This action might not be possible to undo. Are you sure you want to continue?

BooksAudiobooksComicsSheet Music### Categories

### Categories

### Categories

Editors' Picks Books

Hand-picked favorites from

our editors

our editors

Editors' Picks Audiobooks

Hand-picked favorites from

our editors

our editors

Editors' Picks Comics

Hand-picked favorites from

our editors

our editors

Editors' Picks Sheet Music

Hand-picked favorites from

our editors

our editors

Top Books

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Audiobooks

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Comics

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Sheet Music

What's trending, bestsellers,

award-winners & more

award-winners & more

Welcome to Scribd! Start your free trial and access books, documents and more.Find out more

J.S. MILNE

Abstract. These are the notes for the second part of Math 594, University of Michigan, Winter 1994, exactly as they were handed out during the course except for some minor corrections. Please send comments and corrections to me at jmilne@umich.edu using “Math594” as the subject. v2.01 (August 21, 1996). First version on the web. v2.02 (May 27, 1998). About 40 minor corrections (thanks to Henry Kim).

Contents 1. Extensions of Fields 1.1. Deﬁnitions 1.2. The characteristic of a ﬁeld 1.3. The polynomial ring F [X] 1.4. Factoring polynomials 1.5. Extension ﬁelds; degrees 1.6. Construction of some extensions 1.7. Generators of extension ﬁelds 1.8. Algebraic and transcendental elements 1.9. Transcendental numbers 1.10. Constructions with straight-edge and compass. 2. Splitting Fields; Algebraic Closures 2.1. Maps from simple extensions. 2.2. Splitting ﬁelds 2.3. Algebraic closures 3. The Fundamental Theorem of Galois Theory 3.1. Multiple roots 3.2. Groups of automorphisms of ﬁelds 3.3. Separable, normal, and Galois extensions 3.4. The fundamental theorem of Galois theory 3.5. Constructible numbers revisited 3.6. Galois group of a polynomial 3.7. Solvability of equations

Copyright 1996 J.S. Milne. You may make one copy of these notes for your own personal use.

i

1 1 1 2 2 4 4 5 6 8 9 12 12 13 14 18 18 19 21 23 26 26 27

ii

J.S. MILNE

4. Computing Galois Groups. 4.1. When is Gf ⊂ An ? 4.2. When is Gf transitive? 4.3. Polynomials of degree ≤ 3 4.4. Quartic polynomials 4.5. Examples of polynomials with Sp as Galois group over Q 4.6. Finite ﬁelds 4.7. Computing Galois groups over Q 5. Applications of Galois Theory 5.1. Primitive element theorem. 5.2. Fundamental Theorem of Algebra 5.3. Cyclotomic extensions 5.4. Independence of characters 5.5. Hilbert’s Theorem 90. 5.6. Cyclic extensions. 5.7. Proof of Galois’s solvability theorem 5.8. The general polynomial of degree n Symmetric polynomials The general polynomial A brief history 5.9. Norms and traces 5.10. Inﬁnite Galois extensions (sketch) 6. Transcendental Extensions

28 28 29 29 29 31 32 33 36 36 38 39 41 42 44 45 46 46 47 49 49 52 54

FIELDS AND GALOIS THEORY

1

1. Extensions of Fields 1.1. Deﬁnitions. A ﬁeld is a set F with two composition laws + and · such that (a) (F, +) is an abelian group; (b) let F × = F − {0}; then (F × , ·) is an abelian group; (c) (distributive law) for all a, b, c ∈ F , (a + b)c = ac + bc (hence also a(b + c) = ab + ac). Equivalently, a ﬁeld is a nonzero commutative ring (meaning with 1) such that every nonzero element has an inverse. A ﬁeld contains at least two distinct elements, 0 and 1. The smallest, and one of the most important, ﬁelds is F2 = Z/2Z = {0, 1}. Lemma 1.1. A commutative ring R is a ﬁeld if and only if it has no ideals other than (0) and R. Proof. Suppose R is a ﬁeld, and let I be a nonzero ideal in R. If a is a nonzero element of I, then 1 = a−1a ∈ I, and so I = R. Conversely, suppose R is a commutative ring with no nontrivial ideals; if a = 0, then (a) = R, which means that there is a b in F such that ab = 1. Example 1.2. The following are ﬁelds: Q, R, C, Fp = Z/pZ. A homomorphism of ﬁelds α : F → F is simply a homomorphism of rings, i.e., it is a map with the properties α(a + b) = α(a) + α(b), α(ab) = α(a)α(b), α(1) = 1, all a, b ∈ F.

Such a homomorphism is always injective, because the kernel is a proper ideal (it doesn’t contain 1), which must therefore be zero. 1.2. The characteristic of a ﬁeld. The map Z → F, n → 1F + 1F + · · · + 1F (ntimes),

is a homomorphism of rings. Case 1: Kernel = (0); then n · 1F = 0 =⇒ n = 0 (in Z). The map Z → F extends to a homomorphism Q → F , m → (m · 1F )(n · 1F )−1 . Thus F contains a copy of Q. In this case, n we say that F has characteristic zero. Case 2: Kernel = (0), i.e., n · 1F = 0 some n = 1. The smallest such n will be a prime p (else F will have nonzero zero-divisors), and p generates the kernel. In this case, {m · 1F | m ∈ Z} ≈ Fp , and F contains a copy of Fp . We say that F has characteristic p. The ﬁelds Fp , p prime, and Q are called the prime ﬁelds. Every ﬁeld contains a copy of one of them. Remark 1.3. The binomial theorem m m−1 m m−r r a b + ··· + a b + · · · + bm (a + b)m = am + 1 r holds in any ring. If p is prime, then p| p for all r, 1 ≤ r ≤ p − 1. Therefore, when F has r characteristic p, (a + b)p = ap + bp . Hence a → ap is a homomorphism F → F , called the Frobenius endomorphism of F . When F is ﬁnite, it is an isomorphism, called the Frobenius automorphism.

2

J.S. MILNE

1.3. The polynomial ring F [X]. I shall assume everyone knows the following (see Jacobson Chapter II, or Math 593). (a) Let I be a nonzero ideal in F [X]. If f(X) is a nonzero polynomial of least degree in I, then I = (f(X)). When we choose f to be monic, i.e., to have leading coeﬃcient one, it is uniquely determined by I. There is a one-to-one correspondence between the nonzero ideals of F [X] and the monic polynomials in F [X]. The prime ideals correspond to the irreducible monic polynomials. (b) Division algorithm: given f(X) and g(X) ∈ F [X] with g = 0, we can ﬁnd q(X) and r(X) ∈ F [X] with deg(r) < deg(g) such that f = gq + r; moreover, q(X) and r(X) are uniquely determined. Thus the ring F [X] is a Euclidean domain. (c) Euclid’s algorithm: Let f and g ∈ F [X] have gcd d(X); the algorithm gives polynomials a(X) and b(X) such that a(X) · f(X) + b(X) · g(X) = d(X), deg(a) ≤ deg(g), deg(b) ≤ deg(f). Recall how it goes. Using the division algorithm, we construct a sequence of quotients and remainders: f g r0 rn−2 rn−1 Then rn = gcd(f, g), and rn = rn−2 − qn rn−1 = rn−2 − qn (rn−3 − qn−1 rn−2 ) = · · · = af + bg. Maple knows Euclid’s algorithm—to learn its syntax, type “?gcdex;”. (d) Since F [X] is an integral domain, we can form its ﬁeld of fractions F (X). It consists of quotients f(X)/g(X), f and g polynomials, g = 0. 1.4. Factoring polynomials. It will frequently be important for us to know whether a polynomial is irreducible and, if it isn’t, what its factors are. The following results help.

c Proposition 1.4. Suppose r = d , c, d ∈ Z, gcd(c, d) = 1, is a root of a polynomial

= = = ··· = =

q 0 g + r0 q 1 r0 + r1 q 2 r1 + r2 qn rn−1 + rn qn+1 rn .

am X m + am−1 X m−1 + · · · + a0, Then c|a0 and d|am . Proof. It is clear from the equation

ai ∈ Z.

am cm + am−1 cm−1 d + · · · + a0dm = 0 that d|am cm , and therefore, d|am . The proof that c|a0 is similar. Example 1.5. The polynomial X 3 −3X−1 is irreducible in Q[X] because its only possible roots are ±1 (and they aren’t). Proposition 1.6. Let f(X) ∈ Z[X] be such that its coeﬃcients have greatest common divisor 1. If f(X) factors nontrivially in Q[X], then it factors nontrivially in Z[X]; moreover, if f(X) ∈ Z[X] is monic, then any monic factor of f(X) in Q[X] lies in Z[X].

FIELDS AND GALOIS THEORY

3

Proof. Use Gauss’s lemma (see Jacobson, 2.16, or Math 593). Proposition 1.7. (Eisenstein criterion) Let f = am X m + am−1 X m−1 + · · · + a0, suppose that there is a prime p such that: p does not divide am , p divides am−1 , ..., a0, p2 does not divide a0. Then f is irreducible in Q[X]. Proof. We may remove any common factor from the coeﬃcients f, and hence assume that they have gcd = 1. Therefore, if f(X) factors in Q[X], it factors in Z[X]: am X m + am−1 X m−1 + · · · + a0 = (bn X n + · · · + b0 )(cr X r + · · · + c0 ), bi , ci ∈ Z, n, r < m. Since p, but not p2 , divides a0 = b0 c0 , p must divide exactly one of b0 , c0 , say p divides b0. Now from the equation a 1 = b 0 c1 + b 1 c0 , we see that p|b1 . Now from the equation a 2 = b 0 c2 + b 1 c1 + b 2 c0 , we see that p|b2. By continuing in this way, we ﬁnd that p divides b0, b1 , . . . , bn , which contradicts the fact that p does not divide am . The above three propositions hold with Z replaced by any unique factorization domain. Proposition 1.8. There is an algorithm for factoring a polynomial in Q[X]. Proof. Consider f(X) ∈ Q[X]. Multiply f(X) by an integer, so that it is monic, and then replace it by Ddeg(f )f( X ), D = a common denominator for the coeﬃcients of f, to obtain D a monic polynomial with integer coeﬃcients. Thus we need consider only polynomials f(X) = X m + a1X m−1 + · · · + am , ai ∈ Z. From the fundamental theorem of algebra (see later), we know that f splits completely in C[X]:

m

ai ∈ Z;

f(X) =

i=1

(X − αi ),

αi ∈ C.

From the equation f(αi ) = 0, it follows that |αi | is less than some bound M depending on a1, . . . , am . Now if g(X) is a monic factor of f(X), then its roots in C are certain of the αi , and its coeﬃcients are symmetric polynomials in its roots. Therefore the absolute values of the coeﬃcients of g(X) are bounded. Since they are also integers (by 1.6), we see that there are only ﬁnitely many possibilities for g(X). Thus, to ﬁnd the factors of f(X) we (better Maple) only have to do a ﬁnite amount of checking. One other observation is sometimes useful: Suppose that the leading coeﬃcient of f(X) ∈ Z[X] is not divisible by the prime p; if f(X) is irreducible in Fp [X], then it is irreducible in Z[X]. Unfortunately, this test is not always eﬀective: for example, X 4 − 10X 2 + 1 is reducible1 modulo every prime, but it is irreducible in Q[X].

1

I don’t know an elementary proof of this. One proof uses that its Galois group is ≈ (Z/2Z)2.

are linearly independent (see section 1. X. . which implies that any ﬁnite-dimensional vector space over Q is countable. Let f(X) ∈ F [X] be a monic polynomial of degree m. Then. 1. and write x for the image of X in F [X]/(f(X)). I claim that {ei j } is a basis for L over F.10. (d) The ﬁeld F (X) has inﬁnite degree over F . if L is of ﬁnite degree over F . Thus. (c) The ﬁeld of Gaussian numbers Q(i) =df {a + bi ∈ C | a.e. We write [E : F ] for the dimension (possibly inﬁnite) of E as an F -vector space. (b) The ﬁeld of real numbers R has inﬁnite degree over Q.9. but R is not countable. MILNE Maple knows how to factor polynomials in Q[X] and in F [X]. Then: (a) The map P (X) → P (x) : F [X] → F [x] . Construction of some extensions. i. will ﬁnd the factors of X 2 + 3X + 3 modulo 7. some αj ∈ E. Such an E can be regarded (in an obvious fashion) as an F -vector space. we ﬁnd that γ= aij ei j . in which case [L : F ] = [L : E][E : F ]. Extension ﬁelds. . one can ﬁnd real numbers α such that 1. . for each j. i. More explicitly. we need not concern ourselves with the problem of factorizing polynomials in Q[X] or Fp [X]. ..4 J. (It contains the F -subspace F [X]. and call [E : F ] the degree of E over F . α. . γ= αj j . Moreover. and the linear independence of the ei’s now shows that each aij = 0. (a) The ﬁeld of complex numbers C has degree 2 over R (basis {1. some aij ∈ F . α2 . 1.S. then it is certainly of ﬁnite degree over E.6. Let L ⊃ E ⊃ F (all ﬁelds). Example 1. Proof.9 below)).5.) Proposition 1. Conversely. We often say that E is ﬁnite over F when it has ﬁnite degree over F. degrees. On putting these together. i}). Then L/F is of ﬁnite degree ⇐⇒ L/E and E/F are both of ﬁnite degree. will ﬁnd the factors of 6X 2 + 18X − 24. i}).e. and >Factor(X^2+3*X+3) mod 7. which has the inﬁnite basis {1. being a subspace of a ﬁnite dimensional F -space. For example >factor(6*X^2+18*X-24). (We know Q is countable. in F7 [X]. and let (f) be the ideal generated by f. Assume that L/E and E/F are of ﬁnite degree. αj = aij ei . E. Consider the quotient ring F [X]/(f(X)). and let {ei} be a basis for E/F and { j } a basis for L/E. x is the coset X + (f(X)). A ﬁeld E containing a ﬁeld F is called an extension (ﬁeld) of F . }. b ∈ Q} has degree 2 over Q (basis {1. X 2 . I ﬁrst show that it spans L. Let γ ∈ L. and because {ei } spans E as an F -vector space. A linear relation rewritten j ( i aij ei) j = 0.. . The linear independence of the j ’s now shows that i aij ei = 0 for each j. because { j } spans L as an E-vector space. is also ﬁnite dimensional. aij ei j = 0 can be Next I show that {ei j } is linearly independent.

is obvious. This is irreducible over Q. We usually write i for x and C for R[x]. (c) The addition of two elements. In fact. F [x] = F [X]/(f(X)) is a ﬁeld of degree m over F . Example 1. Hence each element of F [x] can be written uniquely as a sum a0 + a1x + · · · + am−1 xm−1 . set α = g(x) where g(X) is a polynomial of degree ≤ m − 1. We call it the subring of E generated by F and S. Hence b(x) is the inverse of g(x). x. On replacing X with x in the equation. multiplication: (a + bx)(a + b x) = (aa − bb ) + (ab + a b)x. and let S be a subset of E. multiply in the usual way. + 28 ) 111 = 1. Because X 3 − 3X − 1 is irreducible.e. we ﬁnd b(x)g(x) = 1. gcd(X 3 − 3X − 1. addition: obvious. Let β = x4 + 2x3 + 3 ∈ Q[x]. To ﬁnd the inverse of an element α ∈ F [x]. and use the relation f(x) = 0 to express the monomials of degree ≥ m in x in terms of lower degree monomials. i. and we write it F [S]. b ∈ R. Conclusion: For any monic irreducible polynomial f(X) ∈ F [X].12.7. write α in the form (*).FIELDS AND GALOIS THEORY 5 is a surjective homomorphism. Then R[x] has: elements: a + bx. we have f(x) = 0. Example 1. . 3X 2 + 7X + 5) = 1. Euclid’s algorithm (courtesy of Maple) gives (X 3 − 3X − 1)( −7 X + 37 Hence 7 (3x2 + 7x + 5)( 111 x2 − 26 x 111 29 ) 111 7 + (3X 2 + 7X + 5)( 111 X 2 − 26 X 111 + 28 ) 111 = 1. x2} as a Q-vector space. and so Q[x] has basis {1. (b) From the division algorithm. (*). 1. The intersection of all the subrings of E containing F and S is again a subring of E (containing F and S). then we know how to compute in F [x]. (e) Now assume f(X) is irreducible. Let E be an extension ﬁeld of F . we ﬁnd that β = 3x2 + 7x + 5. In our case. Generators of extension ﬁelds. Moreover. ai ∈ F. a..11. 3 Then using that x − 3x − 1 = 0. we have found the inverse of β. written in the form (*). if we know how to compute in F . Then use Euclid’s algorithm in F [X] to obtain polynomials a(X) and b(X) such that a(X)f(X) + b(X)g(X) = d(X) with d(X) the gcd of f and g. (d) To multiply two elements in the form (*). we know each element g of F [X]/(f) is represented by a unique polynomial r of degree < m. Let f(X) = X 3 − 3X − 1 ∈ Q[X]. Let f(X) = X 2 + 1 ∈ R[X]. d(X) is 1 because f(X) is irreducible and deg g(X) < deg f(X).

When S = {α1. Proof.13. 1 n ai1 ···in ∈ F. Let E be an extension ﬁeld of F . then we write F · F for F (F )(= F (F )). and that any ring containing F and S contains R. αn] for F [S]. . . f(α) = 0. we write F [α1.. αn) for F (S) when S = {α1 . The ring F [S] consists of all the elements of E that can be written as ﬁnite sums of the form ai1 ···in αi1 · · · αin . Then we have a homomorphism f(X) → f(α) : F [X] → E.. . it is easy to check that R is a ring containing F and S. Case 1: The kernel of the map is (0).. Example 1. The map x → αx : R → R is an injective F -linear map. An extension E of F is said to be simple if E = F (α) some α ∈ E. Let R be the set of all such elements. g(X).6 J. i.14 shows that F [S] is sometimes already a ﬁeld. .S. It is the smallest subﬁeld of E containing both F and F . αn}. When F and F are subﬁelds of E. . MILNE Lemma 1.. Algebraic and transcendental elements. αn}.. h(π) = 0. Let α be a nonzero element of R—we have to show that α is invertible... Lemma 1. can be written uniquely as a ﬁnite sum a0 + a1π + a2π 2 + · · · . b ∈ Q.. . αi ∈ S.14. Thus: F [α1.15. and is the smallest such ﬁeld). An element of Q[π]. and we call it the composite of F and F . Example 1.. π = 3. αn ] consists of all elements of E that can be expressed as polynomials in the αi with coeﬃcients in F . In particular. h(X) ∈ Q[X]. An element of Q(π) can be expressed as a quotient g(π)/h(π). Note that the expression of an element in the form (*) will not be unique in general. then it is a ﬁeld. . There are two possibilites. . f(X) ∈ F [X] =⇒ f(X) = 0. (Everything considered in C. It is equal to the ﬁeld of fractions of F[S] (since this is a ﬁeld containing F and S. . We write F (α1. .. . in which case F (S) = F [S].16.. αn ) consists of all elements of E that can be expressed as quotients of two such polynomials. If R is ﬁnite-dimensional when regarded as an F -vector space.. and is therefore surjective. and F (α1. Let E ⊃ R ⊃ F with E and F ﬁelds and R a ring. An element of Q[i] can be written uniquely in the form a + bi..e.. The ring Q[i] is already a ﬁeld. . there is an element β ∈ R such that αβ = 1. Lemma 1.) Let E again be an extension ﬁeld of F and S a subset of E. ai ∈ Q. .14159.8. 1.. a. For example. and let α ∈ E.. Q(π) and Q[i] are simple extensions of Q. The subﬁeld F (S) of E generated by F and S is the intersection of all subﬁelds of E containing F and S. therefore R equals F [S]. (*) Proof.

. .. . and f(α) = 0. Then factor(X^4+4. . and β −1 = 7 α2 111 − 26 α 111 + 28 . α2 } is a basis for Q[α] over Q. then F [α1] is ﬁnite over F (because α1 is algebraic over F ). In other words. α. g(α) = 0 and g ∈ F [X] =⇒ f|g.. (b) Let E = F [α1. Let f(X) be the monic polynomial generating the kernel of the map. Moreover. then E is algebraic over F. Since F [x] ≈ F [α]. then β = 3α2 + 7α + 5. Now type: alias(c=RootOf(X^2+2*X+2). . Since the ﬁrst is a ﬁeld. 1. (a) If [E : F ] is ﬁnite. but g(α) = 0 = h(α)). It is irreducible (if f = gh is a proper factorization. f(α) = 0. Hence F [α1.c). . then [E : F ] is ﬁnite.. f is the monic polynomial of least degree such f(α) = 0. arithmetic in F [α] can be performed using the same rules as in F [x]. (b) If E is algebraic over F and ﬁnitely generated (as a ﬁeld). irreducible. An extension E/F is algebraic if all elements of E are algebraic over F . For example. each element of F [α] has a unique expression a0 + a1α + a2α2 + · · · + am−1 αm−1 . α2] is ﬁnite over F . Proposition 1. .18.. factor(X^4+4).20.19. ai ∈ F. Maple has factored X 4 + 4 in Q[c] where c has minimum polynomial X 2 + 2X + 2. where m = deg(f). . If E is algebraic over F then any subring R of E containing F is a ﬁeld. i. returns the factorization (X 2 − 2X + 2)(X 2 + 2X + 2).17. We then say that α is algebraic over F . f is monic. Example 1. αn]. α2] is ﬁnite over F [α1] (because α2 is algebraic over F . Corollary 1. α.e. (a) If α were transcendental over F . αm−1 is a basis for F [α] over F . then 1. 111 Remark 1. returns the factorization (X + c)(X − 2 − c)(X + 2 + c)(X − c). then g(α)h(α) = f(α) = 0. would be linearly independent over F. Case 2: The kernel is = (0). and has α as a root). It is characterized as an element of F [X] by each of the following sets of conditions: f is monic. We call f the minimum polynomial of α over F . The calculations in an example above show that if β is the element α4 + 2α3 + 3 of Q[α]. so also is the second: F (α) = F [α]. and hence F [α1]). F [α1. irreducible. Let α ∈ C be such that α3 − 3α − 1 = 0. Hence [F (α) : F ] = m. Proof. The isomorphism F [X] → F [α] extends to an isomorphism F (X) → F (α).. Maple knows how to compute in Q[α]. . Note that g(X) → g(α) induces an isomorphism F [X]/(f) → F [α]. g(α) = 0 for some nonzero g(X) ∈ F [X]. The minimum polynomial of α over Q is X 3 − 3X − 1 (because this polynomial is monic. The set {1. α2 .. i. α. This argument can be continued.FIELDS AND GALOIS THEORY 7 In this case we say that α transcendental over F . otherwise it is transcendental over F.e.

Therefore α has an inverse in R. 1873: Cantor showed that the set of algebraic numbers is countable. We prove that the analogue of this number in base 2 is transcendental. There are only ﬁnitely many rational numbers with height less than a ﬁxed number N. (This was one of Hilbert’s famous problems) 1994: Euler’s constant n γ = lim ( n→∞ k=1 1/k − log n) has not yet been proven to be transcendental. then count the elements of A(100). The ﬁeld of complex numbers C is algebraically closed. Deﬁne the height h(r) of a rational number to be max(|m|. Every ﬁeld F has an algebraically closed algebraic extension ﬁeld (which is unique up to a nonunique isomorphism). A complex number is said to be algebraic or transcendental according as it is algebraic or transcendental over Q. MILNE Proof. 1 A typical Liouville number is ∞ 10n! —in its decimal expansion there are increasingly n=0 long strings of zeros. 1. then / αβ is transcendental. Proof.S. but they have not even been proved to be irrational! Proposition 1. All these statements will be proved later. [Thus almost all numbers are transcendental. every nonconstant polynomial in F [X] has a root in F . First some history: 1844: Liouville showed that certain numbers (now called Liouville numbers) are transcendental.22. Count the elements of A(10). |n|). Let α ∈ R. and so on.9. The set of algebraic numbers is countable. Equivalent condition: the only irreducible polynomials in F [X] are of degree one. where r = m/n is the expression of r in its lowest terms. Example 1. then count the elements of A(1000). . Then A(N) is ﬁnite for each N. The set of all complex numbers algebraic over Q is an algebraically closed ﬁeld. A ﬁeld F is said to be algebraically closed if E algebraic over F implies E = F . 1. 1994: The numbers e + π and e − π are surely transcendental. but it is usually very diﬃcult to prove that a particular number is transcendental. 1873: Hermite showed that e is transcendental. and β ∈ Q.8 J. α = 0.] 1882: Lindemann showed that π is transcendental. Transcendental numbers. but that R is not countable. then F [α] is a ﬁeld and F [α] ⊂ R.21. 1934: Gelfond-Schneider showed that if α and β are algebraic. Let A(N) be the set of algebraic numbers whose minimum equation over Q is of degree ≤ N and has coeﬃcients of height < N.

b. Suppose we are given a length. Let f(X) = i=1 N 1 and choose a nonzero integer D such that Df(X) ∈ Z[X].23.) Clearly xN ∈ Q. and so (X − αi ). a straight-edge. Constructions with straight-edge and compass. αi ∈ C. and let xN = f(ΣN ). The F -plane is F × F ⊂ R × R. except possibly α. d be the minimum polynomial of α over Q. Thus [Q[α] : Q] = d. (In fact α is obviously nonrational because its expansion to base 2 is not periodic. c ∈ F.FIELDS AND GALOIS THEORY 9 Theorem 1. This led them to three famous problems that they were unable to solve: is it possible to duplicate the cube. We make the following deﬁnitions: A line in the F -plane is a line through two points in the F -plane. and • circles with centre a point already constructed and radius a constructed length. but ΣN = α. Let F be a subﬁeld of R. ∞ where M = max |αi |. trisect an angle. ai ∈ Q. A number (better a length) is constructible if it can be constructed by forming successive intersections of • lines drawn through two points already constructed. a. and so xN = 0. and a compass (device for drawing circles). which we call 1. Let ΣN = n=0 2n! . . Suppose not. Such a line is given by an equation: ax + by + c = 0. or square the circle by straight-edge and compass constructions? We’ll see that the answer to all three is negative. The Greeks understood that integers and the rational numbers. |xN | = and |α1 − ΣN | = Hence |(2N !)d DxN | ≤ 2 · because 2(N +1)! 2d·N ! N! |ΣN − αi | ≤ |α1 − ΣN |(M + ΣN )d−1 . is not rational. √ They were surprised to ﬁnd that the length of the diagonal of a square of side 1. it has no rational root.10. so that ΣN → α as N → ∞. On the other hand. They thus realized that they needed to extend their number system. 1. α1 = α. in fact (2N ! )d DxN ∈ Z. Because f(X) is irreducible in Q[X]. The number α = Proof. f(X) = X d + a1X d−1 + · · · + ad . |(2N ! )d DxN | ≥ 1. and let 1 2n! is transcendental. i=1 2 1 ≤ (N +1)! 2n! 2 n=N +1 2d·N !D · (M + ΣN )d−1 → 0 as N → ∞ 2(N +1)! = 2d 2N +1 → 0. They then hoped that the “constructible” numbers would suﬃce. namely 2. We have a contradiction.

Lemma 1. (a) If c and d are constructible. Appendix 3. (For more details. For (b).) Theorem 1. . Proof. draw a circle of radius c+1 about ( c+1 . Rotman. ar ]. By an astute choice of the triangles. ar ] : Q] = 2r . Let L = L be F -lines. Proof. We know that [Q[α] : Q] divides [Q[ a1 . it follows from Lemma 1. (a) Immediate from (a) of Lemma 1. and √ (b) If c > 0 is constructible. and √ √ it follows from (a) and Lemma 1. . Proof. (a) L ∩ L = ∅ or consists of a single F -point.10 to obtain: Corollary 1. then so also are c ± d. 0). . and draw a vertical line through the point A = (1. 0) to meet the circle 2 2 √ at P . and so [Q[ 2] : Q] = 3. . If α is constructible. (b) From (a) we know that the set of constructible numbers is a ﬁeld containing Q. (b) A number α is constructible if and only if it is contained in ﬁeld of the form √ √ √ √ Q[ a1 . a. one constructs cd and c−1 .25. Now we can apply the (not quite elementary) result Proposition 1. Galois Theory. . First show that it is possible to construct a line perpendicular to a given line through a given point. Corollary 1. It is impossible to duplicate the cube by straight-edge and compass constructions. and then a line parallel to a given line through a given point. The problem is to construct a cube with volume 2. see for example. Lemma 1. cd. . . ar ] is constructible. ai−1 ]. .26. c d (d = 0).10 J. The points in the intersection are found by solving the simultaneous equations. (a) The set of constructible numbers is a ﬁeld. . . and let C = C be F -circles. This requires constructing = a root of the polynomial X 3 − 2 √ 0. .29. Conversely.28. and hence by solving (at worst) a quadratic equation with coeﬃcients in F . then α is algebraic over Q.S. Proof. (c) C ∩ C = ∅ or consists of one or two points in the F [ e]-plane. . But this polynomial is irreducible (by Eisenstein’s 3 criterion for example).25 that every number in Q[ a1. . The length AP is c. √ √ Proof. some e ∈ F . Such a circle is given by an equation: (x − a)2 + (y − b)2 = c2 .25. c ∈ F. it is impossible to trisect an angle by straight-edge and compass constructions. then so also is c. √ (b) L ∩ C = ∅ or consists of one or two points in the F [ √ e]-plane. . some e ∈ F. and [Q[α] : Q] is a power of 2.24 that every constructible number is in a ﬁeld of the √ √ form Q[ a1. Hence it is possible to construct a triangle similar to a given one on a side with given length. Corollary 1. . . b. . MILNE A circle in the F -plane is a circle with centre an F -point and radius an element of F . ai ∈ Q[ a1.24. . .27. . . In general. ar ].

But 2r + 1 can only be a prime if r is a power of 2. X .. shows that it is ≤ 2. take 3α = 60. and we don’t know of any more Fermat primes. This success encouraged him to become a mathematician. Hence [Q[cos 2π ] : p Thus if the regular p-gon is constructible. that of constructing a regular polygon. Since π is the transcendental. Lemma 1.FIELDS AND GALOIS THEORY 11 Proof. because otherwise r has an odd factor t. hence Q[e2πi/p] has degree p − 1 over Q. Note that X m − 1 is not irreducible. Therefore. we have to construct a solution to cos 3α = 4 cos3 α − 3 cos α. p 2πi p . then p = 22 + 1 for some k. 5. k . then (p − 1)/2 = 2k some k (later. and it is not 1 because Q[e Q] = p−1 . are prime but Euler showed that 232 + 1 = 641 · 6700417. We now consider another famous old problem. Knowing an angle is equivalent to knowing the cosine of the angle. we need to construct cos 2π . p an odd prime. Thus if the regular p-gon is constructible. to construct α. and claimed to show that this is true for k ≤ 5—for this reason primes of this form are called Fermat primes. 17. 65537. But p ] ⊃ Q[cos 2π ] ⊃ Q. which imples p = 2k+1 + 1.31.. f(X+1) is irreducible by Eisenstein’s criterion. i = 1. For example. we have to solve 8x3 − 6x − 1 = 0. . in fact X m − 1 = (X − 1)(X m−1 + X m−2 + · · · + 1). Since p|ai . For 0 ≤ k ≤ 4. f(X + 1) = 2πi p with ai = Q[e 2πi p p i+1 In order to construct a regular p-gon. 2 ] is not contained in R. Y t + 1 = (Y + 1)(Y t−1 − Y t−2 + · · · + 1).30. Corollary 1. The degree of Q[e p α2 − 2 cos ] over Q[cos 2π ] is 2—the equation p α=e 2πi p 2π · α + 1 = 0. If p is prime then X p−1 + · · · + 1 is irreducible. we shall see a converse). so also is π. the numbers p = 3. 257. p−2.. which is irreducible. A square with√ same area as a circle of radius r has side πr. Gauss showed that √ √ √ √ 1 1√ 2π 1 1 =− + cos 17 + 34 − 2 17 + 17 + 3 17 − 34 − 2 17 − 2 34 + 2 17 17 16 16 16 8 when he was 18 years old. and for t odd. Fermat conjectured k that all numbers of the form 22 + 1 are prime. to trisect 3α. Proof. Consider (X + 1)p − 1 = X p−1 + · · · + a2X 2 + a1 X + p. It is impossible to square the circle by straight-edge and compass constructions. √ Proof.

2. ϕ(α) is transcendental over F . Maps from simple extensions.S. g(X) → g(γ). let γ ∈ Ω be a root of f(X). The map F [X] → Ω.e. Proposition 2. (b) Assume α is algebraic over F . Note that if E and E have the same ﬁnite degree over F . (a) Let γ ∈ Ω. (b) Let f(X) = aiX i . Splitting Fields. then the map ϕ → ϕ(α) deﬁnes a one-to-one correspondence {extensions ϕ : F (α) → Ω of ϕ0 } ↔ {elements of Ω transcendental over ϕ0(F )}.12 J. In particular. (b) Assume α is algebraic over F . ai1 ···im ∈ F . Thus an F -homorphism maps a polynomial ai1 ···im αi1 · · · αim . Conversely. MILNE 2. (a) Assume α is transcendental over F . Proof. Proposition 2. On applying ai ϕ(α)i = 0. Algebraic Closures 2. When composed with the inverse of the isomorphism F [X]/(f(X)) → F [α]. An F -isomorphism is a bijective F -homomorphism. to 1 m ai1 ···im ϕ(α1 )i1 · · · ϕ(αm )im . Let F (α) be a simple ﬁeld extension of a ﬁeld F . the number of such maps is the number of distinct roots of ϕ0f in Ω. then the map ϕ → ϕ(α) deﬁnes a one-to-one correspondence {extensions ϕ : F [α] → Ω of ϕ0 } ↔ { distinct roots of (ϕ0f)(X)in Ω}. it becomes a homomorphism F [α] → Ω sending α to γ. with minimum polynomial f(X). ϕ(α) is a root of f(X) in Ω. Let F (α) be a simple ﬁeld extension of a ﬁeld F . then for any F homomorphism ϕ : F [α] → Ω. and the map ϕ → ϕ(α) deﬁnes a one-to-one correspondence {F -homomorphisms ϕ : F [α] → Ω} ↔ { distinct roots of f(X) in Ω}. we obtain the equation γ =df ϕ(α) is a root of f(X) in Ω. then for any F -homomorphism ϕ : F (α) → Ω. which shows that ϕ to the equation aiαi = 0. and let Ω be a second ﬁeld containing F .1. To say that α is transcendental over F means that F [α] is the ring of polynomials in α (as variable).. (a) Assume α is transcendental over F . which is so if and only if γ is transcendental. . the number of such maps is the number of distinct roots of f in Ω. This extends to F (α) if and only if all nonzero elements of F [α] are sent to invertible (i. By the universal property of polynomial rings. and consider an F -homomorphism ϕ : F [α] → Ω.1. Let E and E be ﬁelds containing F . there is a unique F -homomorphism ϕ : F [α] → Ω sending α to γ. and the map ϕ → ϕ(α) deﬁnes a one-to-one correspondence {F -homomorphisms ϕ : F (α) → Ω} ↔ { elements of Ω transcendental over F }. nonzero) elements of Ω. We shall need a slight generalization of this result. An F -homomorphism is a homomorphism ϕ : E → E such that ϕ(a) = a for all a ∈ F . and let ϕ0 : F → Ω be a homomorphism of F into a second ﬁeld Ω. with minimum polynomial f(X). In particular. factors through F [X]/(f(X)). then an F -homomorphism is automatically an F -isomorphism.

then it is called a splitting ﬁeld for f. (b) Let f(X) = X 3 + aX 2 + bX + c ∈ Q[X] be irreducible. Example 2. α2 ] : Q] = 3 or 6.e. We’ll see later that the degree is 3 if and only if the discriminant of f(X) is a square in F . the remainder are α + 1. Note that [Q[α1] : Q] = 3 and that [Q[α1. . f= ai X i =⇒ ϕ0f = ϕ(ai )X i . α1 = X + (g1 ). Let f ∈ F [X]. and so F automatically contains all the pth roots of 1. . .. Then α1 is a root of f(X) in F1. Then Q[α1. any ﬁeld generated by a root of f is a splitting ﬁeld (if α is one root. and let Ω ⊃ F be a second ﬁeld splitting f. Therefore. any ﬁeld generated by a root of f is a splitting ﬁeld (if ζ is one root. (b) Suppose F is of characteristic p... By continuing in this fashion. and so [Q[α1. α2] is a splitting ﬁeld for f(X).FIELDS AND GALOIS THEORY 13 Proof. then X p − 1 = (X − 1)p . Let g1 be an irreducible factor of f(X). . if F contains all the nth roots of 1.. then F [α] is a splitting ﬁeld for X n − a. we obtain a splitting ﬁeld. Then f1 ∈ F1 [X]. α3 be its roots in C. Whether or not there exist polynomials of degree n in F [X] whose splitting ﬁeld has degree n! depends on F . . i. the remainder are ζ 2 .4. Proposition 2. .e. and let f = X p − X − a.3. and let F1 = F [X]/(g1(X)) = F [α1]. Every polynomial has a splitting ﬁeld. ζ 3 . (c) If α is one root of X n −a. nor for n > 2 if F = R. α2 . then the remaining roots are all of the form ζα. (a) Let f = (X p − 1)/(X − 1). Proof. α + p − 1).6. However. the discriminant of X 3 + bX + c is −4b3 − 27c2 . where ζ n = 1. By ϕ0 f we mean the polynomial obtained by applying ϕ0 to the coeﬃcients of f. 2 √ Example 2. if f(X) = (X − αi ) with αi ∈ E.. α2 . later we shall see how to write down large numbers (in fact inﬁnitely many) polynomials of degree n in Q[X] whose splitting ﬁelds have degree n!. [F1 : F ] ≤ n. [F2 : F1 ] ≤ n − 1. i. For example. .e. and let α1 . (a) There exists at least one F -homomorphism ϕ : E → Ω. there don’t for n > 1 if F = C or Fp . The proof is essentially the same as that of the preceding proposition. Note that if f(X) = fi (X)mi . Let f be a polynomial with coeﬃcients in F . and let E be a splitting ﬁeld for f. Splitting ﬁelds.7. Let n = deg f.. ζ p−1 ). Let f ∈ F [X]. and so the splitting ﬁeld of X 3 + 10X + 1 has degree 6 over Q. then a splitting ﬁeld for fi (X) is also a splitting ﬁeld for f (and conversely). For example. and we deﬁne f1 (X) to be the quotient f(X)/(X − α1 ) (in F1[X]).. α2 ] : Q[α1]] = 1 or 2. i. Note that if p is the characteristic of F . and let α = b2 − 4ac. (a) Let f(X) = aX + bX + c ∈ Q[X] be irreducible. α3 ] = Q[α1. If E is also generated by the αi . Remark 2. and the same construction gives us a ﬁeld F2 = F1[α2 ] with α2 a root of f1 . if X n − 1 splits in F [X]. and so the degree of the splitting ﬁeld over F is ≤ n!. Proposition 2.2. In the proof.5. 2. then the subﬁeld Q[α] of C generated by α is a splitting ﬁeld for f. A ﬁeld E containing F is said to split f if f splits in E[X]...

10. then we obtain the reverse inequality also. then we know there is an F -isomorphism E → E . Recall that Ω is said to be algebraically closed if every nonconstant polynomial f(X) ∈ Ω[X] has a root in Ω (and hence splits in Ω[X]). 2. Next. if the only irreducible polynomials in Ω[X] are those of degree 1. Warning! If E and E are splitting ﬁelds of f(X) ∈ F [X]. According to Proposition 2. Recall also that a ﬁeld Ω containing F is said to be an algebraic closure of F if it is algebraic over F and it is algebraically closed. with equality holding when f has deg(f) distinct roots in Ω. Let Ω be a splitting ﬁeld for f regarded as an element of L[X].S.1 shows that there exists an F -homomorphism ϕ1 : F [α1] → Ω. with the αi the distinct roots of f(X). Lemma 2. . Proof. with equality holding when f (hence also f1 ) has distinct roots in Ω. Write E = F [α1. .9. The proposition shows that there is an F -homomorphism E → Ω.. Algebraic closures. MILNE (b) The number of F -homomorphisms E → Ω is ≤ [E : F ]. α2] → Ω.3. any two splitting ﬁelds for f are F -isomorphic. . and replace E with the subﬁeld of E generated by F and all the roots of f(X). note that... If Ω is also a splitting ﬁeld. then Ω is an algebraic closure of F. . each ϕ1 extends to a homomorphism ϕ2 : F [α1. and = [E : F ] if f has deg(f) distinct roots in Ω. α2] : F ]. On combining these statements we conclude that there exists an F -homomorphism ϕ : F [α1. and the number of extensions is ≤ deg(f2 ) = [F [α1. and the number of such homomorphisms is ≤ [F [α1. Let E and L be extension ﬁelds of F . Error and confusion can result if you simply identify the ﬁelds. α2] → Ω. we obtain (a) and (b). . Thus E is now the splitting ﬁeld of f(X) ∈ F [X]. Let E be a splitting ﬁeld of f =df fi regarded as an element of E[X]. Corollary 2.2.14 J. then there exists an extension ﬁeld Ω of L and an F -homomorphism E → Ω. (c) If Ω is also a splitting ﬁeld for f. For (c). and let fi be the minimum polynomial of αi over F . then each F -homomorphism E → Ω is an isomorphism. We want to show that (assuming the axiom of choice) every ﬁeld has an algebraic closure. but there will in general be no preferred such isomorphism. we will have that E and L are subﬁelds of Ω.8. After replacing E by its (isomorphic) image in Ω. After repeating the argument m times. In particular. As f (hence f1 ) splits in Ω. the minimum polynomial of α2 over F [α1] is an irreducible factor f2 of f(X) in F [α1][X]. Write E = F [α1. and the number of ϕ1’s is ≤ deg(f1 ) = [F [α1] : F ]. equivalently. The minimum polynomial of α1 is an irreducible polynomial f1 dividing f. We therefore have equality. m ≤ deg(f). α2] : F [α1]]. αm]. . Remark 2. The following criterion suggests how this might be done. with equality holding when f (hence also f2 ) has distinct roots in Ω. with E ﬁnite over F . we must have [E : F ] ≤ [Ω : F ]. Suppose that Ω is algebraic over F and every polynomial f ∈ F [X] splits in Ω[X]. αm ]. because an F -homomorphism E → Ω is injective. This will allow us to assume that E and L are subﬁelds of a common ﬁeld. Proof. and so any F -homomorphism E → Ω is an isomorphism. Proposition 2.

Write f = an X n +· · ·+a0.e. F [a1. b ∈ Ω∗ . then A ⊗F B is a ring containing F . 2Every ﬁeld has an algebraic closure. . it suﬃces to construct an extension in which every polynomial in F [X] splits.FIELDS AND GALOIS THEORY 15 Proof. The preceding lemma shows that if every polynomial in F [X] splits in Ω[X]. . ai ∈ Ω. We know how to do this for a single polynomial. then E = {α ∈ Ω | α algebraic over F } is a ﬁeld. . . it is obviously an algebraic closure of F .6) how to construct a ﬁnite extension E of Ω containing a root α of f. We want to show that α in fact lies in Ω.. . and there are F -homomorphisms A. . Proof. αβ. this polynomial splits in Ω[X]. . β] is algebraic over F . an . an ][X]. .. Because each ai is algebraic over F . and consider the sequence of ﬁelds F ⊂ F [a1.11. . Let f ∈ Ω[X]. Proof. . . then ⊗F Ai is a ring containing F . Thus to construct an algebraic closure of F . . If α and β are algebraic over F . Every element of F [α. i. (Second proof of 2. . write a ∼ b if a = b when regarded as elements of the larger of Ei or Ej . it is the root of a polynomial with coeﬃcients in F . and so all its roots lie in Ω. Ei is the splitting ﬁeld of fi over Ei−1 . Remark 2. . and because f ∈ F [a1. It is deﬁned to be the quotient of the F -vector space with basis ΠAi by the subspace generated by elements of the form: • (xi) + (yi ) − (zi ) with xj + yj = zj for one j ∈ I and xi = yi = zi for all i = j. β] is of ﬁnite degree over F . We sketch three proofs of the general result. an ] ⊂ F [a1. α/β. but does assume that F is countable.e. i. But. . The ﬁrst doesn’t assume the axiom of choice. This proves the theorem for such ﬁelds. verify that ∼ is an equivalence relation. and let Ω = Ω∗ / ∼. and there are F -homomorphisms Aj → ⊗F Ai for each j ∈ I. by assumption. and so is a ﬁeld (see 1. . . Let Ω ⊃ F . α is algebraic over F [a1. B → A ⊗F B. (First proof of 2. . α ∈ Ω. Therefore α lies in a ﬁnite extension of F . 2 Results marked with an asterisk require the axiom of choice for their proof. Once we have proved the fundamental theorem of algebra.12) If A and B are rings containing a ﬁeld F . forming the union requires explanation: deﬁne Ω∗ to be the disjoint union of the Ei . .12 (*). . .12) Because F is countable. . . it follows that F [X] is countable. then we will know that the algebraic closure in C of any subﬁeld F of C is an algebraic closure of F .13. The ﬁeld E constructed in the lemma is called the algebraic closure of F in Ω. if (Ai )i∈I is some family of rings each of which contains F . Proof. including α ± β. . then E is an algebraic closure of F . . Since the Ei are not subsets of a ﬁxed set. . an ] is a ﬁnite ﬁeld extension of F . f2 (X). .14). Lemma 2. let a. Theorem 2. α]. • (xi) − a(yi ) with xj = ayj for one j ∈ I and xi = yi for all i = j. Note that E0 ⊂ E1 ⊂ E2 ⊂ · · · . then F [α. Deﬁne Ω = ∪Ei . that C is algebraically closed. In particular. an ]. and is therefore algebraic over F . say a ∈ Ei and b ∈ Ej . but passing from there to all polynomials causes set-theoretic problems. We know (see §1. Deﬁne the ﬁelds Ei inductively as follows: E0 = F . . . More generally. we can list its elements f1 (X).

.2) to extend ϕ to M[α]. Suppose ﬁrst that E is countably generated over F . an element s such that s ≤ s =⇒ s = s ). and so on. anti-symmetric. (See Jacobson. +. ]. ·) contains F as a subﬁeld and is algebraic over it. Then there exists an element α ∈ E. Let S be the set triples (E. Suppose that M = E.. Alg`bre.S.e. · ) if the ﬁrst is a subﬁeld of the second. ϕ ) is an upper bound for T in S.1). a ≤ b and b ≤ a =⇒ a = b). Then we can extend the inclusion map F → Ω to F [α1] (map α1 to any root of its minimal polynomial in Ω). and it is proper because if 1 ∈ J then 1 ∈ I for some I in T . it must also split in the larger ﬁeld Ω.e...) There do exist naturally occurring ﬁelds. / . there exists an s ∈ S such that t ≤ s for all t ∈ T ). ≤) be a nonempty partially ordered set (reﬂexive.. Proof. ϕM ) with M a ﬁeld F ⊂ M ⊂ E and ϕM an F -homomorphim M → Ω. MILNE It can be made into a ring in an obvious fashion (see Bourbaki.14 (Zorn’s). Let S be the set of pairs (M. +. Thus J is an upper bound for T . For example. α2 ]. ϕM ) ≤ (N. ·) a ﬁeld structure on E such that (E. and identify F with a subset of Ξ. ·) ≤ (E . not contained in C. Every nonzero commutative ring A has a maximal ideal. t ∈ T . choose a splitting ﬁeld Ef . Ω itself is an algebraic closure of F. Let Ω be an algebraic closure of F . Then Ω is a ﬁeld (see 1. Apply Zorn’s lemma to show that S has maximal elements. . Lectures in Algebra. + . Chapt 3. and let Ω = (⊗f Ef )/M where M is a maximal ideal in ⊗f Ef —Zorn’s lemma implies that M exists (see below). Zorn’s lemma is equivalent to the Axiom of Choice. Let T be a totally ordered subset of S. then any such map is an isomorphism.e. The algebraic closure of F in Ω is therefore an algebraic closure of F . . Let S be the set of all proper ideals in A. p144 for the details. . Lemma 2. If T is a totally ordered set of ideals. III. partially ordered by inclusion. e For each polynomial f ∈ F [X].. Since α is algebraic over M. E = F [α1. α ∈ M. Then S has a maximal element (i. The pair (M . Hence Zorn’s lemma gives us a maximal element (M. then J = I∈T I is again an ideal. Proof. Then M = ∪M ∈T M is a subﬁeld of E. Since f splits in Ef . ai ∈ F . then to F [α1. Appendix). (Third proof of 2. for any ﬁeld F there is a ring F [[T ]] of formal power series i≥0 aiT i . ϕ) in S. ϕN ) if M ⊂ N and ϕN |M = ϕM .. and then show that a maximal element is algebraically closed.12) First show that the cardinality of any ﬁeld algebraic over F is the same as that of F . and its ﬁeld of fractions is uncountable even if F is ﬁnite. for all s.16 (*). Write (M. If E is also an algebraic closure of F . then there is an F -homomorphism E → Ω. Suppose that every totally ordered subset T of S (i. and let E be an algebraic extension of F .e. i..16 J. αn. either s ≤ t or t ≤ s) has an upper bound in S (i. and we can deﬁne a homomorphism ϕ : M → Ω by requiring that ϕ (x) = ϕM (x) if x ∈ M.e. Theorem 2. (Actually.) Lemma 2. ·) with E ⊂ S and (+.15 (*). transitive. The uncountable case is a straightforward application of Zorn’s lemma. Let (S. which is a maximal ideal in A. Write (E. that are uncountable. This makes S into a partially ordered subset. i. Now Zorn’s lemma implies that S has a maximal element. and there are F -homomorphisms Ef → Ω (which must be injective) for each f ∈ F [X]. +. Next choose an uncountable set Ξ of cardinality greater than that of F . Proof.. we can apply (2.

there will exist uncountably many isomorphisms from one algebraic closure to a second. Thus it is (uncountably) sloppy to say that the algebraic closure of F is unique. given two algebraic closures Ω. Hence M = E. then every polynomial f ∈ F [X] splits in E and hence in ϕ(E). then. Even for a ﬁnite ﬁeld F . f(X) splits in ϕ(E)[X]. αi ∈ ϕ(E). but. none of which is to be preferred over any other. . i. there exists an F -isomorphism Ω → Ω . Remark 2.e. and the proof of the ﬁrst statement is complete. If E is algebraically closed. f(X) = (X − αi).FIELDS AND GALOIS THEORY 17 contradicting the maximality of M. as we just observed.. All one can say is that. Let α ∈ Ω. The above proof is a typical application of Zorn’s lemma: once we know how to do something in a ﬁnite (or countable) situation. Because of unique factorization. Ω of F .17. thanks to the axiom of choice. this implies that α ∈ ϕ(E). and let f(X) be the minimum polynomial of α. Then X − α is a factor of f(X) in Ω[X]. Zorn’s lemma allows us to do it in general.

A root of multiplicity one is said to be simple.g. and let f(X) = (X − αi )mi . The Fundamental Theorem of Galois Theory In this section. Let f be a (monic) irreducible polynomial in F [X]. g). In fact. It then has multiple roots in the subﬁeld of Ω generated by its roots. be a splitting of f over some large ﬁeld Ω ⊃ F . If f has multiple factors in F [X].18 J. then the proposition shows that f can only have multiple roots if one of the fi has multiple roots. we prove the fundamental theorem of Galois theory. Let f. The following statements are equivalent: . this doesn’t happen—gcd’s don’t change when the ﬁeld is extended. The proposition allows us to write gcd(f. Thus it remains to examine irreducible polynomials for multiple roots. say f = fi (X)mi with some mi > 1. Let f and g be polynomials in F [X].S. Certainly rF (X)|rΩ (X) in Ω[X]. Hence rΩ = rF . g) = 1 (in F [X]). Even when f and g have no common factor in F [X]. if f and g are monic and irreducible and f = g. note that the hypotheses imply that gcd(f. g ∈ F [X]. For the second statement. F = Fp (T ). and assume that F has an element a that is not a pth -power (e. which gives a one-toone correspondence between the subﬁelds of the splitting ﬁeld of a separable polynomial and the subgroups of the Galois group of f. In particular. Let f ∈ F [X]. The Euclidean algorithm shows that there are polynomials a and b in F [X] such that a(X)f(X) + b(X)g(X) = rF (X). If r(X) is the gcd of f and g computed in F [X].1. this shows that f will have roots of multiplicity > 1 in every ﬁeld containing F in which it splits. αi distinct. Thus an irreducible polynomial can have multiple roots. Example 3. Multiple roots. without reference to a ﬁeld. Proposition 3. If f = fi with the fi distinct monic irreducible polynomials. Hence they can’t have a common factor X − α in any extension ﬁeld. and because any two splitting ﬁelds are F -isomorphic. this agrees with the usual deﬁnition. Since rΩ (X) divides f and g in Ω[X]. but note that the derivative of X p is zero in characteristic p. 3.3. We then say that αi is a root of multiplicity mi . Proposition 3. MILNE 3. The usual rules for diﬀerentiating sums and products still hold. We say that f has multiple roots if it has roots of multiplicity > 1 in some big ﬁeld Ω. Let F be of characteristic p. a = T ).. and therefore also the right. you might expect that they could acquire a common factor in Ω[X] for some Ω ⊃ F .2. then obviously it will have multiple roots. Let rF (X) and rΩ (X) be the greatest common divisors of f and g in F [X] and Ω[X] respectively. and let Ω ⊃ F . When We deﬁne the derivative f (X) of a polynomial f(X) = F = R. but X p − a = (X − α)p in its splitting ﬁeld. ai X i to be iaiX i−1 . then it is also the gcd of f and g in Ω[X]. it divides the left-hand side of the equation. then they do not have a common root in any extension ﬁeld of F. Proof.1. Then X p − a is irreducible in F [X].

all the roots of f are multiple. A polynomial f ∈ F [X] is said to be separable if all its irreducible factors have simple roots. or • it has characteristic p and F = F p (i. (b) If F has characteristic p. A ﬁeld F is said to be perfect if all polynomials in F [X] are separable. Proof. which must be surjective if F is ﬁnite). Hence every root of f(X) has multiplicity at least p.. and (b) at least one of the irreducible factors of f is a polynomial in X p .FIELDS AND GALOIS THEORY 19 (a) (b) (c) (d) f has at least one multiple root (in a splitting ﬁeld). then every polyonomial in X p is a pth power— ai X p = ( bi X)p if ai = bp—and so can’t be irreducible. then F = F (X) is not perfect (because X is not a pth 0 0 power). A ﬁeld F is perfect if and only if it either • has characteristic 0. Groups of automorphisms of ﬁelds. X p − a. (a) =⇒ (b). then F [X] contains a nonseparable polynomial. Proposition 3. and let g(X) = Then (X p − ai)mi = f(X) = g(X p ) = (X − ai)mi in some splitting ﬁeld.5. every element of F is a pth power). Note that the preceding discussion shows that f is not separable if and only if (a) the characteristic of F is p = 0. Suppose f(X) = g(X p ). automorphisms σ : E → E such that σ(a) = a for all a ∈ F . (c) =⇒ (d). Every root multiple =⇒ at least one root multiple (I hope). .e. Definition 3. Proof. (X − αi )pmi where αp = ai (in some big ﬁeld).e. ⇐= : If char F = p and F = F p. gcd(f. and write f = (X − α)m g(X). We write Aut(E/F ) for the group of F -automorphisms of E. f ) = 1. in some splitting ﬁeld. gcd(f. some g ∈ F [X]. i Example 3.6.. and so gcd(f. f ) = 1. =⇒ : If char F = p and it contains an element a that is not a pth power. (b) =⇒ (c).4. f ) = 1 =⇒ f = 0 =⇒ f = g(X p ). Hence f (α) = 0. m > 1. Then f (X) = m(X − α)m−1 g(X) + (X − α)m g (X). (a) All ﬁnite ﬁelds are perfect (because a → ap is an injective homomorphism F → F .2. Let α be a multiple root of f. Consider ﬁelds E ⊃ F . i (d) =⇒ (a). 3. F has characteristic p = 0 and f(X) = g(X p ). any ﬁeld algebraic over Fp is perfect. i. namely. In fact. Since f is irreducible and deg(f ) < deg(f).

Xn )/C) is the group of birational automorphisms of Pn . We have maps G → Inv(G) F → Aut(E/F ). Lectures III.9. The splitting ﬁeld of f is the same as the splitting ﬁeld of fi . Thus. all σ ∈ G}. Then p Aut(E/F ) = 1. and the set of subgroups H of G. Let f = fimi . we write E G = Inv(G) = {α ∈ E | σα = α. Hence we may assume f is a product of distinct monic separable irreducible polynomials. The splitting ﬁeld of f = X p − a is F [α] where α is the unique root of f. it is essential that E be the splitting ﬁeld of a separable polynomial. . called the subﬁeld of G-invariants of E or the subﬁeld of E ﬁxed by G. (a) Let E = √ [α]. (The group Aut(C(X1 . C (c) The group Aut(C(X1 . the group is unknown. we shall be concerned with the groups Aut(E/F ) when E is a ﬁnite extension of F . For example. MILNE Example 3. then F √ 3 Aut(E/F ) = 1. Analysts will note that this is the same as the automorphism group of the Riemann sphere. f(α) = 0. in the proposition. then Aut(E/F ) has order [E : F ]. This is not a coincidence: the ﬁeld of meromorphic functions on the Riemann sphere P1 is C(z) ≈ C(X).8. ad − bc = 0 cX+d (Jacobson. with the fi monic irreducible and distinct. They are all noncontinuous and (I’ve been told) nonmeasurable—hence they require the Axiom of Choice for their construction. one can construct uncountably many more. F ⊂ M ⊂ E.) In this section. Thus. Proof. Proposition 3. (a) There are two obvious automorphisms of C. it is essential that E be a splitting ﬁeld. . When there are more X’s. in the proposition. and let a be an element of F that is not a th power. and hence has deg f distinct roots in E. and so there C is a map Aut(P1 ) → Aut(C(z)/C). When G is a group of automorphisms of a ﬁeld E. if f has no other root in E than α. the identity map and complex conjugation. X2 )/C) is quite complicated—there is a map PGL3 (C) = Aut(P2 ) → Aut(C(X1 . We’ll see later (last section) that by using the Axiom of Choice.20 J. M → Aut(E/M) give a one-to-one correspondence between the set of intermediate ﬁelds M. (b) Let E = C(X). Then Aut(E/C) consists of the maps X → aX+b . they are automatically isomorphisms. namely. It is a subﬁeld of E. if 2 denotes the real cube root of 2. Now Proposition 2.S. . (b) Let F be a ﬁeld of characteristic p = 0. Its study is part of algebraic geometry. It is C called the Cremona group. which one can show is an isomorphism. p158). X2 )/C). . If E is a splitting ﬁeld of a monic separable polynomial f ∈ F [X]. C but this is very far from being surjective. then Aut(Q[ 3 2]/Q) = 1. and so Aut(E/C) = PGL2(C). . Goal: Show that when E is the splitting ﬁeld of a separable polynomial in F [X] and G = Aut(E/F ).7b shows that there are [E : F ] distinct F -homomorphisms E → E.7. Example 3. then H → Inv(H).

ci − σk (ci ). . and Galois extensions. ) is also a solution to the system of equations. . for each α ∈ E. After renumbering the αi ’s. . but has more zeros than the ﬁrst solution (look at the ﬁrst coordinate)— contradiction. we ﬁnd that (1. . . Choose a nontrivial solution (c1 . . . Let G = {σ1 = 1. On apply σk to the equations σ1 (α1)c1 + · · · + σ1 (αn )cn = 0 ··· ··· σm (α1 )c1 + · · · + σm (αn )cn = 0 and using that {σk σ1. An algebraic extension E/F is normal if the minimum polynomial of every element of E splits in E. . ). and let α1 . . . . and then (after multiplying by a scalar) that c1 = 1. . E/F normal =⇒ f splits in E Therefore. For example. the minimum polynomial of α has [F [α] : F ] distinct roots in E. We shall show that the αi are linearly dependent over F . if every irreducible polynomial in F [X] having a root in E is separable.3. we may suppose that c1 = 0. In the system σ1 (α1 )x1 + · · · + σ1 (αn )xn = 0 ··· ··· σm (α1 )x1 + · · · + σm (αn )xn = 0 there are m equations and n > m unknowns. . σk σm } is a permutation of {σ1 . With these normalizations. σk (ci ). doesn’t have multiple roots (in a splitting ﬁeld). Proof. Thus if f ∈ F [X] is irreducible of degree m and has a root in E. if every irreducible polynomial f ∈ F [X] having a root in E splits in E. we obtain a solution (0. . . and in particular has characteristic p = 0. . equivalently. which is nonzero (look at the ith coordinate). . . equivalently. k. normal. Artin). Hence the ﬁrst equation (recall σ1 = 1) α1c1 + · · · + αn cn = 0 shows that the αi are linearly dependent over F . and let F = E G . . . cn ) with the fewest nonzero elements. . Let G be a ﬁnite group of automorphisms of a ﬁeld E. . . Thus E/F is inseparable if and only if (a) F is nonperfect. . 3..FIELDS AND GALOIS THEORY 21 Lemma 3. On subtracting it from the ﬁrst.e. .10 (E. . σm }. then σk (ci ) = ci for some i. . Separable. . σm }. and (b) there is an element α of E whose minimal polynomial is of the form g(X p ). An algebraic extension E/F is said to be separable if the minimum polynomial of every element of E is separable. then [E : F ] ≤ (G : 1). i. . . . αn be n > m elements of E. g ∈ F [X]. E/F is normal and separable if and only if. . we’ll see that all ci ∈ F . If not all ci are in F . and hence there are nontrivial solutions (in E). . . then E/F separable =⇒ roots of f distinct =⇒ f has m distinct roots in E. . E = Fp (T ) is an inseparable extension of Fp (T p).

The following statements are equivalent: (a) E is the splitting ﬁeld of a separable polynomial f ∈ F [X]. and let F = E G ⊃ F . • E is the splitting ﬁeld of a separable polynomial in E G [X] (because b =⇒ a).12. we conclude that F = F . Let α ∈ E and let f be the minimum polynomial of α. in particular. Hence Proposition 3. (b) The ﬁeld Fp (T ) is normal but not separable over Fp (T p)—it is the splitting ﬁeld of the inseparable polynomial X p − T p.11. Because E has ﬁnite degree over F . and so g(X) ∈ F [X]. MILNE Any σ ∈ G merely permutes the αi . Let E be an extension ﬁeld of F . (a) The ﬁeld Q[ 3 2]. Finally. if E is as in (a). Theorem 3. . we ﬁnd that σai = ai for all i. we have to show that if G is a ﬁnite group acting on a ﬁeld E.. αm]. where 3 2 is the real cube root of 2.S. because each αi is a root of f(X) (if αi = σα. . (b) =⇒ (c). Moreover. say. αi ∈ E. Since Aut(E/F ) = Aut(E/F ) = G.. Because E is normal over F . and. it is ﬁnite. and so f(X) splits into distinct factors in E.. then applying σ to the equation f(α) = 0 gives f(αi ) = 0). and g(α) = 0. Proof. over F. . 22 J. E = F [α1. is separable but not normal over Q (X 3 − 2 doesn’t split in Q[α]). Let G = Aut(E/F ). and of ﬁnite degree. and f is still separable when regarded as a polynomial over F . Since the ai are symmetric polynomials in the αi . We conclude that f(X) = g(X). But also g(X)|f(X). Because E is separable over F . By Artin’s lemma. and let g(X) = (X − αi ) = X m + a1X m−1 + · · · + am . Now the inequalities [E : E G ] ≤ (G : 1) ≤ (Aut(E/E G ) : 1) = [E : E G ] must be equalities. • G ⊂ Aut(E/E G ).. and so (by 3. αi algebraic over F . then F = E Aut(E/F ). and so F = E Aut(E/F ).8 shows that [E : F ] = # Aut(E/F ) [E : F ] = # Aut(E/F ). and so f(X)|g(X) (see p7). Let fi be the minimum polynomial of αi over F . each fi splits in E.√ √ Example 3. We know that: • [E : E G ] ≤ (G : 1) (Artin). (b) F = E G for some ﬁnite group of automorphisms of E. we have to prove that f splits into distinct factors in E. if G and F are as in (b) then G = Aut(E/F ). and so G = Aut(E/E G ). then G = Aut(E/E G ). f is separable. It is monic.. and so E is the splitting ﬁeld of f = fi . αm} be the orbit of α under G. it is generated over F by a ﬁnite number of elements. Then E is also the splitting ﬁeld of f ∈ F [X]. (a) =⇒ (b).8) the order of Aut(E/E G ) is [E : E G ]. we know that [E : F ] ≤ (G : 1).. Let {α1 = α. (c) =⇒ (a). (c) E is normal and separable.

FIELDS AND GALOIS THEORY 23 An extension of ﬁelds E ⊃ F satisfying the equivalent conditions of the proposition is called a Galois extension. [E : F ]sep = [E : M]sep[M : F ]sep. Let E ⊃ M ⊃ F . 3. When we drop the assumption that E is separable over F . Theorem 3. In the course of the proof of the the above theorem we showed that the minimum polynomial of α is (X − αi ). it is also the splitting ﬁeld of f regarded as an element of M[X]. When E ⊃ M ⊃ F (ﬁnite extensions). H1 ⊃ H2 ⇐⇒ E H1 ⊂ E H2 .13. Next let M be an intermediate ﬁeld. The elements of E separable over F form a subﬁeld E of E that is separable over F . The fundamental theorem of Galois theory. Let E be Galois over F with Galois group G. i. E σHσ = σ(E H ). Every ﬁnite separable extension E of F is contained in a ﬁnite Galois extension. If Ω is an algebraically closed ﬁeld containing F . An element α ∈ E is said to be separable over F if its minimum polynomial over F is separable. if E is Galois over F . . We ﬁrst have to show that Gal(E/E H ) = H.. and let α ∈ E. then there are exactly [E : F ]sep F -homomorphisms E → Ω. Corollary 3. Corollary 3. The elements α1 = α. −1 The group σHσ −1 ↔ σM.e. in which case Gal(E H /F ) = G/H. We know E is the splitting ﬁeld of some f ∈ F [X].16.. and let H = Gal(E/M). Note that we have shown that F = E Gal(E/F ).15. i. and let G = Gal(E/F ). we can still say something.12. αm of the orbit of α are called the conjugates of α.. Let H be a subgroup of G.. The maps H → E H and M → Gal(E/M) are inverse bijections between the set of subgroups of G and the set of intermediate ﬁelds between E and F : {subgroups of G} ↔ {intermediate ﬁelds F ⊂ M ⊂ E}. Remark 3.e. (H1 : H2 ) = [E H2 : E H1 ].. Thus we have proved that Inv(·) and Gal(E/·) are inverse bijections.17 (Fundamental theorem of Galois theory). α2 . Indexes equal degrees. and Theorem 3. Let E = F [α1. then it is Galois over M. We have to show that E H = M.. Moreover: (a) (b) (c) (d) The correspondence is inclusion-reversing. In particular. αm]... Let E be a Galois extension of F . Remark 3. but this is again proved in Theorem 3. . Gal(E/σM) = σ Gal(E/M)σ −1 .e. E is separable over F ⇐⇒ E is separable over M and M is separable over F. i. But we have already observed that E is Galois over E H . Proof. Proof.12 shows that Gal(E/E H ) = H. Proof. Let fi = minimum polynomial of αi over F ..4. write [E : F ]sep = [E : F ] (separable degree of E over F ). .14. Let E be a ﬁnite extension of F . and take E to be the splitting ﬁeld of fi over F . The group H is normal in G ⇐⇒ E H is normal (hence Galois) over F . and Aut(E/F ) is called the Galois group Gal(E/F ) of E over F .

assume that M is normal over F .. Therefore Gal(E/M1 · · · Mr ) = H1 ∩ . and the map σ → σ|E : Gal(EL/L) → Gal(E/E ∩ L) is an isomorphism. For example let M1 . ∩ Hr .e.24 J. and let M = E H . 2. Then Q[ζ] is the splitting ﬁeld of the irreducible polynomial X6 + X5 + X4 + X3 + X2 + X + 1 (see 1. But Gal(E/E Hi ) = Hi . . Assume L and E are contained in some large ﬁeld Ω. .. Let E1 /F and E2 /F be Galois. the ﬁrst equality follows from (3. σαi is a root of the minimum polynomial of αi over F . hence it must correspond to the largest subgroup contained in all Hi . and so lies in M. we must have σM = M for all σ ∈ G. Because σHσ −1 = H for all σ ∈ G.31). and write M = F [α1. all α ∈ M}. Remark 3. i.. Hi = Gal(E/Mi )). σ2) | σ1 |E1 ∩ E2 = σ2 |E1 ∩ E2 }. Hence σM = M. where ζ is the primitive 7th root of 1. σHσ −1 = Gal(E/σM). Mr be intermediate ﬁelds.S. (b) In the case H2 = 1. Using this we can read oﬀ more results.12). all α ∈ M}. Then E1 E2 is Galois over F . then σHσ −1 = {τ ∈ G | τ σα = σα. i. The theorem shows that there is an order reversing bijection between the intermediate ﬁelds of E/F and the subgroups of G. We investigate the subﬁelds of Q[ζ] corresponding to the subgroups < σ 3 > and < σ 2 >. Then EL is Galois over L.e.. and σ → (σ|E1.. 2.8) and (3. 1). Then (by deﬁnition) M1 M2 · · · Mr is the smallest ﬁeld containing all Mi . The general case follows.18. M2 . Let σ be the element of Gal(Q[ζ]/Q) such that σζ = ζ 3 .. For any σ ∈ G. 6. . and this implies that σHσ −1 = H (by (c)). Let G be the image. say ζ = e2πi/7. 1 ≤ i ≤ 6. the action of G on E stabilizes M.. For σ ∈ G. and so M is Galois over F with Galois group G (by Theorem 3. We therefore have a homomorphism σ → σ|M : G → Aut(M/F ) with kernel H. (d) Assume H to be normal in G. (c) If H = {τ ∈ G | τ α = α. which is Hi . . We mention two further results (they are not diﬃcult to prove): 1. with E1 and E2 subﬁelds of some ﬁeld Ω. Let E/F be Galois. σζ = ζ i .e. αm].12). . MILNE (a) We have the obvious implications: H1 ⊃ H2 =⇒ E H1 ⊂ E H2 =⇒ Gal(E/E H1 ) ⊃ Gal(E/E H2 ). H = Gal(E/M). E is Galois over E ∩ L. Example 3. Then σ generates Gal(Q[ζ]/Q) because the class of 3 in (Z/7Z)× generates it (the powers of 3 mod 7 are 3. and the map σ → i deﬁnes an isomorphism Gal(Q[ζ]/Q) → (Z/7Z)× ..19. σ|E2) : Gal(E1 E2 /F ) → Gal(E1 /F ) ×Gal(E2/F ) is injective with image {(σ1 . 4. Conversely.e. and so is Galois of degree 6 over Q. i. and let Hi be the subgroup corresponding to Mi (i. . using that (H1 : 1) = (H1 : H2 )(H2 : 1) and [E : E H1 ] = [E : E H2 ][E H2 : E H1 ].. and let L be any ﬁeld containing F . some i. Then F = M G . 5. We analyse the extension Q[ζ]/Q.

Hence the ﬁeld ﬁxed by < σ > is Q[ −7]. being generated by the class of 2. and let σ be a generator of N. We have the picture: Q[ζ. which is cyclic. . The minimum polynomial of cos 2π = 7 α1 2 is therefore g(2X) = X 3 + X 2 /2 − X/2 − 1/8.FIELDS AND GALOIS THEORY 25 ¯ Note that σ 3 ζ = ζ 6 = ζ (complex conjugate of ζ). we have H ∩ N = 1 (see 3. We compute the Galois group of the splitting ﬁeld E of X 5 − 2 ∈ Q[X]. and α is a root of X 5 − 2. α]. For example. . α] N Q[ζ] G/N Q The degrees [Q[ζ] : Q] = 4. Direct calculation shows that 6 αi = i=1 6 2 5 3 4 3 ζ i = −1. The conjugates of α1 = ζ + ζ 3 −3 2 −2 are α3 = ζ + ζ .18). 2 2 Then (β − β ) = −7. . α1α2 + α1 α3 + α2α3 = −2. which we can take to be ζα (after possibly replacing σ by a power). 7 ¯ ¯ Q[ζ + ζ] is Galois over Q. α] : Q] = 20. α]/Q) has order 20. N is normal in G. ). We have H ≈ G/N ≈ (Z/5Z)× . Because Q[ζ] · Q[α] = Q[ζ. and the subgroups N and H corresponding to Q[ζ] and Q[α] have orders 5 and 4 respectively (because N = Gal(Q[ζ. α2 = ζ + ζ . ¯ Hence the minimum polynomial of ζ + ζ is g(X) = X 3 + X 2 − 2X − 1. we could take E to be the splitting ﬁeld of X 5 − 2 in C. The subﬁeld of Q[ζ] corresponding to 3 ¯ and ζ + ζ = 2 cos 2π . with ζ = e2πi/5 and α equal to the real 5th root of 2. Thus σ(α) is another root of X 5 − 2. τ στ −1 = σ 2. Q[α] H [Q[ζ. ¯ < σ > is Q[ζ + ζ]. Since < σ 3 > is a normal subgroup of < σ >. Example 3. Hence: τ ζ = ζ2 σζ = ζ τα = α σα = ζα. with Galois group < σ > / < σ 3 > . Hence G = Gal(Q[ζ. Note that τ στ −1(α) = τ σα = τ (ζα) = ζ 2 α and it ﬁxes ζ. α]/Q[ζ] . Let τ be the generator of H corresponding to 2 under this isomorphism. Let β = σβ.20. Recall (from the Homework) that E = Q[ζ. Because Q[ζ] is normal over Q (it is the splitting ﬁeld of X 5 − 1). [Q[α] : Q] = 5. Because 4 and 5 are relatively prime. therefore τ στ −1 = σ 2. and so G = N θ H. 8 by The subﬁeld of Q[ζ] corresponding to < σ 2 > is generated √ β = ζ + ζ 2 + ζ 4. Thus G has generators σ and τ and deﬁning relations σ 5 = 1. τ 4 = 1. α1 α2α3 = (ζ+ζ )(ζ +ζ )(ζ +ζ ) = (ζ+ζ +ζ 4 +ζ 6)(ζ 3 +ζ 4 ) = (ζ 4 +ζ 6 +1+ζ 2 +ζ 5+1+ζ+ζ 3 ) = 1. α] where ζ is a primitive 5th root of 1.

is constructible if and only if p is a Fermat prime. Proof. abelian. i. An extension E ⊃ F is called a cyclic. .22. . Q ⊂ E1 ⊂ E2 ⊂ · · · ⊂ Er = E with Ei of degree 2 over Ei−1 . . there will be a sequence of ﬁelds. we know there will be a sequence of groups {1} ⊂ G1 ⊂ G2 ⊂ · · · ⊂ Gr = G with Gi /Gi−1 of order 2. For example.. 2π without (happily) having to exhibit an explicit formula for cos 65537 . In particular α constructible =⇒ [Q[α] : Q] = 2s some s. and so E = F [ b2 − 4c]. Thus Gf can . MILNE The subgroup H has ﬁve conjugates. 3. We know elements of Gal(E/F ) map roots of f to roots of f.24. and let G = Gal(E/Q). . Correspondingly. Thus a regular p-gon. αn }. Suppose α ∈ E where E is Galois over Q of degree 2r . they deﬁne permutations of {α1 . Then √ E = F [ d] for some d ∈ F . Let E/F be a quadratic extension of ﬁelds of characteristic = 2. . If the polynomial f ∈ F [X] is separable. α= 2 Corollary 3.e. If α is contained in a Galois extension of Q of degre 2r then it is constructible.. 3. α2. . .. α2 . α ∈ F . every quadratic extension is obtained by extracting a square root. .. of the r form 22 + 1. This proves the theorem. αn]. and we call Gal(E/F ) the Galois group Gf of f. .. Now we can prove a partial converse to this last statement. Since they are automorphisms. which has order p − 1 = 2k . αn }.23. σ iHσ −i ↔ σ i Q[α] = Q[ζ iα]. p Proof.. we have proved that the regular 65537-polygon is constructible. αn } into itself.21.√ let X 2 + bX + c be the minimum polynomial of α. they map the set {α1. Proof. . then its splitting ﬁeld E is Galois over F . Let α ∈ E. As E = F [α1.5... then cos 2π is constructible. solvable Galois group.. i. Constructible numbers revisited. .6. an element of Gal(E/F ) is uniquely determined by its action on {α1 .. From a theorem on the structure of p-groups. n Let f = i=1 (X − αi ) in the splitting ﬁeld E.. . p prime. . we showed that a number α is constructible if and only if it is contained in a ﬁeld √ √ Q[ a1] · · · [ ar ]. Definition 3. The / and √ −b± b2 −4c .26 J. . which correspond to the ﬁve ﬁelds Q[ζ i α].S. If p is a prime of the form 2k + 1. Galois group of a polynomial. α2 . 1 ≤ i ≤ 5. Lemma 3. Theorem 3. Earlier. . abelian. But (see below). solvable extension if it is Galois with cyclic. The ﬁeld Q[e2πi/p] is Galois over Q with Galois group G ≈ (Z/pZ)× . and we know that square roots can be constructed using only a ruler and compass.e.

Note that (Gf : 1) ≤ deg(f)!.25. and f is required to be separable then the theorem becomes true in characteristic p. .FIELDS AND GALOIS THEORY 27 be identiﬁed with a subset of Sym({α1 . This gives a description of Gf without mentioning ﬁelds or abstract groups (neither of which were available to Galois). . Let f be a polynomial.7. αmi ∈ Fi−1 . We shall prove this later. αn }) ≈ Sn . αn ) = 0 =⇒ P (σα1. Also we shall exhibit polynomials f(X) ∈ Q[X] with Galois group Sn . . i (b) Fm contains a splitting ﬁeld for f. Theorem 3. αn } with the property P ∈ F [X1. . We say the equation f(X) = 0 is solvable (by extracting radicals) if there is a tower F = F0 ⊂ F1 ⊂ F2 ⊂ · · · ⊂ Fm such that (a) Fi = Fi−1 [αi]. The equation f = 0 is solvable if and only if the Galois group of f is solvable. . . (ii) X p − X − a is not solvable by radicals. . . . . P (α1 . Xn ]. one sees that Gf consists of the permutations σ of {α1 . From the deﬁnitions. . . .26. If the deﬁnition of solvable is changed to allow extensions of the type in (ii) in the chain. which therefore are not solvable when n ≥ 5. If F has characteristic p. . . . Remark 3. 1832) Let F be a ﬁeld of characteristic zero. and so not have a Galois group. 3. . then the theorem fails for two reasons: (i) f may not be separable. Solvability of equations. . . σαn ) = 0. α2. . (Galois. α2. .

the proof of Proposition 4. (a) We know that D(f) is an element of Ff ﬁxed by Gf =df Gal(Ff /F ). Proposition 4. Note that D(f) = 0 if f has a only simple roots.3. In this section. we investigate general methods for computing Galois groups. ∆(f) = 0.6). . and so the formula σ∆(f) = sign(σ)∆(f) shows that σ ﬁxes ∆(f) ⇐⇒ σ ∈ An . The formulas for the discriminant rapidly become very complicated.e. one can put any cubic polynomial in this form. Therefore it lies in F (by the Fundamental Theorem of Galois Theory). 4. so that Gf = Gal(Ff /F ).. When is Gf ⊂ An ? Consider a polynomial and let f(X) = f(X) = X n + a1X n−1 + · · · + an n i=1 (X − αi ) in some splitting ﬁeld. Fortunately. then Gf is not contained in An . Gf is not necessarily contained in An . Corollary 4. For example: D(aX 2 + bX + c) = b2 − 4ac D(X 3 + bX + c) = −4b3 − 27c2 . This can be proved more directly by noting that complex conjugation will act on the roots as the product of s transpositions (cf. (b) The subﬁeld of Ff corresponding to An ∩ Gf is F [∆(f)]. Computing Galois Groups. . Then D(f) will not be a square if it is negative. Hence Gf ⊂ An ⇐⇒ ∆(f) ∈ F ⇐⇒ D(f) is a square in F. Suppose F ⊂ R. The ﬁrst equation follows immediately from the deﬁnition of the signature of σ (see Groups. (b) Because f has simple roots. Assume f is separable. and so Gf ∩ An = Gf ⇐⇒ F [∆(f)] = F .X). The discriminant of f can be expressed as a universal polynomial in the coeﬃcients of f—we shall prove this later. It is known that the sign of D(f) is (−1)s where 2s is the number of nonreal roots of f in C. αn }) (as in §3. Therefore Gf ∩ An is the subgroup of Gf corresponding to F [∆(f)]. that for 5 X + aX 4 + bX 3 + cX 2 + dX + e has about 60 terms. 1≤i<j≤n ∆(f) = D(f) = ∆(f)2 = 1≤i<j≤n (αi − αj )2. (a) σ∆(f) = sign(σ)∆(f). Remark 4. .13). .1. p31). MILNE 4. and the second equation is obtained by squaring the ﬁrst. Set (αi − αj ).S.28 J. i.1. Proof. Let Ff be a splitting ﬁeld for f. By completing the cube. Identify Gf with a subgroup of Sym({α1 . and let σ ∈ Gf . for example. Maple knows them: the syntax is “discrim(f. Of course the converse is not true: when s is even. (a) The discriminant D(f) ∈ F . Thus if s is odd.2. (b) σD(f) = D(f). where sign(σ) is the signature of σ. Let f(X) ∈ F [X] be of degree n and have only simple roots. if f is separable with no multiple factors. Proof.” where f is a polynomial in the variable X. .

5. the other roots of f(X).. and let α be one of its roots.3.] → Ff (see 2. and assume that the roots of f are simple.4). On the other hand.. Proof. . its discriminant is −4(−3)3 − 27 = 81 = 92 .2. and we see that we must have f(X) = g(X).4). When is Gf transitive? Proposition 4. . Polynomials of degree ≤ 3. and so there is a natural F -isomorphism F [α] → F [β]. 4. Then f is inseparable ⇐⇒ F has characteristic 3 and f(X) = X 3 − a some a ∈ F \ F 3. Thus Gf is a transitive subgroup of Sn whose order is divisible by n. ⇐= : Let g(X) ∈ F [X] be an irreducible factor of f. Let E be the splitting ﬁeld of f. If β is a second root of f. F [α] ≈ F [X]/(f(X)) ≈ F [β]. for otherwise we are essentially back in the previous case. Now the equation 0 = σg(α) g(X)∈F [X] = g(σα) shows that β is also a root of g. If f is separable.FIELDS AND GALOIS THEORY 29 4. α2. and let f(X) = (X − αi ) in E. Let f(X) ∈ F [X] be a polynomial of degree 2. . If f is separable.] with α1 = α and α2 . Let f(X) ∈ F [X] have only simple roots. Example 4. We can assume f to be irreducible. =⇒ : If α and β are two roots of f(X) in a splitting ﬁeld Ff for f. (12)(34). X 3 +3X +1 ∈ Q[X] is also irreducible (apply 1. . then (by assumption) β = σα for some σ ∈ Gf . Then f(X) is irreducible if and only if Gf permutes the roots of f transitively. There are only two possibilities: Gf = A3(=< (123) >) or S3 according as D(f) is a square in F or not. .4. Example 4. and so its Galois group is A3 .6. and so its Galois group is S3. then n|(Gf : 1) because [F [α] : F ] = n and [F [α] : F ]|[Ff : F ] = (Gf : 1)..7). Quartic polynomials. then they both have f(X) as their minimum polynomial. For example. X 3 − 3X + 1 ∈ Q[X] is irreducible (apply 1. Then the F -isomorphism F [α] → F [β] extends (step by step) to a homomorphism F [α1.4. 4. then Gf = 1(= A2) or S2 according as D(f) is a square in F or not. α ↔ X ↔ β. Write Ff = F [α1. In order to determine Gf we shall exploit the fact that S4 has V = {1. Let f(X) ∈ F [X] be a polynomial of degree 3. but its discriminant is −135 which is not a square in Q. which must be an isomorphism. (14)(23)} as a normal subgroup—it is normal because it contains all elements of type 2+2—see Groups p34. α2 . (13)(24). namely. then Gf is a transitive subgroup of S3 whose order is divisible by 3. Let f(X) be a quartic polynomial. Note that when f(X) is irreducible of degree n. α3 . Then f is inseparable ⇐⇒ F has characteristic 2 and f(X) = X 2 − a for some a ∈ F \ F 2.. We identify the .

They are distinct elements of E because the αi are distinct. α2. By the Sylow theorems. and. and hence has order 8. Compute everything in terms of the αi ’s.10. If f = X 4 +bX 3 +cX 2 +dX+e. For example. It follows that V is the intersection of the three Sylow 2-subgroups. because they are conjugate and it is normal. α4 }) permutes {α. the stabilizer of β is < (1234). Proof. α3 . α4 }). MILNE Galois group Gf of f with a subgroup of the symmetric group S4 = Sym({α1 . γ. γ] corresponds to Gf ∩ V .. Each Sylow 2-subgroup stabilizes exactly one of α. Therefore (by the Fundamental Theorem) g(X) has coeﬃcients in F . β. it must be contained in all three. and therefore their intersection V is the subgroup of S4 ﬁxing α. Lemma 4.30 J. we have: G∩V Lemma 4. or γ. Groups of order 8 in S4 are Sylow 2-subgroups. then g = X 3 −cX 2 +(bd−4e)X−b2e+4ce−d2. The ﬁrst statement follows from the above discussion.) Now let f be an irreducible separable quartic. α − β = α1 (α2 − α3 ) + α4 (α3 − α2) = (α1 − α4 )(α2 − α3 ). β. There are three of them. Picture: E 1 | | M G∩V G/G ∩ V | | F G Let g(X) = (X − α)(X − β)(X − γ) ∈ M[X]—it is called the resolvant cubic of f. with Galois group G/G ∩ V . β. (Cf. and γ. and the second follows from the Fundamental Theorem of Galois Theory. Hungerford. γ} transitively. The stabilizer of each of α. Hence M is Galois over F . any element of Gf ) merely permutes α.4. V.8. (13) >. β. e.g.S. β. The discriminants of f and g are equal. γ must therefore be a subgroup of index 3 in S4. V is contained in a Sylow 2subgroup. α3 . and so ﬁxes g(X). Any permutation of the αi (a fortiori. Consider the partially symmetric elements α = α1 α2 + α3 α4 β = α1 α3 + α2 α4 γ = α1 α4 + α2 α3 . β. Proof. all isomorphic to D4 . α2. More explicitly. The ﬁeld M = F [α. The group Sym({α1 . There are the following possibilities: G (G ∩ V : 1) (G : V ∩ G) S4 4 6 A4 4 3 V 4 1 D4 4 2 C4 2 2 .7. Then G = Gf is a transitive subgroup of S4 whose order is divisible by 4.

because G/V ∩ G = Gal(M/F ). Then we’ll have (2 3). factors over M. Only the ﬁrst group acts transitively on the roots of f. . After replacing σ with the power. and so (from 4.20). Let the p-cycle be σ = (i1 · · · ip ). Consider f(X) = X 4 − 10X 2 + 4 ∈ Q[X]. We can compute (G : V ∩ G) from the resolvant cubic g. Note that G can’t. Proof. in the group generated by σ and τ . . Let f be an irreducible polynomial of prime degree p in Q[X].4) we see that (in this case) G = D4 or C4 according as f is irreducible or not in M[X]. jp ). we may assume that the transposition is τ = (12). To learn the syntax. Lemma 4. Therefore Gf ⊂ Sp contains a transposition and a p-cycle. (4 5). Example 4. Then Sp is generated by any transposition and any p-cycle.5. Note that f. then G ∩ V = V or C2. If f splits in C and has exactly two nonreal roots. hence Gf = C4 . Once we know (G : V ∩ G). Examples of polynomials with Sp as Galois group over Q. [Q[α] : Q] = deg f = p. for example. be the group generated by (12) and (34) because this is not transitive. then Gf = Sp . Let p be a prime number. Consider f(X) = X 4 + 4X 2 + 2 ∈ Q[X]. we may suppose σ = (1 2 j3 . we may choose to write σ so that 1 occurs in the ﬁrst position. and these elements generated Sp . . One can check directly (using 1. Proposition 4. Let σ be complex conjugation on C. . After renumbering. Maple knows how to factor polynomials over Q and over Q[α] where α is a root of an irreducible polynomial. analyze the equation as in (3. Let E ⊂ C be the splitting ﬁeld of f. Example 4. Hence Gf = V . .10. Consider f(X) = X 4 − 2 ∈ Q[X]. and its resolvant cubic is (X + 10)(X + 4)(X − 4).12. thus M = Q[ 2]. 4.9. Now some power of σ will map 1 to 2 and will still be a p-cycle (here is where we use that p is prime). (3 4). It is irreducible by Eisenstein’s √ criterion.13. and so is the whole of Sp . we may suppose σ = (1 2 3 . If [M : F ] = 2. p). but the only elements of order p in Sp are p-cycles (here we use that p is prime again). . type: ?Factor. Alternatively. Proof. (G : V ∩ G) = [M : F ]. and its resolvant cubic is (X − 4)(X 2 − 8). and the groups of type C4 are those generated by cycles of length 4. and after renumbering again. and its resolvant cubic is g(X) = X 3 + 8X. Because f is irreducible. Hence M = Q[i 2]. The groups of type D4 are the Sylow 2-subgroups discussed above. Then σ transposes the two nonreal roots of f(X) and ﬁxes the rest. One can check that f is irreducible over M. σ = (1 i2 · · · ip). Therefore Gf contains an element of order p (Cauchy’s theorem). The next lemma gives a criterion for a subgroup of Sp to be the whole of Sp.11.FIELDS AND GALOIS THEORY 31 (G ∩ V : 1) = [E : M]. and M is the splitting ﬁeld of g. . and so its Galois group is D4 .6) that it is irreducible. and so p|[E : Q] = (Gf : 1). when regarded as a polynomial in X 2 . Example 4. we can deduce G except in the case that it is 2. It is irreducible by Eisenstein’s √ criterion. . and let α ∈ E be a root of f.

No harm results if we identify Fp with this subﬁeld of E. m ≥ 0.2. and all elements of E (including 0) are roots of X q − X. and so any two ﬁelds with q elements are isomorphic.17. then (a − b)q = aq − bq = a − b. For each power q = pn there is a ﬁeld Fq with q elements. It has order n. The ﬁeld Fq is Galois over Fp because it is the splitting ﬁeld of a separable polynomial. Finite ﬁelds. and E is Galois over that ﬁeld. The polynomial g(X) certainly has exactly two nonreal roots. 4. Corollary 4. Proof.3) that σ = (x → xp ) is an automorphism of Fq . .14.6. Corollary 4. Fq is Galois over Fp with cyclic Galois group generated by the Frobenius automorphism σ(a) = ap. which is relatively prime to f(X) (in fact. and so the ﬁxed ﬁeld of < σ > is Fp . Then E contains exactly one ﬁeld with pm elements for each m|n. Its graph crosses the x-axis exactly p − 2 times. Example 4. q = pn . Then S is obviously closed under multiplication and the formation of inverses.(X − np−2 ). The derivative f (X) = −1. This proves that < σ >= Gal(Fq /Fp ). namely. where g(X) = (X 2 + m)(X − n1 ). Let p≥ 5 be a prime number. and a ∈ Fq is ﬁxed by σ if and only if ap = a. Proof. to every polynomial). Proposition 4. We noted in (1. Hence E is a splitting ﬁeld for X q − X.15. In particular. Let E be a ﬁeld with pn elements. then all ai are even. the ﬁeld of p elements. We know that E is Galois over Fp and that Gal(E/Fp ) is the cyclic group of order n generated by σ. Hence the graph of f(X) = g(X) − 2 also crosses the x-axis exactly p − 2 times. {m1E | m ∈ Z}. The ﬁxed ﬁeld of < σ m > is Fpm .. But Fp consists exactly of such elements. and so f(X) has q distinct roots in E. Hence S is a ﬁeld. and hence any two such ﬁelds are isomorphic. and so S = E. MILNE It remains to construct polynomials satisfying the conditions of the Proposition. Moreover. Then E has q = pn elements. and so E × is a group of order q − 1.16.S. Let E be a ﬁeld of degree n over Fp . Let Fp = Z/pZ. any other ﬁeld E of characteristic p contains a copy of Fp . Every extension of ﬁnite ﬁelds is simple. Hence Eisenstein’s criterion implies that f(X) is irreducible.32 J. Let f(X) = g(X) − 2. but it is also closed under subtraction: if aq − a = 0 and bq − b = 0. Only the ﬁnal statement remains to be proved. When we write f(X) = X p +a1X p−1 +· · ·+ap.. and ap = −(m ni )−2 is not divisible by 4. Hence the nonzero elements of E are roots X q−1 − 1. Choose a positive even integer m and even integers n1 < n2 < · · · < np−2 . It is the splitting ﬁeld of X q − X. Let S be the set of its roots. The subgroups of < σ > are the groups < σ m > with m|n. Now let E be the splitting ﬁeld of f(X) = X q − X. E has pn elements. As we noted in §1. and its maxima and minima all have absolute value > 2 (because its values at odd integers have absolute value > 2).

. . Lemma 4. For S ⊂ {1.20. then its factors in the cycle decomposition of σ correspond to the distinct orbits of σ. a generator for F× ). the factors of X p −X are distinct because it has no common factor with its derivative.”. . then σ has a cycle decompostion of type m1 + · · · + mr = m. Proof. then f factors as f = f1 · · · fr with fi irreducible of degree mi . . Then F contains one ﬁeld Fpn for each integer n ≥ 1— n it consists of all roots of X p − X—and Fpm ⊂ Fpn ⇐⇒ m|n. Let P be a prime ideal in R.FIELDS AND GALOIS THEORY 33 Proof. Let α1. then clearly E = Fp [ζ]. then f(X) has a root in a ﬁeld of degree d over n Fp . . .. n Maple factors polynomials modulo p very quickly. 4. If f(X) is irreducible of degree d. Thus.19. . Computing Galois groups over Q. and their reductions αi modulo P are the roots of f . m} stable under G. and let f ¯ be a monic polynomial in R[X]. We sketch a practical method for computing Galois groups over Q and similar ﬁelds. to obtain a list of all monic polynomials of degree 1. Proposition 4. Consider E ⊃ F . But the splitting ﬁeld of X p − X contains a copy of every ﬁeld of degree d over Fp with n n d|n. When we regard σ as a permutation of the roots of f. The partially ordered set of ﬁnite subﬁelds of F is isomorphic to the set of integers n ≥ 1 partially ordered by divisibility. . . Hence. αm be the distinct roots of f(X). Assume neither f nor f ¯ lie in R. . Moreover Gf¯ ⊂ Gf when both ¯ ¯ ¯ are identiﬁed with subgroups of Sym{α1. consider fS = i∈S (X − αi). . . . First.4. The syntax is “Factor(f(X)) mod p. αm } = Sym{α1 . m2. for example. . . . Each monic irreducible polynomial of degree d|n in Fp [X] occurs exn actly once as a factor of X p − X. If ζ generates E × as a multiplicative group. . and Fq used to be denoted GF (q) (it still is in Maple). and it is ﬁxed under the action of Gf (and hence has coeﬃcients in F ) if and only if S is stable under Gf . . For example. . αm }. . 2. Then the roots α1 . ask Maple to factor X 625 − X.. mr . . . Let f(X) be a monic separable polynomial in F [X] of degree m with distinct roots. but such sets S are precisely the orbits of G in {1. and let E be the splitting ﬁeld of f. . or 4 over F5 . Corollary 4. and suppose that Gf ⊂ Sm has r orbits with m1 . . Hence some root of X p − X is also a root of f(X). . with pn elements say. . . if the degrees of the distinct irreducible factors of f are m1 .18. To start it. . αm of f in (R/P )[X]. . Now suppose F is ﬁnite. . it can ﬁnd a primitive element for Fq (i.e. . Maple contains a “Galois ﬁeld package” to do computations in ﬁnite ﬁelds. . 2. type: q readlib(GF). This polynomial divides f(X) in Ff [X]. m}.7. . and hence is cyclic (see Exercise 3). Then E × is a ﬁnite subgroup of the multiplicative group of a ﬁeld. Our ﬁrst result generalizes Proposition 4. The n Galois group of E over F is generated by the Frobenius automorphism σ : x → xp . . m}. Finite ﬁelds were sometimes called Galois ﬁelds. . Therefore the irreducible factors are the polynomials fS corresponding to minimal subsets S of {1. Proof. . mr elements respectively (so that m = m1 + · · · + mr ). and let f be the image of f ¯ has a multiple root. Let F be an algebraic closure of Fp . Let R be a unique factorization domain with ﬁeld of fractions F . and therefore f(X)|X p − X. .

the eﬀective Chebotarev density theorem. we obtain the following theorem. Consider X 5 −X −1. On combining these results. Let f(X) ∈ Z[X] be a monic polynomial of degree m. Therefore Gf = S5. the transposition can be transformed into (in). Then (i) Gf is transitive (it contains an n-cycle because f ≡ f1 mod 2).21 (Dedekind). (2n). n − 1) be the (n − 1)-cycle. (n − 1 n). Select monic polynomials of degree n. Hence Gf contains (12345) and (ik)(lmn). This lists all transitive subgroups of Sn . MILNE Proof. f3 with coeﬃcients in Z such that: (a) f1 is irreducible modulo 2. then look at its action on the set of subsets of r roots for some r.24. and hence also ((ik)(lmn))3 = (ik). Algebra 11 (1983). (ii) Gf contains a cycle of length n − 1 (because f ≡ f2 mod 3). .34 J. (b) f2 = (degree 1)(irreducible of degree n − 1) mod 3. then this will be seen by looking modulo a set of prime numbers of positive density.23. Hence Gf is Sn . We choose them to have distinct roots. . and let p be a prime such that f mod p has simple roots (equivalently. Proof. See. and so it contains the product of a transposition with a commuting element of odd order. some 1 ≤ i ≤ n − 1.S. The transitive groups of degree up to eleven. Lemma 4. we are left with the transposition). §61 (second edition) or Math 676 (Algebraic Number Theory). Theorem 4. Then Gf contains an element whose cycle decomposition corresponds to the partition: m = m1 + · · · + mr . Factor f modulo a sequence of primes p not dividing D(f) to determine the cycle types of the elements in Gf —a diﬃcult theorem in number theory. and modulo 3 it is irreducible. Now look up a table of transitive subgroups of Sn with order divisible by n and their cycle type. f2 . D(f) is not divisible by ¯ p).22. . Suppose that f = fi with fi irreducible of degree mi in Fp [X]. n ≤ 11. says that if a cycle type occurs in Gf . Modulo 2. on raising this to an appropriate odd power. (c) f3 = (irreducible of degree 2)(product of 1 or 2 irreducible polys of odd degree) mod 5. and will occur for a prime less than some bound. Example 4. (iii) Gf contains a transposition (because f ≡ f3 mod 5. This gives the following strategy for computing Galois groups over Q. 863–911. Butler and McKay. . Omitted—see van der Waerden. f1. Modern Algebra. By virtue of the transitivity. Example 4. and these elements obviously generate Sn . If this doesn’t suﬃce to determine the group. Now the (n − 1)-cycle and its powers will transform this into (1n). Comm. I. Let (123 . . A transitive subgroup of H ⊂ Sn containing a transposition and an (n − 1)cycle is equal to Sn . . this factors as (X 2 +X +1)(X 3 +X 2 +1). Take f = −15f1 + 10f2 + 6f3 . and gives the cycle types .

but my eﬀorts to obtain a manual for it have been unsuccessful.. See also. 20 (1985) 273–281. Soicher and McKay. J. Computing Galois groups over the rationals. these invariants are suﬃcient to determine the subgroup up to isomorphism. Maple can compute Galois groups for polynomials of degree ≤ 7 over Q. Magma (the replacement for Cayley) probably knows much more. Number Theory.FIELDS AND GALOIS THEORY 35 of their elements and the orbit lengths of the subgroup acting on the r-sets of roots. To learn the syntax. . type ?galois. with few exceptions.

and have β as a root: g(β) = 0. . It is clear that this argument should work much more generally.. X = α1−αj . where √ √ √ √ σ √2 = − 2 2 τ √2 = √ √ . Let α1 = α. βt. . √ √ Consider for example Q[ 2. . . j = 1.. αs be the roots of f in some ﬁeld containing E. The polynomials g(X) and f(γ − cX) have coeﬃcients in F [γ][X].. Primitive element theorem. because the roots of g are β1. Theorem 5. αr are separable over F (but not necessarily α1 ). We say that an element α algebraic over a ﬁeld F is separable over F if its minimum polynomial over F has no multiple roots. Thus let E = F [α. and assume that α2 . Applications of Galois Theory In this section. Let f and g be the minimum polynomials of α and β over F . we proved this in (4. . It suﬃces to prove the statement for r = 2. 3] = Q[ 2 + 3]. Therefore gcd(g(X). βt be the roots of g. √ √ √ √ These all diﬀer from 2 + 3. 3 Note that √ √ √ √ σ(√2 + √3) = − 2 + 3. and so the the equation αi + Xβj = α1 + Xβ1 .. . namely.36 J. Hence we may assume F to be inﬁnite. We know (see Exercise 13) that its Galois group over Q is a 4-group < σ. 3]/Q. β] with β separable over F [α]. and so only the identity element of Gal(Q[ 2. i 1 has exactly one solution. . In fact. and let β1 = β. and we chose c so that γ − cβj = αi unless i = 1 = j. Then there is an element γ ∈ E such that E = F [γ]. MILNE 5.. then αi + cβj = α + cβ unless i = 1 = j. I claim that γ = α + cβ generates E over F. αr ] be a ﬁnite extension of F .1. 3]/Q) ﬁxes √ √ the √ √ elements of Q[ 2 + 3]. .. βj = β1. Thus a ﬁnite extension E of F is separable if and only if all its elements are separable over F . If we choose a c diﬀerent from any of these β −β solutions (using that F is inﬁnite). we apply the Fundamental Theorem of Galois Theory to obtain other results about polynomials and extensions of ﬁelds.S. this implies that 2 + 3 is a primitive element: √ √ √ √ Q[ 2.. β2 . 5. Let E = F [α1. √ √ τ (√2 + √3) = 2 3. For ﬁnite ﬁelds. For j = 1.16). Recall that a ﬁnite extension of ﬁelds E/F is simple if E = F [α] for some element α of E. We shall show that (at least) all separable extensions have primitive elements. Such an α is called a primitive element of E.1. Proof.. . . σ 3 = τ 3 = − 3. β is their only common root. τ >. f(γ − cX)) computed in some ﬁeld . f(γ − cβ) = f(α) = 0. . According to the Fundamental Theorem. √− √ (στ )( 2 + 3) = − 2 − 3..

Therefore. of the permutations of the roots). These remarks make it very easy to write down primitive elements. . and so M = M : M is generated by the coeﬃcients of g(X).3 Zariski showed that there is even an intermediate ﬁeld M that is not isomorphic to F (X. The proof shows that γ can be chosen to be of the form γ = α1 + c2 α2 + · · · + cr αr . Let M be the subﬁeld of E generated over F by the coeﬃcients of g(X). If E = F [α1. where k is an algebraically closed ﬁeld of characteristic p.2. . Y p ) = F . p508). Remark 5. Y ] ⊃ F [X + cY ] ⊃ F with the degree of each extension equal to p. Then g(X) divides f(X) in M[X]. and let g(X) be the minimum polynomial of γ over M. αr ] is Galois over F . Y ) ⊃ k(X p . F ⊂ M ⊂ E.3. (c) The simplest nonsimple extension is k(X. . F ⊂ M ⊂ E. ci ∈ F. and hence also in E[X]. Let M be such a ﬁeld.e. (a) Note that the proposition in fact gives a description of all the intermediate ﬁelds: each is generated over F by the coeﬃcients of a factor g(X) of f(X) in E[X]. This gives another proof of the theorem when E is separable over F . then F [X + cY ] would contain both X and Y . and consequently only ﬁnitely many possible M’s. (b) The proposition has a converse: if E is a ﬁnite extension of F and there are only ﬁnitely many intermediate ﬁelds M. Before proving this. Hence there are inﬁnitely many distinct intermediate ﬁelds. Remark 5. Proof. Y ) = F [X. that there is an inﬁnite sequence of intermediate ﬁelds.. because Galois theory shows that there are only ﬁnitely many intermediate ﬁelds in this case (embed E in a Galois extension of F ). Our hypotheses are minimal: if two of the α’s are not separable. c = c . For any c ∈ k. Y ). 1977. ﬁxed by some. then E is a simple extension of F (see Dummit. The coeﬃcients of such a g(X) are partially symmetric polynomials in the roots of f(X) (i. no two of which are isomorphic. Assume F to be inﬁnite.4. using the methods of algebraic geometry. Y ) : F ] = p2 . and then α = γ − cβ also lies in F [γ]. . there are only ﬁnitely many possible g’s. In fact. but not necessarily all. Proposition 5. Then there are only ﬁnitely many intermediate ﬁelds M. Let E = F [γ] be a simple algebraic extension of F . Hence [E : M ] = deg g = [E : M]. and Piotr Blass showed in his UM thesis. Let f(X) be the minimum polynomial of γ over F . we need another result.FIELDS AND GALOIS THEORY 37 splitting fg is X − β. If F [X + cY ] = F [X + c Y ]. but (equally clearly) g(X) is the minimum polynomial of γ over M .1) that the gcd of two polynomials has coeﬃcients in the same ﬁeld as the coeﬃcients of the polynomials. which is impossible because [k(X. then an element of this form will be a primitive element provided it is moved by every element of Gal(E/F ) except 1. all but a ﬁnite number of elements of this form will serve. Clearly M ⊂ M. but we have seen (Proposition 3. 3 . we have k(X. Hence β ∈ F [γ]. then the extension need not be simple.

Both are immediate consequences of the Intermediate Value Theorem. He gave the ﬁrst rigorous proof in 1816. In fact. and can therefore be obtained by extracting the square root of an element of C (see 3. and so has a root in R. and so E is Galois over R.5. I prefer the proof based on Liouville’s theorem in complex analysis to the more algebraic proof given in the text: if f(z) is a polynomial without a root in C. then (c + di)2 = α. The Fundamental Theorem was quite a diﬃcult theorem to prove. 4 . then it has a subgroup N of index 2. • Every polynomial of odd degree with real coeﬃcients has a real root. and choose c. (a) The ﬁeld C is the algebraic closure of R. Let f(X) ∈ R[X]. The ﬁeld C of complex numbers is algebraically closed. It therefore has degree 1. and let E be a splitting ﬁeld for f(X)(X 2 + 1)—we have to show that E = C. Theorem 5. thus C = R[i]. d = . Corollary 5. which is a contradiction. Let G be its Galois group. and so M = R and G = H. then f(z)−1 will be bounded and holomorphic on the whole complex plane. (b) The set of all algebraic numbers is an algebraic closure of Q. d to be real numbers such that √ √ (a + a2 + b2) (−a + a2 + b2) 2 2 . with a. Thus E = C. Part (a) is obvious from the deﬁnition of “algebraic closure”. Let M = E H . The minimum polynomial of α over R has odd degree. Because it is not strictly a theorem in algebra: it is a statement about R whose construction is part of analysis. We ﬁnally prove the misnamed4 fundamental theorem of algebra. b ∈ R. unlike the rationals. We now have that Gal(E/C) is a 2-group.) We ﬁrst show that every element of C has a square root. (Intuitively. Gauss gave a proof in his doctoral dissertation in 1798 in which he used some geometric arguments which he didn’t justify. and let H be a Sylow 2-subgroup of G. c = 2 2 Then c2 − d2 = a and (2cd)2 = b2 . The ﬁeld E N has degree 2 over C.38 J.10) that every f(X) ∈ R[X] has a root in C. and let i be a root of X 2 + 1 in C. MILNE 5. We have to show (see 2. this says that. and hence (by Liouville) constant.2.23). Since R has characteristic zero. Fundamental Theorem of Algebra. and (b) follows from the discussion on p15. but we have seen that all such elements already lie in C.6. the real line has no “holes”.S. the polynomial is separable. and M = R[α] some α (Theorem 5. Proof.1). Hence E N = C. The two facts we need to assume about R are: • Positive real numbers have square roots. The elegant argument given here is a simpliﬁcation by Emil Artin of earlier proofs. Then M is of odd degree over R. Deﬁne C to be the splitting ﬁeld of X 2 + 1 ∈ R[X]. Proof. which says that a continuous function on a closed interval takes every value between its maximum and minimum values (inclusive). Write α = a + bi. If it is = 1. If we choose the signs of c and d so that cd has the same sign as b.

and X6 − 1 = X 2 − X + 1. Φ3 (X) = X 2 + X + 1. the order of (Z/nZ)× . Thus Φn = (X − ζ) (product over the primitive nth roots of 1). then its image is {1}. Proposition 5.3. Let E be the splitting ﬁeld of X n − 1. Let F be a ﬁeld of characteristic 0 or characteristic p not dividing n. For example. Φ6 (X) = . Gal(C/R) is ¯ generated by complex conjugation ι. It has degree ϕ(n). The nth roots of one form a ﬁnite subgroup of E × . We shall prove that the map is surjective for all n when F = Q. In particular. A primitive nth root of 1 in F is an element of order n in F × . The map σ → i : Gal(F [ζ]/F ) → (Z/nZ)× need not be surjective. The polynomial X n − 1 has some obvious factors in Q[X]. and the map σ → i mod n is injective because ζ generates E over F . The quotient of X n − 1 by all these factors for d < n is called the nth cyclotomic polynomial Φn . For example. and F [ζ] contains them all.7. we see that Xn − 1 = d|n Φd (X). Φ1 (X) = X − 1. On the other hand.31) that [Q[ζ] : Q] = p − 1. (b) If ζ is a primitive nth root of 1 in E. It obviously is a homomorphism (and is independent of the choice of ζ). Because X n − 1 has coeﬃcients in Z and is monic. Alternatively ask Maple by typing: with(numtheory). we saw in (1. it equals ζ i for some i relatively prime to n. 2 + X + 1) (X − 1)(X + 1)(X This gives an easy inductive method of computing the cyclotomic polynomials. then E = F [ζ]. and so the map is surjective. Since every nth root of 1 is a primitive dth root of 1 for exactly one d dividing n. the polynomials X d − 1 for any d|n. and the map Gal(E/F ) → (Z/nZ)× sending σ to [i] if σζ = ζ i is injective. cyclotomic(n. (c) The ﬁeld E is Galois over F . (b) The roots of X n − 1 are the powers of ζ. namely.FIELDS AND GALOIS THEORY 39 5. Any generator will have order n. and hence will be a primitive nth root of 1. because its derivative nX n−1 has only zero as a root (we use here the condition on the characteristic). and so E contains n distinct nth roots of 1. and so (see Exercise 3) they form a cyclic group. (a) The roots of X n − 1 are distinct.. (c) If ζ is one primitive nth root of 1. Cyclotomic extensions. when n = p is prime. Since σζ is again a primitive nth root of 1 for any automorphism σ of E. and ιζ = ζ = ζ n−1 . Φ2 (X) = X + 1. (a) There exists a primitive nth root of 1 in E. it is {±1} (n = 2)—because F [ζ] = C. the cyclotomic polynomials lie in Z[X]. if F = C.6)). Proof. then the remaining primitive nth roots of 1 are the elements ζ i with i relatively prime to n. Such an element can exist only if F has characteristic 0 or characteristic p not dividing n.X). any monic factor of it has coeﬃcients in Z (see (1. and if F = R.

8. MILNE Lemma 5. Write Φn (X) = f(X)g(X). Then ζ p is a root g(X). and so get a topological space. This proof is very old—in essence it goes back to Dedekind in 1857—but its general scheme has recently become very popular: take a statement in characteristic zero. which is true if and only if the injection Gal(F [ζ]/F ) → (Z/nZ)× is onto. Let F be a ﬁeld of characteristic 0 or p not dividing n. Remark 5. For example. and exploit the existence of the Frobenius automorphism a → ap to obtain a proof of the original statement..10. Xn) ∈ Z[X1 . but we saw in the proof of 5. Its roots will be primitive nth roots of 1.. Xn]. . ζ p is not a root of f. The following are equivalent: (a) the nth cyclotomic polynomial Φn is irreducible. the minimum polynomial of ζ divides Φn . Hence ¯ so gcd(f ¯ ¯ n X − 1 (regarded as an element of Fp [X]) has multiple roots.. g (X)p ) = 1. g(X p )) = 1 =⇒ gcd(f(X).. and we have to show they include all primitive nth roots of 1. Write h(X) → h(X) for the map Z[X] → Fp [X]. Contradiction. Again (1. . let f(X1 .7 that it doesn’t. Since f(X) and g(X p ) have a common root. ¯ p p But g (X ) = g (X) (use the mod p binomial theorem and that ap = a for all a ∈ Fp ). Proof. and look at the solutions of f = 0 in Fpn . (c) the homomorphism Gal(F [ζ]/F ) → (Z/nZ)× is an isomorphism. and there are theorems about complex manifolds5 that have only been proved by reducing things to characteristic p. their greatest common ¯ divisor (in Q[X]) is = 1 (see 3. There are some beautiful and mysterious relations between what happens in characteristic 0 and in characteristic p. Let f(X) be a monic irreducible factor of Φn in Q[X]. but that for some prime p not dividing n.. and let ζ be a primitive nth root of 1 in some extension ﬁeld. n) = 1. Suppose ζ is a root of f. Because ζ is a root of Φn . Proof. Theorem 5. and so it suﬃces to show that ζ a root of f(X) =⇒ ζ p a root of f(X) for all primes p n. and ¯ ¯ ¯(X). The nth cyclotomic polynomial Φn is irreducible in Q[X]. ¯ (ii) reduce mod p.S.. We can (i) look at the solutions of f = 0 in C. It is equal to it if and only if [F [ζ] : F ] = ϕ(n).40 J.6) implies that f(X) and g(X) lie in Z[X]. and note that ¯ gcd(f(X). Such an i is a product of primes not dividing n.1). g (X p )) = 1.9. commutative algebraists use this method to prove results about commutative rings. (b) the degree [F [ζ] : F ] = ϕ(n). which implies that ζ is a root of g(X p ). . 5 This is from my old notes—I no longer remember what I was thinking of. For this it suﬃces to show that ζ a root of f(X) =⇒ ζ i a root of f(X) for all i such that gcd(i. reduce modulo p (where the statement may no longer be true). which implies that f (X) and g (X) have a common factor. For example.

The limits of the present work exclude this demonstration here. The regular n-gon is constructible if and only if cos 2π (or ζ = e2πi/n ) is conn structible.4. Since χ2 (g) − χ1(g) = 0. Remark 5..12. we obtain the equation a1χ1(g)χ1 (x) + a2χ1(g)χ2(x) + · · · + am χ1 (g)χm(x) = 0. This leads some people to credit him with the above proof of the irreducibility of Φn . . a2χ2 + · · · + am χm = 0. We suppose a1χ1(x) + a2χ2(x) + · · · + am χm (x) = 0 for all x ∈ G. 5. The induction hypothesis now shows that ai = 0 for all i ≥ 2. On multiplying the ﬁrst equation by χ1 (g) and subtracting it from the second. and the induction hypothesis shows that all the remaining ai ’s are also zero.. .13 (Dedekind’s theorem on the independence of characters). Independence of characters.27 and 3. Proof.. Theorem 5. and let G be a group (monoid will do). . χm } of homomorphisms G → F × is linearly independent over F .. and that the regular n-gon is constructible when n is as in the Theorem. The regular n-gon is constructible if and only if n = 2k p1 · · · ps where the pi are distinct Fermat primes. . In it Gauss proves that the nth roots of 1 form a cyclic group. Assume it for m − 1. and before Galois). Proof. The ﬁnal section of Gauss’s. and this is a power of 2 if and only if n has the required form.. n= pn(p) .11. but we issue this warning lest anyone attempt to achieve geometric constructions for sections other than the ones suggested by our theory. it’s obvious. Then any ﬁnite set {χ1. Let F be a ﬁeld.” 6 . “Whenever n − 1 involves prime factors other than 2. we must have a2 = 0.. all x ∈ G. aiχi = 0 (as a function G → E) =⇒ a1 = 0. . we are always led to equations of higher degree. am = 0.WE CAN SHOW WITH ALL RIGOR THAT THESE HIGHER-DEGREE EQUATIONS CANNOT BE AVOIDED IN ANY WAY NOR CAN THEY BE REDUCED TO LOWER-DEGREE EQUATIONS. and show that this implies the ai to be zero. Since χ1 = χ2. On replacing x with gx in the equation. .and so spend his time uselessly. But (see Groups 3. . χ1(g) = χ2 (g) for some g ∈ G. we obtain the equation ai = ai (χi(g) − χ1 (g)). Disquisitiones Arithmeticae (1801) is titled “Equations deﬁning sections of a Circle”.10) ϕ(n) = p|n (p − 1)pn(p)−1 .FIELDS AND GALOIS THEORY 41 The Weil conjectures (Weil 1949. and so (according to 1. proved by Grothendieck and Deligne 1973) assert that the Betti numbers of the space in (i) control the cardinalities of the sets in (ii). . I’m sticking with Dedekind. Theorem 5. i.22) ζ is constructible if and only if [Q[ζ] : Q] is a power of 2.e. If m = 1. Induction on m. but in the absence of further evidence. that X n − 1 is solvable (this was before the theory of abelian groups had been developed. We know that Q[ζ] is Galois over Q. He also claimed to have proved the converse statement6.

. Let E be a Galois extension of F . Thus we can deﬁne {crossed homomorphisms} H 1 (G.15.S. then (E. MILNE Corollary 5.. with Galois group G. we obtain a crossed homomorphism by putting f(σ) = σx − x.16. Example 5. Then f(σ 2) = f(σ) + σf(σ). m. and let σ1. 5. (∗) Conversely. Note that the condition implies that f(1) = f(1 · 1) = f(1) + f(1).. then a crossed homomorphism is simply a homomorphism. Let F1 and F2 be ﬁelds. if x ∈ M satisﬁes (*). (b) (στ )(m) = σ(τ m) for all σ. A G-module is an abelian group M together with an action of G. then a crossed homomorphism f : G → M is determined by f(σ) = x. (c) 1m = m for all m ∈ M. (a) Consider a crossed homomorphism f : G → M. all σ ∈ G.. Example 5. until f(σ n ) = f(σ) + σf(σ) + · · · + σ n−1 f(σ). i. Thus.. Thus. and the sum of two principal crossed homomorphisms is again principal. M) = . σm = m for all σ ∈ G and m ∈ M. f(σ 3 ) = f(σ · σ 2 ) = f(σ) + σf(σ) + σ 2 f(σ) and so on. A crossed homomorphism is a map f : G → M such that f(στ ) = f(σ) + σf(τ ).5. +) and E × are G-modules. a map G × M → M such that (a) σ(m + m ) = σm + σm for all σ ∈ G. and x satisﬁes the equation x + σx + · · · + σ n−1 x = 0. then the formulas f(σ i ) = x + σx + · · · + σ i−1 x deﬁne a crossed homomorphism f : G → M. σm be distinct homomorphisms F1 → F2.14. σm are linearly independent over F2. and there are no nontrivial principal crossed homomorphisms. In this case we have a one-to-one correspondence {crossed homs f : G → M} f →f (σ) ↔ {x ∈ M satisfying (∗)}.42 J.e. Such a crossed homomorphism is called a principal crossed homomorphism. Apply the theorem to χi = σi |F1 . Hilbert’s Theorem 90. m ∈ M. (b) For any x ∈ M. × Proof.. and so f(1) = 0. to give an action of G on M is the same as to give a homomorphism G → Aut(M) (automorphisms of M as an abelian group). i. Let M be a G-module. (c) If G acts trivially on M. .. m ∈ M. The sum of two crossed homomorphisms is again a crossed homomorphism.e. and let σ ∈ G. if G is a cyclic group of order n generated by σ.. {principal crossed homomorphisms} . τ ∈ G. Let G be a ﬁnite group. Then σ1 . .

Dedekind’s theorem implies that f(τ )τ : E → E is not the zero map. but Emil Artin has been quoted as saying that it set number theory back thirty years—it wasn’t suﬃciently abstract for his taste. E × ) = 0. Many point to it as a book that made a fundamental contribution to mathematical progress. then α = β/σβ for some β ∈ E. Let π : X → X be the universal covering space of a topological space X. f(στ ) = f(σ) · σ(f(τ )). and we have to ﬁnd a γ ∈ E × such that f(σ) = σγ/γ for all σ ∈ G. for σ ∈ G. this becomes the isomorphism H 1 (X. all elements with norm 1 are of this form. then Nm(α) = 1. 1897. Let f be a crossed homomorphism G → E × . σβ = τ ∈G σ(f(τ )) · στ (α) = τ ∈G β σ(β) f(σ)−1 f(στ ) · στ (α) = f(σ)−1 τ ∈G f(στ )στ (α) = f(σ)−1 β.19 (Hilbert’s theorem 90). i. For example. Our next result show that. In multiplicative notation. We are interested in determining the kernel of this homomorphism. every crossed homomorphism G → E × is principal. the √ √ norm map C× → R× is α → |α|2 and the norm map Q[ d]× → Q× is a + b d → a2 − db2 . a Γ-module M will deﬁne a sheaf M on X. The map α → Nm α : E → F × is a homomorphism. σ.e. Example 5.. We deﬁne the norm of an element α ∈ E to be Nm α = σα. σ∈G Then. For example. M) have been deﬁned for all n ∈ N. when M = Z with the trivial action of Γ. The theorem is Satz 90 in Hilbert’s book. this means. Clearly if α is of the form τββ . Z) ≈ H 1 (Γ. τ (Nm α) = σ∈G × τ σα = Nm α. i. Let E be a Galois extension of F with group G.. and let Γ be the group of covering transformations. if NmE/F α = 1. Theorem 5. M). it will not be discussed in this course. Theorie der Algebraische Zahlk¨rper. τ ∈ G. Under some fairly general hypotheses. but since this was not done until the twentieth century.FIELDS AND GALOIS THEORY 43 The cohomology groups H n (G. Z) = Hom(Γ. M) ≈ H 1 (Γ.e. and H 1 (X. Let E be a Galois extension of F with Galois group G. for τ ∈ G. 7 . 7Let E be a ﬁnite cyclic extension of F with Galois group < σ >.17. there exists an α ∈ E such that β= τ ∈G f(τ )τ α = 0. Z). But then. which laid the o foundations for modern algebraic number theory. and so Nm α ∈ F . Because the f(τ ) are nonzero.18. for cyclic extensions. Corollary 5. Proof. then H 1 (G. which shows that f(σ) = and so we can take β = γ −1 .

if (i) a is not a pth power for any p dividing n. and let S(E) = {a ∈ F × | a becomes an nth power in E}. Theorem 16. and let σ generate Gal(E/F ). (Cf. i. Under the same assumption on F . Hence the splitting ﬁeld of X n − a is F [α]. there is an element β ∈ E such that σβ = ζβ. Remark 5. according to Hilbert’s Theorem 90. then there is an element β ∈ E such that E = F [β] and b =df β n ∈ F . On the other hand σβ n = ζ n β n = β n . Remark 5.21. [E : F ] = (S(E) : F ×n ). and when it is irreducible. Above we gave a description of all Galois extensions of F with Galois group cyclic of order n in the case that F contains a primitive nth root of 1. Then σ iβ = ζ i β. it is possible to give a description of all the Galois extensions of F with abelian Galois group of exponent n. σ) → a F In particular. VIII. Exercise 5 for the case n = 2. then the other roots are the elements of the form ζα with ζ an n th root of 1. then X p − X − a is irreducible in F [X] unless a = bp − b for some b ∈ F . The condition on α is that α · σα · · · σ m−1 α = 1. if E is cyclic of degree n over F . Then S(E) is a subgroup of F × containing F ×n . Let F be a ﬁeld containing a primitive nth root of 1.16a) there is a crossed homomorphism f :<σ>→ E × with f(σ) = α. and its Galois group is of order n.e. / See Lang. 5..) . We are now able to classify the cyclic extensions of degree n of a ﬁeld F in the case that F contains n nth roots of 1.22 (Kummer theory). Let E be such an extension of F . (a) If α is one root of X n − a. MILNE Proof. The theorem now shows that f is principal. Then Nm ζ = n ζ = 1. (b) If F has characteristic p (hence has no pth roots of 1 other than 1).S. and the map E → S(E) deﬁnes a oneto-one correspondence between abelian extensions of E of exponent n and groups S(E). there is a perfect pairing σa S(E) : ×n × Gal(E/F ) → µn (group of nth roots of 1). (a) The Galois group of X n − a is cyclic of order dividing n. its Galois group is cyclic of order p (generated by α → α + 1 where α is a root). hence E is the splitting ﬁeld of X n − b. Theorem 5. every extension of F which is cyclic of degree p is the splitting ﬁeld of such a polynomial.44 J. The ﬁeld E is recovered from S(E) as the splitting ﬁeld of (X n − a) (product over a set of representatives for S(E)/F ×n ).20. (a) Under the hypothesis of the theorem X n − a is irreducible. and so β n ∈ F. such that (S(E) : F ×n ) < ∞. Moreover. Cyclic extensions. which means that there is a β with f(σ) = β/σβ. §9. (b) Let ζ be a primitive nth root of 1 in F . (b) Conversely. Proof. Algebra. and so. F × ⊃ S(E) ⊃ F ×n .6. a quotient of (Z/nZ)r for some r. Let m = [E : F ]. The map σ → σα is α an injective homomorphism of Gal(F [α]/F ) into the cyclic group <ζ> . (a. and so (see 5. and so only the identity element of Gal(E/F [β]) ﬁxes β—we conclude by the Fundamental Theorem of Galois Theory that E = F [β]. Moreover. (ii) if 4|n then a ∈ −4k 4 .

The map σ → σ|E is an injection Gal(E /F ) → Gal(E/F ). . αm ]. . Proof of Galois’s solvability theorem. (f solvable =⇒ Gf solvable).24. . we see that if E = F [α1. Recall that a polynomial f(X) ∈ F [X] is said to be solvable if there is a tower of ﬁelds F = F0 ⊂ F1 ⊂ F2 ⊂ · · · ⊂ Fm such that (a) Fi = Fi−1 [αi].17). we need a lemma. . By (5. Theorem 5. σ ∈ G.. We are given that f splits in an extension Fm of F with the following property: Fm = F [α1. . and let . . The Galois closure E of E in Ω is the smallest subﬁeld of Ω containing E that is Galois over F . Let G = Gal(Ω/F ) and H = Gal(Ω/E). A polynomial f ∈ F [X] is solvable if and only if its Galois group Gf is solvable. αm] is a splitting ﬁeld of f over F . Let E be a splitting ﬁeld for f over F . .23. Let Ω be a Galois extension of F . and σHσ −1 corresponds to σE. Then E will be the subﬁeld of Ω corresponding to the largest normal subgroup of G contained in H (Galois correspondence 3. Proof. Then E = F [α1. Let g(X) be the minimum polynomial of γ over F . we need to make some observations. This means that there is a sequence of subgroups G = Gm ⊃ Gm−1 ⊃ · · · ⊃ G1 ⊃ G0 = {1} such that each Gi is normal in Gi+1 and Gi+1 /Gi is cyclic (even of prime order. Then the Galois group of f as an element of F [X] is a subgroup of that of f as an element of F [X]. Before proving the suﬃciency.1) i we know Fm = F [γ] for some γ. . . and hence is solvable. there exists an mi such that αmi ∈ F [α1. Let F = F [ζ] where ζ is a primitive nth root of 1 for some large n—for example. Before proving the necessity.7. Let f ∈ F [X] be separable. (Gf solvable =⇒ f solvable). Hence it suﬃces to ﬁnd a Galois extension E of F with Gal(E/F ) solvable and such that f(X) splits in E[X].18) E is the composite of the ﬁelds σE.FIELDS AND GALOIS THEORY 45 5.10). αm ] and. Let F be a ﬁeld of characteristic 0. Fi = Fi−1[αi ] with [F :F ] αi i i−1 ∈ Fi−1 . but this is σ∈G σHσ −1 (see Groups 4. The lemma shows that the Galois group G of f as an element of F [X] is a subgroup of Gf . We have a sequence of ﬁelds F ⊂ F [ζ] = F ⊂ F1 ⊂ F2 ⊂ · · · ⊂ Fm = E with Fi Galois over Fi−1 with cyclic Galois group. This shows that f is solvable. i (b) Fm splits f(X). According to (5.20b). αi−1 ]. but we don’t need this). Hence (see 3. and let Fi = E Gi . Any element of Gal(E /F ) permutes the αi and so maps E into itself. . n = (deg f)! will do. αm be the roots of f(X) in E . . . σ ∈ G. for all i. . Lemma 5. In particular. . Proof. . and let E be an extension of F contained in Ω. Let f ∈ F [X] have solvable Galois group. and let F be an extension ﬁeld of F . It suﬃces to show that Gf is a quotient of a solvable group. . Let E be a splitting ﬁeld of f(X) over F . where αmi ∈ Fi−1 for some mi . and let α1 .. Proof. . then E is generated over F by the elements σαi..

We can identify Fm with a subﬁeld of Ω. When we adjoin these elements one by one. we get a sequence of ﬁelds F ⊂ F [ζ] ⊂ F [ζ.. These particular polynomials are called the elementary symmetric polynomials. σ2 αm .e. We deﬁne the general polynomial of degree n to be f(X) = X n − t1X n−1 + · · · + (−1)n tn ∈ F [t1. its Galois group is Sn . i. b. Xn) in R[X1 . We shall show that.. . . For example p1 p2 p3 pr pn = = X1 + X2 + · · · + Xn . .e. the formula gives the roots of the particular equation.46 J.. It is therefore solvable. pn ]. MILNE Ω be a splitting ﬁeld of g(X)(X n − 1) for some suitably large n. .20a) and (5. if P (Xσ(1) .. Xn ) ∈ R[X1 . αm . Xσ(n) ) = P (X1 . c. . α2 . We shall prove in this section that there is no similar formula for the roots of the “general polynomial” of degree ≥ 5. . According to (5. .23 proves the above remark (at least on characteristic zero).Xir ··· = X1 X2 · · · Xn all σ ∈ Sn . . and so G has a normal series with cyclic quotients. . .. . . Every symmetric polynomial P (X1 .7). . . .. . Xn ] is said to be symmetric if it is unchanged when its variables are permuted. . . i Xi = = X1 X2 + X1 X3 + · · · + X1 Xn + X2 X3 + · · · + Xn−1 Xn . . .. = X1 X2 X3 + · · · i<j<k ··· = i1 <···<ir Xi1 .25 (Symmetric polynomials theorem)... . σ2 . . } be the Galois group of Ω/F and let ζ be a primitive nth root of 1 in Ω. tn ). . . Xn ). . A polynomial P (X1 .. According to the above remarks. each of these extensions is Galois with cyclic Galois group. .S.... . σ2α1 . . ... .8. Xn] is equal to a polynomial in the elementary symmetric polynomials with coeﬃcients in R. . .. 5. Symmetric polynomials. are all symmetric. E is generated over F by the elements ζ. σ3α1 ... The general polynomial of degree n. c as variables: for any particular values of a. . i. Theorem 5. α1. . Choose E to be the Galois closure of Fm[ζ] in Ω. . Let R be a commutative ring (with 1). . P ∈ R[p1 . Then Theorem 5. When we say that the roots of aX 2 + bX + c are √ −b ± b2 − 4ac 2 we are thinking of a. α1] ⊂ · · · ⊂ F ⊂ F ⊂ · · · ⊂ E such that each ﬁeld F is obtained from its predecessor F by adjoining an rth root of an element of F . . i<j Xi Xj = Xi Xj Xk . because pr is the sum of all monomials of degree r made up out of distinct Xi ’s. Let G = {σ1 = 1.. when we regard f as a polynomial in X with coeﬃcients in the ﬁeld F (t1. b. tn][X] where the ti are variables.

. pn ). is symmetric. and so the theorem implies that every symmetric polynomial in the roots of f(X) lies in R. . Both Hf and H are polynomials. the discriminant (αi − αj )2 D(f) = i<j of f lies in R. . Hence k1 ≥ k2 ≥ · · · ≥ kn ..12). . f(X) = (X − αi ) in S[X].. Therefore P (X1 .. pn) with Galois group Sn (acting by permuting the Xi ).. Xn) by permuting the Xi ’s. For example. . but is+1 > js+1 . an = ±pn (α1. . . g Proof. Xn )Sn . . and after a ﬁnite number of steps. Proof. . The highest monomial in pi is X1 · · · Xi . k k Let X1 1 · · · Xnn be the highest monomial occurring in P with a coeﬃcient c = 0. . . Then a1 = −p1 (α1 . . αn be the roots of f(X) in some ring S containing R. We have shown that F (p1. a2 = p2 (α1 . Thus the elementary symmetric polynomials in the roots of f(X) lie in R. . . . . . . . .. h ∈ F [X1. . Let f(X) = X n + a1X n−1 + · · · + an ∈ R[X]. Xn ) − cpk1 −k2 pk2 −k3 · · · pkn < P (X1 . and so this follows from (3. .. i. ﬁxed by all σ ∈ Sn . . H Corollary 5. For example. . and it follows that the highest monomial in pd1 · · · pdn is 1 n d d d X1 1 +d2 +···+dn X2 2 +···+dn · · · Xnn . αn ). The general polynomial.27... the ﬁeld of invariants is F (p1. Xn ). Because k k P is symmetric.26 (Symmetric functions theorem). .. .. pn . Hence their quotient f = Hf lies in F (p1 . Theorem 5. Then H = σ∈Sn σh is symmetric. . .. The Galois group of the general polynomial of degree n is Sn . . . i. . .e.FIELDS AND GALOIS THEORY 47 Proof.. and let α1 . .e. . . . . . pn ]. Xn ]. Xn ) is Galois over F (p1 . ... When Sn acts on E = F (X1 . . and so therefore is Hf. g. .28. 3 2 2 X1 X2 X3 > X1 X2 X3 > X1 X2 X3 . . it contains all monomials obtained from X1 1 · · · Xnn by permuting the X’s. we will arrive at a representation of P as a polynomial in p1 . . Theorem 5. . The ﬁeld F (X1. .. . pn ). . . αn ). . i1 = j1 . for some s. . pn ) = F (X1. . αn ). . . .. and therefore lie in F [p1. . . 1 2 n We can repeat this argument with the polynomial on the left. We deﬁne an ordering on the monomials in the Xi by requiring that j j i i i j X11 X22 · · · Xnn > X11 X22 · · · Xnn if either i1 + i2 + · · · + in > j1 + j2 + · · · + jn or equality holds and. . . . is = js . Suppose f = h .

pn ). pn ]... and Serre. . . 1992. .. .. . Hence the lower horizontal map is injective. . tn] are polynomial rings. (This is quite a diﬃcult theorem—see Serre.. and therefore induces an isomorphism on the ﬁelds of fractions F (t1. . . . Xn] and F [t1. . . Therefore the last corollary shows that the Galois group of g is Sn . . The diagram commutes because ti = pi (α1 . Swan proved in 1969 that the conjecture is false for C47. . . Xn ] ∪ F [p1. αn be the roots of f in E. tn) (for variables ti ). and every ﬁnite group occurs as a subgroup of some Sn . . . . the polynomial you obtain in Q[X] has Galois group G. . Consider the diagram E ⊃ F [α1. then F (X1 . . tn ][X] and prove that for inﬁnitely many values of the ti .. . . .. . We shall prove shortly that this is an isomorphism. Lectures on the Mordell-Weil Theorem. .29. . it is an isomorphism. .. . realize E as the splitting ﬁeld of a polynomial f(X) ∈ k[t1. . we’ll discuss algebraic independence.] . then it occurs inﬁnitely often as a Galois group over Q. . Xn) by permuting the Xi . . . which simpliﬁes the proof.. tn ] → F [p1. . . Xn )[X].. pn )[X]. and so F (X1. . . . .. . see Serre. αn ] ∪ F [t1.. but it doesn’t exclude other approaches. . . since it is obviously surjective. . ibid. Topics in Galois Theory. tn) (the ti are variables). pn ] is an isomorphism.48 J. Chapter 9. tn ] → F [p1. .. Since Sn occurs as a Galois group over Q. ←− ti →pi −→ F [X1 .. tn ] αi←Xi ti → pi . [For more information on the problem. . Remark 5. MILNE Proof. For the proof. The top and bottom maps are well-deﬁned because F [X1. . . . . .30. .. It remains to show that the homomorphism ti → pi is an isomorphism. f(X) corresponds to g(X) = X n − p1 X n−1 + · · · + (−1)n pn . ..S.. αn). and let α1 . . tn ) be a splitting ﬁeld of f. . . ti → pi . Let E ⊃ F (t1. ... . . Chapter 10. . Xn ) is the splitting ﬁeld of g(X) ∈ F (p1. which must also be the Galois group of f... . Under this isomorphism.. .. but does every ﬁnite Galois group occur as a Galois group over Q itself? The Hilbert-Noether program for proving this was the following. tn][X]. Consider the homomorphism F [t1. . Then it will be obvious that the map ti → pi : F [t1. tn) → F (p1. . . it follows that every ﬁnite group occurs as a Galois group over some ﬁnite extension of Q. .. Let f(X) be the general polynomial of degree n. . . . .) Noether conjectured the following: Let G ⊂ Sn act on F (X1 . pn ]. . Xn )G ≈ F (t1. f(X) = X n − t1X n−1 + · · · + (−1)n tn ∈ F [t1.. Hence this approach can not lead to a proof that all ﬁnite groups occur as Galois groups over Q. Hilbert proved that if G occurs as the Galois group of an extension E ⊃ Q(t1. Unfortunately. . and. . Remark 5. . But g(X) = (X − Xi ) in F (X1. In the ﬁnal section of this course. .

. until. Now let E be a ﬁnite extension of F of degree n. . det(αβ) = det(α) det(β). Nm is homomorphism (E × . The trace of a square matrix is the sum of its diagonal elements. 5. Norms and traces. Tr(aij ) = i aii . regarded as a polynomial in X. x → αx. ×) → (F × . As far back as 1500 BC. . a direct computation shows that Tr(AB) = Tr(BA). Since Tr(UAU −1 ) = Tr(A). . It is a beautiful complex manifold S of dimension n. . The discriminant of X n − T1X n−1 + · · · + (−1)n Tn . Ruﬃni and Abel proved that there are none. t1. there are exactly n points of S. is a polynomial D(f) ∈ C[T1.9. Similarly. (x. the Babylonians (at least) knew a general formula for the roots of a quadratic polynomial. Consider the projection π : S → Cn . +) → (F.31. Over the next 275 years there were many fruitless attempts to obtain similar formulas for higher degree polynomials. n = dim V . tn ). Ferrari (about 1545 AD) found a general formula for the roots of quartic polynomial (he introduced the resolvant cubic. Tr(α + β) = Tr(α) + Tr(β). then cα(X) = det(XIn − A). cα(X) = cαL (X). in about 1820. For α and β endomorphisms of a ﬁnite-dimensional F -vector space V . Cardan (about 1515 AD) found a general formula for the roots of a cubic polynomial. we can deﬁne the determinant and characteristic polynomial of α to be the determinant and characteristic polynomial of the matrix of α with respect to any basis of V . Take F = C. +). Tn ]. . an ) is the set of roots of the polynomial X n − a1X n−1 + · · · + (−1)n an . Let ∆ be the zero set of D(f) in Cn . Then over each point of Cn \ ∆. and S \ π −1(∆) is a covering space over Cn \ ∆ with group of covering transformations Sn . The characteristic polynomial of α can be deﬁned to be cα (X) = X n + c1 X n−1 + · · · + cn . Its ﬁbre over a point (a1 . . . Tr is a homomorphism (E. . . . c1 = − Tr(A) and cn = (−1)n det A. and regard E as an F -vector space. . and consider the subset of Cn+1 deﬁned by the equation X n − T1X n−1 + · · · + (−1)n Tn = 0. NmE/F (α) = det(αL ).FIELDS AND GALOIS THEORY 49 Remark 5. det(α) ∈ F . and used Cardan’s result). Deﬁne: TrE/F (α) = Tr(αL ). In a little more detail. ci = (−1)i Tr(α|Λi V ). If A is the matrix of α with respect to some basis for V . Then α ∈ E deﬁnes an F -linear map αL : E → E. in particular. A brief history. . we have Tr(α) ∈ F . ×). we can deﬁne the trace of an endomorphism α of a ﬁnite-dimensional vector space V to be the trace of the matrix of α with respect to any basis of V . . which shows that Tr(UAU −1 ) = Tr(A) and hence Tr(α) is well-deﬁned. . . tn ) → (t1 . .

. . Because F [α] is a ﬁeld. write V for E regarded as an F -vector space. NmE/F α = i=1 αi . and the injectivity of E → EndF -linear (E) then implies that cα (α) = 0. Then n n m Tr(α) = m i=1 αi . NmC/R(α) = |α|2 .8) is the same as the minimum polynomial of αL (in the sense of linear algebra). We shall see a quicker way of computing them presently. Note that αβi γk = aji βj γk . i] be the splitting ﬁeld of X 8 − 2. .. is f(X). Example 5. Therefore the matrix of αL in End(E) breaks up into n × n blocks with A’s down the diagonal and zero matrices elsewhere. .. n and so TrE/F (α) = −ma1 = m αi . and so b a TrC/R(α) = 2 (α). The endomorphism αL of V deﬁnes an action of F [X] on V (see Math 593). βn be a basis for F [α] over F . r = [E : F ]. (a) Consider the ﬁeld extension C ⊃ R.8). . and in fact. Corollary 5. (b) For α ∈ F .S.33.. we have to show that cα (X) = f(X). Tr(α) = rα. Therefore its characteristic polynomial is f(X)m . Hence the characteristic polynomial of α acting on V is the mth power of its characteristic polynomial acting on F [α]. Write αβi = according to the ﬁrst case proved. In this case. i is . γm be a basis for E over F [α]. (X − αi ). Proof. then A = (aij ) has characteristic polynomial f(X) for E over F . What are the norm and the trace of α? The deﬁnition requires us to compute a 16 × 16 matrix. {βiγk } is a basis aji βj . For the general case. As we saw in the proof of (1. Proposition 5. Write the minimum polynomial of α as f(X) = X n + a1 X n−1 + · · · + an = Then cα (X) = (f(X))m = X mn + ma1X mn−1 + · · · + am . Suppose ﬁrst that E = F [α]. Proof. and let γ1 .. V is a free F [α]-module. and let f(X) be the minimum polynomial of α ∈ E (in the sense of Section 1. and so the minimum polynomial of α (in the sense of Section 1.34.10). Suppose that the roots of the minimum polynomial of α are α1 . the matrix of αL ... and this action factors through F [X]/(f(X)) = F [α]. which. MILNE Note that α → αL is an injective ring F -homomorphism from E into the ring of endomorphisms of E as a vector space over F . they must be equal.50 J.32. Alternatively. Let β1. Consider a ﬁnite ﬁeld extension E/F . according to case already proved. and that [E : F [α]] = m. Then cα (X) = f(X)[E:F [α]]. α = a + bi. (c) Let E = Q[α. . Nm(α) = αr . But f(X)|cα (X) because cα (αL ) = 0 (Cayley-Hamilton theorem). we can be more explicit. V ≈ F [α]m with m = [E : F [α]] (count dimensions over F ). αn (in some splitting ﬁeld containing E). Since the polynomials are monic of the same degree. a −b relative to the basis 1. .

(b) Let E = Q[α. then cα (X) = f(X) = X 2 − 2 (α)X + |α|2 . When E = F [α]. .. i] be the splitting ﬁeld of X 8 − 2 (see Exercise 16). and let Ω be an algebraic closure of F . 2. Remark 5. Proposition 5.1b) that the σiα are the roots of the minimum polynomial f(X) of α over F . Assume E is separable over F . σ∈G NmE/F α = Proposition 5.38. TrE/Q α = 0. For example. Then TrE/F α = NmE/F α = σi α σi α. Let f(X) ∈ F [X] factor as f(X) = df ﬁeld. σr be the distinct embeddings of E into Ω. σrα are still roots of f(X) in Ω. . then this can be proved fairly easily using the descriptions in the above remark. Compute that disc f(X) = i<j df (αi − αj )2 ( i j=i = (−1)m(m−1)/2 · = (−1)m(m−1)/2 · (αi − αj )) f (αj ) = (−1)m(m−1)/2 NmF [α]/F (f (α)). . Then.. and so TrQ[α]/Q α = 0. we have TrE/M ◦ TrM/F = TrE/F . with f = dX . but now each root of f(X) occurs [E : F [α]] times (cf. Example 5. Proof.37.36. let σ1.. Proof. If α ∈ R. NmE/M ◦ NmM/F = NmE/F . In the general case..7). NmQ[α]/Q α = −2. If E is separable over F . NmE/Q α = 4. For ﬁnite extensions E ⊃ M ⊃ F . (a) Consider the extension C ⊃ R. then TrE/F α = σ∈G σα σα. the proof of 2. and let α = α1.FIELDS AND GALOIS THEORY 51 and NmE/F (α) = (−1)mn am = ( n αi )m . We omit the proof in the general case. we have m i=1 (X − αi ) in some splitting disc f(X) = (−1)m(m−1)/2 NmF [α]/F f (α). If α ∈ C \ R. The minimum √ polynomial of α = 8 2 is X 8 − 2. this follows from the observation (cf. then cα (X) = (X − a)2. σ1 α...35. if E is Galois over F with Galois group G.

α= −nb . Equivalently. 5. we obtain the equation nαn−1 = −na − nbα−1 . Let Gal(Ω/F ) = Aut(Ω/F ). γ + (n − 1)a from which it is clear that F [α] = F [γ]. i. [E : F ] < ∞. Hence Solving for α gives γ = nαn−1 + a = −(n − 1)a − nbα−1 . disc(X n + aX + b) = (−1)n(n−1)/2(nn bn−1 + (−1)n−1 (n − 1)n−1 an ). and consider the map σ → (σ|E) : Gal(Ω/F ) → Gal(E/F ) (product over the ﬁnite Galois extensions E of F contained in Ω). Similarly. Recall that we deﬁned a ﬁnite extension Ω of F to be Galois over F if it is normal and separable. We compute its norm. Let α be a root of f(X).52 J.39. We compute the discriminant of f(X) = X n + aX + b. and let γ = f (α) = nαn−1 + a. We give each ﬁnite group Gal(E/F ) the discrete topology and Gal(E/F ) the product topology.e. Therefore Nm γ is (−1)n times the constant term of this polynomial. and so we ﬁnd that Nm γ = nn bn−1 + (−1)n−1 (n − 1)n−1 an .S. This map is injective. MILNE Example 5. it must be the minimum polynomial of γ. and we give Gal(Ω/F ) the subspace topology. assumed to be irreducible and separable. Inﬁnite Galois extensions (sketch). Finally we obtain the formula. disc(X 5 + aX + b) = 55 b4 + 44 a5. if every irreducible polynomial f ∈ F [X] having a root in Ω has deg f distinct roots in Ω. we deﬁne an algebraic extension Ω of F to be Galois over F if it is normal and separable. . Thus the subgroups Gal(Ω/E). If we write P (X) −nb )= . which is something Maple doesn’t know (because it doesn’t understand symbols as exponents). a ﬁeld Ω ⊃ F is Galois over F if it is a union of subﬁelds E ﬁnite and Galois over F . a. b ∈ F. Since P (X) = (X + (n − 1)a)n − na(X + (n − 1)a)n−1 + (−1)n nn bn−1 is monic of degree n. For example. and so the minimum polynomial of γ over F has degree n also.10. because Ω is a union of ﬁnite Galois extensions.. On multiplying the equation αn + aα + b = 0 by nα−1 and rearranging. form a fundamental system of neighbourhoods of 1 in Gal(Ω/F ). f( X + (n − 1)a Q(X) then P (γ) = f(α) = 0.

43. Remark 5. Therefore (σ|Fpn )ai is independent of i provided it is suﬃciently large.. j > N. then Gal(Ω/Q) is one of the most basic. and the trivial Cauchy sequences form a subgroup. form a fundamental system of neighbourhoods of 0. To number theorists it will seem quite natural— the Chinese remainder theorem implies that it is isomorphic to p prime Zp where Zp is the ring of p-adic integers. and intractible. and we can deﬁne σ α ∈ Gal(Ω/Fp ) to be such that. σ α|Fpn = (σ|Fpn )ai for all i suﬃciently large (depending on n). G is the closure of <σ>. We do however understand Gal(F ab/F ) when F ⊂ C is a ﬁnite extension of Q and F ab is the union of all ﬁnite abelian extensions of F contained in C. ai ∈ Z. It has a ring structure.41. The remaining assertions in the Fundamental Theorem of Galois Theory carry over to the inﬁnite case provided that one requires the subgroups to be closed. We can deﬁne Z to be the quotient of the ﬁrst group by the second. satisfying the following condition: for all n ≥ 1. For example.42. Artin. The restriction of σ to Fpn has order n. Note that. i. and the map sending m ∈ Z to the constant sequence m. Let Ω be the algebraic closure of Q in C. objects in mathematics. Proof.. Gal(E/F ) is compact. The map α → σ α : Z → Gal(Ω/Fp ) is an isomorphism. m. and every.FIELDS AND GALOIS THEORY 53 By the Tychonoﬀ theorem. E. M → Gal(Ω/M) deﬁne a one-to-one correspondence between the closed subgroups of G and the intermediate ﬁelds M. A Cauchy seqence in Z is a sequence (ai )i≥1 . As for any topological group.) . Omit—it is not diﬃcult given the ﬁnite case. A ﬁeld M is of ﬁnite degree over F if and only if Gal(Ω/M) is open in Gal(Ω/F ). n ≥ 1.40.. Algebraic Numbers and Algebraic Functions. . i. Let Ω be Galois over F with Galois group G. Let Ω be an algebraic closure of a ﬁnite ﬁeld Fp . The Cauchy sequences form a commutative group. as far as we know. it is a little weird—its connected components are one-point sets. Call a Cauchy sequence in Z trivial if ai → 0 as i → ∞. .e. (This is abelian class ﬁeld theory—see Math 776. for each n. and it certainly has Sn as a quotient group for every n (and every sporadic simple group. Endow Z with the topology for which the groups nZ. and it is easy to see that the image of Gal(Ω/F ) is closed—hence it is compact and Hausdorﬀ. σ = (a → ap). Thus two integers are close if their diﬀerence is divisible by a large integer. The maps H → ΩH . Gal(Qab/Q) ≈ Z× . Example 5. and it is generated by it as a topological group. Let α ∈ Z be represented by the Cauchy sequence (ai). The group Z is uncountable. we can complete Z for this topology.e. it could have every ﬁnite group as a quotient. Example 5. Theorem 5. . See for example. there exists an N such that ai ≡ 0 mod n. identiﬁes Z with a subgroup of Z.. To most analysts. Then G = Gal(Ω/Fp ) contains a canonical Frobenius element.). p103. if for all n ≥ 1. there exists an N such that ai ≡ aj mod n for i. m.

. there exists an f(X1 . 1 n Note the similarity with linear independence. . . Xn) → F (α1.. αn ∈ F [A] such that α0 β n + · · · + αn = 0.28) the polynomial f(X) = X n − α1 X n−1 + . .. A set B is algebraically dependent on A if each element of B is algebraically dependent on A. if f is required to be homogeneous of degree 1. . . F (α1. This isomorphism then extends to the ﬁelds of fractions..αin = 0 =⇒ ai1 .. (a) A single element α is algebraically independent over F if and only if it is transcendental over F.in = 0 all i1 .... there exist α1. . . . αn of Ω are said to be algebraically dependent over F if there is a nonzero polynomial f(X1 . When these conditions hold. ... but this has not been proved. The theory in this section is logically very similar to a part of linear algebra.... there exist α0.. Let β ∈ Ω and let A ⊂ Ω. . Xn) → f(α1 .. .. . αn are algebraically independent over F . αn ∈ F (A) such that β n + α1 β n−1 + · · · + αn = 0... . . For example..in αi1 . . . .. we could have C ⊃ Q. . Otherwise. . αn). . To say that α1 .. Elements α1 .. we say that β is algebraically dependent on A (over F ). Xm . αn) = 0. Xn] such that f(α1 .. MILNE 6.54 J.. Thus they are algebraically independent if ai1 .. and hence an isomorphism. Xn] → F [α1. . . . . is the same as to say that the map f(X1 .... ... Y ) = 0 but f(a1 .... In fact. .. αn] is an injection.. am . am. ... am ∈ F such that f(a1. . . .1.. then the deﬁnition becomes that of linear independence. β) = 0..... Xn) ∈ F [X1 ..2. Xm .. . ai1 . αn) : F [X1 . (−1)n αn has Galois group Sn over F (α1.. in... . . Remark 6. αn) In this case. Transcendental Extensions In this section we consider ﬁelds Ω ⊃ F with Ω much bigger than F . . . αn) is called a pure transcendental extension of F .in ∈ F.. The following conditions are equivalent: (a) (b) (c) (d) β is algebraic over F (A). Then (see 5. Y ] and a1 .. ...S.. (b) The complex numbers π and e are almost certainly algebraically independent over Q.. Y ) ∈ F [X1 .. It is useful to keep the following correspondences in mind: Linear algebra Transcendence linearly independent algebraically independent A ⊂ span(B) A algebraically dependent on B basis transcendence basis dimension transcendence degree Example 6. An inﬁnite set A is algebraically independent if every ﬁnite subset of A is algebraically independent.. . . . . .. .. the elements are said to be algebraically independent over F . Xi → αi : F (X1 . .

.. then γ is algebraic over F (if a1. αm−1. Since αk+1 is algebraically dependent on {α1 . Proof. . .. f(α1 . Apply this with F (A ∪ B) for E and F (A) for F ..6.. .. . Because β is algebraically dependent on {α1 .. Xm . Xm−1. Assume (a) A is algebraically independent (over F ). .3 (Fundamental result)... αm−1. such that αk+1 is algebraically dependent on {α1 . Let k be the number of elements that A and B have in common..7..... . . Lemma 6.. ai0 (α1... βj } but not {α1 .. Proof. If k = m. αm }. . Let {α1 .. . . . A transcendence basis for Ω over F is an algebraically independent set A such that Ω is algebraic over F (A). The argument in the proof (2. Because β is not algebraically dependent on {α1 . αm−1. Lemma 6.. γ] has ﬁnite degree over F ).. αm. Suppose that k < m. . β) = 0. and m ≤ n... Xm... say ai0 .. . Since f(α1 . .. (of the theorem).. βn} but not on {α1 . αm−1}.. If C is algebraically dependent on B. . Y )Xm f(α1 . Write f(X1 .. then C is algebraically dependent on A. .. βk+1. Y ) = 0. αm} and B = {β1. . and observe that. . αn} be a subset of Ω. Eventually we’ll achieve k = m. and B is algebraically dependent on A.. there will be a βj . there exists a polynomial f(X1 . .... is not the zero polynomial. αk . then the ﬁeld F [a1.. βj−1}. ... β) = 0. αm . (b) A is algebraically dependent on B (over F ). . Y ) = 0... Definition 6. Therefore B is algebraically dependent on B1 ... at least one of the polynomials ai (α1. Let A = {α1 . and so A is algebraically dependent on B1 (by the last lemma). Proof.. βn}. then it is a transcendence basis for Ω over F . and write B = {α1 .. . repeat the argument with A and B1... If k + 1 < m. . Lemma 6.. αk.. ... β) = 0.. Proof. . Then m ≤ n..FIELDS AND GALOIS THEORY 55 Theorem 6.. then αm is algebraically dependent on {α1 . αk }. this shows that αm is algebraically dependent on {α1.. .. if β is algebraically dependent on {α1 . and certainly m ≤ n. αm.. .5 (Transitivity of algebraic dependence). . . . Y ) with coeﬃcients in F such that f(α1 . The exchange lemma then shows that βj is algebraically dependent on B1 =df B ∪ {αk+1 } − {βj }.. k + 1 ≤ j ≤ n. . . αm−1..4 (The exchange property). Y ). because f(α1 . αm−1}. β}. . ... . αm−1. .. βk+1 . . then A ⊂ B... αm.. . an . . βn} be two subsets of Ω. αk.. αk .. β}. and A is minimal among subsets of Ω with this property. αm} but not on {α1 ... We ﬁrst prove a lemma.. .. an are the coeﬃcients of the minimum polynomial of γ over E. We now prove the theorem... Xm .. .. βk+1.10) shows that if γ is algebraic over a ﬁeld E which is algebraic over a ﬁeld F . βk+1 . Therefore. β) is not the zero polynomial.. . Y ) = i i ai (X1 . If Ω is algebraic over F (A).....

. .. an) = 0. and so {p1 . some distinct a1. . Let p1 .. . p2 . Lemma 6.. .. The second statement follows from Theorem 6.9. . The cardinality of a transcendence basis for Ω over F is called the transcendence degree of Ω over F. A ∪ {β} is algebraically dependent. . . and it follows from (6. .8. Y does occur in f. an) ⊂ F (A).10. But the maximality implies that.. pn ). Theorem 6. . Proposition 6. Because A is algebraically independent..12.. . . Example 6. gi ∈ F [X1.. Xn ) has transcendence degree n.... pn } contains a transcendence basis for F (X1. but that A∪{β} is algebraically dependent.. Hence. . . . . Proof. Xn. MILNE Proof. . In fact.56 J. .. Y ) ∈ F [X1. . with equality holding if M is embeddable in some projective space. . Therefore f = g0 Y m + g1 Y m−1 + · · · + gm . ... Proof.. Xn . . and Ω is a pure transcendental extension of C ⇐⇒ M is isomorphic to the Riemann sphere (c) If M has complex dimension n. Xn. Because F (X1. . Xn]. .. then Ω has a ﬁnite transcendence basis over F . Xn ) has transcendence degree n over F . . . in this case. Because β is a root of f = g0 (a1.. . .. an. g0 = 0. .. any minimal subset A of A such that Ω is algebraic over F (A ) will be a transcendence basis. . Y ] such that f(a1. an). Then β is algebraic over F (A). Example 6. Ω is a pure transcendental extension of C of transcendence degree 1. . Moreover. . pn be the elementary symmetric polynomials in X1 . . . then one is algebraically dependent on the remainder.3. an ∈ A. The ﬁeld F (X1. Proof. it is algebraic over F (a1. . . (b) If M is a Riemann surface. the pi ’s must themselves be a transcendence basis.11.. Let Ω be the ﬁeld of meromorphic functions on a compact complex manifold M. Every maximal algebraically independent subset of Ω is a transcendence basis for Ω over F .. then the transcendence degree is ≤ n. . We have to prove that Ω is algebraic over F (A) if A is maximal among algebraically independent subsets. . As g0 = 0 and the ai are algebraically independent. .... the pure transcendental extension F (X1. If α1 . If there is a ﬁnite subset A ⊂ Ω such that Ω is algebraic over F (A). .5) that Ω will still be algebraic over F (A) after it has been dropped from A.. . then the transcendence degree of Ω over C is 1. For example. .. Suppose that A is algebraically independent. The hypothesis is that there exists a nonzero polynomial f(X1 . Xn ) is algebraic over F (p1 . an)X m + g1 (a1 . Xn ). and they all have the same number of elements. g0 (a1... . for every β ∈ Ω. Theorem 6. β) = 0.13 (*). (a) The only meromorphic functions on the Riemann sphere are the rational functions in z. every transcendence basis is ﬁnite. . . .. m ≥ 1. Every ﬁeld Ω containing F has a transcendence basis over F. . an)X m−1 + · · · + gm (a1.S. and so the lemma shows that β is algebraic over F (A). αm ∈ A are not algebraically independent.

15. On the other hand. We can partially order it by inclusion. If not. then it is easy to construct many: choose any transcendence basis A for C over Q. there are probably only two. without the axiom of choice. Finally. See. Y ) : E] < ∞ and F is algebraically closed of characteristic zero. i. then α deﬁnes an isomorphism Q(A) → Q(A) that can be extended to an automorphism of C. Remark 6. and it is known that the axiom of choice is required to construct nonmeasurable functions.30. But such a subset will be contained in one of the sets in T . the cardinality of Ω is the same as the cardinality of a transcendence basis over F . Any subﬁeld E of F (X) containing F but not equal u to F is a pure transcendental extension of F. Then the two ﬁelds in question are algebraic closures of the same ﬁeld. that B is algebraically independent. What are the automorphisms of C? If we assume the axiom of choice.) Theorem 6. Choose transcendence bases A and A for the two ﬁelds. the identity map and complex conjugation. Proposition 6. I claim that B ∈ S. .. (I have been told that any other is nonmeasurable. Remark 6.16). and choose a bijection ϕ : A → A . and choose any permutation α of A. The idea of the proof is as follows. The best that is true is that if [F (X. and hence are isomorphic (Theorem 2. Jacobson. which is a contradiction. Let F and F be the prime subﬁelds of Ω and Ω . Any two algebraically closed ﬁelds with the same uncountable cardinality and the same characteristic are isomorphic. and let B = ∪{A | A ∈ T }. there exists a ﬁnite subset B of B that is not algebraically independent. apply the proposition. Remark 6.17 (L¨roth’s theorem). Use this isomorphism to identify F (A) with F (A ). This fails when there is more than one variable—see the footnote on p38 and Noether’s conjecture 5. Any two algebraically closed ﬁelds with the same transcendence degree over F are F -isomorphic.14. Proof. we can identify F with F . then E is a pure transcendental extension of F (Theorem of Zariski.16. Let T be a totally ordered subset. Proof. p157. It is possible to show that any two (possibly inﬁnite) transcendence bases for Ω over F have the same cardinality.FIELDS AND GALOIS THEORY 57 Proof. Lectures in Abstract Algebra III.18. Then ϕ extends uniquely to an F -isomorphism ϕ : F [A] → F [A ]. Let S be the set of algebraically independent subsets of Ω. Then show that when Ω is uncountable. 1958).e. and hence to an isomorphism of the ﬁelds of fractions F (A) → F (A ). Now we can apply Zorn’s lemma to obtain a maximal algebraically independent subset A.

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue listening from where you left off, or restart the preview.

scribd