Skip to main content
Logo image

Section 23.1 Field Automorphisms

Our first task is to establish a link between group theory and field theory by examining automorphisms of fields.

Proof.

If σ and τ are automorphisms of F, then so are στ and σ1. The identity is certainly an automorphism; hence, the set of all automorphisms of a field F is indeed a group.

Proof.

We need only show that the set of automorphisms of E that fix F elementwise is a subgroup of the group of all automorphisms of E. Let σ and τ be two automorphisms of E such that σ(α)=α and τ(α)=α for all αF. Then στ(α)=σ(α)=α and σ1(α)=α. Since the identity fixes every element of E, the set of automorphisms of E that leave elements of F fixed is a subgroup of the entire group of automorphisms of E.
Let E be a field extension of F. We will denote the full group of automorphisms of E by Aut(E). We define the Galois group of E over F to be the group of automorphisms of E that fix F elementwise; that is,
G(E/F)={σAut(E):σ(α)=α for all αF}.
If f(x) is a polynomial in F[x] and E is the splitting field of f(x) over F, then we define the Galois group of f(x) to be G(E/F).

Example 23.3.

Complex conjugation, defined by σ:a+biabi, is an automorphism of the complex numbers. Since
σ(a)=σ(a+0i)=a0i=a,
the automorphism defined by complex conjugation must be in G(C/R).

Example 23.4.

Consider the fields QQ(5)Q(3,5). Then for a,bQ(5),
σ(a+b3)=ab3
is an automorphism of Q(3,5) leaving Q(5) fixed. Similarly,
τ(a+b5)=ab5
is an automorphism of Q(3,5) leaving Q(3) fixed. The automorphism μ=στ moves both 3 and 5. It will soon be clear that {id,σ,τ,μ} is the Galois group of Q(3,5) over Q. The following table shows that this group is isomorphic to Z2×Z2.
idστμididστμσσidμτττμidσμμτσid
We may also regard the field Q(3,5) as a vector space over Q that has basis {1,3,5,15}. It is no coincidence that |G(Q(3,5)/Q)|=[Q(3,5):Q)]=4.

Proof.

Let
f(x)=a0+a1x+a2x2++anxn
and suppose that αE is a zero of f(x). Then for σG(E/F),
0=σ(0)=σ(f(α))=σ(a0+a1α+a2α2++anαn)=a0+a1σ(α)+a2[σ(α)]2++an[σ(α)]n;
therefore, σ(α) is also a zero of f(x).
Let E be an algebraic extension of a field F. Two elements α,βE are conjugate over F if they have the same minimal polynomial. For example, in the field Q(2) the elements 2 and 2 are conjugate over Q since they are both roots of the irreducible polynomial x22.
A converse of the last proposition exists. The proof follows directly from Lemma 21.32.

Proof.

We will use mathematical induction on [E:F]. If [E:F]=1, then E=F and there is nothing to show. If [E:F]>1, let f(x)=p(x)q(x), where p(x) is irreducible of degree d. We may assume that d>1; otherwise, f(x) splits over F and [E:F]=1. Let α be a root of p(x). If ϕ:F(α)E is any injective homomorphism, then ϕ(α)=β is a root of p(x), and ϕ:F(α)F(β) is a field automorphism. Since f(x) has no repeated roots, p(x) has exactly d roots βE. By Proposition 23.5, there are exactly d isomorphisms ϕ:F(α)F(βi) that fix F, one for each root β1,,βd of p(x) (see Figure 23.8).
EψEF(α)ϕF(β)FidentityF
Figure 23.8. Commutative diagram for Theorem 23.7
Since E is a splitting field of f(x) over F, it is also a splitting field over F(α). Similarly, E is a splitting field of f(x) over F(β). Since [E:F(α)]=[E:F]/d, induction shows that each of the d isomorphisms ϕ has exactly [E:F]/d extensions, ψ:EE, and we have constructed [E:F] isomorphisms that fix F. Finally, suppose that σ is any automorphism fixing F. Then σ restricted to F(α) is ϕ for some ϕ:F(α)F(β).

Proof.

Let p be the characteristic of E and F and assume that the orders of E and F are pm and pn, respectively. Then nk=m. We can also assume that E is the splitting field of xpmx over a subfield of order p. Therefore, E must also be the splitting field of xpmx over F. Applying Theorem 23.7, we find that |G(E/F)|=k.
To prove that G(E/F) is cyclic, we must find a generator for G(E/F). Let σ:EE be defined by σ(α)=αpn. We claim that σ is the element in G(E/F) that we are seeking. We first need to show that σ is in Aut(E). If α and β are in E,
σ(α+β)=(α+β)pn=αpn+βpn=σ(α)+σ(β)
by Lemma 22.3. Also, it is easy to show that σ(αβ)=σ(α)σ(β). Since σ is a nonzero homomorphism of fields, it must be injective. It must also be onto, since E is a finite field. We know that σ must be in G(E/F), since F is the splitting field of xpnx over the base field of order p. This means that σ leaves every element in F fixed. Finally, we must show that the order of σ is k. By Theorem 23.7, we know that
σk(α)=αpnk=αpm=α
is the identity of G(E/F). However, σr cannot be the identity for 1r<k; otherwise, xpnrx would have pm roots, which is impossible.

Example 23.10.

We can now confirm that the Galois group of Q(3,5) over Q in Example 23.4 is indeed isomorphic to Z2×Z2. Certainly the group H={id,σ,τ,μ} is a subgroup of G(Q(3,5)/Q); however, H must be all of G(Q(3,5)/Q), since
|H|=[Q(3,5):Q]=|G(Q(3,5)/Q)|=4.

Example 23.11.

Let us compute the Galois group of
f(x)=x4+x3+x2+x+1
over Q. We know that f(x) is irreducible by Exercise 17.5.20 in Chapter 17. Furthermore, since (x1)f(x)=x51, we can use DeMoivre’s Theorem to determine that the roots of f(x) are ωi, where i=1,,4 and
ω=cos(2π/5)+isin(2π/5).
Hence, the splitting field of f(x) must be Q(ω). We can define automorphisms σi of Q(ω) by σi(ω)=ωi for i=1,,4. It is easy to check that these are indeed distinct automorphisms in G(Q(ω)/Q). Since
[Q(ω):Q]=|G(Q(ω)/Q)|=4,
the σi’s must be all of G(Q(ω)/Q). Therefore, G(Q(ω)/Q)Z4 since ω is a generator for the Galois group.

Subsection Separable Extensions

Many of the results that we have just proven depend on the fact that a polynomial f(x) in F[x] has no repeated roots in its splitting field. It is evident that we need to know exactly when a polynomial factors into distinct linear factors in its splitting field. Let E be the splitting field of a polynomial f(x) in F[x]. Suppose that f(x) factors over E as
f(x)=(xα1)n1(xα2)n2(xαr)nr=i=1r(xαi)ni.
We define the multiplicity of a root αi of f(x) to be ni. A root with multiplicity 1 is called a simple root. Recall that a polynomial f(x)F[x] of degree n is separable if it has n distinct roots in its splitting field E. Equivalently, f(x) is separable if it factors into distinct linear factors over E[x]. An extension E of F is a separable extension of F if every element in E is the root of a separable polynomial in F[x]. Also recall that f(x) is separable if and only if gcd(f(x),f(x))=1 (Lemma 22.5).

Proof.

First assume that charF=0. Since degf(x)<degf(x) and f(x) is irreducible, the only way gcd(f(x),f(x))1 is if f(x) is the zero polynomial; however, this is impossible in a field of characteristic zero. If charF=p, then f(x) can be the zero polynomial if every coefficient of f(x) is a multiple of p. This can happen only if we have a polynomial of the form f(x)=a0+a1xp+a2x2p++anxnp.
Certainly extensions of a field F of the form F(α) are some of the easiest to study and understand. Given a field extension E of F, the obvious question to ask is when it is possible to find an element αE such that E=F(α). In this case, α is called a primitive element. We already know that primitive elements exist for certain extensions. For example,
Q(3,5)=Q(3+5)
and
Q(53,5i)=Q(56i).
Corollary 22.12 tells us that there exists a primitive element for any finite extension of a finite field. The next theorem tells us that we can often find a primitive element.

Proof.

We already know that there is no problem if F is a finite field. Suppose that E is a finite extension of an infinite field. We will prove the result for F(α,β). The general case easily follows when we use mathematical induction. Let f(x) and g(x) be the minimal polynomials of α and β, respectively. Let K be the field in which both f(x) and g(x) split. Suppose that f(x) has zeros α=α1,,αn in K and g(x) has zeros β=β1,,βm in K. All of these zeros have multiplicity 1, since E is separable over F. Since F is infinite, we can find an a in F such that
aαiαββj
for all i and j with j1. Therefore, a(ββj)αiα. Let γ=α+aβ. Then
γ=α+aβαi+aβj;
hence, γaβjαi for all i,j with j1. Define h(x)F(γ)[x] by h(x)=f(γax). Then h(β)=f(α)=0. However, h(βj)0 for j1. Hence, h(x) and g(x) have a single common factor in F(γ)[x]; that is, the minimal polynomial of β over F(γ) must be linear, since β is the only zero common to both g(x) and h(x). So βF(γ) and α=γaβ is in F(γ). Hence, F(α,β)=F(γ).