. consider G together with the homomorphism g 7! ig W G ! Aut.˛/x. A subgroup H of G is admissible or Ainvariant if x 2 H H) ˛ x 2 H . and so the theory of groups with operators includes the theory of groups without operators.G/. let P be a Sylow psubgroup of G.. suppose every maximal proper subgroup of G is normal. S UBNORMAL S ERIES .H /: An Ahomomorphism (or admissible homomorphism) of Agroups is a homomorphism W G ! G 0 such that .
In this case an admissible subnormal series is a sequence of subgroups G D G0 G1 G2 Gs D f1g.47). to give an action of G on Gi =Gi C1 .H \ N / 7! hH is an admissible isomorphism H=H \ N ! HN=N: N T HEOREM 6. a normal series.25 For any admissible homomorphism W G ! G 0 of Agroups..e. 1.Groups with operators 91 Almost everything we have proved for groups also holds for groups with operators.G/ .! G 0 : ' def T HEOREM 6. HN is an admissible subgroup of G.45.28 (a) Consider G with G acting by conjugation.G/ be a group with A operating. . Then H \ N is a normal admissible subgroup of H . Deﬁne similarly an admissible composition series. In each case.'/ and the set of subgroups of G admissible subgroups. Let 'W A ! Aut. an admissible subnormal series is a sequence G D G0 G1 G2 Gs D f1g with each Gi a characteristic subgroup of G. the Theorems 1.27 Let 'W G ! G be a surjective admissible homomorphism of Agroups. they have no normal admissible subgroups apart from the obvious two. and an admissible injection: G G=N ! . (b) Consider G with A D Aut. i. with each Gi normal in G. an admissible isomorphism. In this case. and factors in a natural way into the composite of an admissible surjection. An admissible subnormal series is a chain of admissible subgroups of G G G1 G2 Gr with each Gi normal in Gi 1 .e. In this case the isomoro phisms between the corresponding quotients of two composition series are admissible. The quotients of two admissible composition series are isomorphic as Ggroups. i. E XAMPLE 6. In particular. and 1.. The action of G on Gi by conjugation passes to the quotient. The proof is the same as that of the original theorem.26 Let G be a group with operators A.46. / is an admissible normal subgroup of G. .G/ is an admissible subgroup of G 0 . N D Ker. and let H and N be admissible subgroups with N normal. admissible subgroups correspond to ing Ker. the proof is the same as before except that admissibility must be checked. and the quotients of an admissible composition series are simple Agroups.G/groups. The JordanH¨ lder theorem continues to hold for Agroups.G/ as operator group. The quotients of an admissible subnormal series are Agroups. which we have noted also hold for Agroups. T HEOREM 6.47 hold for groups with operators. and h. and the quotients of two admissible composition series are isomorphic as Aut. because it uses only the isomorphism theorems. N Under the onetoone correspondence H $ H between the set of subgroups of G containN (see 1.
It follows that one must be cyclic of order p n . Then G D G1 Gr HrC1 Fp . then the subgroups in the statement of the theorem will all be characteristic.1. See Rotman 1995. this is obvious from the classiﬁcation. i. i. ascending and descending sequences of subgroups of G become stationary. E XAMPLE 6. : : : . Ht : G ' G1 Gs . If g is an element of G of highest order. (c) Every ﬁnite group can be written as a direct product of indecomposable groups (obviously). one shows that hgi is a direct factor of G. T HEOREM 6. Conversely. and that the other is trivial. Then H and H 0 must be pgroups.e. let Aut.1. 1/i. H2 D h.31 Let G D Fp Fp .G/ operate on G. 1/i: G2 D h.30 (K RULL S CHMIDT ) Suppose that G is a direct product of indecomposable subgroups G1 . but it is not difﬁcult to prove it directly. and there is a reindexing such that Gi arrange the numbering so that G D G1 P ROOF. S OLVABLE AND N ILPOTENT G ROUPS KrullSchmidt theorem A group G is indecomposable if G ¤ 1 and G is not isomorphic to a direct product of two nontrivial groups. if G H H 0 H) H D 1 or H 0 D 1: E XAMPLE 6.92 6. Let G1 D h. H2 . we can Ht : 2 Then s D t. G hgi H . and suppose that G H H 0 . S3 is indecomposable but has C3 as a normal subgroup. (b) The KrullSchmidt theorem also holds for groups with operators. 6. S UBNORMAL S ERIES . G ' H1 Ht : Hi . which is a contradiction.e. and think of it as a twodimensional vector space over H1 D h. : : : . G D H1 R EMARK 6. (c) When applied to a ﬁnite abelian group. Of course.36.0. 1/iI H2 . 0/i. For example. suppose that G is commutative and indecomposable. Gs and of indecomposable subgroups H1 . m < n. Let G be cyclic of order p n . and they can’t both be killed by p m .29 (a) A simple group is indecomposable. G D G1 G2 .. For example.1. . the theorem shows that the groups Cmi in a decomposition G D Cm1 ::: Cmr with each mi a prime power are uniquely determined up to isomorphism (and ordering). but an indecomposable group need not be simple: it may have a normal subgroup.. Since every ﬁnite commutative group is (obviously) a direct product of pgroups with p running over the primes. (b) A ﬁnite commutative group is indecomposable if and only if it is cyclic of primepower order. given r. Moreover.32 (a) The KrullSchmidt theorem holds also for an inﬁnite group provided it satisﬁes both chain conditions on subgroups. G is a pgroup.
0 0 0 0 62 If G1 and G2 are groups such that G1 G2 and G1 =G1 G2 =G2 .) .Exercises 93 Exercises 61 Let G be a group (not necessarily ﬁnite) with a ﬁnite composition series G D G0 G1 Gn D 1. are G1 and G2 necessarily isomorphic? (Here 0 denotes the commutator subgroup. Show that N D N \ G0 N \ G1 N \ Gn D 1 becomes a composition series for N once the repetitions have been omitted. and let N be a normal subgroup of G.
.
en be a basis for A as an F vector space. In particular. and it is semisimple if it is isomorphic to a direct sum of simple modules. (b) Let G D Sn . called the structure constants of A relative to the basis . (c) Let G D Cn D h i. All Amodules are ﬁnite dimensional when regarded as F vector spaces. We do not assume A to be commutative. i. An F algebra is a ring A containing F in its centre and ﬁnite dimensional as an F vector space. I.C/ of the quaternion group Q D 1 p0 ha. / be the matrix obtained from the identity matrix by using to permute the rows. bi sending a to and b to 0 1 .ei /i .22).C HAPTER 7 Representations of Finite Groups Throughout this chapter. In fact. The opposite Aopp of an F algebra A is the same F algebra as A but with the multiplication reversed. E XAMPLE 7.e. Then. Let e1 . As every ﬁnite group embeds into Sn for some n (Cayley’s theorem. then ei ej D P k k k aij ek for some aij 2 F . An Amodule M is simple if it is nonzero and contains no submodules except 0 and M . : : : . it is faithful. then there is representation i 7! i W C ! GL . once a basis has been chosen.18).A.F /.ba/0 .. If F contains a nth root of 1. see 1. All vector spaces are ﬁnite dimensional. for any n n matrix A. for example. For each 2 Sn . Matrix representations A matrix representation of degree n of G over F is a homomorphism G ! GLn . Thus a faithful representation identiﬁes G with group of n n matrices. that is how we originally deﬁned 10 1 0 Q in (1. 0 / D I. A could be the matrix algebra Mn . For an Amodule V . C. /A is obtained from A by using to permute the rows. The representation is said to be faithful if the homomorphism is injective. /I. Clearly. mV denotes the direct sum of m copies of V . let I. G is a ﬁnite group and F is a ﬁeld.F /. Aopp D . the algebra A is uniquely determined by its structure constants.F / D F . say . 0 /. 0 / with a 0 b D ba. there is a onetoone correspondence a $ a0 W A $ Aopp which is an isomorphism of F vector spaces and has the property that a0 b 0 D . and so 7! I. this shows that every ﬁnite group has a faithful matrix representation. I. In other words. The representation is faithful if and only if has order n 1 95 .1 (a) There p a Á is representation Q ! GL2 . / is a representation of Sn .
Then . no inﬁnite ﬁnitely generated group with ﬁnite exponent has a faithful representation over C. n 1 in F . 37). and we can always arrange that j n .1) that we have deﬁned the notion of a group G acting a set. from a linear action of G on V . Roots of 1 in ﬁelds As the last example indicates. by replacing a subﬁeld F of C with F Œ where D e 2 i=n . . The nth roots of 1 in a ﬁeld F form a subgroup n . If the characteristic of F divides n.C/.g 1 /V . When the set is an F vector space V .96 7. We call a homomorphism G ! GL.3 (a) Let G D Cn D h i. or by replacing F with F ŒX =.X 1/p . is linear for each g 2 G. In this case. and so the minimum polynomial of L divides X n 1. R EPRESENTATIONS OF F INITE G ROUPS exactly n. n /L D 1. If n D p is prime and F has characteristic p. . E XAMPLE 7.F / of F .V /. : : : . every such homomorphism deﬁnes a linear action of G on V . say .F /j < n. and assume that F contains a primitive nth root of 1. X n 1 has distinct roots (a multiple root would have to be a root of its derivative nX n 1 ). we say that the action is linear if the map gV W V ! V . Otherwise. then X p 1 D . Then gV has inverse the linear map . To say that F contains a primitive nth root of 1. every such direct sum decomposition of G arises from a representation of G. and g 7! gV W G ! GL.X / is an irreducible factor of X n 1 not dividing X m 1 for any proper divisor m of n: An element of order n in F is called a primitive nth root of 1. but there is a faithful representation Â Ã 1 i i 7! W Cp ! GL2 . As X n 1 has n distinct roots 0 . Note that a linear representation of G on F n is just a matrix representation of degree n.F /: 0 1 A SIDE 7.X // where g.g. then j n . and so 1 is the only pth root of 1 in F .F / is a cyclic group of order n and that generates it (and it implies that either F has characteristic 0 or it has characteristic a prime not dividing n). Burnside proved that the problem has a positive answer for subgroups of GLn . x 7! gx. Thus. conversely. 0Äi Än 1 Conversely.V / a linear representation of G on V . the vector space V decomposes into a direct sum of eigenspaces M def V D Vi . Let G ! GL. the representation is trivial.2 Recall that the Burnside problem asks whether every ﬁnitely generated group with ﬁnite exponent is ﬁnite (see p. L /n D . Linear representations Recall (4. for example. we obtain a homomorphism of groups G ! GL. means that n .56). the representations of a group over a ﬁeld F depend on the roots of 1 in the ﬁeld. Vi D fv 2 V j v D i vg.V / be a linear representation of G.V / is a homomorphism.F /j D n by extending F . Therefore. which is cyclic (see 1.
and let W be a subspace of V . and this complementary subspaces are 0 01 those of the form F a . where p is the characteristic of F . and the general case follows easily (decompose V with respect to the action of one cyclic factor of G.˛v/ for all g 2 G. v. and let acts on V D F 2 as the matrix 1 1 (see 7. Let v 2 W ? and let g 2 G.V / be a linear representation of G on an F vector space V .w.w. P ROOF. v/ D 0 because W is Ginvariant. it sufﬁces to show that there exists a Ginvariant nondegenerate symmetric bilinear from W V V ! F . v/ because is Ginvariant. Maschke’s theorem Let G ! GL. and assume that F contains a primitive nth root of 1. there exists a Ginvariant subspace W 0 such that V D W ˚ W 0 .4 (M ASCHKE ) Let G ! GL. An F linear map ˛W V ! V 0 of vector spaces on which G acts linearly is said to be Ginvariant if ˛.5 Let be a symmetric bilinear form on V . then decompose each summand with respect to the action of a second cyclic factor of G.3b). b ¤ 0. v/ D 0 for all w 2 W g. Note that the theorem always applies when F has characteristic zero. this is a restatement of (a). If def and W are Ginvariant.G. then so also is W ? D fv 2 V j ..gv. . v 2 V: Finally. gv/ D .gv/ D g.g 1 w. none of them is Ginvariant. If the characteristic of F does not divide jGj.v. V D fv 2 V j v D . gv 0 / D . F / D Hom. / is Ginvariant. Therefore. _ 2G When G is cyclic.G. a bilinear form W V V ! F is said to be Ginvariant if . /vg. n . we present two proofs of Maschke’s theorem. This shows that gv 2 W ? .F // To give a representation of G on a vector space V is the same as to give a direct sum decomposition M def V D V . i. P ROOF OF M ASCHKE ’ S THEOREM ( CASE F D R OR C) L EMMA 7. the subspace . then every Ginvariant subspace W of V has a Ginvariant complement.e. 2 Recall from linear algebra that if is nondegenerate. v 0 2 V: T HEOREM 7. A subspace W of V is said to be Ginvariant if gW W for all g 2 G. v 0 / for all g 2 G.V / be a linear representation of G. b Because of the importance of the ideas involved. and so on). and . Let G _ D Hom. then V D W ˚ W ? . For any w 2 W .Maschke’s theorem 97 (b) Let G be a commutative group of exponent n. The condition on the characteristic is certainly necessary: let G D h i be the cyclic group of order p. .g 1 w. in order to prove Maschke’s theorem.
98
7. R EPRESENTATIONS OF F INITE G ROUPS on V , .gv; gw/
L EMMA 7.6 For any symmetric bilinear form X def N .v; w/ D
g2G
is a Ginvariant symmetric bilinear form on V . P ROOF. The form is obviously bilinear and symmetric, and for g0 2 G, X def N .g0 v; g0 w/ D .gg0 v; gg0 w/;
g2G
which equals
P
g2G
.gv; gw/ because, as g runs over G, so also does gg0 .
2
Unfortunately, we can’t conclude that N is nondegenerate when is (otherwise we could prove that all F ŒGmodules are semisimple, with no restriction on F or G). L EMMA 7.7 Let F D R. If also is N . is a positive deﬁnite symmetric bilinear form on V , then so
P ROOF. If N is positive deﬁnite, then for any nonzero v in V , X N .v; v/ D .gv; gv/ > 0:
g2G
2
This completes the proof of Maschke’s theorem when F D R, because there certainly exist positive deﬁnite symmetric bilinear forms on V . A similar argument using hermitian forms applies when F D C (or, indeed, when F is any subﬁeld of C).
A SIDE 7.8 A representation of a group G on a real vector space V is unitary if there exists a Ginvariant positive deﬁnite symmetric bilinear form on V . Lemma 7.6 shows that every unitary representation is semisimple, and Lemma 7.7 shows that every real representation of a ﬁnite group is unitary.
P ROOF OF M ASCHKE ’ S THEOREM ( GENERAL CASE )
An endomorphism of an F vector space V is called a projector if 2 D . The minimum polynomial of a projector divides X 2 X D X.X 1/, and so V decomposes into a direct sum of eigenspaces, V D V0 . / ˚ V1 . /, where V0 . / D fv 2 V j v D 0g D Ker. / V1 . / D fv 2 V j v D vg D Im. /:
Conversely, a decomposition V D V0 ˚ V1 arises from a projector .v0 ; v1 / 7! .0; v1 /. Now suppose that G acts linearly on V . If a projector is Ginvariant, then V1 . / and V0 . / are obviously Ginvariant. Thus, to prove the theorem it sufﬁces to show that W is the image of a G–invariant projector . We begin by choosing an F linear projector with image W , which certainly exists, and we modify it to obtain a Ginvariant projector N with the same image. For v 2 V , let N .v/ D 1 X g g2G jGj .g
1
v/ :
The group algebra; semisimplicity
99
This makes sense because jGj 1 2 F , and it deﬁnes an F linear map N W V ! V . Let w 2 W ; then g 1 w 2 W , and so 1 X 1 X N .w/ D g.g 1 w/ D w D w: (28) g2G g2G jGj jGj The image of N is contained in W , because Im. /
def
W and W is Ginvariant, and so
N 2 .v/ D N . N .v// D N .v/ for any v 2 V . Thus, N is a projector, and (28) shows that Im. N / W , and hence Im. N / D W . It remains to show that N is Ginvariant. For g0 2 V 1 X 1 X N .g0 v/ D g .g 1 g0 v/ D g0 .g0 1 g/ .g 1 g0 v/ ; g2G g2G jGj jGj which equals g0 N .v/ because, as g runs over G, so also does g0 1 g.
.28/
The group algebra; semisimplicity
The group algebra F ŒG of G is deﬁned to be the F vector space with basis the elements of G endowed with the multiplication extending that on G. Thus, P ˘ an element of PŒG is a sum P cg g, cg 2 F , F g2G 0 0 ˘ two elements g2G cg g and g2G cg g of F ŒG are equal if and only if cg D cg for all g, andÁ Á P P P P 0 00 00 0 ˘ cg g cg g D g2G cg g; cg D g1 g2 Dg cg1 cg2 : g2G g2G A linear action g; v 7! gvW G V !V of G on an F vector space extends uniquely to an action of F ŒG on V , X X cg gvW F ŒG V ! V; cg g; v 7!
g2G g2G
which makes V into an F ŒGmodule. The submodules for this action are exactly the Ginvariant subspaces. Let G ! GL.V / be a linear representation of G. When V is simple (resp. semisimple) as an F ŒGmodule, the representation is usually said to be irreducible (resp. completely reducible). However, I will call them simple (resp. semisimple) representations. P ROPOSITION 7.9 If the characteristic of F does not divide jGj, then every F ŒGmodule is a direct sum of simple submodules. P ROOF. Let V be a F ŒGmodule. If V is simple, then there is nothing to prove. Otherwise, it contains a nonzero proper submodule W . According to Maschke’s theorem, V D W ˚ W 0 with W 0 an F ŒGsubmodule. If W and W 0 are simple, then the proof is complete; otherwise, we can continue the argument, which terminates in a ﬁnite number of steps because V has ﬁnite dimension as an F vector space. 2 As we have observed, the linear representations of G can be regarded as F ŒGmodules. Thus, to understand the linear representations of G, we need to understand the F ŒGmodules, and for this we need to understand the structure of the F algebra F ŒG. In the next three sections we study F algebras and their modules; in particular, we prove the famous Wedderburn theorems concerning F algebras whose modules are all semisimple.
100
7. R EPRESENTATIONS OF F INITE G ROUPS
Semisimple modules
In this section, A is an F algebra. P ROPOSITION 7.10 Every Amodule V admits a ﬁltration V D V0 V1 Vs D f0g
such that the quotients Vi =Vi C1 are simple Amodules. If V D W0 W1 Wt D f0g of f1; : : : ; sg such that
is a second such ﬁltration, then s D t and there is a permutation Vi =Vi C1 W .i / =W .i /C1 for all i .
P ROOF. This is a variant of the JordanH¨ lder theorem (6.2), which can be proved by the o same argument. 2 C OROLLARY 7.11 Suppose V V1 ˚ ˚ Vs W1 ˚ ˚ Wt of
with all the Amodules Vi and Wj simple. Then s D t and there is a permutation f1; : : : ; sg such that Vi W .i / .
P ROOF. Each decomposition deﬁnes a ﬁltration, to which the proposition can be applied.2 P ROPOSITION 7.12 Let V be an Amodule. If V is a sum of simple submodules, say P V D i 2I Si (the sum need not be direct), then for any submodule W of V , there is a subset J of I such that M V DW ˚ Si :
i 2J
def P P ROOF. Let J be maximal among the subsets of I such the sum SJ D j 2J Sj is direct and W \ SJ D 0. I claim that W C SJ D V (hence V is the direct sum of W and the Sj with j 2 J ). For this, it sufﬁces to show that each Si is contained in W C SJ . Because Si is simple, Si \ .W C SJ / equals Si or 0. In the ﬁrst case, Si W C SJ , and in the second SJ \ Si D 0 and W \ .SJ C Si / D 0, contradicting the deﬁnition of I . 2
C OROLLARY 7.13 The following conditions on an Amodule V are equivalent: (a) V is semisimple; (b) V is a sum of simple submodules; (c) every submodule of V has a complement. P ROOF. The proposition shows that (b) implies (c), and the argument in the proof of (7.9) shows that (c) implies (a). It is obvious that (a) implies (b). 2 C OROLLARY 7.14 Sums, submodules, and quotient modules of semisimple modules are semisimple. P ROOF. Each is a sum of simple modules.
2
Note that Mn . k (as an F vector space) and with the multiplication determined by i 2 D a. b/ is an F algebra. 1/ is the usual quaternion algebra. Thus a division algebra satisﬁes all the axioms to be a ﬁeld except commutativity (and for this reason is sometimes called a skew ﬁeld).a. E XAMPLE 7. i. Nj D 0 ) . right. 80 19 0 > ˆ 0 0 ˆ > <B = 0C C B0 0 L.15 We consider the matrix algebra Mn . Clearly.n/ of minimal left ideals. when n D 4. For example.15a) shows that the algebra Mn . mi0 j0 ¤ 0. and suppose that I contains a nonzero matrix M D . If D is a division algebra. Then H. b/ be the F algebra with basis 1. ij D k D j i (so ik D i ij D aj etc. and so is simple. In particular. we see that I contains all the matrices eij and so equals Mn . therefore.M N /j can be arbitrary: (29) For 1 Ä i Ä n.F /.F / is simple. then f is injective. i. the j th column of M N is M Nj where Nj is the j th column of N . 0 0 0 It follows from (29) that L. let H. We shall make frequent use of the following observation: The kernel of a homomorphism f W A ! B of F algebras is an ideal in A not containing 1. if F D R. a division algebra has no nonzero proper ideals.F /. We have shown that Mn .17 For a.a.F /.D/ is simple..F / is a direct sum Mn . (b) For M. 1.1/ ˚ ˚ L. As ei i0 M ej0 j D mi0 j0 eij and ei i0 M ej0 j 2 I .F / D L. then the argument in (7. if A is simple.F /. for a given matrix N . or twosided. (a) Let I be a twosided ideal in Mn . called a quaternion algebra over F .e. . let L. b 2 F . then H.3/ D @ M4 . left.i / be the set of matrices whose j th columns are zero for j ¤ i and whose i th column is arbitrary.F /. Therefore.Simple F algebras and their modules 101 Simple F algebras and their modules An F algebra A is said to be simple if it contains no twosided ideals except 0 and A.). j 2 D b. One can show that H. N 2 Mn .F /. b/ is either a division algebra or it is isomorphic to M2 . there exists a b such that ab D 1 D ba. For example. E XAMPLE 7.16 An F algebra is said to be a division algebra if every nonzero element a has an inverse.i / is a minimal left ideal in Mn . say. j /th position and zeros elsewhere. E XAMPLE 7. Let eij be the matrix with 1 in the .M N /j D 0 Nj ¤ 0 ) .a. it is simple.mij / with.F /: 0A> ˆ 0 0 ˆ > : . j.i.
j aij eij .F /. (c) Let A be the set of diagonal matrices in Mn .F / is set of scalar matrices F In . E XAMPLE 7. and so ai l D 0 for i ¤ l.a elm ˛ D j amj elj .1/ D 'a . then ˛elm D elm ˛. Thus C. Right multiplication x 7! xa on A A by an element a of A is an Alinear endomorphism of A A. Let eij be the matrix with 1 in the . (as F algebras). EndA .A/ D Mn . amj D 0 for j ¤ m. every Alinear map 'W A A ! A A is of this form with a D '. The centralizer of A in B is CB . which we now compute. the usual arguments show that a ﬁnitely generated module V over a division algebra D has a basis. If ˛ is in the centre of Mn . 7. the centralizers are taken in Mn . Then C. Moreover. Then ˛ D i. In this case.F /. (a) Let A be the set of scalar matrices in Mn . j /th position and zeros elsewhere.A A/ ' A (as F vector spaces).32 below).C. Thus. In particular. Clearly.F /. and that all bases have the same number n of elements — n is called the dimension of V . so that eij elm D ei m 0 P if j D l if j ¤ l: P Let ˛ DP ij / 2 Mn .20 In the following examples. Notice that in all three cases. For example.F /. EndA .F /.e. (b) Let A D Mn .V / ' Aopp for any Amodule V that is free of rank 1. and EndA . A D F In .V / ' Mn . C. It follows that the centre of Mn . . and let A A denote A regarded as a left Amodule.i. C.19 Let A be an F algebra.F /.'a ı 'a0 /.F /.102 7.Aopp / for any free Amodule V of rank n (cf.A/ is the centre of Mn .a0 / D a0 a D 'a0 a .1/. Then . and so EndA .F /. all ﬁnitely generated Dmodules are free. def C ENTRALIZERS Let A be an F subalgebra of an F algebra B. and so ˛elm D i ai l ei m and . i.'a0 .18 Much of linear algebra does not require that the ﬁeld be commutative.1// D 'a . and al l D amm . 7.A/ D fb 2 B j ba D ab for all a 2 Ag: It is again an F subalgebra of B.A/ D A. R EPRESENTATIONS OF F INITE G ROUPS 7. Let 'a be the map x 7! xa. C.A A/ ' Aopp More generally.A/ D F In .1/..A// D A.
: : : .A// D A (centralizers taken in EndF . any minimal left ideal of A. Let D D C.21). shows that the centralizer of Mn . Let D be the centralizer of A in the F algebra EndF . 7.C. i. b. i. We now apply the case n D 1 of the lemma to A.D/ in EndF . a. Let W be the direct sum of n copies of V with A acting diagonally.V /). i.A/ and let B D C.v/ D v if and only if v 2 Av1 . 2 .e. An argument as in Example 7. the centralizer of A in EndF .23 Every simple F algebra is isomorphic to Mn . such that ij 2 D (cf.19). and the reverse inclusion follows from the next lemma when we take v1 .S /.D/.V /. Let W V ! V be the map . : : : .22 For any v1 . S D n .. and the vector .A/ is Mn . a 2 A. 2 L EMMA 7. Choose a simple Amodule S. : : : .e. there exists an a 2 A such that av1 D bv1 . Therefore S is a free Dmodule (7. ij 2 EndF . We now prove the general case. : : : . The centralizer of A in EndF .S / is A. We ﬁrst prove this for n D 1. and so bv1 2 Av1 . such that . 2 T HEOREM 7. and so (see 7. Then A acts faithfully on S .21 (D OUBLE C ENTRALIZER T HEOREM ) Let A be an F algebra.av1 . It is Alinear. P ROOF.D/ for some n and some division F algebra D.av1 . for example. avn D bvn : P ROOF. According to the double centralizer theorem (7.ı/ with ı in the ij th position and zeros elsewhere.a ij / for all a 2 A. avn /.14). hence lies in D. vn / to complete the proof.S / Mn .S / will be a twosided ideal of A not containing 1. Schur’s lemma (7. vn / D .D opp / (see 7.D/. using the matrices eij .v1 . because the kernel of A ! EndF . and let V be a faithful semisimple Amodule.18).j Än . the centralizer of D in EndF .bv1 / D b. 0/ (projection onto Av1 ). and has the property that . av2 D bv2 . In other words.W / consists of the matrices .S / of F linear maps S ! S . vn 2 V and b 2 B. vn to generate V as a F vector space. Clearly A B.24 below) implies that D is a division algebra. Note that Av1 is an Asubmodule of V . ij a/ D . vi 2 V: Then W is again a semisimple Amodule (7.v1 .e. w/ 7! . Now . and so EndD . v1 / D bv1 .W / consists of the diagonal matrices 0 1 ˇ 0 0 B0 ˇ 0C B C B : : :: :C @: : : :A : : : 0 0 ˇ with ˇ 2 B. and hence is 0. Then C.av1 . :::. A D EndD ..20(b). as required. : : : .13) there exists an Asubmodule W of V such that V D Av1 ˚ W . P ROOF.32). ij /1Äi. W . say.Simple F algebras and their modules 103 T HEOREM 7..
P ROOF. and D is uniquely determined up to isomorphism. V is a sum of simple submodules each isomorphic to S0 .D/ are those of the form L. Let be an Alinear map S ! S. P ROOF. P ROOF.28 The integer n in Theorem 7.. P ROOF.a1 .24 (S CHUR ’ S L EMMA ) For every F algebra A and simple Amodule S. Obvious from the Theorem. we may assume that A D Mn . it has an inverse that is also Alinear. 2 C OROLLARY 7. R EPRESENTATIONS OF F INITE G ROUPS L EMMA 7.D/ for some division algebra D. then D EndA . the submodules of A A are the left ideals in A. ar / 7! ai ei n S0 for some realizes V as a quotient of a direct sum of r copies of A A.S / for any simple Amodule S. We saw in (7. and so it is either S or 0. and the simple submodules of A A are the minimal left ideals. The proposition shows that A A n.26.ALGEBRAS For any F algebra A. P ROOF. 2 T HEOREM 7. and any two Amodules having the same dimension over F are isomorphic. / is an Asubmodule of S. is zero.104 7. The map X .23.fj g/. Then every Amodule is isomorphic to a direct sum of copies of S . After Theorem 7. er be a set of generators for V as an Amodule. and hence as a quotient of nrS0 . P ROPOSITION 7.S / is a division algebra. L Clearly A D 1Äj Än L.fj g/ is isomorphic to D n with its natural action of Mn . Then Ker.D/. 2 C OROLLARY 7. : : : . 2 M ODULES OVER SIMPLE F . and A A is a direct sum of its minimal left ideals. and let S be a simple Amodule. 2 .fj g/ and each L.27 Let A be a simple F algebra.e. i. and so Proposition 7. : : : .23 is uniquely determined by A. Then any two simple Amodules are isomorphic.25 Any two minimal left ideals of a simple F algebra are isomorphic as left Amodules.D/.16) that the minimal left ideals in Mn . the statement follows from Theorem 7.26 Let A be a simple F algebra. EndA . In the ﬁrst case.12 shows that V mS0 for some m. Let S0 be a minimal left ideal of A. Thus. and in the second it is an isomorphism. Let e1 . Thus. If A Mn .
32 Let A be an F algebra. 7. . Let D be division algebra over F . A division algebra over F whose centre is F is said to be central (formerly normal). the isomorphism class of F is an identity element. For xj 2 Mj . and hence is isomorphic to Mr .26 shows that simple F algebras are semisimple. this shows that A is semisimple. called (by Hasse and Noether) the Brauer group1 of the ﬁeld.30 The classiﬁcation of the isomorphism classes of division algebras over a ﬁeld F is one the most difﬁcult and interesting problems in algebra and number theory. ym /: tensor product D ˝F D 0 of two central simple algebras over F is again a central simple algebra. and consider modules M D M1 ˚ N D N1 ˚ ˚ Mn ˚ Nm : Let ˛ be an Alinear map M ! N . For any ˛ 2 D. Deﬁne 1 The ŒDŒD 0 D ŒD 00 : This product is associative because of the associativity of tensor products. E XAMPLE 7. Then every minimal left ideal of a simple factor of A is a simple Asubmodule of A A. Before stating the main result of this section. Semisimple F algebras and their modules An F algebra A is said to be semisimple if every Amodule is semisimple. P ROOF. A A is a direct sum of simple Amodules.31 Let A be a ﬁnite product of simple F algebras. the only division algebra is the usual quaternion algebra. The statements in the last paragraph show that the Brauer groups of algebraically closed ﬁelds and ﬁnite ﬁelds are zero. For F D R. xj . 2 A SIDE 7. The Brauer groups of Q and its ﬁnite extensions were computed by Albert. the only division algebra with centre F is F itself (theorem of Wedderburn). and ŒD opp is an inverse for ŒD.y1 . Theorem 7. and Noether in the 1930s as a consequence of class ﬁeld theory. 0/ D . to classify the simple algebras over F .9). Brauer showed that the set of isomorphism classes of central division algebras over a ﬁeld form a group. we recall some elementary module theory.23. the only division algebra over F is F itself. Therefore. Hasse. and Maschke’s theorem shows that the group algebra F ŒG is semisimple when the order of G is not divisible by the characteristic of F (see 7. and so is semisimple. and the Brauer group of R has order 2.0.Semisimple F algebras and their modules 105 C LASSIFICATION OF THE DIVISION ALGEBRAS OVER F After Theorem 7. 0. Therefore ˛ 2 F. : : : . Brauer. let ˛. it remains to classify the division algebras over F . P ROPOSITION 7. For F ﬁnite. : : : . Since every Amodule is a quotient of a direct sum of copies of A A. : : : .D 00 / for some central simple algebra D 00 .29 When F is algebraically closed. 0. the F subalgebra F Œ˛ of D generated by ˛ is a ﬁeld (because it is an integral domain of ﬁnite degree over F ).
we can write V i ri Si where the Si are simple Amodules.C. every such matrix . ˛ deﬁnes an m n matrix whose ij th coefﬁcient is an Alinear map Mj ! Ni .16).x1 / C C ˛i n .Sj . then the centralizer of A in EndF .xn / xn ˛m1 ˛mj ˛mn xn Thus. T HEOREM 7. The centralizer of A in EndF .M / ' Mm .M0 /).A// is a product of simple algebras.27).x1 / C C ˛1n .V / EndA . According to Theorem 7. say. At with the Ai . Ni / 1Äj Än. namely. According to Schur’s lemma (7. Thus.24).Si /. Because HomA . which we denote ˛ij . c/m D bm: T HEOREM 7.Di / is a simple F algebra (see 7.ri Si / by (30) Yi ' Mri .V /. P ROOF. for example. if M is a direct sum of m copies of M0 . 1 10 1 0 0 1 0 ˛11 . this becomes an isomorphism of F algebras.106 7. Di is a division algebra. ri Si / ' EndA . If V is semisimple as an Amodule.V /. and therefore Mri .C. Choose an Amodule V on which A acts faithfully.b. By assumption. M Y EndA . let Si be a simple Ai module (cf. For each Ai .35 Let A be a semisimple F algebra.A// in EndF . 7.Di / by (31) i (31) where Di D EndA . A D A1 simple. and so C.M0 // (m m matrices with coefﬁcients in the ring EndA . N / ' HomA .x1 / C C ˛mn . i ri Si /.EndA .34 Every semisimple F algebra is isomorphic to a product of simple F algebras.˛ij / deﬁnes an Alinear map M ! N . no two of which are isomorphic.xn / C : C CB C B B C B C B:C B : : : : CB : C B : : : A@ : A @ :A @ : A @: : : : : : ˛m1 . Si / D 0 for i ¤ j . we see HomA . For example. then EndA . V D A A. A Bmodule M becomes an Amodule with .21). Conversely. 1Äi Äm (30) (isomorphism of F vector spaces). and L EndA . When M D N .Mj .xn / x1 ˛11 ˛1j ˛1n x1 C B:C B : : : : CB : C B : : : CB : C B C B:C B : : : : CB : C B C B:C B : def Bxj C 7! B ˛i1 ˛ij ˛j n C Bxj C D B ˛i1 .V / is a product of simple F algebras (hence it is a semisimple F algebra). and let A be an F subalgebra of EndF . R EPRESENTATIONS OF F INITE G ROUPS Then xj 7! yi is an Alinear map Mj ! Ni .M.33 Let V be a ﬁnite dimensional F vector space. Then A is equal to its double centralizer C. C.33. L P ROOF.V / (see 7. 2 T HEOREM 7. 2 Modules over a semisimple F algebra Let A D B the action C be a product of F algebras.A/ is semisimple.V / is EndA .
Recall that. The action of G on F ŒG corresponds to the action .29). Each of these is a matrix algebra over a division algebra (7.g 1 a/ of g 2 G on f W G ! F .36 The dimension of the centre of F ŒG as an F vector space is the number of conjugacy classes in G. Let S be a simple Amodule.38 The group algebra F ŒG is isomorphic to a product of matrix algebras over F . for each i . Such functions are called class functions. P ROOF. and that no L two of them are isomorphic. Let C1 . and let x be a nonzero element of S . In this way. and the second part follows from (7. we assume that F is an algebraically closed ﬁeld of characteristic zero (e. In the above proof. i. 2 .g. P ROOF. let ci be the element P a2Ci a in F ŒG. ct are obviously linearly independent.. and so its restriction to some Si in A A is nonzero. L L (b) Every Amodule is isomorphic to ri Si for some ri 2 N. C) P ROPOSITION 7. if and only if a2G ma a 2 i F ci . but the only division algebra over an algebraically closed ﬁeld is the ﬁeld itself (7. and so is a product of simple algebras (7.11).37 An element a2G ma a of F ŒG can be regarded as a map a 7! ma W G ! F . F ŒG ' Map.35).25) that A A ri Si for some ri 2 N. we showed that the elements of the centre of F ŒG correspond exactly to the functions f W G ! F that are constant on each conjugacy class. and. 2 The representations of G P ROPOSITION 7. It follows from (7.The representations of G 107 (a) Each Si is a simple Amodule.23). : : : . Á X X g ma gag 1 : ma a g 1 D a2G a2G The coefﬁcient of a in the right hand sum is mg 1 ag . and hence an isomorphism. it sufﬁces to show that they span the centre. : : : . and two modules ri Si L 0 and ri Si are isomorphic if and only if ri D ri0 for all i. (a) It is obvious that each Si is simple when regarded as an Amodule. We shall prove the stronger statement: centre of F ŒG D F c1 ˚ ˚ F ct (32) As c1 .gf /.a/ D f .9) implies that F ŒG is semisimple. F /.. when F has characteristic zero. and so Á X X g ma a g 1 D m 1 a: a2G a2G g ag P This shows that a2G ma a lies in the centre of F ŒG if and only if the function a 7! ma is P P constant on conjugacy classes.e. 2 P R EMARK 7. Then the map a 7! axW A A ! S is surjective. Maschke’s theorem (7. (b) The ﬁrst part follows from (a) and the deﬁnition of a semisimple ring. P ROOF. P For any g 2 G and a2G ma a 2 F ŒG. In the remainder of this chapter.G. Ct be the conjugacy classes in G. and every simple Amodule is isomorphic to exactly of of the Si .
F ŒG F ŒG/ is called the regular representation. and the regular character reg is that deﬁned by the regular representation. GL. V ˚V 0 D V C V 0: .F / Mft . The centre of a product of F algebras is the product of their centres. called the character of .40 For any Gmodules V and V 0 . one see that reg . (b) With the notations of (7.g/ D TrV . Mf . and that V is a class function.F / D dimF F ŒG: 1Äi Ät 2 The characters of G P Recall that the trace TrV . and so the centre of F ŒG is isomorphic to tF .gV /.F / for some integers f1 .˛/ of an endomorphism ˛W V ! V of a vector space V is ai i where .f /. L EMMA 7. From each representation of g 7! gV W G ! GL.108 7. (b) The multiplicity of any simple representation S in the regular representation is equal to its degree dimF S . R EPRESENTATIONS OF F INITE G ROUPS The representation G ! GL. which we know equals the number of conjugacy classes of G. In this case.g/ is the number of elements a of G such that ga D a.g/ D 1 for all g 2 G). On computing reg .39 (a) The number of isomorphism classes of simple F ŒGmodules is equal to the number of conjugacy classes in G. V is a homomorphism G ! F .35.1/ ˚ ˚ L.15).aij / is the matrix of ˛ with respect to some basis for V . and let fi D dimF Si .F / ' L. : : : . When V has dimension 1.g/ D .g/. Therefore. Then X fi2 D jGj: 1ÄiÄt P ROOF.V / ' F . the number of isomorphism classes of simple F ŒGmodules is the number of factors t. Note that V depends only on the isomorphism class of the F ŒGmodule V . According to Theorem 7. ft . The character is said to be simple (or irreducible) if it is deﬁned by a simple F Gmodule. It is independent of the choice of the basis (the traces of conjugate matrices are equal). F ŒG Mf1 . (c) The equality is simply the statement X dimF Mfi .g/ by using the elements of G as a basis for F ŒG. The principal character 1 is that deﬁned by the trivial representation of G (so 1 .g/ D jGj if g D e 0 otherwise. and so V . Therefore t is the dimension of the centre of F . (c) Let S1 . V . : : : . St be a set of representatives for the isomorphism classes of simple F Gmodules. and so reg . and so this deﬁnition of “linear character” essentially agrees with the earlier one. (a) Under our hypothesis. we obtain a function V on G. the character V of G is said to be linear.V /. T HEOREM 7.
Therefore V determines the multiplicity with which each Si occurs in V .F /. : : : .P i 2 N. and so the virtual m characters are exactly the class functions that can be expressed mi i .. the character of the representation of G on pV is identically zero. then its character is P V D 1Äi Ät ci i . ci 2 N. if c1 . for any representation G ! GL. and let 1 .1c) is not trivial. Therefore the simple characters certainly generate the Zmodule of virtual characters. and hence it determines the isomorphism class of V . 2 from which the claim follows. The proposition is false even when the characteristic of F doesn’t divide the order of the group. but it has the same character 0 1 as the trivial representation. i. : : : . 2 A SIDE 7.F / Mft .ej / D c j fj . : : : .41 The functions 1 . P ROOF.e. : : : .41 shows that they are linearly independent over Z (even over F ). P ROOF. Any function G ! F that can be expressed as a Zlinear combination of characters is called a virtual character. For example. P the characters of G are exactly Then the class functions that can be expressed in the form mi i . 0. P ROPOSITION 7.g/ D 0 for all g 2 G. because. and let ej D . 0. and Proposition 7. 2 Some . Write F ŒG Mf1 . : : : . 0/.ej / D fj D dimF Sj 0 X i if i D j otherwise. Compute the matrix of gL with respect to a basis of V ˚ V 0 that is made by combining a basis for V with a basis for V 0 . : : : .2 P ROPOSITION 7. We have already observed that the character of a representation depends only L on its isomorphism class.43 The proposition is false if F is allowed to have characteristic p ¤ 0.44 The simple characters of G form a Zbasis for the virtual characters of G. : : : . 2 Let S1 .F / of (7.42 Two F ŒGmodules are isomorphic if and only if their characters are equal. mi 2 Z. P ROPOSITION 7. t be the corresponding characters. t be the simple characters of G. 2 authors call it a generalized character. the Â Ã 1 i representation i 7! W Cp ! GL2 . St be a set of representatives for the isomorphism classes of simple F Gmodules with S1 chosen to be the trivial representation. ci i . 1. Conversely. and (33) shows that ci D V . then the ci are all zero. if V D 1Äi Ät ci Si .V /.ei /=fi .0. but this is to be avoided: there is more than one way to generalize the notion of a character. and so i . t are linearly independent over F . Let 1 . P ROOF. Then ej acts as 1 on Sj and as 0 on Sj for i ¤ j . (33) Therefore. ct 2 P F are such that i ci i .The characters of G 109 P ROOF.
f1 . then gv D g v0 D . V G denotes the submodule of elements ﬁxed by G: V G D fv 2 V j gv D v for all g 2 Gg 1 P L EMMA 7. if v 2 V G . P ROOF. they form an F vector space of dimension t . P ROPOSITION 7.f1 C f2 jf / D . . As the simple characters are a set of t linearly independent elements of this vector space. As this set has t elements.a/f2 . For any F ŒGmodule V . V D V and so 1 X 1 X ga D aD .46 The pairing . 2 We now assume that F is a subﬁeld of C stable under complex conjugation c 7! c.cf1 jf2 / D c. for any F ŒGV is a projector.f2 jf / for all class functions f1 . g D from which it follows that 2 module V . f2 / for c 2 F and class functions f1 . Conversely. a2G a2G jGj jGj V is (34) D (in the F algebra F ŒG). f2 . 2 All of these are obvious from the deﬁnition.45 The simple characters of G form an F basis for the class functions on G. . P ROOF. For an F ŒGmodule V .a/: a2G jGj L EMMA 7. .f1 jf2 / for all class functions f1 . N For class functions f1 and f2 on G.110 7. f2 . P ROOF.f2 jf1 / D .34/ v0 D v 1 jGj and so v lies in V G . dimF V G D 1 X a2G jGj V .f1 jf2 / D 1 X f1 . We have to check: ˘ ˘ ˘ ˘ . Therefore. the obviously v D v is in the image of . The class functions are the functions from the set of conjugacy classes in G to F .48 For any F ŒGmodule V . For any g 2 G. j / is an inner product on the F space of class functions on G.f jf / > 0 for all nonzero class functions f . a projector with image V G . say v D v0 .a/: P a2G av D v. they must form a basis.47 Let be the element jGj a2G a of F ŒG. f . If v is in its image. R EPRESENTATIONS OF F INITE G ROUPS P ROPOSITION 7. f2 . deﬁne .f1 jf / C . and so 2 .
On the other hand. and we showed that the latter is V G .g/ D f1 .The character table of a group 111 P ROOF. TrV . P ROOF. Therefore.47. C OROLLARY 7.V. ' 2 HomF .g/f2 . V is the direct sum of its 0eigenspace and its 1eigenspace. when we identify c 2 k with the constant function.49 For any F ŒGmodules V and W. Because V is a projector. Let S be a ﬁnite set.V. A becomes a kalgebra. W / of F linear maps V ! W by the rule. Show that fM 2 Mn . and HomF . V / D dimF V G . because the trace is a linear function. g 2 G. V j W /.a/: 2 T HEOREM 7. Therefore the simple characters form an orthonormal basis for the space of class functions on G.g'/. The character table of a group To be written. . W /. j 0/ D 1 if D 0 0 otherwise. Exercises 71 Let C be an n r matrix with coefﬁcients in a ﬁeld F . v 2 V. then .V.50 If and 0 2 are simple characters.f1 f2 /. W /G D HomF G . dimF HomF ŒG . 72 This exercise shows how to recover a ﬁnite group G from its category of representations over a ﬁeld k.V. and that every left ideal is of this form for some C . W /.F /. and let A be the set of maps S ! k.V. W / D .'. g 2 G: Moreover.F / j M C D 0g is a left ideal in Mn . V/ D 1 X 1 X TrV . f1 .gv//. .aV / D a2G a2G jGj jGj V . Examples To be written. f2 2 A. Let be as in Lemma 7.g/. The group G acts on the space HomF . (a) Show that A becomes a commutative ring with the product .v/ D g. TrV .
˛f / . and Richard Brauer”. V ˝W D V ˝ W I ii) for k with its trivial representation. g/ for all g 2 G.rA .R/ such that A satisﬁes the conditions of (d). rV /. R EPRESENTATIONS OF F INITE G ROUPS s2S ks (product of copies of k indexed by the elements of S ).1˝˛/ı D ı˛.g/f /. g 0 2 G is a representation of G (this is the regular representation). Issai Schur. f 2 A.] V VI D rV . emphasizing the work of “the four principal contributors to the theory in its formative stages: Ferdinand Georg Frobenius.f / D .e/f for all f 2 A. for each ﬁnite dimensional representation . . g2 2 G.R/.g/ for all V . g2 / D f .g1 g2 / D g1 . k D idk .g1 . W . [Hint: Deduce from (b) that there exists a bijection W G ! G such that .g/ D g .V. Deduce that ˛. iii) for all Gequivariant maps ˛W V ! W .g 0 / D f . Deduce that Endkalg .] (f) Show that the following maps are Gequivariant eWk ! A mWA ˝ A ! A WA ! A ˝ A (trivial representation on k. k/. W ı ˛ D ˛ ı then there exists a unique g 2 G.S /. Now take S D G.f1 . From the hypothesis on ˛.s2 /. (c) Let .g1 g2 /.g/ D f . 1 ˝ rA on A ˝ A/: (g) Suppose that we are given.f1 .s1 /f2 . If the family .S S.s1 . Show that.gg 0 /. see Curtis 1999.g2 / for all g1 . a klinear map V . V / satisﬁes the conditions i) for all representations V .A/ ' Sym. Hence .A/. deduce that . and that the ks are exactly the minimal ksubalgebras of A. f2 / 2 A A act on S S by . (d) Show that the map rA W G ! Endklinear . rA on A/ (rA ˝ rA on A ˝ A. show that this deﬁnes a bijection A ˝ A ' Map. rA on A/ (rA on A. (e) Deﬁne W A ! A ˝ A by . g.f /. [Hint: show that N OTES For a historical account of the representation theory of ﬁnite groups. .e/ for all g 2 G. s2 / D f1 . William Burnside.f / D gf for all f 2 A.112 (b) Show that A' Y 7. for any homomorphism ˛W A ! A of kalgebras such . f2 /. there exists a unique element g 2 G such that ˛.
Prove or disprove the following statement: if H is a subgroup of an inﬁnite group G.a1 : : : ak /. Suppose that the group G is generated by a set X. then for all x 2 G.] 42. P is a Sylow psubgroup. Prove that there is no simple group of order 616. (c) If g is an element of G such that gHg 1 H . Prove. 7. (b) the coset gH consists of elements of G of order n. xH x 1 H H) x 1 H x H . then the commutator subgroup of G is generated by the set of all elements xyx 1 y 1 for x.A PPENDIX A Additional Exercises 34. and H NG . (b) Show that if x 2 D 1 for all x 2 X . Let H be the subgroup of Sn generated by the cycle ..2p 1/ is commutative. Let H be a ﬁnite normal subgroup of a group G. 43. p prime. Let n and k be integers 1 Ä k Ä n. 113 . and let g be an element of G. then g 2 NG . : : :/. then the subgroup H of G generated by the set of all elements xy for x. 38. [The centralizer of H is the set of g 2 G such ghg 1 D h for all h 2 H . what are the possible orders of the product ab? 37. It is again a subgroup of G. then H \ P is a Sylow psubgroup of H. 39. Prove or disprove the following: (a) If G is ﬁnite and P is a Sylow psubgroup. 40. then NG . 35. Let H be a subgroup of a group G.P /.H / D H . p D 3. If a and b both have order 146 and ab D ba. g 2 G. 41. Then ﬁnd the order of the normalizer of H in Sn . or disprove by example. Suppose p 3 and 2p 1 are both prime numbers (e. Let a and b be two elements of S76 .g. Prove that a ﬁnite group G having just one maximal subgroup must be a cyclic pgroup. Find the order of the centralizer of H in Sn . Suppose that g has order n and that the only element of H that commutes with g is 1. Show that: (a) the mapping h 7! g 1 h 1 gh is a bijection from H to H . y 2 X has index 1 or 2.H /. 31. (a) Show that if gxg 1 2 X for all x 2 X. 19. (b) If G is ﬁnite. y 2 X. that every group of order p.
and P a Sylow psubgroup. and let Y be the set of orbits of N in X. am D 1.x/ D fy 2 Gjxy D yxg. then ˛x D x for all x in G. then the normalizers of the stabilizers Stab. then (a) the normalizer of H in G is H .xy/4 D 1 is a ﬁnite solvable group. Show that if G is its own commutator subgroup.y/ of x and y have the same order.x/ and Stab. Prove that if all Sylow subgroups of a ﬁnite group G are normal and abelian. Describe all subgroups of G in terms of (b) and (d). For what values of m and n can G be inﬁnite.F / of the subgroup H of diagonal matrices. 45. Determine the normalizer N in GLn . Let G be a group with generators x and y and deﬁning relations x 2 .114 A. Show that G is solvable. (Hint: A ﬁrst step is to ﬁnd a normal subgroup of order 11 using the Sylow theorems. Show that if a permutation in a subgroup G of Sn maps x to y. 52. 56. b n D 1 where m and n are positive integers.G W H / Á 1 (mod p). and that the order of G=H is a power of 2. Let n D . 49. A DDITIONAL E XERCISES 44.xy/4 . . (b) Show by example that if N is not normal then its orbits need not have the same cardinality. Describe all cyclic subgroups of G. Describe all conjugacy classes of G. G 000 .p prime).G W 1/=2. (a) (b) (c) (d) (e) (f) Show that G has a cyclic subgroup of order n. Let G be a ﬁnite group with generators s and t each of order 2. then the group is abelian. Let N be a normal subgroup of G. A group is generated by two elements a and b satisfying the relations: a3 D b 2 . 47. Show that the group G generated by elements x and y with deﬁning relations x 2 D y 3 D . G 00 .) 54. 50. Suppose that ˛ is an endomorphism of the group G that maps G onto G and commutes with all inner automorphisms of G. What is the index in G of the commutator group G 0 of G. 53. Verify that any two psubgroups of G are conjugate . (b) . Describe all subgroups of G of the form C. 46. Let G be a ﬁnite group. 55. and ﬁnd the order of G and its successive derived subgroups G 0 . Let G be a group of order 33 25. Let G be a ﬁnite group. 51. . y 5 . A group G is generated by a normal set X of elements of order 2. Show that the commutator subgroup G 0 of G is generated by all squares of products xy of pairs of elements of X . Prove that: (a) There is a natural action of G on Y which is transitive and shows that every orbit of N on X has the same cardinality. Now assume n odd.P / H G. and H the subgroup generated by the elements of odd order. Let G act transitively on a set X . and prove that N=H is isomorphic to the symmetric group Sn . Show that H is normal. 48. Show that if H is a subgroup of G such that NG . x 2 G.
c j bc D cb. Let H be a maximal subgroup of G. (b) Exhibit a nonabelian group of order 16 whose centre is not cyclic. has order p. p prime. p prime. Prove or disprove that direct products of metacyclic groups are metacylic. Let X be a set with p n elements. determine its order. Determine 69. Prove that every Sylow psubgroup of G acts transitively on X.) 67. (b) In case G is ﬁnite. then either N is contained in Gx or it acts transitively on X.˛ˇ/3 D 1 is isomorphic with the symmetric group S3 of degree 3 by giving. a4 D b 2 D c 2 D 1. and let G be the group with generators x. Let G be a group of order pq. Prove that every maximal subgroup of a ﬁnite pgroup is normal of prime index . and y ¤ 1 has 61. y and relations xyx x 3 D 1. aca the order of G and ﬁnd the derived series of G. Let G be a group acting doubly transitively on X . an explicit isomorphism. b. Do not assume Sylow’s theorems in this problem. y be elements of a group G such that xyx odd order. Let t 2 Z.115 57. 1 D y 5 . and let G be a ﬁnite group acting transitively on X . 59. x has order 3. . aba 1 D bci. (b) Prove that G is nilpotent ” G is abelian ” G is cyclic. (b) Every group of order 30 is nilpotent.p is prime). Find (with proof) the order of y. 64. 66. and let A be a normal subgroup of H and such that the conjugates of A in G generate it. Prove or give a counterexample: (a) Every group of order 30 has a normal subgroup of order 15. D c. with proof.G/ ¤ 1. (b) If N is a normal subgroup of G. (a) Find necessary and sufﬁcient conditions on t for G to be ﬁnite. and let x 2 X. 58. 63. then G=M is a simple group. (a) Prove that the centre of a nonabelian group of order p 3 . 1 1 D yt . 65. 70. Let x. 68. (b) Let M be the intersection of the conjugates of H in G. (a) Prove G is solvable. Prove that: (a) The stabilizer Gx of x is a maximal subgroup of G. Let N be a nontrivial normal subgroup of a nilpotent group G. then either N H or G D NA. 62. (c) Is G always nilpotent? (Prove or ﬁnd a counterexample. Prove that if G is equal to its commutator subgroup and A is abelian. (a) Prove that if N is a normal subgroup of G. A group G is metacyclic if it has a cyclic normal subgroup N with cyclic quotient G=N . Prove that subgroups and quotient groups of metacyclic groups are metacyclic. Show that the group with generators ˛ and ˇ and deﬁning relations ˛ 2 D ˇ 2 D . Let G D ha. Prove that N \ Z. p ¤ q primes. 60.
] (b) Let G be a ﬁnite group. Prove that every group G of order 56 can be written (nontrivially) as a semidirect product. (c) There exist at least two maximal subgroups that are abelian.G/ D ' m . Let G be a group of order p 2 q 2 for primes p > q. then N K. Find (with proofs) two nonisomorphic nonabelian groups of order 56. and let L be a subgroup of K such that L ıK D K. (b) Prove that G is the semidirect product of the subgroups Ker ˛ and Im ˛. (b) An element x of a group G is said to be an FCelement if its centralizer CG . 77. Prove that L D K. 71. (c) Prove that K D ŒK. Let G be a ﬁnite noncyclic pgroup. (a) Prove that there is an integer n 0 such that ' n . 74. Prove that the following are equivalent: (a) . (c) Prove that Im ˛ is normal in G or give a counterexample. Let G be a ﬁnite group. (a) Prove that there is a unique normal subgroup K of G such that (i) G=K is solvable and (ii) if N is a normal subgroup and G=N is solvable. 78. . Prove that the set of all F C elements in G is a normal. Show that S H H) H D G. Prove that H is isomorphic to a direct sum of cyclic groups of order p for some prime p.Fp /. K and that K D 1 or K is nonsolvable. and P a Sylow psubgroup of G.116 A. (a) Let K be a ﬁnite nilpotent group. Let S be a set of representatives for the conjugacy classes in a ﬁnite group G and let H be a subgroup of G.G/ for all integers m Let ˛ D ' n . 79. Let G be a ﬁnite group and ' W G ! G a homomorphism. 1 1 ::: C B0 1 C is a Sylow psubgroup of (b) Prove that the group of n n matrices B A @ ::: 0 1 GLn . Suppose H is a normal subgroup of a ﬁnite group G such that G=H is cyclic of order n. If G has a subgroup H such that both G=ıH and H are nilpotent. 75. prove that G is nilpotent. n. Let H be a minimal normal subgroup of a ﬁnite solvable group G. 80. (b) Every maximal subgroup of G is abelian. 72. where n is relatively prime to . Prove 1 that there exists an x 2 G such that xP x 0\ H is a Sylow psubgroup of H . (b) Show that K is characteristic. 76.G W 1/. 73. where ıK is the derived subgroup. Prove that G has a normal subgroup of order p n for some n 1. A DDITIONAL E XERCISES (a) Let H be a subgroup of a ﬁnite group G. (c) Indicate how (a) and (b) can be used to prove that any ﬁnite group has a Sylow psubgroup.G W Z. Prove that G is equal to the semidirect product H S with S a cyclic subgroup of G of order n. (a) Prove that subgroups A and B of a group G are of ﬁnite index in G if and only if A \ B is of ﬁnite index in G.G// Ä p 2 .x/ has ﬁnite index in G. [You may assume that a ﬁnite group is nilpotent if and only if every maximal subgroup is normal.
But . Then G is the union of the conjugates of H .. Snr . 4. g 1 g of G. (b) Show by induction that 0 1n 0 1 1 a b 1 na nb C n. Since G is a disjoint union of these sets. or 8. abab D e. this result goes back to Jordan in the 1870s. and so g n 2 N . and they overlap (at least) in f1g.GW H / of these. The remaining subgroups have orders 1. Thus c commutes with all elements of Q. i. and assume that H contains at least one element from each conjugacy class of G. sg of D3 . and so t 3 D t … f1. For the second statement. consider the subgroup f1. 11 By inspection. 16 (a) Let a. . 14 The symmetric group Sn contains a subgroup that is a direct product of subgroups Sn1 . sg. It has index 3 in D3 . . Since gcg 1 also has order 2. Students were expected to write out complete solutions.gN /n D g n N . in which case it has 1 element. and hence at least one element of order 2. .ab/2 D e. but the element t has order 2. but there are at most . In particular. b 2 G. gcg 1 D c for all g 2 Q. it must equal c.) 117 . cg is a normal subgroup of Q.A PPENDIX B Solutions to the Exercises These solutions fall somewhere between hints and complete solutions. 15 Because the group G=N has order n. We are given that a2 D b 2 D .n2 1/ ac @0 1 c A D @0 1 A: nc 0 0 1 0 0 1 17 Let H be a subgroup of a ﬁnite group G. there must be a (nonzero) even number of sets with 1 element. .27). Â Ã Â Ãn Â Ã 1 1 1 1 1 n 12 The product ab D .e. On multiplying this on right by ba. and f1.gN /n D 1 for every g 2 G (see 1. and so this can happen only if H D G. and are automatically normal (see 1. 0 1 0 1 0 1 13 Consider the subsets fg. (According to Serre 2003. Each set has exactly 2 elements unless g has order 1 or 2. . the only element of order 2 is c D a2 D b 2 . and D . we ﬁnd that ab D ba.36a).
As H1 \ H3 H1 \ H2 \ H3 . srsr D 1i with that for G suggests putting r D ab and s D a. Since gb 2 g 1 also has order 2 for any g 2 Qn . b 2 D 1. han i. From acac 1 D 1 this gives a2 D 1. 27 The hint gives ab 3 a 1 D bc 3 b 1 . and so a and xi ). if an b 1. Because G satisﬁes the cancellation law. Since c 4 D 1. and relations a2 D 1. we see that gb 2 g 1 D b 2 . an x D x for all x 2 G. So c 3 D 1. 26 (a) A comparison of the presentation Dn D hr. (b) Omit. Check (using 2. the reduced form of Œxj . [Check that it is the full centre. then H1 \ H2 \ H3 is of ﬁnite index in H1 \ H2 and in H2 \ H3 (and therefore also in H1 and H3 ). r 7! ab. So a D 1. any map fa1 .10a). Apply (1.] n 2 The quotient group Qn =hb 2 i has generators a and b. and so it sufﬁces to prove transitivity.118 B. : : : . and the only commutative free groups are 1 and C1 . 19 By assumption. then the word a ab 1 b 1 is reduced and ¤ 1. x ¤ 1. 1 D a 1 . G ! Dn . s 7! a. For j ¤ i . But b 3 D 1. By counting. s 2 . and so we can apply (1. then a D b. has length at least n.8). 23 (a) If a ¤ b. (c) The reduced form of x n .8 again) are identity maps. 21 The key point is that hai D ha2 i product. it also has ﬁnite index in each of H1 and H3 . this forces c D 1. Using this. We shall use that if a subgroup H of a group G has ﬁnite index in G. and so the group G in the problem is commutative. so let a 2 G. Therefore. a D xj axj a xj don’t commute. and therefore (2. (c) Suppose a is a nonempty reduced word in x1 . 25 The unique element of order 2 is b 2 . s j r n . the set G is nonempty. Therefore. and some power of this permutation is the identity permutation. : : : . b 7! s 1 r. According to (2. and so G has the universal property that characterizes the free abelian group on the generators ai . xn . then H \ G 0 has ﬁnite index in G 0 for any subgroup G 0 of G (because the natural map G 0 =H \ G 0 ! G=H is injective). n D 24 (a) Universality. But a3 D 1. and so b 2 lies in the centre. a 7! s. an g ! A with A commutative extends uniquely to homomorphism G ! A. say a D xi (or def 1 1 1 can’t be empty.8) that there are homomorphisms: Dn ! G. which is a presentation for D bab 2n 2 (see 2. (b) C1 C1 is commutative.9). (b) is similar. for some n 1. the map x 7! axW G ! G is a permutuation of G. The composites Dn ! G ! Dn and G ! Dn ! G are the both the identity map on generating elements.50) to see that D2n breaks up as a 22 Note ﬁrst that any group generated by a commuting set of elements must be commutative. S OLUTIONS TO THE E XERCISES 18 Commensurability is obviously reﬂexive and symmetric. it follows that if H1 and H3 are both commensurable with H2 . one sees that every element has a left inverse. . and so an is a left neutral element. The ﬁnal relation then gives b D 1.
119 28 The elements x 2 , xy, y 2 lie in the kernel, and it is easy to see that hx; yjx 2 ; xy; y 2 i has order (at most) 2, and so they must generate the kernel (at least as a normal group — the problem is unclear). One can prove directly that these elements are free, or else apply the NielsenSchreier theorem (2.6). Note that the formula on p. 34 (correctly) predicts that the kernel is free of rank 2 2 2 C 1 D 3 29 We have to show that if s and t are elements of a ﬁnite group satisfying t 1 s 3 t D s 5 , then the given element g is equal to 1. Because the group is ﬁnite, s n D 1 for some n. If 3jn, the proof is easy, and so we suppose that gcd.3; n/ D 1. But then 3r C nr 0 D 1, some r; r 0 2 Z; and so s 3r D s. Hence t Now, gDs
1 1 1 3r 1 3
st D t
1
s t D .t
1
s t /r D s 5r :
1
.t
s
1
t /s.t
st / D s
s
5r
ss 5r D 1;
as required. [In such a question, look for a pattern. Note that g has two conjugates in it, as does the relation for G, and so it is natural to try to relate them.] 31 Let N be the unique subgroup of order 2 in G. Then G=N has order 4, but there is no subgroup Q G of order 4 with Q \ N D 1 (because every group of order 4 contains a group of order 2), and so G ¤ N Q for any Q. A similar argument applies to subgroups N of order 4. 32 For any g 2 G, gM g 1 is a subgroup of order m, and therefore equals M . Thus M (similarly N ) is normal in G, and MN is a subgroup of G. The order of any element of M \ N divides gcd.m; n/ D 1, and so equals 1. Now (1.51) shows that M N MN , which therefore has order mn, and so equals G. 33 Show that GL2 .F2 / permutes the 3 nonzero vectors in F2 space over F2 ). F2 (2dimensional vector
34 The following solutions were suggested by readers. We write the quaternion group as Q D f˙1; ˙i; ˙j; ˙kg: (A) Take a cube. Write the six elements of Q of order 4 on the six faces with i opposite i , etc.. Each rotation of the cube induces an automorphism of Q, and Aut.Q/ is the symmetry group of the cube, S4 . (B) The group Q has one element of order 2, namely 1, and six elements of order 4, namely, ˙i , ˙j , ˙k. Any automorphism ˛ of Q must map 1 to itself and permute the elements of order 4. Note that ij D k, j k D i , ki D j , so ˛ must send the circularly ordered set i; j; k to a similar set, i.e., to one of the eight sets in the following table: i j k i j k i j k i j k i k j i k j i k j i k j Because ˛. 1/ D 1, ˛ must permute the rows of the table, and it is not difﬁcult to see that all permutations are possible.
120 35 The pair
B. S OLUTIONS TO THE E XERCISES
80 80 19 19 < 1 0 b = < a 0 0 = N D @0 1 c A and Q D @0 a 0 A : ; : ; 0 0 1 0 0 d satisﬁes the conditions (i), (ii), (iii) of (3.8). For example, for (i) (Maple says that) 1 10 10 0 a 0 b 1 0 b a 0 b @ 0 a c A @0 1 c A @ 0 a c A 0 0 d 0 0 1 0 0 d
1
1 0 @0 1 D 0 0
0
b d c d
1 1 C d .b C ab/ 1 C d .c C ac/ A 1
It is not a direct product of the two groups because it is not commutative. 36 Let g generate C1 . Then the only other generator is g 1 , and the only nontrivial automorphism is g 7! g 1 . Hence Aut.C1 / D f˙1g. The homomorphism S3 ! Aut.S3 / is injective because Z.S3 / D 1, but S3 has exactly 3 elements a1 ; a2 ; a3 of order 2 and 2 elements b; b 2 of order 3. The elements a1 ; b generate S3 , and there are only 6 possibilities for ˛.a1 /, ˛.b/, and so S3 ! Aut.S3 / is also onto. 37 (a) The element g o.q/ 2 N , and so has order dividing jN j. (c) The element g D .1; 4; 3/.2; 5/, and so this is obvious. (d) By the ﬁrst part, ..1; 0; : : : ; 0/; q/p D ..1; : : : ; 1/; 1/, and .1; : : : ; 1/ has order p in .Cp /p . (e) We have .n; q/.n; q/ D .nn 1 ; qq/ D .1; 1/: 38 Let n q 2 Z.G/. Then .n q/.1 q 0 / D n qq 0 0 /.n q/ D q 0 nq 0 1 q 0 q .1 q and .n q/.n0 1/ D nq n0 q 1 q .n0 1/.n q/ D n0 n q all q 0 2 Q H) n 2 CN .Q/ q 2 Z.Q/
1 0
n0 2 N
H) n
n n D q n0 q
1
:
The converse and the remaining statements are easy. 41 Let 'W G=H1 ! G=H2 be a Gmap, and let '.H1 / D gH2 . For a 2 G, '.aH1 / D a'.H1 / D agH2 . When a 2 H1 , '.aH1 / D gH2 , and so agH2 D gH2 ; hence g 1 ag 2 H2 , and so a 2 gH2 g 1 . We have shown H1 gH2 g 1 . Conversely, if g satisﬁes this condition, the aH1 7! agH2 is a welldeﬁned map of Gsets. 42 Let H be a proper subgroup of G, and let N D NG .H /. The number of conjugates of H is .G W N / Ä .G W H / (see 4.8). Since each conjugate of H has .H W 1/ elements and the conjugates overlap (at least) in f1g, we see that ˇ[ ˇ ˇ ˇ ˇ gHg 1 ˇ < .G W H /.H W 1/ D .G W 1/: For the second part, choose S to be a set of representatives for the conjugacy classes. 43 According to 4.17, 4.18, there is a normal subgroup N of order p 2 , which is commutative. Now show that G has an element c of order p not in N , and deduce that G D N hci, etc..
121 44 Let H be a subgroup of index p, and let N be the kernel of G ! Sym.G=H / — it is the largest normal subgroup of G contained in H (see 4.22). If N ¤ H , then .H W N / is divisible by a prime q p, and .G W N / is divisible by pq. But pq doesn’t divide pŠ — contradiction. 45 Embed G into S2m , and let N D A2m \ G. Then G=N ,! S2m =A2m D C2 , and so .G W N / Ä 2. Let a be an element of order 2 in G, and let b1 ; : : : ; bm be a set of right coset representatives for hai in G, so that G D fb1 ; ab1 ; : : : ; bm ; abm g. The image of a in S2m is the product of the m transpositions .b1 ; ab1 /; : : : ; .bm ; abm /, and since m is odd, this implies that a … N . 46 The set X of kcycles in Sn is normal, and so the group it generates is normal (1.38). But, when n 5, the only nontrivial normal subgroups of Sn are An and Sn itself. If k is odd, then X is contained in An , and if k is even, then it isn’t. 47 (a) The number of possible ﬁrst rows is 23 1; of second rows 23 2; of third rows 23 22 ; whence .G W 1/ D 7 6 4 D 168. (b) Let V D F3 . Then jV j D 23 D 8. Each line 2 through the origin contains exactly one point ¤ origin, and so jX j D 7. (c) We make a list of possible characteristic and minimal polynomials: 1 2 3 4 5 6 Characteristic poly. X3 C X2 C X C 1 X3 C X2 C X C 1 X3 C X2 C X C 1 X 3 C 1 D .X C 1/.X 2 C X C 1/ X 3 C X C 1 (irreducible) X 3 C X 2 C 1 (irreducible) Min’l poly. X C1 .X C 1/2 .X C 1/3 Same Same Same Size 1 21 42 56 24 24 Order of element in class 1 2 4 3 7 7
Here size denotes the number of elements in the conjugacy class. Case 5: Let ˛ be an endomorphism with characteristic polynomial X 3 C X C 1. Check from its minimal polynomial that ˛ 7 D 1, and so ˛ has order 7. Note that V is a free F2 Œ˛module of rank one, and so the centralizer of ˛ in G is F2 Œ˛ \ G D h˛i. Thus jCG .˛/j D 7, and the number of elements in the conjugacy class of ˛ is 168=7 D 24. Case 6: Exactly the same as Case 5. Case 4: Here V D V1 ˚ V2 as an F2 Œ˛module, and EndF2 Œ˛ .V / D EndF2 Œ˛ .V1 / ˚ EndF2 Œ˛ .V2 /: Deduce that jCG .˛/j D 3, and so the number of conjugates of ˛ is 168 D 56. Case 3: Here 3 CG .˛/ D F2 Œ˛ \ G D h˛i, which has order 4. Case 1: Here ˛ is the identity element. Case 2: Here V D V1 ˚ V2 as an F2 Œ˛module, where ˛ acts as 1 on V1 and has minimal polynomial X 2 C 1 on V2 . Either analyse, or simply note that this conjugacy class contains all the remaining elements. (d) Since 168 D 23 3 7, a proper nontrivial subgroup H of G will have order 2; 4; 8; 3; 6; 12; 24; 7; 14; 28; 56; 21; 24, or 84: If H is normal, it will be a disjoint union of f1g and some other conjugacy classes, and so P .N W 1/ D 1 C ci with ci equal to 21, 24, 42, or 56, but this doesn’t happen. 48 Since G=Z.G/ ,! Aut.G/, we see that G=Z.G/ is cyclic, and so by (4.19) that G is commutative. If G is ﬁnite and not cyclic, it has a factor Cpr Cps etc..
1j /. of order p m say. 413 Since Stab. and so they must be even (i.12/ .13/.67/ . n D 6. show that if the partition of n deﬁned by consists of distinct integers. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Cycle Size Parity Splits in A7 ‹ C7 .3456/ 504 O .x/=CG .x0 / then H Stab.67/ . if two cycles have the same odd length k.13/.34/. Now Stab.1i /.x/ W CG . and commutes with .x// D kc: 411 (a) The ﬁrst equivalence follows from the preceding problem.567/ 210 E N .x0 /g 1 .x/ \ H D CH .. b 7! .x//.x/. and so H=CH .34/.12/ 412 According to Gap.345/ 420 O . one can ﬁnd a product of k transpositions which interchanges them. 410 Note that CG . and so there can be only one. Prove each jKj D .4567/ 420 O . 51 Let p be a prime dividing jGj and let P be a Sylow psubgroup.12/.14/.1/ 1 E N .9) shows that G is a product of its Sylow subgroups.G W CG . conversely.34/.122 B.e.x//. parity.3456/ 630 E N . Hence any subgroup containing . therefore P must have an element of order p m . contrary to hypothesis.67/ .34/ 105 E N . Then H CG .456/ 280 E N . and it has at most 1CpC C pm 1 D pm 1 < pm p 1 elements of order dividing p m 1 .12345/ 504 E N .x// D .G W H CG . / contains . note that commutes with all cycles in its decomposition.26/. (b) List of conjugacy classes in S7 .x/ for all x. then commutes only with the group generated by the cycles in its cycle decomposition.12/ 21 O . : : : contains all transpositions.123/.x0 / D G. .12/.12/. and so H D 1.45/. For the second. if H Stab. and so H Stab.1234567/ 720 E Y 720 doesn’t divide 2520 .12/.1234/ 210 O .56/.12/.x0 / is maximal.12/.123/ 70 E N .25/.x/ class has the same number c of elements. which shows that H acts transitively. and we know Sn is generated by transpositions.ij / D .123/.1j /.12/.56/ 105 O .gx0 / D g Stab.123456/ 840 O . Now (5. S OLUTIONS TO THE E XERCISES 49 Clearly . and (when the parity is even) whether it splits in A7 .36/ .H CG . . and so it is cyclic. Each Sylow psubgroup has exactly p m elements of order dividing p m . have odd length).12/. their size. The elements of P all have order dividing p m . a 7! .
D4 and the quaternion group have isomorphic commutator subgroups and quotient groups but are not isomorphic. .123 62 No. Similarly. Sn and An C2 are not isomorphic when n 5.
.
(b) Describe the group with generators x and y and deﬁning relations yxy 1 D x xyx 1 D y 1 .123/ is A5 itself. Which of the following statements are true (give brief justiﬁcations for each of (a). where X is a set with n elements. You may use results proved in class or in the notes. If G is commutative and acts transitively on X.G W 1/ Ä n. b 3 D 1 H) . Show that if every commutative quotient of G is cyclic. 6. (a) If a and b are elements of a group.A PPENDIX C TwoHour Examination 1. 1.ab/6 D 1. How many Sylow 11subgroups can a group of order 110 D 2 5 11 have? Classify the groups of order 110 containing a subgroup of order 10. (a) Describe the group with generators x and y and deﬁning relation yxy 1 D x 1 . Let G be a ﬁnite nilpotent group. then G H . (b) For each m 1. (d) The only subgroup of A5 containing . 125 . then G itself is cyclic. 5. (c). 2. give a correct set of implications for (e)). Is the statement true for nonnilpotent groups? 4. and let P be a subgroup of H . : 3 4 5 6 7 2 1 2 3 1 5 6 7 4 (c) If G and H are ﬁnite groups and G A594 H A594 . Deduce that . show that each element g ¤ 1 of G moves every element of X .P /. (d). n (c) Show that a commutative subgroup of Sn has order Ä 3 3 . Assume that every automorphism of H is inner. but you should indicate clearly what you are using. Must every group of order 110 contain a subgroup of order 10? 3. (a) Let G be a subgroup of Sym. Let H be a normal subgroup of a group G. (b) The following two elements are conjugate in S7 : Â Ã Â Ã 1 2 3 4 5 6 7 1 2 3 4 5 6 7 . Prove that G D H NG . ﬁnd a commutative subgroup of S3m of order 3m . (b). (e) Nilpotent H) cyclic H) commutative H) solvable (for a ﬁnite group).X /. then a2 D 1.
From the two relations get yx D x 1 y. 2. bja2 . 4. i.Oi /. T WO H OUR E XAMINATION S OLUTIONS 1.C1 / D ˙1.! Sn . and so y 4 D 1. Then G has quotient H of class 2. (a) It’s the group . 6. Hence there is only one Sylow 11subgroup P . : : : .126 C. and let Gi be the image of G in Sym. No! In fact. (123)(4567). Fix an x 2 X. the cycle decompositions are (1357)(246). then g 7! gx W G ! X is injective. Or be the orbits of G. H D C10 or D5 : Now have to look at the maps Â W H ! Aut.G1 W 1/ . b 3 i. Alternatively. it’s not even true for solvable groups (e. and yx D x 1 y. which is a contradiction.3465/ and . the group it generates is proper. Consider 1 ! Z. Have GDP Â H. Then G .1287/.C11 / D C10 . Alternatively. (d) False.] (b) Partition the set into subsets of order 3. Let g 2 G. ab has inﬁnite order.! G1 Gr . the ToddCoxeter algorithm shows that it is the subgroup of S8 generated by .P /. and let G D G1 Gm . which shows that G=Z. and so (by induction).1584/.2673/.. (a) If gx D x. then ghx D hgx D hx. Then gP g 1 D hP h 1 . Suppose G has class > 1. . the elements can be written uniquely in the form x i y j . D y 2. : : : and divides 10. (b) It’s the quaternion group. n D . P D C11 . (b) True. G D hxi hyi D C1 Â C1 with ÂW C1 ! Aut.G/ is cyclic.g. 12. and let h 2 H be such that conjugation by h on H agrees with conjugation by g. (c) True. 3.G W 1/. (a) False: in ha. . by induction. by the SchurZassenhaus lemma. (e) Cyclic H) commutative H) nilpotent H) solvable.Gr W 1/ Ä 3 n1 3 3 nr 3 D 33 : n 5.H / ! 1: Then H is commutative by (4.H / ! H ! H=Z. and hence cyclic. Therefore G is commutative. The second relation implies xy 2 x 1 Dy 2 .G W 1/ Ä . and so g D 1. (c) Let O1 .17). [Note that Cayley’s theorem gives an embedding G . use the KrullSchmidt theorem. and so h 1 g 2 NG . Yes. Alternatively. j 2 Z. Hence g ﬁxes every element of X. S3 ). The number of Sylow 11subgroups s11 D 1. yx D xy 1 and so x 2 D y 2 .
1897. New York. 13:775–1029. Basel. Cambridge. 1999. E. A new look at the FeitThompson odd order theorem.. F EIT.Y. W. (2) 137:203–220. C URTIS . G LAUBERMAN . SpringerVerlag. American Mathematical Society.Bibliography A LPERIN . NJ. Schur.mat. Algebraic topology: An introduction. 16:73–92. RI. Mat. 1963. M. volume 15 of History of Mathematics. C. N. Brace & World. On groups of even order. AND S MITH . 112 of Mathematical Surveys and Monographs. 1995. of Math. Vol. M ASSEY.. D. B. H UMPHREYS . American Mathematical Society. 1998). II. R. 15th School of Algebra (Portuguese) (Canela. Harcourt.. The groups of order at most 2000. W. volume 29 of Cambridge Studies in Advanced Mathematics. Math. M. W. Providence. 1. The classiﬁcation of quasithin groups. J. H. H ALL . Enumerating ﬁnite groups of given order. B URNSIDE . Prentice Hall Inc. In Proceedings of the International Congress of Mathematicians. A. Ann. Inc. Electron. Groups and representations..unb. and Brauer. Paciﬁc J. I.. Soc. 2001. G. J R . B ESCHE . G. J. 1955. 1993. Cambridge. Contemp. Birkh¨ user. Available at www. u a F EIT. volume 111. E. 2 (Z¨ rich. Structure of strongly quasithin Kgroups. J. RI. A RTIN . U. 2004. volume 162 of Graduate Texts in Mathematics. B. L. 1967. K. 1991. 127 . A. E ICK . Burnside. of Math. A SCHBACHER . New York. Announc. L. The theory of groups. The Macmillan Co. 17–24. AND F OWLER . (2) 62:565–583. Cambridge: at the University Press. Theory of groups of ﬁnite order. AND T HOMPSON .ps. R. 1990. M. 1995. AND O’B RIEN . On the work of Eﬁm Zelmanov. P YBER . 1999. S. Cambridge University Press. Ann. pp.. Englewood Cliffs. AND B ELL . 7:1–4 (electronic). B RAUER . W.br/~matcont/16_5. Providence. 1959. Algebra. Pioneers of representation theory: Frobenius. Amer. S. New York. 1994). W. Reﬂection groups and Coxeter groups. Solvability of groups of odd order. Res. Math.
S ERRE . On a theorem of Jordan.) 38:315–352. J. 2003. Math. 2001. Trees.128 B IBLIOGRAPHY RONAN . SpringerVerlag. Amer. Math. J. M. J. 1872. 1980. New York. Soc. E. fourth edition. Bull. Soc. Math. SpringerVerlag. P. Symmetry and the monster.S. . A characterization of inner automorphisms. e e W ILD . S ERRE . A brief history of the classiﬁcation of the ﬁnite simple groups. ROTMAN . 2006. 101:226–228. 1987. Amer. Proc. Soc. Math. S OLOMON . Amer. M. (N.S. Math. S CHUPP. M. Amer.P. Monthly 112:20–31. 5:584–594.P. An introduction to the theory of groups. volume 148 of Graduate Texts in Mathematics. Ann. S YLOW. One of the greatest quests of mathematics. (N. R. Th´ or` mes sur les groupes de substitutions. 2005. Translated from the French by John Stillwell. The groups of order sixteen made easy. L. Berlin. Bull. Oxford.) 40:429–440. 1995. Oxford University Press. J.
42 Coxeter. 41 of a bilinear form. 54 exact sequence. 38. 102 disjoint cycles disjoint. 101 semisimple. 55 effective. 16 right. 24. 56 of subgroup. 55 trivial. 55 left. 53 free. 102 centre of a group. 8 elementary divisors. of a subalgebra. 99 quaternion. 62 Burnside. 53 linear. 66 automorphism inner. 101 algorithm ToddCoxeter. 26 equivariant map. 48 faithful representation. 58 commutator. 37 commutative. 53 faithful. 48 isomorphic. 38 cycle. 62 dimension. 54 transitive. 95 ﬂag full. 87 class equation class. 7 alternating. 12 class nilpotency. 16 Coxeter system. 37 extension central. 14 A. 35 group. 13 generators of a group. 48 split. 41 basis for a commutative group. 53 algebra division. 90 abelian. 76 Frattini’s argument. 54 Gset. 113 centralizer. 105 simple. 10 of a group. 48 exponent of a group. 69 centralizer of element. 55 imprimitive. 53 generates. 54 coset left. 62 element neutral. 41 outer. 96 primitive. 7 4. 30. 49 extension. of groups. 68 right. 9 complete. 89 Gmap. 68 kfold transitive. 83 conjugacy class. 25 block. 35 composition factors.Index action doubly transitive. 9 additive. 101 group. 38 129 .
9 primitive. 20 Klein Viergruppe. 61 presentation of a group. 8. 95 monoid free. 64 stabilized. 37 free. 35 problem Burnside. 79 of order pq. 37 word. 77 of order p. 98 quotients. 54 of symmetries. 95 simple. 37 product direct. 90 of groups. 95 orbit. 68 quaternion. 14 homogeneous. 92 isotropy. of Gsets. 21 groups of order 12. 59 of order p 3 . 10 with operators. 37 restricted Burnside. 10 p. 9 orthogonal. factor. 17 invariant factors. 86 length of a cycle. 19 soluble. m Ä 3. 8 normalizer of a subgroup. 15 index of a subgroup. 39 ﬁnitely presented. 14 generalized. 39 simple. 23 knit. 13 dihedral. 8 of an element. 81 solvable. 21 reﬂection. 68 permutation even. 54 module semisimple. 9. 35 quotient. 55 metabelian. of a permutation.. 90 group. 78 of small order. 8 permutation. 81 rank of a commutative group.130 cyclic. 80 of order 99. 33 free abelian. 14 length of a subnormal series. 26 inverse. 80 of order 2p. 31 negative. 61 I NDEX isomorphism of Gsets. 62 map. 87 metacyclic. 78 of order 60. 35 general linear. 48 semidirect. 15 kernel. 34 . 87 of rigid motions. 8 inversion. 44 ZappaSz´ p. 55 homomorphism admissible. 56 opposite algebra. 79 of order 2m p n . of a normal series of a normal series. 48 e projector. 17 of order p 2 . 38 of a free group. 83 special linear. 20 symplectic. 54 order of a group. 61 odd. 7 nilpotent. 66. 10 indecomposable. 26 of a Coxeter system. 115 multiplicative. 58 of order 30. 9 partition of a natural number. 83 solvable. 13 ﬁnite reﬂection. 54 of groups.
97 NielsenSchreier. 39 relations. 10 subset normal. 31 reduced. 55 stable subset stable. 90 characteristic. 83 subnormal. 35 representation linear. 90 admissible. 95 series admissible subnormal. 86 normal.Index reduced form of a word. 32 words equivalent. 13 normal. 26 structure of Coxeter groups. 12 Ainvariant. 73 torsion. 92 Lagrange. 81 solvable. 103 factorization of homomorphisms. 25 correspondence. 81 without repetitions. 81 signature. 81 derived. 39 Sylow I. 35 deﬁning. 77 transposition. 18 normal generated by a set. 59 commutative groups have bases. 17 Maschke. 22 FeitThompson. 87 composition. 43 commutator. 54 subgroup. 91 ascending central. 56 of an element. 85 generated by a set. 34 nilpotency condition. 101 stabilizer of a subset. 65 131 isomorphism. 84 Galois. 61 skew ﬁeld. 75 Sylow subgroups of subgroups. 69 SchurZassenhaus. 32 reﬂection. 22 double centralizer. 89 primitivity condition. 62 table multiplication. 16 centre of a pgroup. 86 Sylow p. 33 . 11 theorem Cauchy. 82 o KrullSchmidt. 20 second derived. 58 Cayley. 49 structure of commutative groups. 20 support of a cycle. 73 Sylow II. 85 ﬁrst derived. 14 word. 96 matrix. 22 JordanH¨ lder.
This action might not be possible to undo. Are you sure you want to continue?
Use one of your book credits to continue reading from where you left off, or restart the preview.