1 Definition of a Group Action
Let G be a group and X a set. A map G×X→X, written (g,x)↦g⋅x, is called a (left) action of G on X if the following two conditions hold:
e⋅x=x for all x∈X.
(gh)⋅x=g⋅(h⋅x) for all g,h∈G and x∈X.
When G acts on X, we call X a G-set.
2 Fundamental Examples
Setting g⋅x=x for all g∈G and x∈X defines a group action, called the trivial action.
The group G acts on itself by left multiplication: g⋅x=gx. Axiom (1): e⋅x=ex=x. Axiom (2): (gh)⋅x=(gh)x=g(hx)=g⋅(h⋅x).
G acts on itself by conjugation: g⋅x=gxg−1. Axiom (1): e⋅x=exe−1=x. Axiom (2): (gh)⋅x=(gh)x(gh)−1=g(hxh−1)g−1=g⋅(h⋅x).
G acts on the set of subgroups of G by g⋅H=gHg−1.
For H≤G, the group G acts on the set of left cosets G/H={gH∣g∈G} by g⋅(aH)=(ga)H.
Sn acts on {1,2,…,n} by σ⋅i=σ(i).
3 Orbits and Stabilizers
Let G act on X. The orbit of x∈X is Orb(x)=G⋅x={g⋅x∣g∈G}.
The stabilizer (or isotropy subgroup) of x∈X is Stab(x)=Gx={g∈G∣g⋅x=x}.
Stab(x) is a subgroup of G.
Since e⋅x=x, we have e∈Stab(x). If g,h∈Stab(x), then (gh−1)⋅x=g⋅(h−1⋅x)=g⋅x=x, where the equality h−1⋅x=x follows from h⋅(h−1⋅x)=(hh−1)⋅x=e⋅x=x. Hence gh−1∈Stab(x). □
The relation x∼y⟺y∈Orb(x) is an equivalence relation on X, and the orbits form a partition of X: X=i⨆Orb(xi)(choosing one representative from each orbit).
Reflexivity: x=e⋅x∈Orb(x). Symmetry: if y=g⋅x, then x=g−1⋅y∈Orb(y). Transitivity: if y=g⋅x and z=h⋅y, then z=(hg)⋅x∈Orb(x). □
4 The Orbit–Stabilizer Theorem
Let G act on X, and let x∈X. Then ∣Orb(x)∣=[G:Stab(x)].
In particular, if G is finite, ∣Orb(x)∣=∣G∣/∣Stab(x)∣.
We construct a bijection f:G/Stab(x)→Orb(x) by setting f(gStab(x))=g⋅x.Well-definedness. If gStab(x)=hStab(x), then h−1g∈Stab(x), so (h−1g)⋅x=x, whence g⋅x=h⋅x.Injectivity. If g⋅x=h⋅x, then (h−1g)⋅x=x, so h−1g∈Stab(x), giving gStab(x)=hStab(x).Surjectivity. Every g⋅x∈Orb(x) is the image of gStab(x). □
S3 acts naturally on {1,2,3}. For x=1: Orb(1)={1,2,3} (the full set), and Stab(1)={e,(23)}. Indeed, ∣Orb(1)∣=3=6/2=∣S3∣/∣Stab(1)∣.
5 Counting Orbits: Burnside's Lemma
For g∈G, the fixed-point set of g is Fix(g)=Xg={x∈X∣g⋅x=x}.
Let G be a finite group acting on a finite set X. The number of orbits ∣X/G∣ is given by ∣X/G∣=∣G∣1g∈G∑∣Fix(g)∣.
Count the elements of the set S={(g,x)∈G×X∣g⋅x=x} in two ways. Summing over g gives ∑g∈G∣Fix(g)∣. Summing over x gives ∑x∈X∣Stab(x)∣.By the orbit–stabilizer theorem, ∣Stab(x)∣=∣G∣/∣Orb(x)∣, so x∈X∑∣Stab(x)∣=x∈X∑∣Orb(x)∣∣G∣=∣G∣x∈X∑∣Orb(x)∣1.
Elements in the same orbit contribute the same value 1/∣Orb(x)∣, and each orbit of size k contributes k⋅1/k=1. So the right-hand side equals ∣G∣⋅∣X/G∣. □
6 Conjugacy Classes and the Class Equation
Applying the orbit–stabilizer theorem to the conjugation action g⋅x=gxg−1 yields powerful structural results.
The orbit of an element a under the conjugation action is called its conjugacy class: C(a)={gag−1∣g∈G}.
The stabilizer of a under conjugation is called the centralizer of a: CG(a)={g∈G∣gag−1=a}={g∈G∣ga=ag}.
The center of G is Z(G)={z∈G∣zg=gz for all g∈G}=a∈G⋂CG(a).
One readily verifies that Z(G)⊴G.
Let G be a finite group. Decomposing G into conjugacy classes gives ∣G∣=∣Z(G)∣+i∑[G:CG(ai)],
where the sum runs over one representative ai from each conjugacy class of size greater than 1 (i.e. ai∈/Z(G)). Each summand [G:CG(ai)]≥2 and divides ∣G∣.
G decomposes as a disjoint union of conjugacy classes. The size of the conjugacy class of a is ∣C(a)∣=[G:CG(a)] (by the orbit–stabilizer theorem). A conjugacy class has size 1 precisely when a∈Z(G). Collecting the size-1 classes and the others: ∣G∣=∣Z(G)∣+i∑∣C(ai)∣=∣Z(G)∣+i∑[G:CG(ai)].
□
7 Properties ofp-Groups
For a prime p, a finite group G with ∣G∣=pn (n≥1) is called a p-group.
Let G be a p-group with ∣G∣=pn, n≥1. Then Z(G)={e}.
In the class equation ∣G∣=∣Z(G)∣+∑i[G:CG(ai)], each term [G:CG(ai)] divides ∣G∣=pn (by Lagrange's theorem) and is at least 2, so p∣[G:CG(ai)]. Since p∣∣G∣: ∣Z(G)∣=∣G∣−i∑[G:CG(ai)]≡0(modp).
Since e∈Z(G), we have ∣Z(G)∣≥1, and combined with p∣∣Z(G)∣, we get ∣Z(G)∣≥p. In particular, Z(G)={e}. □
If ∣G∣=p2, then G is abelian.
Since Z(G)={e}, we have ∣Z(G)∣=p or p2. If ∣Z(G)∣=p2, then Z(G)=G and G is abelian.Suppose ∣Z(G)∣=p. Then ∣G/Z(G)∣=p, so G/Z(G) is cyclic; say G/Z(G)=⟨gZ(G)⟩. Every element of G can be written as giz for some integer i and some z∈Z(G). For any two elements a=giz1 and b=gjz2 (with z1,z2∈Z(G)): ab=giz1gjz2=gi+jz1z2=gjz2giz1=ba,
since elements of Z(G) commute with everything. This shows G is abelian, so Z(G)=G, contradicting ∣Z(G)∣=p. □
Let G be a p-group acting on a finite set X. Let XG={x∈X∣g⋅x=x for all g∈G} denote the set of fixed points. Then ∣XG∣≡∣X∣(modp).
In the orbit decomposition X=XG⊔⨆iOrb(xi) (where ∣Orb(xi)∣≥2), each orbit size ∣Orb(xi)∣=[G:Stab(xi)] divides ∣G∣=pn and is at least 2, hence is divisible by p. Therefore: ∣X∣=∣XG∣+i∑∣Orb(xi)∣≡∣XG∣(modp).
□
8 Cayley's Theorem
Every group G is isomorphic to a subgroup of a symmetric group. If G is finite with ∣G∣=n, then G is isomorphic to a subgroup of Sn.
Let G act on itself by left multiplication: g⋅x=gx. For each g∈G, define σg:G→G by σg(x)=gx. This is a bijection (with inverse σg−1), so σg∈SG.Define φ:G→SG by φ(g)=σg. This is a homomorphism: φ(gh)(x)=σgh(x)=(gh)x=g(hx)=σg(σh(x))=(φ(g)∘φ(h))(x).
Furthermore, kerφ={g∈G∣gx=x for all x∈G}={e}, so φ is injective. By the first isomorphism theorem, G≅Imφ≤SG. □
9 Summary and Next Steps
In this chapter we have covered:
The definition of a group action and its fundamental examples.
Orbits, stabilizers, and the orbit–stabilizer theorem.
Burnside's lemma for counting orbits.
Conjugacy classes, centralizers, the center, and the class equation.
The center of a p-group is nontrivial; groups of order p2 are abelian.
The fixed-point theorem for p-groups.
Cayley's theorem.
With the class equation and the fixed-point theorem for p-groups in hand, we are now ready to prove one of the pinnacles of finite group theory: the Sylow theorems (existence, conjugacy, and the counting theorem for Sylow p-subgroups).