Ioan TOFAN

Aurelian Claudiu VOLF

Ring Arithmetic, Field Extensions and Applications

Table of Contents

Foreword..............................................................................................7 I. Arithmetic in integral domains ....................................................10 I.1 Divisibility...................................................................................10 I.2 Euclidian domains .......................................................................22 I.3 Euclidian rings of quadratic integers...........................................26 I.4 Principal ideal domains ...............................................................31 I.5 Unique factorization domains .....................................................36 I.6 Polynomial ring arithmetic..........................................................43 Exercises ...........................................................................................50 II. Modules.........................................................................................58 II.1 Modules, submodules, homomorphisms....................................58 Exercises ...........................................................................................72 II.2 Factor modules and the isomorphism theorems.........................74 Exercises ...........................................................................................80 II.3 Direct sums and products. Exact sequences ..............................81 Exercises .........................................................................................102 II.4 Free modules ............................................................................104 Exercises .........................................................................................116 III. Finitely generated modules over principal ideal domains ....118 III.1 The submodules of a free module...........................................118 Exercises .........................................................................................128 III.2 Finitely generated modules over a principal ideal domain.....130 Exercises .........................................................................................138 III.3 Indecomposable finitely generated modules ..........................140 Exercises .........................................................................................149

6

III.4 The endomorphisms of a finite dimensional vector space ..... 150 Exercises ........................................................................................ 167 IV. Field extensions ........................................................................ 169 IV.1. Algebraic extensions ............................................................. 169 Exercises ........................................................................................ 190 IV.2 Roots of polynomials. Algebraically closed fields ................ 193 Exercises ........................................................................................ 207 IV.3 Finite fields ............................................................................ 209 Exercises ........................................................................................ 221 IV.4 Transcendental extensions ..................................................... 223 Exercises ........................................................................................ 230 V. Galois Theory ............................................................................. 232 V.1 Automorphisms ....................................................................... 233 V.2 Normal extensions................................................................... 238 Exercises ........................................................................................ 242 V.3 Separability.............................................................................. 244 Exercises ........................................................................................ 259 V.4 The Fundamental Theorem of Galois Theory ......................... 262 Exercises ........................................................................................ 271 VI. Applications of Galois Theory ................................................ 273 VI.1 Ruler and compass constructions........................................... 273 VI.2 Trace and norm ...................................................................... 289 Exercises ........................................................................................ 296 VI.3 Cyclic extensions and Kummer extensions ........................... 297 Exercises ........................................................................................ 312 VI.4 Solvability by radicals............................................................ 313 VI.5 Discriminants, resultants........................................................ 320 Exercises ........................................................................................ 326 Appendices ...................................................................................... 331 1. Prime ideals and maximal ideals................................................ 331 2. Algebras. Polynomial and monoid algebras............................... 333 3. Symmetric polynomials ............................................................. 348 4. Rings and modules of fractions.................................................. 354 5. Categories, functors.................................................................... 362 6. Solvable groups.......................................................................... 373 Index ................................................................................................ 379 Bibliography.................................................................................... 387

Foreword
The book is aimed at undergraduate students in Mathematics, Computer Science or technical universities having a background of standard courses of Abstract Algebra and Linear Algebra as well as at general Mathematics readers with an interest in Algebra. The topics covered by the book are mandatory algebraic background of any mathematics graduate: arithmetic in integral domains, module theory basics, the structure of the finitely generated modules over a principal ideal domain with applications in abelian groups and Jordan forms of matrices, field extensions and Galois theory with applications. Throughout the book the reader is motivated by concrete applications and exercises. The book is reasonably self contained, in the sense that the reader is assumed to be familiar with general notions on algebraic structures (monoids, groups, rings, fields), factor rings and isomorphism theorems, vector spaces and bases, matrices, polynomials, permutation groups basics, elementary arithmetic of cardinals. Some topics less likely to appear in standard general Algebra courses are presendted in the Appendix. The language of Category Theory is used beginning with the chapter on Modules, the reader being invited to refer to the appendix (or to a book on Categories) when categorical notions appear in the text. We

7

Some notations used in the text: . if A is finite).| A | denotes the cardinal of the set A (the number of elements of A.1.! marks the end or the absence of a proof. Ruler and compass constructions. The chapters on Modules (II. 2.8 think that the Categories are particularly useful for a better understanding and for unifying many algebraic concepts and proofs. The section VI.N is the set of positive natural numbers. . is not a prerequisite for the other sections in chapter VI. III) can be read independently of the rest of the book. …} * . {0. 2.N is the set of natural numbers. …} Z is the set of integers .x := y means “x is equal by definition to y” (where y is already defined) or “we denote y with x" . 1. {1. .

9 .

Arithmetic in integral domains The set Z of integers. three classical and important classes of domains are studied (Euclidian domains. is the prototype for the familiar concept of ring. Field Extensions and Galois Theory. Such a generalization is interesting by itself and it also illuminates and yields nontrivial results on the divisibility in Z.1 Divisibility The classical definition for the relation of divisibility in the ring of integers Z generalizes easily to an arbitrary ring R: 10 . I. The material in this chapter is at the very basis of all Algebra. principal ideal domains and unique factorization domains). The definitions of these classes of rings originate in fundamental arithmetic properties of Z. endowed with the operations of addition and multiplication. and vital in Algebraic Number Theory. After a general study of the divisibility in integral domains.I. the integral domains. The classical divisibility theory in Z (the arithmetic of Z) can be extended with outstanding results to a large class of rings.

For instance. 1.2 Definition.: d ≠ 0. for short) if it has identity (denoted by 1). The fact that a | b can be also expressed by writing b ! a (read “b is divisible by a”). y ∈ F are nonzero. One says “a divides b in R” because the ring R plays an essential role here. in a ring without identity element. if d ∈ Z is squarefree (i. Other ways of reading a | b are: “a is a divisor of b” or “b is a multiple of a”.. xy = 0 implies x = 0 or y = 0.3 Examples. other difficulties arise if R is not commutative or if R has zero divisors. though (cf. a) Any field F (for instance Q. A ring R is called an integral domain (a domain. If the ring R lacks some natural properties. Write a . 0 denotes the zero element of the domain and 1 its identity element. the theory of divisibility in R can be very poor or very peculiar when compared to the classical theory in Z.1 Definition. This can also be said: for any nonzero x. . contradiction. 2 | 3 in Q. b) Every subring of a domain is itself a domain. it is commutative and has no zero divisors: for any x. then xy = 0 implies (by −1 multiplying with x .I. y ∈ R. In what follows. that y = 0. . which exists in F since F is a field and x ≠ 0). all rings we consider are domains. b ∈ R. This motivates the following definition: 1. Indeed. Let R be a ring and let a.. y ∈ R. we have xy ≠ 0. In particular.1 Divisibility 11 1. an element may not divide itself. For instance. but of course not in Z! We shall omit any reference to R in the notation a | b if the ring R is clear from the context. R. We say that a divides b in R (and write a | b) if there exists c ∈ R such that b = ac. All subrings considered are unitary (they contain the identity element of the ring). C.e.7). if x.) is a domain. every subring of a field is a domain.b if a does not divide b. So. The theory of divisibility in fields is trivial. 1.

Since R is a domain. Let R be a domain and let a. Then: a) For any a ∈ R.4 Proposition. Let R be a domain. c ∈ R such that a | b and a | c. One easily checks that " [ d ] consists of the complex numbers of the form a + b d . is a domain. simply. Notation: a ∼ b. c ∈ R. with c ≠ 0. b ∈ Z. If ac = bc. c ∈ R such that a | b and b | c. one can simplify the nonzero factors: 1. the subring of C generated by 1 and d . b. with Z as a model (and a particular case). The ring " [ −1 ] is called the ring of Gauss integers. . The elements a. is a domain. In a domain. b in R are called associated in divisibility (or. b) For any a.1 c) If R is a domain and n ∈ N*. a − b = 0 or c = 0. ! Properties a) and b) say that the divisibility relation on R is reflexive and transitive... famous German mathematician. b. d) For any x.12 I. 1. we have a | 0 and 1 | a. with a. c) For any a ∈ R. we have a | (bx + cy). We have ac = bc ⇔ ac − bc = 0 ⇔ (a − b)c = 0. But c ≠ 0. R[X1. 1 Carl Friedrich Gauss (1777 − 1855). then a = b.6 Definition. denoted by " [ d ] . we have a | c. Xn]. so a − b = 0. Arithmetic in integral domains d ≠ 1 and d is not divisible by the square of any integer greater than 1)..5 Proposition. a | a.. Proof. The proof of the following properties is an easy exercise: 1. y ∈ R and a. b. then the polynomial ring in n indeterminates with coefficients in R. associated) if a | b and b | a. ! We will develop a theory of divisibility in R.

b) For any a. U(K[X]) = { f ∈ K[X] | deg f = 0} = K* (we identify nonzero elements in K with the polynomials of degree 0). we have: a ∼ b ⇔ there exists u ∈ R such that a = bu. An invertible element u ∈ R is called a unit of R. The relation "∼" defined above is an equivalence relation on R (exercise!) and it is very important when studying the arithmetic of R: two elements associated in divisibility have exactly the same divisors and the same multiples. One can say they are indistinguishable as far as divisibility is concerned. d is called a proper divisor of a if d | a and d is neither invertible in R. b) If K is a field. b ∈ Z. we have: u ∈ U(R) ⇔ u ∼ 1 ⇔ u | a.8 Examples. knowing the group U(R) is a first step. where σ(α) is the conjugate of α. Let R be a domain.I. defined as : Proof. It is useful to define the “norm” N : R → Z. Let U(R) denote the set of all invertible elements of R: U(R) = {x ∈ R | (∃) y ∈ R such that xy = 1}. c) If d ∈ Z is squarefree. b ∈ Z. 1. in the study of divisibility in R. for any a. nor associated with a. a) U(Z) = {−1. then: U{Z[ d ]} = { a + b d | a. a − db = ±1} by N(α) = ασ(α). because u ∼ 1 (so u behaves just like 1 from the divisibility standpoint). 1. b ∈ R.1 Divisibility 13 For d. 2 2 σ (a + b d ) = a − b d . very important. Let R = Z[ d ]. . a ∈ R. ! For a given domain R. (∀) a ∈ R ⇔ uR = R.7 Proposition. 1}. U(R) is a group with respect to the ring multiplication (as a straightforward checking shows) and is called the group of units of R. Then : a) For any u ∈ R.

Arithmetic in integral domains So N (a + b d ) = a 2 − db2 ..... . we use the divisibility relation itself to order the common divisors. . . An easy computation N(α)N(β) = N(α)N(β).... If 1 is a GCD of a1 and a2. it follows that m | e.. an ∈ R. if it exists.. n ∈ N* and a1.. . [a1. . an) ii) For any e ∈ R such that e | a1. We call the element d ∈ R a greatest common divisor (abbreviated GCD) of the elements a1. an). We define the central concepts of greatest common divisor and least common multiple... . Assume R is a domain... We call m ∈ R a least common multiple (LCM for short) of a1. . . 1. an) or GCD(a1. b ∈ Z.9 Definition. This implies: if α... if it exists.. an) denotes a LCM of a1. .. then (a + b d )(a − b d ) = ±1 . Since an order relation like the one on N is not available.. β ∈ R.... b ∈ Z.. an..... . an if: i') a1 | m. e | an. .. an | m... (d is a common divisor of a1. Thus: 2 2 U{Z[ d ]} = { a + b d | a. . Let u = a + b d ∈ U(R).14 I. a − db = ±1} = = {α ∈ Z[ d ] | N(α) = ±1}.. then N(α) | N(β) in Z.. d | an. shows: .. . an. we call a1 and a2 coprime or mutually prime (or relatively prime).. . Similarly. if N(u) = ±1.. an | e. so N(u) = ±1.. β ∈ R with α | β in R. ii') For any e ∈ R such that a1 | e.. it follows that e | d (d is the “greatest” among the common divisors of a1.. We denote by (a1. ∀α..... Conversely. an] or LCM(a1. Then N(u) = a 2 − db 2 divides 1 in Z. . an) a GCD of a1.. ∀ a.. so ± (a − b d ) is the inverse of u.. an if: i) d | a1.

an] = m and exists [ra1.1 Divisibility 15 1.11 Proposition. c) Note that a1. What can you say about the LCM in these cases? If d ≠ 0 and d|a.. . If u is a unit. a) If there exists d = (a1..... Moreover. an ∈ R. .. then e ∼ d.. an). equal to 1... a) Given a1.. A domain R with the property that any two elements x.... then there exists GCD(a. then e is a GCD of a1. Let R be a domain and let a1. ..... an.. but this does not imply that 1 = −1 (of course. ran) =: e. then µ = rm. an) means that d is associated with a GCD of a1. .. . there exists GCD(a.. ran] =: µ... we use occasionally a subscript. For instance. u) = u. b) When the domain R in which we work is not clear from the context.I. an/d have a GCD. . it implies that 1 ∼ −1). .. . we can write 1 = (1.. like in the notation (a1.. a2 are coprime if and only if all their common divisors are units in R. .. c) If there exists [a1. This can lead to some oddities: in Z. For a given domain R and given x. b) If there exists (a1.. then d is uniquely determined up to association in divisibility: if e is also a GCD of a1. Writing d = (a1. . y) may not exist (see the Exercises for some examples). .. r ∈ R \{0}. then a1/d... an]. say d ∈ R. The same remark applies to the LCM. an. ... ran] = r[a1.. . Z is a GCD domain. an.... 0) = a. Thus: [ra1. d) For any domain R and any a ∈ R... a GCD(x. . then e = rd. Thus: (ra1. y ∈ R.....10 Remarks. if e ∼ d. y ∈ R possess a GCD is called a GCD domain. 2) = −1. if there exists a GCD for a1... an... .. 1. ran) = r(a1.. an) =: d and exists (ra1...... a/d denotes the unique element x ∈ R with a = dx. . an.. .. an)R. . an).

Let xi.. b ∈ R. an.. m | aixi . b ∈ R. rai = rdxi = rduyi. It is enough to prove that u | 1. n . . c) Because rm is a common multiple of rai. n . we also get mr | µ. b) = 1 and the preceding result. i = 1. yi ∈ R such that ai = dxi and rai = eyi. µ = raixi. ! d b The next result is ubiquitous in divisibility arguments. we get rd | e. d ≠ 0. d = eb. whose GCD is 1. If ∈ K. = anxn is a common multiple of a1. bi = xit. the definition of the GCD ensures that a | (ac. bi ∈ R. xn. for some xi. with c.. So. Let R be a GCD domain and a. . for some t ∈ R. Arithmetic in integral domains Proof. for some a. i = 1. Since a | ac and a | bc. i = 1. b ≠ 0 and (a. then let e = GCD(c. Also. i = 1. b) = 1 (use the statement a) c a above). ! Although the definitions of the GCD and the LCM are “dual” to each other. so rm = µt. (a. It follows that u is a common divisor of the elements xi. b. u | 1. Simplifying. n . with a. If e ∈ R is a common divisor of x1. Since µ = raixi.....16 I. c ∈ R such that a | bc and (a. Thus. n . bc) = c. Let u ∈ R with e = rdu.12 Corollary. n . a) Let xi ∈ R such that ai = dxi.. For any i = 1. n . i = 1. imply that (ac. . Let R be a GCD-domain and let K be the field of quotients of R. bc) = c.. we have µ | rm. n . Proof. so de | d. We can write m = aibi. by a). Then every element in K can be written as a quotient a . a1x1 = . Moreover. then de is a common divisor of a1.. b) Since rd | rai. b) = 1. with (a. Then a | c. = . This implies e | 1. 1. an.. ! 1.13 Corollary. We have rm = raibi = µt = raixit. i = 1. d ∈ R. Then d c = ea. d). b) = 1. n . i = 1. the situation is not entirely symmetric (the existence of the . part b). b c Proof.

We need only prove that d = (x. The element m = dxy is obviously a common multiple of a and b. y ∈ R with a = dx. Let R be a domain and x. y ∈ R. c) If any two elements in R have a GCD. Since . t ∈ R such that µ = az = dxz and µ = bt = dyt. any n elements a1. Then ers is a common multiple of x and y. Using the properties of the norm N (see 1. There exist z. b = dy. We have dm = xy = e 2 rs = tem . [0. Indeed. y](x. So. and xy ∼ [x. y] = m ∈ R. so e | d..1 Divisibility 17 LCM implies the existence of the GCD. then. Let d. [x. There exist x. (0. but no LCM. d = te. y) = y.). than a GCD of x. Let t ∈ R such that mt = ers. Similarly. in general).14 Proposition. y exists. a. for any n ∈ N. y).. then any two elements in R have a LCM. Proof. we GCD. m = yb. b ∈ Z) be a common 1. (Exercise!). Likewise. an in R have a GCD and a LCM. We have xy = xad. y] exists and it is 0. e | y and pick r. b ∈ R \ {0} and let d = (a. b). b) If any two elements in R have a GCD.c. The definition of the LCM implies m | xy. so m | ers.8. Simplifying by m.I. Suppose now that x and y are nonzero. (x. y exists. Let µ be another common multiple of a and b. but not conversely. c) Induction on n. µy) = µ(x. let d = a + b − 5 (a. s ∈ R such that x = er and y = es. So m divides µy = dxyz and µx = dxyt. Then: a) If a LCM of x. ! divisor of x and y. 1. This shows that m is a LCM of a and b. b) Let a. y). y). x = 1 + − 5 and y = 2 have a get N(d ) | N(x) = 6 and N(d ) | N(y) = 4 in Z. Take e ∈ R with e | x. N(d ) | 2 in Z. d | x. which means m divides also (µx.. so y = ad ⇒ d | y. b ∈ R with xy = md and m = xa. y) = µ.15 Example. In R = Z [ − 5 ] . a) When x = 0. . n > 1.

so N(µ) | 12. If R is a domain. . We proved that any common divisor of x and y is a unit. The element p ∈ R° is called irreducible (in R) if it has no proper divisors. which is impossible (the equation a 2 + 5b 2 = 12 has no solutions in Z).16 Definition. N(µ) divides N(6) = 36 and we get N(µ) = 12. let R° designate the set of nonzero and non-invertible elements in R: R° := R \ {0}\ U(R) In Z. Usually. the (elementary) definition for the notion of prime number is “the natural number p > 1 is prime if its only divisors in N are 1 and p”. a case-by-case inspection leads to the conclusion that a = ±1 and b = 0. so x and y have GCD equal to 1. The generalization to the case of a domain of this definition leads to the notion of irreducible element (also compare with the notion of prime element below). Suppose a LCM µ ∈ R of x and y exists. On the other hand. Then 6 | N(µ) and 4 | N(µ) in Z. The element p ∈ R° is called prime (in R) if. d | p ⇒ d ∼ 1 or d ∼ p. 1. Thus. for any a.18 I. any divisor of p is either a unit or is associated to p: ∀d ∈ R. Thus. In other words. Let R be a domain. so 12 | N(µ) in Z. p | ab ⇒ p | a or p | b. We emphasize that a prime element or an irreducible element is by definition nonzero and non-invertible. µ. Combining with 12 | N(µ). 6 = 2·3= (1 − − 5 )(1 + − 5 ) multiples of and 2(1 + − 5 ) are common multiples of x and y. Arithmetic in integral domains N(d ) = a 2 + 5b2 . prime numbers play a central role in divisibility questions. d is invertible. b ∈ R. so they are common N (2 )N (1 + − 5 ) = 24 in Z.

so p = ux. so d ∼ 1.1 Divisibility 19 A quick argument shows that. if p is prime and p divides a product of m factors in R. Let p ∈ R. If p . An examination of the possible cases shows that d is ±1 or ±2.I. Every prime element is also irreducible. but 2 divides neither factor. The same remark applies to the notion of irreducible element. Thus. On the other hand. 2 or 4. then N(d ) can only be 1. ! The notions of prime element and irreducible element (which coincide for Z. we cannot have d ∼ p (we would get p | x). But p | x means that p ∼ x (since x | p). irreducible and x. the notion of prime element depends heavily on the ring in which it is considered: 2 is prime in Z. In " [ −5 ] . which implies p | d (and we are finished) or p | x. p | xy and (p.19 Proposition. then p divides one of the factors. The GCD domains do not have the peculiarity described in the example above: 1. with u a unit. This approach allows extending classical results on the divisibility . So p | dx. if d is a divisor of 2. x) = 1. ! The notion of divisibility can be translated in the language of ideals. for any m ∈ N*.x. Thus. Then any irreducible element in R is prime in R. Indeed. Proof. 2 is irreducible and it is not prime. So.17 Proposition. but not in " [ −5 ] . If d ∈ R is a divisor of p.18 Example. Let p ∈ R be prime. 2 divides (1 − −5 )(1 + −5 ) = 6 . as we will see) are not the same in general. Proof. Indeed. then the GCD of p and x (which exists!) is 1.13 guarantees that p | y. there exists x ∈ R (nonzero) such that p = dx. 1. 1. ux = dx = p and thus u = d is a unit. y ∈ R such that p | xy. if d | x and d | p. Corollary 1. Let R be GCD domain.

+ Rxn = Ra. Arithmetic in integral domains in Z to much more general classes of rings (for instance..... For a counterexample. . n ∈ N* and a. j ∈ J}. the ideal generated by a is the set {ra | r ∈ R}. For a ∈ R. xn). .... x1. b) a ∼ b if and only if Ra = Rb. . + ) is a subgroup in the additive group (R... 1.. the primary decomposition theory).. Let R be a domain. Then: a) a | b if and only if Ra ⊇ Rb. f ) a is a common divisor of x1. .2 h) a is a common multiple of x1. xn if and only if Rx1 +. 2 . y ∈ I ⇒ x + y ∈ I. The ideal I is proper if I ≠ R. b) ∀x ∈ I. + ): ∀x.. denoted by Ra or aR and is called the principal ideal generated by a. b. The sum of two ideals I and J of R is the ideal: I + J := {i + j | i ∈ I. see the section Principal Ideal domains.. Write I ≤ R if I is an ideal in the ring R. then a = (x1. c) a ∈ U(R) if and only if Ra = R. Recall that a subset I of the commutative ring R is called an ideal of R if: a) (I.+ Rxn is included in Ra. g) If Rx1 +... we have Ra = Rx or Rx = R). The converse is false in general. xn if and only if Rx1 ∩… ∩ Rxn includes Ra. e) a is irreducible in R if and only if Ra is a maximal ideal among the principal proper ideals of R (more precisely: ∀ x ∈ R such that Ra ⊆ Rx. ∀r ∈ R ⇒ rx ∈ I.20 I.20 Proposition.. xn ∈ R. d) a is prime in R if and only if Ra is a prime ideal.

.. Let d ∈ R be another common divisor. rn ∈ R. Then Ra ⊆ Rd. …. then ∃ c ∈ R with ca = 1. so x ∈ Ra or y ∈ Ra ⇔ a | x or a | y. f ) If a is a common divisor of x1... If Rx is a proper principal ideal of R with Ra ⊆ Rx. then x | a.. Because a ∈ Rx1 +… + Rxn. then xy ∈ Ra. Suppose now Rx is maximal among principal proper ideals. . d) Let x. y ∈ R. b) Obvious.. . a is a common divisor of x1. then a | x or a | y. Conversely. one deduces other two fundamental properties of Z: Every ideal of Z is principal (of the form nZ. If Ra is a prime ideal and a | xy. So 1 ∈ Ra ⇒ Ra = R. Rx = Ra.e. ! An essential role in the arithmetic of Z is played by the theorem of division with remainder: For any a. x is associated with a or it is a unit. + rn xn . . ….. . b ≠ 0. Proof. From this theorem. ….. e) Suppose a is irreducible. so Rd = Ra or Rd = R. there exist q. so any element in the ideal Rx1 +. cn ∈ R with a = c1x1 + …+ cnxn. which shows that Ra is prime. g) From f ). ∃ c1. i) are left to the reader. If a is prime. we obtain d | a. So. The other implication is left to the reader. b ∈ Z. xn. Since d | x1. a) a | b ⇔ ∃ c ∈ R with b = ca ⇔ b ∈ Ra ⇔ Rb ⊆ Ra. and d ∈ R is a divisor of a. x ∼ a..I. h)...1 Divisibility 21 i) a = [x1. xn] if and only if Rx1 ∩ …∩ Rxn = Ra. This means d ∼ a or d ∼ 1. x ∈ Ra or y ∈ Ra. such that a = bq + r and | r | < | b | or r = 0. c) If a is invertible. i. But x cannot be a unit because Rx ≠ R. then 1 ∈ Ra. for some n ∈ Z).+ Rxn is divisible by a. d | xn. i. then a | r1 x1 + . Since a has no proper divisors. for any r1. by a). xn.. if Ra = R. We have xy ∈ Ra ⇔ a | xy. e. r ∈ Z.. so there exists c ∈ R such that 1 = ca.

respectively. This result is called „the fundamental theorem of integer arithmetic” or „the unique integer factorization theorem”. b ∈ R. I. We say in this case that R is an Euclidian domain with respect to ϕ. r ∈ R such that: (DRT) a = bq + r and (r = 0 or ϕ(r) < ϕ(b)). Of course. one obtains the notions of Euclidian domain (a domain in which a property analogous to the theorem of division with remainder holds). A domain R is called an Euclidian domain if there exists a mapping ϕ : R \ {0} → N satisfying: For any a. principal ideal domain (a domain whose every ideal is principal) and. unique factorization domain (a domain in which every nonzero non-invertible element can be written as a product of primes). b ≠ 0.1 Definition. The following sections are devoted to the study of these classes of rings. q is called traditionally the quotient and r is called the remainder of the division of a by b. the definition above originates in the division with remainder theorems in Z (where ϕ is the absolute value | | : Z \ {0} → N). there exist q.22 I. Abstracting these properties of Z.2 Euclidian domains 2. respectively in K[X]. The property (DRT) is called the “Division with remainder theorem in R”. with K a field (where ϕ is the . Arithmetic in integral domains Any nonzero non-invertible integer can be written as a finite product of prime integers and this writing is unique up to the order of factors and an association of the factors in divisibility.

the natural number ϕ(r) cannot be < 0. b. the remainder r of the division of 1 to b satisfies 1 = bq + r. the following condition on ϕ is also required: For any a.2 Euclidian domains 23 degree function deg : K[X] \ {0} → N). The Euclidian domains are GCD rings. a. Stop. b := r.4 Theorem (Euclidian Algorithm).9). a | b implies ϕ(a) ≤ ϕ(b). else (Step 1) Find q. Stop. r ∈ R with a = bq + r and (r = 0 or ϕ(r) < ϕ(b)). and ϕ(b) = 0. b ∈ R \ {0}. If b ∈ R. for some q ∈ R. This extra condition is not essential and in fact defines the same class of rings as the definition above (see 4. The converse is false in general (give a counterexample!). These rings are also the most important examples of Euclidian domains. else a := b. so r = 0. This follows from the fact that the Euclidian algorithm can be performed in these rings and guarantees the existence of GCD's: 2. go to Step 1 end . if r = 0 then d := b. b ∈ R. b ≠ 0. with b ≠ 0. Indeed. Sometimes. d) Input: a. b) ∈ R.I. Then a GCD of a and b exists and it can be found by the following algorithm: Algorithm Euclid (R. 2. Let R be an Euclidian domain and let a.3 Remark. then b ∈ U(R). 2. Output: d = GCD(a. begin if b = 0 then d := a. b ∈ R. in the definition of an Euclidian domain.2 Remark. Clearly. and (r = 0 or ϕ(r) < ϕ(b) = 0).

…) and merely uses the fact that these elements exist. (1). . r at step 1 does not imply a description of a concrete procedure (such procedures can be given for particular rings as Z. 3 Proof.. there exist (and can be algorithmically determined) u. These important issues are not discussed here. (n − 1) rn − 3 = rn − 2 qn − 1 + rn − 1 with rn − 1 = 0 or ϕ(rn − 1) < ϕ(rn − 2). The strictly decreasing sequence of natural numbers * ϕ(b) > ϕ(r1) > ϕ(r2) > . “finding” the elements q. Also. then e also divides r1 = a − bq1. The relation (n) shows rn − 1 | rn − 2. addition and multiplication algorithms in R etc.. The algorithm above implies the following sequence of divisions with remainder performed in R: (1) a = bq1 + r1 with r1 = 0 or ϕ(r1) < ϕ(b).. e divides We hope that the algorithm is clear to the reader. . Now let e ∈ R be a common divisor of a and b. (2). Arithmetic in integral domains Moreover. (2) b = r1q2 + r2 with r2 = 0 or ϕ(r2) < ϕ(r1). ri ∈ R with the properties above is guaranteed by the condition (1) in the definition of the Euclidian domain.. must terminate. Q[X].. We do not want to present a rigorous “pseudo-programming language” or use strict syntax rules from a particular language. this algorithm is intended to serve a theoretical purpose. (n) rn − 2 = rn − 1 qn + rn with rn = 0. Using relations (n − 2). (n − 2) rn − 4 = rn − 3qn − 2 + rn − 2 with rn − 2 = 0 or ϕ(rn − 2) < ϕ(rn − 2). we obtain (by induction) that rn − 1 | b and rn − 1 | a. Using (2)... 3 . (3). decision algorithms of the equality of two elements in R. for instance.24 I. implementing this algorithm must take into account computer representations of the elements of R. Moreover. The existence of the elements qi. We must prove that rn − 1 (the last nonzero remainder) is a GCD of a and b. The relation (n − 1) implies rn − 1 | rn − 3. so there exists n ∈ N with rn = 0 (which means that the algorithm terminates in a finite number of steps). v ∈ R such that d = au + bv. (3) r1 = r2q3 + r3 with r3 = 0 or ϕ(r3) < ϕ(r2).

v ∈ R such that d = au + bv. u := 1. q = 0. else a := b. u1 := u1 − qu. r = a satisfy these conditions. by induction exist q. Since a − b < a. Thus. We shall prove (by induction) that for every a ∈ N. By induction. r ∈ N satisfying a = bq + r and r = 0 or r < b. v with d = (a. If a < b. v := 0. a) Z is an Euclidian domain with respect to the absolute value. r ∈ N such that a − b = bq + r and r = 0 or r < b. Let b ∈ Z. Step 1. e | ri for any i < n. The following version of Euclid's algorithm (the extended Euclidian algorithm) outputs d. we note that r1 = a − bq1 is a linear combination of a and b. b := r. u. r ∈ N such that − a = bq + r and r = 0 or r < b. so e | rn − 1. Note that the variables u and v are such that the last remainder in the computation is always au + bv) : Input: a. Stop. Find q. there exist q. t := u. so there exist q. If a < 0. r2 can be written as au' + bv' and so on. v := v1. rn − 1 = (a. suppose that the claim is true for any n < a and we prove it for a. b > 0. u := u1. b) and d = au + bv. This implies a = b(q + 1) + r and the claim is proven. By induction.5 Examples. rn − 1 is of the form au + bv. We leave the case b < 0 to the reader. with r = 0 or |−r | < b. b ∈ R. go to Step 1 end ! 2. r ∈ R with a = bq + r and r = 0 or ϕ(r) < ϕ(b). In order to obtain d = rn − 1 written as au + bv. u1 := t. v1 := v1 − qv. Here is a proof.2 Euclidian domains 25 r2 = b − r1q2. If a ≥ b. replacing r1 in (2) with a − bq1.I. u := 0. b) ∈ R and u. v1 := 0. so a = b(−q) + (−r). . v1 := t. Output : d = GCD(a. begin if b = 0 then d := a. else u1 := 1. t := v. − a ∈ N. stop. if r = 0 then d := b. v := 1. b).

called the ring of integers modulo n. important examples of Euclidian domains are among the rings of the form Z[ d ] = {a + b d | a. h = gq + r. We have a ≡ b (mod n) ⇔ a − b ∈ nZ ⇔ a and b yield the same remainder when divided by n. We prove by induction on n = deg f. with g ≠ 0 (i. h −1 = f − bm an g has degree less than n (the degree n terms cancel out) and. Note that q and r are unique (prove!). there exist q. b ∈ Z}. I. set q = 0. we recall: 2.3 Euclidian rings of quadratic integers Besides Z and the polynomial ring in one indeterminate with coefficients in a field. with deg r < m. f = g q + bm −1an X n − m + r .6 Definition. r ∈ K[X]. We denote this fact by a ≡ b (mod n). If we require that the remainder should be positive. then q and r are unique. bm ≠ 0). If n ≥ m. r = f. Two integers a.e. the quotient q and remainder r of the division of a by b are not unique: for instance. The proof originates in the procedure of long n polynomial division taught in school. b ∈ Z are congruent modulo n if n | a − b. Let f = a0 +… + an X and m g = b0 + … + bm X ∈ K[X]. Thus.26 I. b) The ring K[X] of polynomials in one indeterminate X with coefficients in the field K is Euclidian with respect to deg : K[X] \ {0} → N. . b ∈ Z. Let n ∈ N be fixed. 3 = 2·1 + 1 = 2·2 + (−1). If n < m. Arithmetic in integral domains Given a. Recall that the relation defined here is an equivalence relation on Z and the set of equivalence classes is structured as a ring (which is in fact the factor ring Z/nZ). by induction. often denoted by Zn. ( ) For further reference.

3 Euclidian rings of quadratic integers 27 A somewhat surprising fact is that the ring Z[ (1 + d ) 2 ] = {a + b (1 + d ) 2 | a. A subfield of C that has dimension 2 viewed as a vector space over Q is called a quadratic number field. b ∈ Q}.5 An element of a quadratic number field that is integral over Z is called a quadratic integer.1. Define the norm N : Q[ d ] → Q and the trace Tr : Q[ d ] → Q. For a systematic treatment of the theory of algebraic integers. one can readily show that any quadratic number field is of the form Q[ d ] = {a + b d | a. some elementary facts on these rings are presented. but 1/2 is not. where d ∈ Z is squarefree. b ∈ Z} has sometimes better arithmetic properties than Z[ d ]. In the following. 2 is integral over Z. see for instance LANG [1964]. 3. One can prove that: a quadratic integer is a root of a monic polynomial in Z[X] of degree 2. Definition. If α = a + b d ∈ Q[ d ]. the element α = a − b d is called the conjugate of α. A complex number that is the root of a monic4 polynomial in Z[X] is called integral over Z (or algebraic integer). Definition. For instance. Sometimes. Fix d ∈ Z. 3.I. for avoiding confusions. The unproven statements below are proposed as exercises (some of them in the next chapters). an important topic in (algebraic) number theory.2. squarefree. This fact is closely connected to the theory of the rings of quadratic integers. 5 Because every algebraic integer over Z which is rational (in Q) must be in Z. Prove! 4 . the numbers in Z are called rational integers. N(α) := α α = a2 − db2 A polynomial is called monic if the coefficient of its monomial of maximum degree is 1. Using elementary tools of field extensions theory.

v ∈ Z having the same parity. with u. see the Exercises): 3. d if d ≡ 2 or 3 (mod 4) ! In what follows. ∀α ∈ Z[θ ].3 Proposition. as we saw above. The norm N is multiplicative and the trace Tr is additive: for any α. If d ≡ 1(mod 4). The quadratic integers in Q[ d ] form a ring. Arithmetic in integral domains Tr(α) := α + α = 2a for any α = a + b d ∈ Q[ d ] (a. ∀α ∈ Z[θ ]. Z[ 2 ]. According to the above. squarefree. Z [(1 + 5 ) 2] are examples of rings of quadratic integers. We obtain a mapping: |N| : Z[θ ] → N. We remark also that Q[ d ] = Q[θ ] = {a + bθ | a. Z [(1 + i 3 ) 2]. We have the following (for the proof. for a fixed d ∈ Z. A natural problem arises: . Z[i]. The ring of integers of Q[ d ] is Z[θ ]. This ring is traditionally called a ring of quadratic integers (imaginary if d < 0.28 I. respectively real if d > 0). θ denotes the number above. N(αβ) = N(α)N(β). called the ring of integers of Q[ d ] . |N|(α) = |N(α)|. Tr(α + β) = Tr(α) + Tr(β). b ∈ Q). then Z[θ ] = {a + b (1 + d ) 2 | a. b ∈ Z} and θ=⎨ ⎧ ⎩ (1 + d ) 2 if d ≡ 1 (mod 4). where Z[θ ] = {a + bθ | a. β ∈ Q[ d ]. b ∈ Z} can also be described as the set of complex numbers of the form (u + v d )/2. b ∈ Q}. for any x ∈ Q[ d ] : x is integral over Z (x is a quadratic integer) ⇔ Tr(x) ∈ Z and N(x) ∈ Z. One can prove that. The norm N : Q[ d ] → Q has the property that N(α) ∈ Z.

β ∈ R. Z[θ ] is a (dense) subset of R and no geometric interpretation is available. x = α/β = γ + δ/β. ∃γ. 3.I.3 Euclidian rings of quadratic integers 29 For which d ∈ Z. y ∈ Q[ d ] ). “⇒” Q[ d ] is the field of quotients of R. i. If d < 0. there exists γ ∈ R such that |N|(x − γ) < 1. the representation of the complex numbers in Z[θ ] in the plane yields a grid (rectangular if d ≡ 2 or 3 (mod 4) and oblique if d ≡ 1(mod 4)). since the norm N : Q[ d ] → Q is multiplicative (N(xy) = N(x)N(y).: ∀x ∈ [ d ] . squarefree. Since R is Euclidian with Q respect to |N|. ∃α. δ ∈ R such that α = βγ + δ. Let d ∈ Z. |N|(x − y) is the Euclidian distance between the points x and y. Then R := Z[θ ] is Euclidian with respect to |N| if and only if for any x ∈ Q[ d ] . with |N|(x − γ) = |N|(δ /β) = |N|(δ )/(β) < 1. . |N| : Z[θ ] → N is also multiplicative. for any x. So. Proof. d ≡ 2 or 3 (mod 4) d ≡ 1 (mod 4) If d > 0. y ∈ Q[ d ] . squarefree. Z[θ ] is Euclidian with respect to |N|? We remark first that. with |N|(δ ) < |N|(β) or δ = 0. β ≠ 0 such that x = α/β.4 Lemma. ∀x. moreover.e.

This result was obtained in 1923 by L. Z[(1 + d ) 2] is Euclidian with respect to |N| if and only if d ∈ {−3. ! If d < 0. We must have then ( − d − 3 ) 2 < 1 ⇔ d > −7 − 4 3 ⇔ d ∈ {−3. there exists x ∈ Q[ d ] such that |z − x| < ε. α = βγ + βδ . d = −1 or d = −2. ! 6 7 For any z ∈ C and any ε ∈ R. Arithmetic in integral domains “⇐” Let α. because Q[ d ] is dense 6 in C. We have 1 − d 2 < 1 if and only if d > −3. Proof. −7.5 Proposition. the lemma can be rephrased as follows: The ring Z[θ ] is Euclidian with respect to |N| if and only if any point in the complex plane is situated at a distance smaller than 1 to some point in the grid Z[θ ]. a) If d is congruent with 2 or 3 (mod 4). i. For x = α/β ∈ Q[ d ] . −11}. Thus. 3. The greatest distance to the vertices is attained at the intersection of the diagonals.7 Let d ∈ Z. the grid forms isosceles triangles with base 1 and height − d 2 . b) If d ≡ 1 (mod 4). . Let δ = x − γ ∈ Q[ d ] . An easy geometric reasoning shows that the distance from P to the third vertex is ( − d − 3 ) 2 . b) In this case. at the distance of 1 − d 2 to any vertex. The points inside the triangle have distance less than 1 to some vertex if and only if the circles of radius 1 centered in base vertices intersect in a point P situated at a distance less than 1 to the third vertex. with βδ = α − βγ ∈ R and |N|(βδ ) = |N|(β)·|N|(δ ) < 1. ∃γ ∈ R with |N|(x − γ) < 1. −7.30 I.e. E. β ∈ R. Dickson. a) Z[ d ] is Euclidian with respect to |N| if and only if all points inside a rectangle of the grid are situated at a distance less than 1 from some vertex of the rectangle. ε > 0. d < 0. squarefree. Z[ d ] is Euclidian with respect to |N| if and only if d = −1 or d = −2. −11}.

4 Principal ideal domains 4. Proof. Any Euclidian domain is a principal ideal domain. b ∈ I. so r ∈ I ! and ϕ(r) < ϕ(a) contradict the choice of a. But a. a ≠ 0. 4. such that ϕ(a) = min{ϕ(x) | x ∈ I. The real case d > 0 has no geometric interpretation and is considerably more difficult. We claim that a is a generator of the ideal I. we obtain elements q. I. Any field is a PID 8 . Obviously. " [i 2 ] .1 Definition. x ≠ 0} contains a minimum element: ∃a ∈ I. x ≠ 0} (a may not be unique). there exists a ∈ R such that I = Ra. The most important examples of PIDs are given by the following result. " ⎡(1 + i 3 ) 2 ⎤ .4 Principal ideal domains 31 We obtain that the following rings are Euclidian with respect to |N|: Z[i]. The set of natural numbers {ϕ(x) | x ∈ I. .2 Theorem. In order to prove the opposite inclusion. Ra ⊆ I. 8 Sometimes the fields are not considered principal ideal domains by definition. " ⎡(1 + i 7 ) 2 ⎤ . suppose by contradiction that there exists b ∈ I \ Ra. for every ideal I of R.I. Let R be Euclidian with respect to ϕ and let I be a nonzero ideal of R. Applying the division with remainder in R. In other words. A domain R is called a principal ideal domain (PID) if any ideal of R is principal. " ⎡(1 + i 11 ) 2 ⎤ . r ≠ 0 (since b ∉ Ra) and ϕ(r) < ϕ(a). r ∈ R such that b = aq + r. ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ One can prove that these are all the imaginary Euclidian rings of quadratic integers (not necessarily with respect to |N|).

Thus. Thus. if K is a field. e | b.32 I. d divides a and b. Proof. a generator of I is a polynomial g ∈ I of minimum degree among the degrees of nonzero polynomials in I. d | b. there exist u. 4. any generator d of aR + bR is a GCD of a and b. namely any generator of the ideal aR + bR: 4. e | d. Since d ∈ aR + bR. The proposition above shows that this notation is consistent to the fact that it designates also the GCD of a and b (which is a generator of the ideal aR + bR). dR = eR = aR + bR.4 Example. But e is also a common divisor of a and b. so dR ⊇ aR and dR ⊇ bR. aR + bR. Then the ideal (r. there exists their GCD. b) The element m ∈ R is a LCM of a and b if and only if mR = aR ∩ bR. the element d ∈ R is a GCD of a and b if and only if dR = aR + bR. We have a. If e ∈ R is such that e | a. for a. so d is associated to e. In particular. X) in R[X] is not princi- . Let R be a domain that is not a field and take r ∈ R nonzero and non invertible. for some u. for any a. so d | a. y ∈ R. a) Since R is a PID. d = au + bv . Moreover. v ∈ R. if d is a GCD of a and b.3 Proposition. Then: a) A GCD d of a and b exists and. Arithmetic in integral domains For instance. y ∈ R} is generated by d. ∀x. dR ⊇ aR + bR. b ∈ R. v ∈ R such that d = au + bv. this means d | e. Recall that. Conversely. b ∈ dR. K[X] is a PID. Let R be a PID and let a. The principal ideal domains are GCD-domains. b ∈ R. Thus. b) The proof is similar to the one above and is proposed as an exer! cise. b ∈ R. then e | ax + by. If e is a generator of aR + bR. if I ≠ 0 is an ideal in K[X]. b) is used to designate the ideal generated by a and b. ∃d ∈ R such that the ideal aR + bR = {ax + by | x. the notation (a.

Recall that R° = {x ∈ R | x is nonzero. so f is invertible in R. are not PID's. there exists g ∈ R[X] with X = fg. K[X. It is sufficient to remark that any nonzero prime ideal is principal.5 Proposition. Y]. Proof. so f ∈ R. it follows that 1 = h(0)·r. In a PID R. the GCD of r and X is 1. Since f | X. We obtain deg f = 0.19 and the fact that any PID is a GCD-domain. 4.20. where p is irreducible in R. the rings Z[X]. suppose that for some f ∈ R[X].I.4 Principal ideal domains 33 pal. so f = 1. The element p is irreducible. Any maximal ideal is of the form pR. we obtain: 4. q ∈ R[X] are such that 1 = hr + qX. Then f | r. we have ( f ) = (r. setting X = 0 in this equality of polynomials. But the ideal generated by r and X does not contain 1: if h. In a PID.6 Corollary. p ∈ R is irreducible if and only if pR is a maximal ideal. Indeed. Moreover. The other state! ments are obvious. generated by a prime p (Prop. which means r is invertible. the prime nonzero ideals are maximal ideals. Using Proposition 1.20.e)) the ideal pR is maximal. so (Prop. . In particular.18 shows. Thus. 1.d)). with K a field. The case R = Z of the following proposition is known as the “the fundamental theorem of integer arithmetic”. as Example 1. 1. the domains containing irreducibles that are not primes are not principal ideal domains. such a ring is Z[ − 5 ]. an element is irreducible if and only if it ! is a prime element. so R[X] is not a PID. Thus. a contradiction. X). non-invertible}.

∀i ∈ N. then r0 is too. hence x + y ∈ It ⊆ I. j ∈ N. s1 ∈ R°. 4. j). Since a ∈ I. But R is a PID. x ∈ Ii.e. y ∈ It.7 Theorem. So x. A ring R satisfying the ascending chain condition (ACC): every ascending sequence of ideals of R. there exists a sequence (rn)n ≥0 of elements in R. Since R is a PID.. for any n ∈ N. not associated with r0. This reasoning applies to r1. and r2 is not a product of irreducibles. we obtain a strictly increasing sequence of ideals r0R ⊂ r1R ⊂ … ⊂ rnR ⊂ rn+1R ⊂ …. there exists m ∈ N such that a ∈ Im. then. I0 ⊆ I1 ⊆ …. at least one of them (say r1) is not a product of irreducibles. Arithmetic in integral domains 4. so aR ⊆ Im ⊆ ! I = aR. where t = max(i. We obtain thus r1∈ R° with r1 | r. false. the element itself is a prime). I is an ideal in R: if x. r2 | r1. rx ∈ Ii ⊆ I.34 I. r0 is reducible. so there exists a ∈ R such that I = aR. rn + 1 is a proper divisor of rn. with r1. so r0 = r1s1. (In other words. Let R be a PID and let (In)n ≥ 0 be a sequence of ideals in R such that In ⊆ In + 1. So aR = I = Im + i. y ∈ I. for some i. Suppose by contradiction that there exists r0 ∈ R° such that r0 cannot be written as a finite product of irreducibles. r2 ¿ r1. Since the sequence (In)n ≥ 0 is ascending. In particular.9 Proof.8 Lemma. If r ∈ R. So. . y ∈ Ij. Proof. By induction. as the following lemma shows. so we get r2 ∈ R°. with the property that for any n ∈ N. If r1 and s1 are finite products of irreducibles. for any i ∈ N. is stationary (there ex- 9 The products may contain a single factor (i. Then there exists m ∈ N such that Im = Im + i . every ascending sequence of ideals is stationary). Then every nonzero non-invertible element r in R is a finite product of prime elements. r1 ¿ r and r1 is not a product of irreducibles. Let R be a PID. This is impossible in a PID. the primes in R coincide with the irreducibles in R. Let I be the union of all ideals In. In other words. n ∈ N.

There exist q. but the examples are not easy to find.I. x ≠ 0}. b ∈ R* with a | b. for some unit u. ∀i ∈ N) is called a Noetherian ring. b ∈ R . Since b0 = bu. Note that the proof above can be slightly modified to obtain the theorem: A ring in which every ideal is finitely generated is Noetherian. defined by ψ(x) = min{ϕ(y) | y ∼ x}. a generator of the ideal aR is an element c (which must be associated with a) such that ψ(c) = min{ψ(x) | x ∈ aR. for any a. Let R be a Euclidian domain with respect to ϕ : R*→ N.9 Proposition. r ∈ R such that a = b0q + r and r = 0 or ϕ(r) < ϕ(b0). b ≠ 0 and let b0 be associated with b such that ψ(b) = ϕ(b0). . as shown by Dedekind in 1894. b ∈ R.10 We have shown that every PID is Noetherian. Clearly. Now let a. example is Z ⎡ ⎢ 2 ⎥ ⎦ ⎣ 10 Emmy Noether (1882–1935). since b ∈ aR. 4. We have remarked that sometimes in the definition of the Euclidian domain R with respect to ϕ : R \ {0} → N. ! Are there any principal ideal domains that are not Euclidian? The answer is affirmative. x ∼ y implies ψ(x) = ψ(y). Noetherian rings are an important topic in Commutative Algebra. We claim that R is Euclidian with respect to ψ.4 Principal ideal domains 35 ists m ∈ N such that Im = Im + i. As seen in the proof of 4.2. Let ψ : R* → N. Such an 1 + i 19 ⎤ . German. So ψ(a) =ψ(c) ≤ ψ(b). Proof. b ∈ R*. a | b implies ψ(a) ≤ ψ(b). an extra condition is re* quired: for any a. Let a. a | b implies ϕ(a) ≤ ϕ(b). We show that this condition is not essential. we have a = b(uq) + r and r = 0 or ψ(r) ≤ ϕ(r) < ϕ(b0) = ψ(b). Then there exists ψ : R*→ N such that R is Euclidian with respect to ψ and. ∀x ∈ R. The converse is also true.

! otherwise p would not be irreducible. because it has no nonzero and non-invertible elements. If r has a prime decomposition. n} such that pi ∼ qσ(i). 11 . Such a product is also called a prime decomposition of the element. But this product can have only one factor.7 shows that any principal ideal domain (thus also any Euclidian domain) is a UFD. Every field is a UFD. So. p is itself a prime. 5. Define l(r) as the * smallest n ∈ N such that there exists a prime decomposition of r of length n.5 Unique factorization domains 5.2 Proposition. ….3 Proposition. then this decomposition is unique up to an ordering of the factors and an association in divisibility.36 I. Arithmetic in integral domains I. This means that: if r = p1…pn = q1…qm are two prime decompositions of r. Every irreducible element in a UFD is also a prime.1 Definition. n}. The next Proposition justifies and explains the epithet “unique” in the name of a UFD. …. Products may have a single factor. Let R be a UFD and let p be irreducible in R. then m = n and there exists a permutation σ of the set {1. Theorem 4. Since p ∈ R°. Call the natural number n the length of the decomposition. p is a product of primes. 5. ∀i ∈ {1. Proof. Let R be a domain and r ∈ R°. Proof. A domain R with the property that every nonzero and non-invertible element is a finite product11 of prime elements is called a unique factorization domain (UFD). Let r = p1 … pn be a prime decomposition of r.

But q1 is irreducible. Simplifying by pn. 5. that is. Suppose the claim is true for any x ∈ R° with l(x) < n. simplifying by q1 in q1u = q1 … qm. qm primes. Let R be a domain. 5. pn − 1 are associated with ! q1. …. b) Every element in R° is a finite product of irreducibles and every irreducible is a prime.I. Thus r = q1u. so we have r ∼ q1. qm are invertible. …. The induction hypothesis applied to p1 … pn−1 shows that n − 1 = m − 1 and p1. contradiction. Since r is a prime and r divides the product q1 … qm. Because pn is prime. …. If l(r) = 1. c) Every element in R° has a decomposition in irreducible factors. with p1.5 Unique factorization domains 37 We prove the claim by induction on l(r). 5. . Let r ∈ R. But qi is irreducible. vpn = qi. This is an important property of the UFD's. Several other characterizations are collected in the next theorem. r divides one of the factors. a)⇒b) It follows from Prop. so pn ∼ qi. Proof. …. then let r = p1 = q1 … qm. d)Every element in R° has a decomposition in irreducible factors and every two elements have a GCD. qi + 1. with u invertible. …. unique up to a reordering of the factors and an association in divisibility. q1. A UFD is also a GCD-domain. qi − 1. …. The following statements are equivalent: a) R is a UFD. with p1. pn.4 Theorem.2. we obtain the relation p1 … pn−1 = vq1 … qi−1qi+1 … qm. qm prime. b)⇒c) It follows from Prop. …. with v a unit. possibly in other order.3. q1. qm. If m ≥ 2. say q1 (relabel if necessary). n} such that pn | qi. ∃i ∈ {1. We must prove that m = 1. we obtain q2 … qm = u. r = p1 … pn = q1 … qm. so q2. ….

. and 2 is not associated with 1 + − 5 or with 1 − − 5 . A GCD of a and b can be determined by “taking the common prime factors. n . with the notations in the proof above. at the smallest power”. …. ti) and define d = p11 … pnn . d | b. pn}. contradicting the uniqueness property. This implies c ∈ {p1. one containing c and the other one not. The reader can also check that 6 has two decompositions as a product of irreducibles in Z [ − 5 ] : 6 = 2·3 = (1 + − 5 )(1 − − 5 ) . b ∈ {0} ∪ U(R). sn. then every irreducible c ∈ P that divides e divides also a and b. Z [ − 5 ] is not a UFD. So w w e = p1 1 … pn n q .38 I. and e | b implies wi ≤ ti. u. We have d | a. at the greatest exponent”. tn ∈ N. let P be a system of representatives of the equivalence classes of the irreducible elements in R (with respect to the association in divisibility). their LCM can be determined by taking “all prime factors in the decompositions.18). 1. The implication is now obvious. 1. s1.5 Example. b ∈ R° (if a. 5. t1. pn ∈ P. Property c) assures that there exist and are unique distinct p1. wn ∈ N. since ! R is a GCD domain. since 2 is irreducible and not a prime (cf. …. This means that every irreducible in R is associated to exactly one element in P. The uniqueness claim follows the uniqueness assertion in c). e | b. Ex. If e | a. Let ri r r = min(si. ti). Indeed. define q q qi = max(si. it is trivial to exhibit a GCD of a and b). the element m = p1 1 … pn n is then a LCM of a and b. with w1. q ∈ U(R). …. i = 1. otherwise a (or b) would possess two decompositions in irreducible factors. From e | a it follows that wi ≤ si. d)⇒a) Prop. Arithmetic in integral domains c)⇒d) Let a. Note that in a UFD any two elements a and b have a LCM. To be precise. …. v ∈ U(R) such that s t s t a = p11 … pnn u and b = p11 … pnn v . Given two prime decompositions of a and b. So wi ≤ ri and e | d. ….19 makes sure that any irreducible is a prime.

Then R is a UFD if and only if any prime nonzero ideal of R contains a prime element. so a decomposes in prime factors: a = p1 … pn. Indeed. Consider the set of ideals (ordered by inclusion): J = {I ideal in R | I ∩ S = ∅ and aR ⊆ I }. We will prove that any ideal I. by contradiction. We note that S is a multiplicative system (1 ∈ S and ∀a. Then there exists a prime ideal P in R containing a. First. together with the hypothesis. J is not empty. with P ∩ S = ∅. …. nonzero. Suppose. a is not a unit. b1…bn) = 1. 5. If S = R – {0}. contradiction. Since P ≠ R. b ∈ S. Suppose R is a UFD and P ≠ 0 is a prime ideal.6 Proposition. Let R be a UFD. n ∈ N* and a. If p is such an element. If (a. Let R be a domain. Suppose now that any nonzero prime ideal in R contains a prime element. ab ∈ S). Therefore. implies the existence of a prime element p ∈ P. there exists i such that pi ∈ P. we have p | (a. Because P is a prime ideal. then R is a UFD. contradiction. if ar ∈ S for some r ∈ R. Consider S = {a ∈ R | a ∈ U(R) or a is a product of prime elements}. If ar is a product of primes. The idea is to use Zorn's Lemma. Take a ∈ P. This fact. a ≠ 0.7 Theorem. bn ∈ R. since aR ∈ J. But p ∈ S.5 Unique factorization domains 39 5. Let us prove first that such an ideal exists. So we must have S = R – {0}. is inspired from Commutative Algebra techniques. It is enough to show that there is no prime p that simultaneously divides a and b1…bn. 1 ≤ j ≤ n such that p | bj. we show that . ! p is invertible. for any 1 ≤ i ≤ n. due to Kaplansky. with a ∉ S. bj) = 1. bi) = 1. The UFDs have many characterizations. Proof. b1. If ar is a unit. that there exists a ∈ R. Since p | a. is prime. Proof. contradicting P ∩ S = ∅. The following. then ar is a unit or is a product of primes. We have to prove the existence of P.I. then a is also a unit. then there exists j. then (a. maximal with the properties I ∩ S = ∅ and aR ⊆ I.

consider the ideals P + Rx and P + Ry. If ar = p1 … pn with p1. .40 I. The following important result shows the property of R being a factorization domain is inherited by the polynomial ring R[X]. 5. is bounded above in J by ∪l∈L Il. for some c. there exist some elements s ∈ S ∩(P + ! Rx) and t ∈ S ∩(P + Ry). br = p2 … pn. so ac = p2 … pn. Zorn's Lemma guarantees now that some maximal element P ∈ J exists. we determine the units of R[X]. Then U(R[X]) = U(R) and R[X]° = R° ∪ {f ∈ R[X] | deg f ≥ 1}. Arithmetic in integral domains a ∈ S. The checking is straightforward and is left to the reader. then the polynomial ring in one indeterminate R[X] is a UFD. it is not obvious that every polynomial in n indeterminates with coefficients in Z decomposes uniquely in irreducible factors. y ∈ R. J is inductively ordered: every chain in J. indexed by some set L. contradiction. which are also interesting on their own.8 Theorem. The induction hypothesis shows that b ∈ S. First. The ideal P is prime: let x. …. (Il)l∈L. In particular. If p1| r. If R is a UFD. If ar = p. If. with p a prime implies p ∼ r (so a is a unit) or p ∼ a (so a ∈ S). The proof of this theorem needs some preparations. by induction on the number of factors. Again by induction. then st ∈ S ∩ P. simplifying. This has far from trivial consequences: for instance. a ∈ S. let r = cp1. if K is a field. xy ∈ P. then a = bp1. with x ∉ P and y ∉ P. which strictly include P. Let R be a domain. If p1 | a. then p1| a or p1| r. so a ∈ S.9 Proposition. pn primes. 5. U(K[X]) = K* and K[X]° = {f ∈ K[X] | deg f ≥ 1}. Since P is maximal. by absurd.

The GCD of the coefficients a0. an is called the content of f. If f is a unit R[X]. From fg = 1 we deduce that f ∈ U(R). such that p . so deg f = 0 = deg g.ai. deg f + deg g = 0. Then the product fg is also primitive. for some primitive g ∈ R[X].11 Proposition. g ∈ R[X]. .fg. Since R is a domain. at least one of which is divisible 12 This is called “Gauss' Lemma”. which means that f. denoted c{ f }. Note that f is primitive if and only if there exists no prime p in R such that p divides all the coefficients of f. too. 5. …. a1. From p .10 Definition.f we deduce that there exists i. g be primitive in R[X]. 0 ≤ i ≤ n.f and p . Take i minimal with p . Also. If p is prime in R. Then c{ f g} = c{ f }·c(g). ! If K is a field.g. g ∈ R. g = b0 + b1X + … + bmX ∈ R[X] such that p . Let us prove that p . there is some g such that fg = 1. then c{ f } = a.12 c) Let R be a UFD and let f. a) Let R be a domain.I. let j be minimal i+j such that p .ai . U(R) ⊆ U(R[X]) is evident. then U(K) = K* and the other claims follow. then p is prime in R[X]. A polynomial with content (associated with) 1 is called a primitive polynomial. 5. b) Let R be a UFD and let f. Let f = a0 + a1X + … n m + anX . Let R be a UFD and let f = a0 + a1X + … + anX ∈ R[X]. aibj is not divisible by p and the other terms are divisible by p (as products of two factors. if f = a·g.5 Unique factorization domains 41 Proof. Then the coefficient of X in fg is ∑ ak bl k + l =i + j n In this sum.bj. Any f ∈ R[X] can be written f = c{ f }·g. a) Note first that: p divides a polynomial in R[X] if and only if p divides all the coefficients of the polynomial. Proof. with a ∈ R and g primitive. Similarly.

But deg g = deg g1 = deg g2 and similarly for h. Let R be a UFD and let K be its field of quotients. which is often easier to accomplish. 5. simplifying by a( = c(g1)c(h1)). we get something like af = g1h1. such that p | fg. Let us prove that f is a prime.b). So. where f1 and g1 are primitive. We will use the characterization in 5. since c{ f } = 1. the coefficient of X is not divisible by p and neither is the polynomial fg . with primitive g2. with g. it is sufficient to prove it is irreducible in R[X]. h1 ∈ R[X] and some a ∈ R. We can give now the Proof of the Theorem 5.8. K its field of quotients and f ∈ R[X]. The previous paragraph implies p | f or p | g. by multiplying with the LCM of the denominators of the coefficients of g and h. Then a polynomial f in R[X] of degree ≥ 1 is irreducible in R[X] if and only if f is primitive and irreducible in K[X]. prime.42 I. Arithmetic in integral domains i+j by p). irreducible. If f | gh. Since f is irreducible in R[X]. If f = gh. Le R be a UFD. we have a = c(g1)c(h1). we obtain f = g2h2. So af = c(g1)·c(h1)·g2·h2. h ∈ K[X]. ! Since f is primitive. for some polynomials g1. f has no non-invertible factors of degree 0. Let us prove that f is irreducible in K[X]. Then f is clearly primitive. Let f be irreducible in R[X]. there exists p ∈ R. g = c(g)·g1. h1 = c(h1)·h2. c) Let f = c{ f }·f1. Conversely.12 Proposition. ! with f1g1 primitive by b). deg g2 = 0 or deg h2 = 0. contradiction.4. so deg g = 0 or deg h = 0. Proof. This result is also important from a practical point of view: in order to prove that some polynomial with coefficients in R is irreducible in K[X]. We have g1 = c(g1)·g2. it has no proper divisors (of degree ≥ 1) in K[X]. b) If fg is not primitive. so it does not have divisors of degree ≥ 1 in R[X]. It is clear now that c{ fg} = c{ f }c(g). h2 ∈ R[X]. with g. Then fg = c{ f }c(g)·f1·g1. h ∈ . Computing the contents. if f is irreducible in K[X].

6 Polynomial ring arithmetic In this section. This implies a·c(g) = c(q). then f ∈ R° and it has a decomposition in irreducible factors in R. Applying the induction for g and h. the following rings are unique factorization domains: Z[X]. We can write q = c(q)·q1. If deg f > 0. We prove this by induction on the degree of f. so there exist a ∈ R. with g. if not. we infer that f1 is a product of irreducibles in R[X].I. Xn]. with f1 primitive. g1 primitive in R[X]. ! Thus. It is sufficient to prove the existence of a decomposition for f1. (David Hilbert's Basissatz). . Recall that R° denotes the set of nonzero and non-invertible elements in R. R designates a domain and K its field of quotients. we have f | g or f | h in K[X]. …. Say f | g in K[X]. having degrees smaller than f. we have g1 = fq1. then R[X] is Noetherian. Next. g = c(g)·g1. with q1. because f is irreducible in K[X] (and thus also prime in K[X]). f1 has a proper divisor in R[X]. factors that are also irreducible in R[X]. such that ag = fq. Thus ac(g)·g1 = f·c(q)·q1. f1 = gh. write f = c{ f }f1. we must show that any nonzero non-invertible polynomial f ∈ R[X] is a product of irreducibles. Z[X1. we are finished. being primitive). q ∈ R[X]. h ∈ R[X].6 Polynomial ring arithmetic 43 R[X]. so f | g in R[X]. If f1 is irreducible. If deg f = 0 and f is not a unit in R[X]. where K is a field. Thus. simplifying by c(q). An analogous result holds for Noetherian rings: if R is a commutative Noetherian ring. I. Xn]. K[X1. …. which must be a polynomial of degree strictly less than deg f ( f1 has no proper divisors in R.

then f ∈ R. ! An important consequence is the “Theorem of Bézout”: 6. then f is invertible and thus reducible. with r = 0 or deg r < deg f . f is irreducible in R[X] if and only if it is irreducible in R. The following statements are equivalent: . in this case. The proof of the following result is left to the reader: 6.if deg f = n > 0. this condition reads “f is primitive”.if deg f = 0. Let f ∈ R[X] and a ∈ R. R[X] is not principal (and certainly not Euclidian). 6. with g. (Integer division theorem) Let f. In this case. In practice.5. deg h < n. h ∈ R[X] and 1 ≤ deg g.3 Theorem. If R = K (R is a field). The fact that f has no non-invertible divisors of degree 0 amounts to saying that the GCD of the coefficients of f exists and is 1. when R is a UFD. A rich collection of irreducibility criteria is thus very useful in such problems.2 Proposition. g ∈ R[X]. If R is a domain and not a field. if f. Arithmetic in integral domains The problem of deciding the irreducibility of a polynomial in R[X] is important and often nontrivial. Recall that. r ∈ R[X] such that f = gq + r. then there exist q. and g has as leading coefficient a unit. Given f ∈ R[X].b)). However. f is irreducible in R[X] ⇔ f is primitive and f is irreducible in K[X]. so the theorem of division with remainder does not hold in R[X]. the following simple facts are worth remembering: . the argument of the proof of the division with remainder theorem for K[X] (K a field) still holds (see Example2. then f is irreducible in R[X] if and only if f has no non invertible divisors of degree 0 and there are no decompositions f = gh.1 Remark.44 I. . If the leading coefficient of g is a unit in R. g ∈ R[X].

Note that X − a divides f if and only if r = 0. 2 . f has no roots in K. where deg r = 0 or r = 0. one concludes that f has a divisor of degree 1.6 Proposition.6 Polynomial ring arithmetic 45 a) a is a root of f. A root of multiplicity 1 is called a simple root. by looking at the degrees of the factors in a decomposition of f. Proof. If f ∈ R[X] and a ∈ R. But f (a) = (a − a)q(a) + r(a) = r. Proof. Nevertheless. if f is reducible and has degree 2 or 3. Let K be a field. 6. There exist q. so f (a) = 0 is equivalent to r = 0.f . 2 2 If the ring R is not a UFD.4 Definition. which has a root in K.7 Examples. a primitive polynomial of degree 2 or 3 in R[X] is irreducible in R[X] if and only if it has no roots in K. we have: 6. the field of quotients of Z. The remaining claims follow from “f is irreducible in R[X] if and only if f is primitive ! and irreducible in K[X]”. r ∈ R[X] such that f = (X − a)q + r. then the criterion above may not work: 6. a) f = (2X + 1) is reducible in Z[X]. b) the polynomial X − a divides f in R[X]. Conversely. Then a polynomial f of degree 2 or 3 in K[X] is irreducible if and only if f has no roots in K. then f is reducible in R[X] (it is divisible with X − a).5 Corollary. if R is a UFD. but has no roots in Z. yet it is obviously reducible in Q[X]. ! This theorem is the basis of the notion of multiple root: 6. ! The converse of this statement is false: (X + 1) has no roots in Q.I. then. deg f > 1. the natural number m is called the multiplicity of the root a. Of course. Since f is irreducible in K[X] and deg f > 1. f has roots in Q. If f has a root a ∈ R. In particular. Let f ∈ R[X]. a is called a multiple root of m m+1 multiplicity m for f if (X − a) | f and (X − a) .

Arithmetic in integral domains b) Let R = {a + 2bi | a. If p/q ∈ Q is a root of f. The polynomial X + 1 is irreducible in R[X] (prove!). n n so p | a0q . we get that none of these numbers is a root. with p and q coprime. ( p. So. with p. By direct testing. q) = 1. where c is a unit in R. Since ( p.9 Example. q) = 1. This means that a polynomial of degree 2 or 3 in R[X] that has roots in Q(R) is not necessarily reducible in R[X]. Writing that f(p/q) = 0 and multiplying with q . Let f = X − X + 2 ∈ Z[X]. (p. 3 n n−1 n Here are some general tricks that may prove useful in irreducibility problems. A quick check shows that R is a 2 subring in Z[i]. d ∈ R. then p | a0 and q | an. q) = 1. 2. So. being primitive). Then f is irreducible if and only if f (cX + d) is irreducible. f ≠ 0.8 Proposition. Proof. f has no rational roots. The following criterion is widely used to find all rational roots of a polynomial in Z[X] (also see Exercise 26). f is irreducible in Q[X] (also in Z[X]. Then f is irreducible if and only if n r{ f } = an + an − 1X + … + a0X n .10 Proposition. b ∈ Z}.46 I. so it is a domain. we also have ( p. so p | a0. −1. the rational roots of f belong to the set {1. −i in Q[i]. ! 6. but has the roots i. Let R be a UFD and f = a0 + a1X + … + n anX ∈ R[X]. 6. a) Let c. q ) = 1 (R is a UFD). Let f = a0 + a1X + … + anX ∈ R[X]. q ∈ R. If p/q ∈ K is a root of f. b) Suppose f(0) = a0 ≠ 0. then p | 2 and q | 1. −2}. A similar proof can be given for q | an. we have: n −a0q = a1 pq + … + an p . We remark first that every element of K can be written as p/q. the field of quotients of R. 6. Since f has degree 3.

note that r ( f ) = X n f ⎛ ⎞ . Since ϕ preserves the degrees and ϕ | R = idR. ∀q ∈ R[X]. The element c is a unit if and only if ϕ is an isomorphism of R-algebras (the homomorphism of R-algebras −1 −1 ψ : R[X] → R[X] that takes X to c X − c d is the inverse of ϕ). p . with nonzero g(0) and h(0). Therefore.an . a) Let ϕ : R[X] → R[X] be the unique homomorphism of R-algebras (that is. we have 1 1 1 r ( gh ) = X m + p ( gh )⎛ ⎞ = X m g ⎛ ⎞ X p h ⎛ ⎞ = r ( g )r (h ) . then f is irreducible in K[X] (thus f is irreducible in R[X] if it is primitive). If there exists a prime ele2 ment p ∈ R such that p | ai. where n = deg f (for a rigorous ⎜ ⎟ ⎝X⎠ argument. Thenψ{ f } = ψ(g)ψ(h) . In other words. f = gh ⇔ ϕ{ f } = ϕ(g)ϕ(h). p -a0. ϕ{ f } is obtained by replacing the indeterminate X in f with cX + d. d) (Eisenstein criterion) Let R be a UFD. h ∈ R[X]. with g. c) Suppose f has no divisors of degree 0 other than units.6 Polynomial ring arithmetic 47 (the reciprocal of f ) is irreducible. ∀ i < n. deg h = p. So. b) If g. ∀g. Since degψ(q) ≤ deg q. If S is a commutative ring and ϕ : R → S is a unitary ring homomorphism such n that ϕ(an) ≠ 0 and ϕ(a0) + ϕ(a1)X + … + ϕ(an)X is an irreducible polynomial in S[X]. if deg g = m. then f is irreducible in R[X]. We must prove that ψ{ f } is irreducible implies f is irreducible. ⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎝X⎠ ⎝X⎠ ⎝X⎠ Because r preserves the degrees and. ψ is a ring homomorphism and ψ |R = ϕ) satisfying ϕ(X) = cX + d. c) Let ψ : R[X] → S[X] be the unique homomorphism of R-algebras such that ψ(X) = X. then r(gh) = r(g)r(h). we have d | f ⇔ r(d) | r{ f }. consider the equality in K(X).I. Suppose f = gh. h ∈ R[X]. for any d ∈ R. the field of quotients of K[X]). we obtain that degψ(g) = deg g and . Proof. one obtains that: f is irreducible if and only if ϕ{ f } is irreducible. the condition ϕ(an) ≠ 0 ensures that degψ(g) + degψ(h) = degψ{ f } = n. h ∈ R[X]. we get the conclusion. 1 Indeed.

there exists a decomposition of f of the form f = gh. The proof is proposed as an exercise. so p | b0c0 and p . If f were reducible. p > 0. The modern symbolic computation software (Maple. h ∈ R[X] are monic. Since f has no 0 degree non-invertible divisors. with f1 primitive.an. Such an algorithm (due to Kronecker) that also outputs a factor of the polynomial if it is not irreducible is described in Exercise 31. Macaulay. A few instances of using the above criteria on. thus there exists some i. Arithmetic in integral domains degψ(h) = deg h. such that p .11 Remark. Then p .bi and p | bj. 0 = degψ(g) = deg g. p does not divide all the bi's. c1. cp ≠ 0. Axiom. The exist algorithms that decide if a given polynomial in Z[X] is irreducible. yields a factorization algorithm (producing a decomposition in irreducible factors) for any polynomial in Z[X] or Q[X]. we can assume f is primitive. …. of degrees > 1.48 I.bic0 and ai = bic0 + ∑ b j ci − j j =1 i −1 is not divisible by p. Mathematica. Suppose p | b0 and p . then: n m p f = a0 + a1X +…+ anX = (b0 + b1X +…+ bmX ) (c0 + c1X +…+ cpX ). bm. so ψ(g) (to make a choice) is a unit. cp ∈ R. b0.b0c0 . It is sufficient now to prove that f is irreducible in R[X]. ! 6. So. We have that f and f1 are associated in K[X]. …. We 2 have b0c0 = a0. where g. d) Write f = c{ f }·f1. of degree 0. thus p divides exactly one of b0 and c0. 1 ≤ i ≤ m.c0. bm ≠ 0. This simple remark is useful in reducibility issues. If f ∈ R[X] is a monic reducible polynomial. concrete cases will give an idea on the strategies of approaching the problem of irreducibility of a polynomial. applied repeatedly. This algorithm. contradicting the hypothesis. Because p . etc. g ∈ U(R). By replacing f with f1. where m > 0. b1. c0. ∀j < i.) have powerful routines that decide polynomial irreduci- . Butψ{ f } is irreducible.

Consider the polynomial p ( X + 1) − 1 = p p X i −1 g = f ( X + 1) = ∑ i X +1−1 i =1 () Remark that the Eisenstein criterion can be applied to g. consider f as a polynomial in Y with coefficients in Z[X]. by a) above. c) Let p be a prime number and let 9 2 f=X p−1 +X p−2 + … + X + 1 ∈ Z[X]. then there exists one or L[X].I. 9 9 7 2 d) The polynomial f = Y + X Y − 3X Y + 2X is irreducible in Z[X. . including polynomials in several indeterminates. Indeed. The polynomial f has no roots in Z2 (easy check: Z2 has 2 elements. g is irreducible and. Note that Z can be replaced with any UFD of characteristic ≠ 2. with K a field. polynomials with coefficients in algebraic extensions of Q or in a finite field. a UFD. so the proper divisors of f can be of degrees 2 or 3. 0 and 1). by the Eisenstein criterion with p = 13.6 Polynomial ring arithmetic 49 bility. Thus. X − p is irreducible in Q[X] and in Z[X] (use Eisenstein again). 6. if there exists a factorization algorithm for K[X]. see SPINDLER [1994]. WINKLER[1996]. For details and developments.12 Examples. Apply Eisenstein with p = X (X is irreducible in Z[X]). 5 2 e) Consider f = X + X + 1 ∈ Z2[X]. A decomposition of f may look only like: 5 2 3 2 2 X + X + 1 = (X + aX + bX + 1)(X + cX + 1). with 1 ≤ i < p. The Eisenstein criterion cannot be applied directly to f. where L is a finitely generated extension of K. it is primitive). One can prove that. since p divides all the binomial coefficients p i ( ) . a) The polynomial 6X + 13X + 26 is irreducible in Q[X] (and in Z[X]. f is irreducible. n b) For any prime p and any n ∈ N*. Y].

b. is injective.c) for a polynomial f with integer coefficients is to “reduce the coefficients modulo n”. The polynomial f reduced 5 2 modulo 2 is X + X + 1 ∈ Z2[X].10. so f is irreducible in Z[X] (also in Q[X]). 3. 1. x & ϕx. Arithmetic in integral domains with a. Identifying the coefficients. (Hint: If R° is finite. Let R be a finite domain. f is irreducible in Z2[X]. then U(R) is infinite. Let S(R°) be the set of all bijections from R° to R°. R is a domain and K is its field of quotients (unless specified otherwise). c ∈ Z2. 5 3 2 f = 7X + 4X − X + 6 X + 9 ∈ Z[X].c) are satisfied. for instance. which is readily seen to have no solutions in Z2. f) A typical use of 6. we obtain a system of equations in a. Exercises Throughout the exercises. which is irreducible. The mapping ϕ : U(R) → S(R°). contradiction. Prove that the set R° of the nonzero non-invertible elements in R is infinite. irreducible by Eisenstein with p = 5. c. Hence. 7 2 g) 10X + 5X + 2 is irreducible in Z[X]: its reciprocal is 7 5 2X + 5X + 10. (denoted with ψ{ f } in the Proof).50 I. More precisely. ∀a ∈ R°. ϕx(a) = xa. Let R be a commutative unitary ring that has zero divisors and let a ∈ R be a zero divisor. If b ∈ R \ {0} is not a zero divisor. 2.) . for some n ∈ N conveniently chosen. The conditions in 6. prove that ax + b = 0 has no solutions in any ring S that includes R as a subring. Let R be an infinite unitary ring. Prove that R is a field. consider the unique ring homomorphism ϕ : Z → Zn and investigate the “polynomial f reduced modulo n”. b.10. Take.

What can you say about the uniqueness of such a representation? 5. Let R be a GCD domain. (b ≠ 0). . is prime if and only if p is a prime element in R.4. (Hint. Write a formula for the norm N : Z[θ ] → Z and determine U(Z[θ ]). Prove that Z[(1 + 5 ) 2] is Euclidian. Deduce a new proof for 5. Let d ∈ Z be squarefree. e) For d < 0. Show that Z[ 2 ] is Euclidian. 9. 8. with a.Exercises 51 4. Use 3. b ∈ R. coprime. d) Show that R := {α ∈ Q[ d ] | α is a quadratic integer} is a subring of Q[ d ] and R = Z[θ ]. pR[X]. b) Suppose known that any quadratic integer is a root of a monic polynomial of degree 2 with integer coefficients. 12. determine explicitly U(R). b ∈ Q. b ∈ Z. Show that the norm N : Z[θ ] → Z is 2 2 N(a + bθ) = a + ab − b .) 10. (Hint.11. 11. 3.3. 7. ∀a. Let θ = (1 + 5 ) 2 . b ∈ Q) is a quadratic integer ⇔ 2 2 2a ∈ Z. Let d ∈ Z. a) Any element in Q[ d ] can be uniquely written as a + b d . Determine U(Z[ d ]). 6. Then any element in K can be written as a/b. The elements 6 and 2 + − 5 have no GCD (and thus no LCM) in Z[ − 5 ]. Let θ = (1 + i 3 ) 2 . (Hint: R[X]/(pR[X]) ≅ (R/pR)[X]). Use 3. Show that: x = a + b d ∈ Q[ d ] (a.4 and Q[ 5 ] = Q[θ ]). d < 0 be squarefree. b ∈ Q) is a quadratic integer ⇔ Tr(x) = 2a 2 2 ∈ Z and N(x) = a − db ∈ Z.a). Show that a commutative ring R is a domain if and only if R[X] is a domain. where θ is given by Prop. c) Show that x = a + b d (with a. 2b ∈ Z and 4a − 4b d ≡ 0 (mod 4). Let p ∈ R°. with a. Prove that the ideal generated by p in R[X].

14. The purpose of the exercise is to determine all primes in Z[i]. then p cannot be written as a sum of two squares. β ∈ Z[ d ] such that αβ ∈ Z. then x ∈ R. d | a + bi in Z[i] ⇔ d | a and d | b in Z. d) If p is a prime in Z and p ≡ 3(mod 4). . b ∈ Z such that α = aγ and ⎯ ⎯ β = b γ ( γ is the conjugate of γ). then ⎛ ⎜ ⎟! ≡ −1(mod p ) ⎝ 2 ⎠ (Hint. b ∈ Z such that p = a + b (p 2 can be written as a sum of two squares) ⇔ the equation x + 1 = 0 has solutions in Zp. Let R be a unitary subring of the commutative ring S. ⇔ p ≡ 1(mod 4). K is its field of quotients and x ∈ K is integral over R. 1 − i and the primes p in Z with p ≡ 3(mod 4). we find the primes in Z that can be written as a sum of two squares. h) Prove that a prime p ∈ Z can be written as a sum of two squares 15. f) Prove that a prime p ∈ Z is also prime in Z[i] if and only if g) Prove that all the prime elements in Z[i] (up to association in divisibility) are: 1 + i. p −1⎞ e) If p is a prime in Z and p ≡ 1(mod 4). Then: (∃) a. Prove that. the field of integers mod p). b) Suppose that p ∈ Z is a prime in Z and in Z[i]. Show that there is some γ ∈ Z[ d ] and a. Show that the 2 equation x + 1 = 0 has no solutions in Zp ( = Z/pZ. Use Wilson's Theorem: (p − 1)! ≡ −1(mod p)). p ≡ 3(mod 4). Let d ∈ Z be squarefree and α. (A domain R with this property is called integrally closed).52 I. if R is a GCD-domain. a) Prove: for any d ∈ Z and any a + bi ∈ Z[i]. An element of S is called integral over R if it is a root of a nonzero monic polynomial in R[X]. 2 2 c) Let p be prime in Z. As a bonus. Arithmetic in integral domains 13.

Take S saturated. 24. 22. Then find y ∈ R such that u does not divide any of y or y ± 1). Let R be an Euclidian domain with respect to the function ϕ and −1 let S be a multiplicatively closed system in R.Exercises 53 16. 19. (Hint.) . a UFD) is inherited by the unitary subrings of R? 20. Let d ≤ −13. Let R be a UFD that is not a field. Is this result a particular case of Proposition 3. K[X]. d ≡ 1 (mod 4). Let R be Euclidian with respect to ϕ. Then S R is an Euclidian domain.5? (Ind. Show that U(R) = { −1. min {ϕ(v) | v ∈ R°} = ϕ(u) for some u ∈ R°. Let d ∈ Z be squarefree.3). …. we have u | x or u | x ± 1. Then. R = Z[θ ]. (Hint. Find such a u for R = Z. d squarefree and denote by R the ring of integers of Q[ d ] (so. −1 Then S R is a UFD. Show that Z contains an infinity of prime elements not associated in divisibility to each other.) 18. (Ind. such that the group of units U(R) is finite. pn are all the primes – up to association in divisibility –. 17. with θ as in Prop. 2 is irreducible and not a prime. Let R be a PID and let S be multiplicatively closed system in R.) 23. let u ∈ R given by the preceding exercise. −1 Then the ring of quotients S R is a PID. Take x = 2 and deduce that u ∈ Z and u = ±2 or ±3. 3. ∃q ∈ R such that x − qu is a unit or 0. Show that R is not Euclidian. Then Z[ d ] is not a GCD-domain. Let R be a UFD and let S be a multiplicatively closed system in R. ∀x ∈ R. then there exists m ≥ 1 m such that 1 + (p1…pn) ∈ R°. Then R contains an infinity of prime elements not associated in divisibility to each other. Use exercise 4. If p1. If R is Euclidian. Does the property of a domain R of being Euclidian (respectively a PID.) 21. Show that there exists u ∈ R nonzero and not a unit with the property: ∀x ∈ R. (Hint. 1}.

xn ∈ K and (not necessarily n . Consider the field of quotients of R and use the previous problem). Write down explicitly the conclusions for an = 1. a) If p/q ∈ K is a root of f.54 I. where p. ∀f ∈ R[X]. (The Lagrange interpolation polynomial) Let K be a field. 28. ∀r ∈ R. b) Any polynomial of degree 1 has at most one root in R. Prove that any nonzero polynomial f ∈ K[X] has at most deg f roots in K (every root is counted with its multiplicity). Then an −1 f ( X ) = g(anX). Is the conclusion still valid if one does not assume that R is infinite? 30. f (x) = f (x) (the value of f in x). Prove that. c) R is a domain. if R is an infinite domain. Let K be a field. give a new proof for the Eisenstein criterion. n n n b) Let g = an −1a0 + an −2a1 X + … + an −1 X n −1 + X n . 29. q | an and (p − qr) | f (r). q) = 1. 27. defined by: ∀x ∈ R. is injective. If f ∈ R[X]. n (Hint. define the polynomial ~ ~ function f : R → R. What connection is between the roots of g and the roots of f? 3 2 c) Find the rational roots of 2X + 5X + 9X − 15 and 3 2 4X − 7X − 7X + 15. q ∈ R and ( p. thenψ { f } = π(an)X = ψ(g)ψ(h). n ≥ 1 an integer. Arithmetic in integral domains 25. Let R be a UFD and let p ∈ R be a prime. fix n + 1 distinct elements x0. Let R be a commutative ring. deg h ≥ 1. then g(0) and h(0) are divisible by p. then p | a0. Let R be a commutative ring. If f = a0 + a1 X + … + an X satisfies the hypothesis of the criten rion and f = gh.) 26. (Hint. Let R be a UFD and f = a0 + a1 X + … + an X ∈ R[X]. If deg g. Prove that the following statements are equivalent: a) Any nonzero polynomial f ∈ R[X] has at most deg f roots in R. …. Using the canonical homomorphism π : R → R/pR and its extension to a homomorphism ψ : R[X] → (R/pR)[X]. then the mapping ~ R ϕ : R[X] → R . ϕ{ f } = f .

…. Show that the following algorithm terminates in a finite number of steps and outputs a nontrivial factor of p of degree ≤ m or proves that p is irreducible: 1. ∀i. e) Suppose there is available a factorization algorithm for R (an algorithm that produces a decomposition of any element in R° in prime factors). If not. …. If there exists d ∈ D with Ld ∈ Z[X] and Ld | p. then Ld is a factor of p and STOP. a) Show that p is reducible in Z[X] ⇔ p has a divisor of degree between 1 and m. If not. primitive. Show that a0 + a1 X + … + an X ∈ Z2[X] (an ≠ 0) has no roots in Z2 if and only if a0(a0 + a1 + … + an) ≠ 0. 33. D is a finite set. 0 ≤ i ≤ n. What properties should R have in order to adapt the algorithm above to R[X]? f) Suppose R is a UFD and there exists a factorization algorithm for R[X]. go to 2. d) Suppose m = 2. ∀i}. 32. then X − xi is a factor of p and STOP. For any d ∈ D. then p is irreducible. If ∃i with p(xi) = 0. …. xm). 31. Propose a choice for (x0. …. m+1 2. (x0. c) Deduce an algorithm of deciding the irreducibility of polynomials in Q[X]. Decide the irreducibility of X + X + 2X − 1 ∈ Z[X]. yn ∈ K. and deg Ld ≤ m. m+1 b) Choose m + 1 distinct integers. Let p ∈ Z[X]. n 4 2 . dm) ∈ Z | di| p(xi). Let D = {d = (d0. xm) ∈ Z . Prove that there exists a unique polynomial L ∈ K[X] satisfying: deg L ≤ n and L(xi) = yi. let Ld ∈ Q[X] be the Lagrange interpolation polynomial with Ld(xi) = di. deg p = n and let m = the largest integer ≤ n/2.Exercises 55 distinct) y0. Then there exists a factorization algorithm for K[X].

then g = 0. with deg(g. d) (Generalization of b) Let R be infinite and let g ∈ R[X1.56 I. 35. Xn]. a2)| > m3 and so on. T] = R[X1. ∀a3 ∈ S(a1. d) Let p ∈ R[ X ] be homogeneous. ….…. A polynomial in n indeterminates p ∈ R[X1. ∀a2 ∈ S(a1). 5 prove that X − X − 1 ∈ Q[X] is irreducible. (∃) S(a1. then g = 0. q c) p is homogeneous of degree q ⇔ p(TX1. Using the equality (in Z2[X]): 5 3 2 2 X + X + 1 = ( X + X + 1)( X + X + 1). n ∀(a1. b) Let g ∈ R[X1. …. homogeneous of degree i. an). Then any divisor of p in R[ X ] is homogeneous.…. Xn] are equal if and only if the associated polynomial functions are equal. 0 ≤ a ≤ m and ga ≠ 0 ≠ gm. It suffices to prove that a = m. respectively b. If g(a1. an) = 0. Let R be a domain.….…. If g(a1. Deduce that two polynomials in R[X1.…. Xn) (equality in R[X1. Arithmetic in integral domains 34. is homogeneous of degree a + b. (∀) a1 ∈ S. Xn. ∀an ∈ S(a1. If f has at least m + 1 roots in R. Xn][ T ]). c) Give an example of a finite field K and distinct polynomials in K[X] that have the same associated polynomial function. a) Let f ∈ R[X]. a2).) . where gi is homogeneous of degree i.…. (∀) a2 ∈ S(a1). an) = 0. (∀) a1 ∈ S. a2) ⊆ R with |S(a1. with pi ∈ R[ X ]. b) The product of two homogeneous polynomials of degrees a.…. deg f = m. write g = ga + … + gm. ∀a1 ∈ S. (∃) S(a1) ⊆ R with |S(a1)| > m2. Suppose that (∃) S ⊆ R with |S| > m1. an) ∈ R . For p = gh. Xn] =: R[ X ] is called homogeneous of degree q if all the monomials in p have total degree q (see the Appendix). (Hint. …. TXn) = T p(X1. Xi) = mi. with R infinite. Xn]. then f = 0. …. …. 36. Show that: a) Any p ∈ R[ X ] can be written uniquely as: p = p0 + p1 + … + pm.….

β ∈ K.1 ≤ i. ∀t.Exercises 57 e) If R is infinite and p ∈ R[ X ]. with a. Y]. i. Y] ⇔ b − 4ac is a square in K ⇔ b − 4ac = 0 or there exist α. 43. The cases char K = 3 and char K ≠ 3 should be treated separately). (Hint. j ≤ n]). 2 2 p = aX + bXY + cY . for g ∈ K[X. c ∈ K. Prove that p is irreducible in K[X. (Hint. Let K be a field of characteristic not equal to 2 (1 + 1 ≠ 0 in K) and p a homogeneous polynomial of degree 2 in K[X. Y] if and only if g is homogeneous g(X. and consider the n×n matrix A = (Xij)1 ≤ i. 41. x1. 1) in K[X]. Y] be homogeneous.1 ≤ i. Y] ⇔ p(X.e. char K ≠ 3 and f = X1 + … + Xn ∈ K[X1. Assume K is a field and p = a0Y + a1 Y X + … + an X homogeneous polynomial of degree n in K[X. X2][X3]. j ≤ n].…. then p is homogeneous of degree q q ⇔ p(tx1. β) ≠ (0. Generalization. n p(X.…. For n = 3. Let K be a field. 0). Then the polynomial: 2 3 3 det A = ∑{X1σ(1)… Xnσ(n) | σ ∈ Sn} is irreducible in Z[Xij. 1) | p(X. Consider n indeterminates Xij. Xn]. Prove that p is reducible in 2 2 K[X. 1 ≤ i.) 42. Use then an induction on n.…. 1) is irreducible in K[X]. txn) = t p(x1.…. n n−1 n is a Let any and 39. 40. apply Eisenstein to f ∈ K[X1. Prove that a commutative ring R is Noetherian if and only if every ideal in R is finitely generated. f) Is it true that any symmetric polynomial in R[ X ] has all its divisors symmetric polynomials in R[ X ]? 37. 38. j ≤ n ∈ Mn(Z[Xij. xn ∈ R. j ≤ n. . Let K be a field and let p ∈ K[X. Y]. (α. with p(α. g | p in K[X. 1) = a0 + a1 X + … + an X ∈ K[X]. b. Write a decomposition in irreducible factors for 3 3 X1 + X2 ∈ K[X1. Prove that. xn). Show that f is irreducible if and only if n ≥ 3. Y]. X2]. β) = 0.

The theory is fundamental in many areas of mathematics: commutative algebra. group representation theory. Also. Modules Module theory can be seen as a generalization of the classic linear algebra (which studies vector spaces over an arbitrary field1).1 Modules. submodules. homomorphisms The notion of a module over a ring can be obtained by replacing the word “field” with the word “ring” in the definition of the vector space. Module theory language and results are indispensable throughout most of modern algebra. homological algebra etc. II. algebraic number theory. notions that can be found in the Appendix).II. Although the volume Algèbre linéaire (1961) of the famous Bourbaki series “Eléments de Mathématique” begins with the definition of the… module. module theory illustrates and uses concepts of category theory (we use some elementary notions from category theory in this chapter. 1 58 . algebra structure theorems.

we say that M is a right R-module. 2 Also called “multiplication of elements in M with scalars in R”. s ∈ R. s ∈ R and x. ii') x(r + s) = xr + xs. for any r. This notational abuse (which is widely used) should not confuse the reader. ii) (r + s)x = rx + sx.2 Remark. i. The usual notation for the “scalar multiplication” in the case of right R-modules is “with the scalars on the right”. like the zero element in M. the scalar multiplication is a function µ : M × R → M. ∀x ∈ M). x) =: rx. satisfying. submodules. ∀x ∈ M.II. +) an Abelian group. y ∈ M: i) r(x + y) = rx + ry. iv') x1 = x. . the zero element in R is denoted by 0. The axioms for the right R-module become in this case: i') (x + y)r = xr + yr. ∀x ∈ M. ∀r ∈ R. r) = xr. iii) (rs)x = r(sx). The addition in R is denoted by + . whereas the + in the RHS denotes the addition in R.1 Definition.1 Modules. homomorphisms 59 1. For instance. iv) 1x = x. the + in the LHS denotes the addition in R. a function µ : R × M → M (notation: µ (r.e. as the addition in M. ∀r. in axiom ii). Also. Let R be a ring with identity (not necessarily commutative) and (M. i. We say that M is a left R-module (or left module over R) if there exists an “external operation on M with operators in R”2. If the axiom iii) is replaced by: iii') (rs)x = s(rx). iii') x(rs) = (xr)s.e. with the notation µ (x. ∀r ∈ R. 1.

c) Any Abelian group (A. (R. The construction above shows that a result that holds for any ring R and any right R-module is valid also for any left R-module. This is the only external operation that endows A with a Z-module structure (exercise!). s ∈ R). xn). y ∈ M.60 II. In the same way. s) & rs. but the multiplicaop tion * in R is defined by r*s = sr. There is a handy notation for the fact that M is a left R-module. R has a (canonical) structure of left R-module. xn + yn). The n-fold Cartesian product n R = {(x1. +. “M is a right R-module” is denoted by MR. For n ∈ Z and a ∈ A. b) If R is a ring with identity. a K-module is exactly a K-vector space. . ∀(x1. If R is an arbitrary ring and M is a right R-module. …. Similarly. then the notions of left R-module and right R-module are the same (look at axiom iii'). 1. yn) = (x1 + y1. Indeed. where (R . s ∈ R and x. Modules for any r. +) is canonically a Z-module. s ∈ R. ∀r. a) If K is a field. namely R M. If R is commutative.3 Examples. then M becomes op op op a left R -module. R is canonically a right R-module. if n < 0. ∀r. xn) | xi ∈ R. 1 ≤ i ≤ n} becomes an R-module if the addition and the scalar multiplication are defined component-wise: (x1. …. all definitions for left modules have a natural correspondent for right modules. denoted RR. yn) ∈ R . and conversely. d) Let R be a ring and let n ∈ N*. +) is an Abelian group. na = a + … + a (n terms). na is defined as the “multiple” of a in the additive group A (if n ∈ N. *) is the opposite of the ring R (R and R have the same underlying Abelian group R. …. …. na = (−a) + … + (−a) (n terms)). xn) + (y1. …. The theory of Abelian groups is thus a particular case of module theory. …. the “external operation” R × R → R is the ring multiplication: (r. n (y1. denoted R R.

If M is a left S-module. ∀U ∈ M2. 1(Z) = M. A = (aij) ∈ Mm. But 3 . and “◦” is the usual map composition: ∀u. ∀x ∈ M. xn) = (rx1. n(R). ∀x ∈ M. then M has a structure of left R-module by “restriction of scalars”: ∀r ∈ R. defining a left R-module structure on M amounts to defining a unitary ring homomorphism λ : R → End(M).1 Modules. S becomes a left R-module (and also a right R-module).◦) is the endomorphism ring of the Abelian group M. v ∈ End(M). submodules. ∀x ∈ M. e) If R is a ring and m. emphasizing that the functions are written at the left of the argument. n ∈ N*. …. ∀r ∈ R.II. AU ∈ M is the usual matrix product. this forces the definition of the composition of functions in the “usual” manner defined above. (u◦v)(x) = u(v(x)). f) Let R := M2(Z) (the ring of 2×2 matrices with entries in Z) and M := M2. For a ring R and an Abelian group M. (u + v)(x) := u(x) + v(x). + . r(aij) := (raij) (multiply every entry of the matrix with r). …. ∀(x1. 1(Z) (the Abelian group of 2×1 matrices with entries in Z). the set Mm. xn) ∈ R . homomorphisms n 61 r(x1. where (End(M). In particular. 3 Sometimes this ring is called the ring of left endomorphisms of M. This example generalizes a situation often encountered in field extensions: any field S is a vector space over any subfield R. rx := ϕ(r)x.n(R) of m×n matrices with entries in R is an Abelian group endowed with usual matrix multiplication and becomes an R-module by defining the “multiplication of matrices with scalars”: for r ∈ R. rxn). Checking the module axioms is straightforward and it boils down to the known properties of matrix operations. …. like u(x). M has a natural structure of left R-module: ∀A ∈ M2(Z) = R. Can you generalize this example? Can M be endowed with a “natural” structure of a right R-module ? g) Let ϕ : R → S be a unitary ring homomorphism. v ∈ End(M). 1.4 Remark. defined as follows: “+” is homomorphism addition: ∀u.

define λ : R → End(M) by λ(r)(x) = rx. if M is a left R-module. All modules are left R-modules (unless specified otherwise). is r(−x). Since M is a group. if one writes (x)u for the value of u at x. The same method shows r0 = 0. y ∈ L ⇒ x − y ∈ L. +). ∀r ∈ R. ! We define the natural notion of submodule: 1. The reader should check the details. Usually one says L is a submodule of M if no confusions can occur. define scalar multiplication R × M → M by (r. 1. we have: a) 0x = r0 = 0. then the composition of u and v is defined as . L ≤ M ). ∀x ∈ L ⇒ rx ∈ L. The.5 Proposition. A non-empty subset L of M is called left R-submodule of M if i) L is a subgroup in (M.6 Definition. the opposite of rx in (M. R denotes a ring with identity. Conversely.62 II. End(M) is called the ring of right endomorphisms of M and it is the opposite of the ring of left endomorphisms of M. b) r(−x) = (−r)x = −(rx). a) 0x = (0 + 0)x = 0x + 0x. by simplification we deduce that 0x = 0. ∀r ∈ R. With this multiplication. x) & rx := λ(r)(x). ∀x ∈ M. Proof. (x)(uv) = ((x)u)v. b) 0 = r0 = r(x + (−x)) = rx + r(−x). Notation: L ≤ R M (or. ii) ∀r ∈ R. ∀x ∈ M. +): ∀x. The fact that L is a right R-submodule of M is written L ≤ MR. for any x ∈ M and r ∈ R. Modules Indeed. So −(rx). What corresponds to a structure of right R-module of M? Throughout this section. a) Let M be a left R-module. simpler. given λ : R → End(M). Let M be an R-module.

x1. not equal to M. {0} is an R-submodule in M. y ∈ L. then rx + 0y = rx ∈ L and 1x + (− 1)y = x − y ∈ L. Also. submodules. it follows that rx + sy ∈ L. xn (r1. c) For any n ∈ N*. Since any submodule L of a module M is a subgroup of the additive group (M. xn ∈ L. s ∈ R and x.…. with r1. …. …. is called a linear combination of x1.II. Thus. rn ∈ R. y ∈ L ⇒ x − y ∈ L and rx ∈ L). r1. a) For any R-module M. b)⇒c) Induction on n (exercise). b)⇒a) If r ∈ R and x. a)⇒ b) If r. +). y ∈ L. L ≤ R M iff any linear combination of elements in L is still in L. rn ∈ R and x1. M is an R-submodule of M.1 Modules. ! Proof. ∀x. …. denoted simply by 0. . Let M be an R-module and ∅ ≠ L ⊆ M. b) For any r. is called a proper submodule of M.7 Proposition. we have 0 ∈ L. rn are called the coefficients of the linear combination). L has a structure of left R-module: the external operation is the restriction at R × L of the external operation of M. b) The left submodules of the canonical module R R are exactly the left ideals of the ring R. …. 1. so L contains also rx − (−s)y = rx − (−sy) = rx + sy. it follows that r1x1 +… + rnxn ∈ L. 1. ! An element of the form r1x1 +… + rnxn. The following statements are equivalent: a) L ≤ RM: (∀r ∈ R. A submodule of M. y ∈ L. then a) implies that rx and (−s)y ∈ L. c) If M is an Abelian group (= Z-module).…. xn ∈ M. the notation I ≤ R R means “I is a left ideal of R”. a Z-submodule of M is the same thing with a subgroup of M.8 Examples. if L ≤ RM. Also. homomorphisms 63 The definition above is the natural generalization of the notion of vector subspace. s ∈ R and x.

Modules 1. So. b) Define the set of linear combinations of elements in X with coefficients in R as the set RX. If X = ∅. If L is a submodule that includes X. x = 1x ∈ RX. So X ⊆ RX. ! This simple result allows us to define the notion of submodule generated by a subset: 1. ….9 Proposition.7. < X > is a submodule and includes X. Proof. For any a ∈ RM. Let (Mi)i∈I be a family of submodules of RM. RX is a submodule. then < X > ⊆ L because L is a member of the family of submodules the intersection of which is L. one says also that X is a system of generators for L (or that X generates L). ! 1. Then: a) < X > is the smallest (inclusion-wise) submodule of M that includes X. The case X = ∅ is trivial.10 Definition. that is. 1. Then the intersection of this family. If X ≠ ∅ and x ∈ X. The difference of two linear combinations in RX and the product of any r ∈ R with a linear combination is still in RX.12 Definition. a) Evidently.64 II. ∩ i∈I Mi . is a submodule of M. x1. If L ≤ R M and < X > = L. ….11 Proposition.c) implies RX ⊆ L. the submodule generated by X is the same as the set of linear combinations of elements in X with coefficients in R. r1. the submodule generated by {a} is Ra = {ra | r ∈ R} and it is also called the cyclic submodule . b) We show that X ⊆ RX and that RX is the smallest submodule that includes X. called the submodule generated by X and denoted by R < X > (or simply < X >). rn ∈ R. xn ∈ X}. Let M be an R-module and X ⊆ M. Let M be an R-module and let X be a subset of M. then x is a linear combination. b) < X > = RX. 1. If L is a submodule in M that includes X. a) The intersection of all submodules of M that include X is a submodule of M. where RX := {r1x1 + … + rnxn| n ∈ N. define R∅ = {0}.

the submodule generated by ∪i∈I Ei is called the sum of the family of submodules (Ei)i∈I. submodules. The submodule generated by E ∪ F is called the sum of the submodules E and F.13 Definition. …. then E1 + … + En = {e1 + … + en | e1 ∈ E1. f ∈ F}. denoted ∑i∈I Ei or ∑I Ei. a) Let S := {e + f | e ∈ E. ! In order to formulate a similar result for the case of the sum of an arbitrary (possibly infinite) family of submodules. ∀e ∈ E. En are submodules of M. A straightforward verification shows that S is a submodule. E + F is just another notation for < E ∪ F >. F be submodules in M. f ∈ F}. The R-module M is called finitely generated if there exists a finite system of generators for M. then the sum of E and F is E + F = {e + f | e ∈ E. S ⊆ L and S is the smallest submodule including E ∪ F. and is denoted E + F. For an arbitrary family (Ei)i∈I of submodules of M.14 Proposition. Let M be an R-module and let E. …. ∀f ∈ F. 1. If L is a submodule containing E and F. The sum of the submodules E1. Thus. homomorphisms 65 generated by a. Proof. While the intersection of a family of submodules is a submodule. an R-module M . 1. En is denoted E1 + … + En or ∑ Ei .e.1 Modules. b) If E1. …. a) If E. we introduce the following notion: for a set I (seen as a set of indexes). then e + f ∈ L. i =1 n The sum of the family of submodules (Ei)i∈I is the smallest submodule of M including all submodules Ei. Thus. F are submodules of RM. the union of a family of submodules is not a submodule in general.II. a finite subset F of M such that < F > = M. en ∈ En}. i.

and inf F = ∩F. ein ∈ Ein } . ∀i ∈ I. (LR(M). 1. ∀i ∈ I . ( ei )i∈I having finite support . i1 . then re = ∑ rei ∈ S. 1. ei1 ∈ Ei1 . we show that S is a submodule: if r ∈ R. define the support of the family (denoted Supp((ei)i∈I)): Supp((ei)i∈I) := {i ∈ I | ei ≠ 0} For any family (ei)i∈I having finite support J ⊆ I. any family of submodules of M). then S ⊆ L. f ∈ S. Modules and a family of elements4 (ei)i∈I. if e. since Ei i∈I is a submodule. and e = ∑ ei ∈ S. Note that a “family (ei)i ∈ I of elements of M” is in fact a function f : I → M .16 Remark. i∈I } As above. in ∈ I . for any B ⊆ A. ∀i ∈ I ).. then e − f ∈ S. there exists sup B (the smallest upper bound of B) and inf B (the largest lower bound of B) in A. then ∑ i∈I Ei = = {ei1 + … + ein n ∈ '.e.… . with ei ∈ Ei. sup F is the sum of the family F. So S = < ∪i∈I Ei>. ∀i ∈ I . If L is another submodule of M that in! cludes all submodules Ei. ∀i ∈ I. furthermore. ⊆) is a complete lattice: 5 for any subset F of LR(M) (i. The set LR(M) of all submodules of the R-module M is ordered by inclusion. its sum is defined as ∑i∈I ei := ∑i∈J ej. clearly Ei ⊆ S. Likewise. 5 An ordered set A is called a complete lattice if.15 Proposition. On the other hand. with ei ∈ M. If (Ei)i∈I is a family of submodules of M. 4 (denoting f (i) = ei. ∀i ∈ I.… . and Supp(ei)i∈I finite.66 II. ( ei )i∈I having finite support = } {∑ e i∈I i ei ∈ Ei . Let S = {∑ e i∈I i ei ∈ Ei . Proof.

j ∈ I. In particular. for some i. homomorphisms 67 1. its maximal elements (if they exist!) are exactly the maximal submodules of M that include L. The submodule L ≤ M is called a minimal submodule if L is minimal among the nonzero submodules: ∀E ≤ R M with E ≠ 0. The following theorem is very important.19 Theorem. Let L ≤ R M. i. The sum of the submodules I.18 Definition. 1. Proof. then. 6 . xn} be a finite generator set of M. The lattice LR(M) always has a greatest element (M itself) and a smallest element (the submodule 0). P is ordered by inclusion. This chain of submodules is bounded above in P by ∪i∈I Ei =: E. If R is a field. Take a chain (Ei)i∈I. Let P be the set of proper submodules of M that include L. J of the left canonical module R R is the same with the sum of the left ideals I and J.: ∀E ≤ R M with E ≠ M. y ∈ E. ∀i ∈ I. 1. x ∈ Ei and y ∈ Ej. note that P ≠ ∅ since L ∈ P. M has a maximal submodule. L ⊆ E implies L = E. E is a submodule6: if x. the submodules of an R-module M are the vector subspaces of M “sum of submodules” means “sum of vector subspaces”. E ⊆ L implies E = L. Let M be a nonzero finitely generated R-module. with Ei ∈ P.1 Modules.17 Remark.e.II. Indeed. submodules. L ≠ M and let {x1. the notions of maximal submodule and minimal submodule are defined as follows. Here is a (singular) situation when the union of a family of submodules is a submodule. We use Zorn's lemma to prove that maximal elements exist in P. The submodule L of the R-module M is called a maximal submodule if L is maximal among the proper (distinct from M) submodules. For this reason. First. …. Then any proper submodule of M is included in some maximal submodule.

so. Thus {x1. We deduce M = < x1. ∀t ∈ {1. A function ϕ : M → N is called a left R-module homomorphism (or simply module homomorphism) or R-homomorphism if it preserves the module operations: ϕ(x + y) = ϕ(x) + ϕ(y). in} such that Eit ⊆ Ej. An R-module homomorphism ϕ : M → M is called an endomorphism of M. 1. ∀x ∈ E. …. xn} ⊆ E = ∪i∈I Ei. Let M and N be left R-modules. Anyway. we have Ei ⊆ Ej or Ej ⊆ Ei. the existence of a maximal submodule in M is proven. contradicting Ej ≠ M (since Ej ∈ P). . linear transformation 8 . R has a maximal left ideal.21 Definition. Modules (Ei)i∈I being a chain. so ∃j ∈ {i1. German mathematician. n}. ϕ(rx) = rϕ(x). we have rx ∈ E. The canonical left R-module R R is finitely generated (by ! {1}). But (Ei)i∈I is a chain. …. xn > ⊆ Ej. ∀r ∈ R. R-morphism.20 Corollary (Krull's Lemma 7 ) Let R be a ring with identity. E ≤ M and E includes L. Similarly. In particular. Suppose E = M. ∀t ∈ {1. …. …. Taking L = 0. So x − y ∈ Ej (because Ej ≤ R M) or x − y ∈ Ei. ∀x ∈ M. …. used mainly for vector space homomorphisms. So. 7 8 Wolfgang Adolf Ludwig Helmuth Krull (1899-1971). Zorn's Lemma provides us with a maximal element of P. there exists it ∈ I such that xt ∈ Eit . ! 1. xn} ⊆ Ej. We must also prove that E ≠ M. …. ∀r ∈ R. Then {x1.68 II. Then every left proper ideal of R is included in some maximal left ideal. y ∈ M. Proof. n}. ∀x. A geometric terminology. x − y ∈ E. Other names: R-linear application.

a) For any R-modules M and N. ∀a ∈ R. b ∈ R.23 Examples. Also: ϕ(− x) = − ϕ(x). 0 : M → N. Conversely. N. Indeed. 1. N) := {ϕ | ϕ : M → N. 0(x) = 0. . 1. if L ≤ M. Note that the commutativity of R is effectively used. b) If M is an R-module and x ∈ M. ϕ is an R-module homomorphism}.for left modules. For any R-modules M. c) If R is commutative ring and r ∈ R. EndR(M) := HomR(M. respectively End(R M) . homomorphisms 69 The first condition in the definition of the module homomorphism ϕ : M → N means that ϕ is an Abelian group homomorphism. is an R-module homomorphism.22 Remark. rx(a + b) = (a + b)x = ax + bx = rx(a) + rx(b). idM(x) = x. rx : R R → M defined by rx(a) = ax. We denote sometimes: Hom(R M. the identity homomorphism of M. then ι : L → M is a module homomorphism. as well as the zero element of N ). ∀x ∈ M.1 Modules. N). ∀a ∈ R. rx(ba) = (ba)x = b(ax) = brx(a). λr : M → M. M). ∀a. then L ≤ M. Let L ⊆ M.24 Definition. ∀x ∈ M. submodules. the “multiplication by x”. ϕ(0) = 0 (0 denotes the zero element of M.II. Thus. is also a homomorphism. ∀x ∈ M. then the “multiplication by r”. λr(x) = rx. ∀x ∈ M. The identity application idM : M → M. If L has an R-module structure such that the canonical inclusion ι : L → M is a module homomorphism. is a module homomorphism. called the zero homomorphism. 1. we denote HomR(M. ∀x ∈ M. is a homomorphism: it is obviously additive and λr(ax) = r(ax) = (ra)x = (ar)x = a(rx) = aλr(x).

(EndR(E).70 II. ϕ(x) = ϕ(r1x1 + … + rnxn) = r1ϕ(x1) + … + rnϕ(xn) = ! r1ψ(x1) + … + rnψ(xn) = ψ(x). G be R-modules and let ϕ : E → F.26 Proposition. xn ∈ S and r1. there exist x1. Moreover. F) is an Abelian group with respect to homomorphism addition. Modules Hom(MR. F be R-modules and let ϕ : E → F. F be R-modules. Then the sum ϕ +η : E → F. Then the restriction of ψ to ! E.for right modules. Then their composition ψ ◦ϕ : E → G is also an R-module homomorphism. the unity element being the identity homomorphism idE. HomR(E. respectively End(MR) . 0(x) = 0. . ψ : E → F module homomorphism. N) is always nonempty. Proof. As a consequence.25 Remark. defined by: (ϕ +η)(x) := ϕ(x) + η(x). defined below: 1. rn ∈ R such that x = r1x1 + … + rnxn. the opposite of ϕ is (−ϕ). F. it contains at least the zero homomorphism 0 : M → N. Then ϕ = ψ iff ϕ|S = ψ|S. Let E. ψ : F → G be R-module homomorphisms. a) Let E. b) Let E. Suppose ϕ|S = ψ|S. S a system of generators of E and ϕ. the zero element is the 0 homomorphism. 1. So. ∀x ∈ E. N) is an Abelian group with respect to homomorphism addition. HomR(M. is an R-module homomorphism. is an R-module homomorphism. ∀x ∈ E. ◦) is a ring. (−ϕ)(x) = −ϕ(x). ….ψ|E : E → G.27 Proposition. c) Let E be a submodule of the R-module F and let ψ : F → G be an R-module homomorphism. ∀x ∈ M. HomR(M. N). η : E → F be R-module homomorphisms. …. If x ∈ E. + . A homomorphism is perfectly determined by its values on a generating set: 1.

x2 ∈ E such that ψ(y1) = x1 and ψ(y2) = x2. so −1 −1 x − y ∈ ϕ (N'). ψ(ry) = rψ(y). If ϕ is an isomorphism. and ϕ(M) is a submodule of N. so rx ∈ ϕ (N') .31 Proposition. Then ϕ(x − y) = ϕ(x) − ϕ(y) ∈ N'. ∀r ∈ R. Let y1. y ∈ ϕ (N'). 1. An R-module homomorphism ϕ : E → F is called an R-module isomorphism if there exists a homomorphism ψ : F → E such that ϕ ◦ψ = idF and ψ ◦ϕ = idE. Let x. y2 ∈ F and x1. Kerϕ is a submodule of M. ψ(y) = x. We have ψ(y1 + y2) = x1 + x2.30 Definition. Let ϕ : M → N be an R-module homomorphism. In particular. Suppose ϕ is a bijective homomorphism. So. An R-module homomorphism is an isomorphism iff it is bijective. We write in this case E ≅ R F (or E ≅ F if it is clear that it is an R-module isomorphism ). and Imϕ is a submodule of N. Consider the subset of M defined by: −1 Kerϕ := {x ∈ M |ϕ(x) = 0} = ϕ (0) Kerϕ is called the kernel of ϕ.II. If ϕ : M → N is an R-module homomorphism −1 and M' ≤ M. homomorphisms 71 1. ! 1.1 Modules. since ϕ(x1 + x2) = ϕ(x1) + ϕ(x2) = y1 + y2. We must prove that ψ is a homomorphism. If r ∈ R. Let Imϕ denote the image of ϕ: Imϕ := {y ∈ N| ∃x ∈ M such that y = ϕ(x)} = {ϕ(x)| x ∈ M}. then ϕ (N') is a submodule of M. ψ(y1 + y2) = x1 + x2 = ψ(y1) + ψ(y2). ∀y ∈ F. Let ϕ : E → F be a homomorphism. Similarly. The R-modules E and F are called isomorphic if there exists an R-module isomorphism ϕ : E → F. Then the inverse of the map ϕ exists. ∀y ∈ F. N' ≤ N. An isomorphism ϕ : E → E is called an automorphism of E.29 Proposition. then ϕ(rx) = rϕ(x) ∈ N'.28 Definition. Proof. ψ : F → E. where x is the unique element in E with ϕ(x) = y. then it is an invertible map. −1 Proof. 1. so it is a bijection. submodules. Recall that.

Then: a) ϕ is injective iff Kerϕ = 0. B. Prove that the commutativity of the addition of a module is a consequence of the other axioms of the definition of the module. distributivity. Study the properties of the operations + and ∩ on LR(M ) (such as commutativity. Which of the following subsets of the K-module K[X] is a K-submodule? a) The set of all polynomials of degree 7. C ≤ M. associativity. 2. Let K be a field. Let M be an R-module and let A. 4. and ! rz = rϕ(x) = ϕ(rx) ∈ ϕ(M'). Then there exist x. d) The set of all polynomials of degree 1 that have the root 1. 5. 1. existence of neutral elements…). . Let M be an R-module. t ∈ ϕ(M') and r ∈ R. If B ≤ A.72 II. then A∩(B + C) = B + A∩C (one says that LR(M ) is a modular lattice). e) The set of all polynomials of even degree. Which of the sets above is a K[X]-submodule? 3. ! b) ϕ is surjective iff Imϕ = N. Let E and F be submodules of the R-module M. We have z − t = ϕ(x) − ϕ(y) = ϕ(x − y) ∈ ϕ(M'). Let ϕ : M → N be an R-module homomorphism. y ∈ M' such that z = ϕ(x). c) The set of all monic polynomials. Show that E ∪ F is a submodule of M iff E ⊆ F or F ⊆ E.32 Proposition. Modules Let z. b) The set of all polynomials of degree at most 7. t = ϕ(y). Exercises 1.

Show that. Show that EndR(R) ≅ R. let M be a left R-module S and let S be a set. B ≤ R M and A'. Study the validity of the statements: a) u(A + B) = u(A) + u(B). e) The projection on a line. HomR(M.1 Modules. +) be an Abelian group and n ∈ N* such that na = 0. Define a left S R-module structure on R . ∀a ∈ G. c) The translation by the vector v = (x. M'). 0). let S be a set and R := {ϕ : S → R}. 11. submodules.) 8. R M'. 10. M' such that the inclusion is strict. S 2 2 . More generally. M') ⊆ HomZ(M. M) ≅ M. b) The rotation of angle α around (0. 7. More generally. for any R M. Try a vector space.. Define a left R-module structure on M = {ϕ : S → M}. Is the converse true? Generalization. A. LR(M ) is not distributive). −1 −1 −1 d) u (A' ∩ B') = u (A') ∩ u (B'). Give examples of R. Let R be a ring. 9. C ≤ M such that A∩(B + C) ≠ A∩B + A∩C (i. 13.e. HomR(R. Then G has a canonical structure of Zn-module. d) The symmetry with respect to a line. B' ≤ R M'. b) u(A ∩ B) = u(A) ∩ u(B). seen as an R-vector space. (Hint. for any R M. Identify the Euclidian plane with R . Which of the following transformations of the plane is a linear 2 2 transformation (an R-module homomorphism) from R to R ? a) The rotation of angle α around (0. B. 12.II. Let (G. 1). −1 −1 −1 c) u (A' + B') = u (A') + u (B'). homomorphisms 73 6. Let R be a ring. Determine all submodules of the R-module R . Let R be a commutative ring. M. y). Give an example of a module M and A. Let u : M → M' be an R-module homomorphism.

with minor modifications. . Modules 14. Let V be a finite dimensional K-vector space and let u ∈ EndK(V). Then: u is injective ⇔ u is surjective ⇔ u is an isomorphism. 2 n II. L is a subgroup in M. …. vn ∈ K . inherited from the R-module structure on M. Write 4 distinct homomorphisms from the R-module R to R.2 Factor modules and the isomorphism theorems The method used to construct the factor ring (given a ring and an ideal in the ring) can be applied to modules. ∀RB ⇔ HomR(B. Prove that A = 0 ⇔ HomR(A. Considering only the Abelian group structure on M. The Abelian group M/L can be endowed with a natural R-module structure. Show that R[X] is not a finitely generated R-module. Give examples of an R-module M and of an endomorphism ϕ : M → M such that: a) ϕ is injective. 15. Find necessary and sufficient conditions for n {v1. ….74 II. Let K be a field. B) = 0. Let A be an R-module. A) = 0. ∀RB. n ∈ N* and v1. Let M be a left R-module and let L be a submodule in M. which is also an Abelian group. so we can construct the factor group M/L. 18. 17. Does the Z-module Q have minimal submodules? 19. …. where vi = (vi1. b) ϕ is surjective. 16. but not surjective. but not injective. Let R be a commutative ring. Which is the general form of such a homomorphism? Generalization. 20. vn} to be a system of generators for the K-module K . vin).

the set {y ∈ M | x ∼ y (mod L)} is x + L. The map π : M → M/L.1 Definition. An easy check shows that this is indeed an equivalence relation and that the equivalence class of x ∈ M. ∀x + L ∈ M/L. A routine exercise shows that M/L becomes a left R-module.e. . with x + L = y + L.e. The R-module M/L defined above is called the factor module of M with respect to L. where x + L := {x + l |l ∈ L} (called “the class of x modulo L”) Define now the set M/L as the set of all equivalence classes: M/L = {x + L | x ∈ M} Make M/L into an Abelian group by putting: (x + L) + (y + L) := (x + y) + L. i. 2. ∀x. so r(x − y) ∈ L (since L ≤ R M). rx + L = ry + L. One proves that: the operation above is well defined (it does not depend on the representatives of the classes modulo L) and (M/L.II. the neutral element is 0 + L (equal to l + L. Thus. y ∈ M. ∀x ∈ M: (r + s)(x + L) = (r + s)x + L = (rx + sx) + L = (rx + L) + (sx + L) = r(x + L ) + s(x + L). ∀r. then x − y ∈ L. π(x) = x + L. ∀x ∈ M. set r(x + L) := rx + L. called the canonical homomorphism or the canonical surjection. s ∈ R. y ∈ M : x ∼ y (mod L) ⇔ x − y ∈ L. ∀x ∈ M. rx − ry ∈ L. For instance. Getting back to the module case. is a surjective module homomorphism. +) is an Abelian group. define an external operation R × (M/L) → M/L: ∀r ∈ R. y ∈ M.2 Factor modules and the isomorphism theorems 75 We briefly recall the construction of the factor group M/L. i. This definition is correct: if r ∈ R and x. ∀l ∈ L). − (x + L) = (− x) + L. Define on M the equivalence relation (also called “modulo L congruence”) by: ∀x.

(n/m − 1)m + nZ}. Then there is a natural one-to-one increasing map between the lattice of the submodules of M that include L and the lattice of submodules of M/L. The following notions are general concepts in category theory. Also. …. If L ≤ R M and π : M → M/L is the canonical homomorphism. any submodule is the kernel of some homomorphism. through the canonical homomorphism π : M → M/L. then Kerπ = {x ∈ M | x + L = 0 + L} = L and Imπ = M/L. L(Z/nZ) = {mZ/nZ | m ∈ N. The above says that L(Z/nZ) is in one-to-one correspondence to the submodules of Z that include nZ. then π (B) = {x ∈ M | x + L ∈ B} is a sub−1 module of M that includes L. If B ≤ R M/L. Modules 2. u◦ϕ = v◦ϕ implies u = v. mZ/nZ = < m + nZ > is the unique submodule having n/m elements in Z/nZ:{0 + nZ. ϕ ◦u = ϕ ◦v implies u = v. v : A → M. for any R A and any homomorphisms u. ! The rest of the proof is left to the reader. 2. applied to the category R-Mod of the left R-modules: 2. the image. Let ϕ : M → N be an R-module homomorphism. m + nZ. ϕ is surjective. We remark that ϕ(A) = π(A). So.5 Definition. given by ϕ : {A ≤ R M | L ⊆ A} → LR(M/L). and ϕ(π (B)) = B. 2. v : N → A. ϕ(A) := {a + L | a ∈ A}. . m|n}. which are {mZ | m ∈ N. −1 ϕ(A) ≤ R M/L. Let M be an R-module and let L be a submodule.76 II. L ⊆ A. m|n}. b) ϕ is called an epimorphism if. To mZ ⊇ nZ corresponds mZ/nZ ≤ Z/nZ. Let n ∈ N. a) ϕ is called a monomorphism if. ∀A ≤ R M. So. of the submodule A. for any R A and any homomorphisms u. Proof. Let us find the submodules of the Z-module Z/nZ (also known as Zn). So.2 Examples.4 Example.3 Proposition. So.

y ∈ M are such that x + Kerϕ = y + Kerϕ.e. For the injectivity.7 Theorem. v ∈ HomR(A. ϕ(u(a)) = ϕ(v(a)). there exists a unique isomorphism ψ : M/Kerϕ → Imϕ such that the following diagram is commutative: M ⎯ϕ ⎯→ N π ↓ ↑ι ψ ⎯→ Im ϕ M Kerϕ ⎯ i. For any a ∈ Kerϕ. ψ is given by ψ(x + Kerϕ) = ϕ(x). Then ≅ Imϕ. Consider ι : Kerϕ → M the inclusion homomorphism and 0 : Kerϕ → M the zero homomorphism. The converse is left to the reader. π = 0. then x − y ∈ Kerϕ ⇔ ϕ(x − y) = 0 ⇔ ϕ(x) = ϕ(y). ϕ ◦ι(a) = ϕ(a) = 0 = ϕ ◦0(a). Thus. a) Suppose ϕ injective. Then: a) ϕ is a monomorphism if and only if ϕ is injective. but only on the equivalence class x + Kerϕ.6 Proposition.II. Let ϕ : M → N be a homomorphism.e. we show that Kerψ = {0 + Kerϕ}: if x ∈ M . ι = 0 ⇔ Kerϕ = 0. (The fundamental isomorphism theorem) Let M ϕ : M → N be an R-module homomorphism. b) ϕ is an epimorphism if and only if ϕ is surjective. ψ is surjective: Imψ = {ψ(x + Kerϕ)| x ∈ M} = {ϕ(x)| x ∈ M} = Imϕ. Since ϕ is injective. we deduce u(a) = v(a). M) such that ϕ◦u = ϕ◦v. where ι is the inclusion and π is the canonical surjection. This means N/Imϕ = 0⇔N = Imϕ. We have π ◦ϕ (x) = ϕ(x) + Imϕ = 0 + Imϕ = 0◦ϕ (x). Kerϕ More precisely. ∀x ∈ M. So. Consider the factor module N/Imϕ and π : N → N/Imϕ (canonical surjection) and 0 : N → N/Imϕ. Proof. Checking that ψ is a homomorphism is simple. Proof. Suppose now ϕ is a monomorphism and let A = Kerϕ. ∀x ∈ M. ψ : M/Kerϕ → Imϕ is well defined: if x. So. ! 2. ∀a ∈ A.2 Factor modules and the isomorphism theorems 77 2. ϕ = ι ◦ψ ◦π. i. b) Let ϕ be an epimorphism. so. Let R A and u. π ◦ϕ = 0◦ϕ. ψ(x + Kerϕ) does not depend on x.

The map ϕ is well defined: if x. then ϕ(x) = 0. The following corollaries illustrate this technique (which is used also for groups. so x ∈ Kerϕ. x + F = 0 + F} = F/E. algebras. Then E+F F ≅ . being the restriction of the canonical homomorphism E + F → (E + F)/E to the submodule F of E + F. ϕ is surjective: ∀(e + f ) + E ∈ (E + F)/E. ! then. ∀x ∈ M. with x + E = y + E. ϕ(x) = x + E. ι ◦ψ ◦π (x) = ψ(x + Kerϕ) = ϕ(x). ∀x ∈ M. F ≤ R M. 2.8 Remark. rings. One looks for a surjective homomorphism ϕ : A → C. Kerϕ = {x + E | x ∈ M. Also. ⇔ x + Kerϕ = 0 + Kerϕ.10 Corollary. Apply now the fundamental isomorphism theorem. then x − y ∈ E. (e + f ) + E = f + E = ϕ( f ). x − y ∈ F and x + F = y + F. If η : M/Kerϕ → Imϕ is a homomorphism such that ϕ = ι ◦η ◦π. Let R M and E. η(x + Kerϕ) = ιηπ(x) = ϕ(x). So. …). Then F/E is a submodule of M/E and: M E M ≅ (R-module isomorphism). F ≤ M such that E ⊆ F. Moreover. ! 2. F E F Proof.78 II. It easy to see that ϕ is a surjective module homomorphism.9 Corollary. ∀x ∈ F. Modules with ψ(x + Kerϕ) = 0. F/E ⊆ M/E = {x + E | x ∈ M}. The mapping ϕ is a module homomorphism. ϕ(x + E) = x + F. Let ϕ : M/E → M/F. so η = ψ. f ∈ F. Let ϕ : F → (E + F)/E. ∀x ∈ M. Let R M and E. Since F/E = {x + E | x ∈ F}. A typical use for the isomorphism theorem is as follows: suppose B ≤ R A and we want to prove that A/B ≅ C. . with e ∈ E. 2. with Kerϕ = B. E E∩F Proof. y ∈ M. The isomorphism theorem provides then the required isomorphism.

Indeed. so ψ(x) = ψ(x'). Let x ∈ E with ψ(x) = 0. η is a homomorphism (standard check) and. Therefore Kerϕ = Kerψ. If Kerϕ ⊆ Kerψ. Also. then η(ϕ(x)) = η'(ϕ(x)). Apply the fundamental isomorphism theorem. b) η is injective if and only if Kerϕ = Kerψ. i.e. x' ∈ E such that ϕ(x) = ϕ(x') = y. η(ϕ(x)) = ψ(x) = 0. so. a) Let y ∈ F. Then η(ϕ(x)) = 0. Since Kerϕ = Kerψ. Since ϕ is surjective. The definition is independent on the choice of x ∈ E with ϕ(x) = y. with ϕ surjective. we get F/(E ∩ F) ≅ (E + F)/E. ∀x ∈ E.2 Factor modules and the isomorphism theorems 79 Kerϕ = {x ∈ F | x + E = 0 + E} = {x ∈ F | x ∈ E} = E ∩ F. which means x ∈ Kerψ. then: a) “ψ factorizes through ϕ”: there exists a unique homomorphism η : F → G such that ψ = η ◦ϕ . Let ϕ : E → F and ψ : E → G be R-module homomorphisms. Define η(y) := ψ(x). E⎯ ⎯→ F ψ η ϕ G Proof. ∀x ∈ E. b) Suppose η is injective. x ∈ Kerϕ.11 Proposition. G) has the property that η'◦ϕ = ψ. x ∈ Kerϕ. if x. ! .II. Since ϕ is surjective. there exists x ∈ E with ϕ(x) = y. then x − x' ∈ Kerϕ ⊆ Kerψ. Suppose now that Kerϕ = Kerψ and take y ∈ F with η(y) = 0. so ϕ(x) = 0. If η´ ∈ HomR(F. η(ϕ(x)) = ψ(x). Then ∃x ∈ E with ϕ(x) = y. so ϕ(x) = y = 0.! The following result is often used in module theory arguments: 2. η = η'.

An R-module is called simple if it has no submodules other than zero or itself. Let M be a finitely generated R-module. Z24. Modules Exercises 1.80 II. Z) = 0 and HomZ(Z. if M is cyclic (generated by one element). 7. . 5. Deduce that. +) has a Zn-module structure? Does there exist a Zn-module structure on (Z. then M is isomorphic to a R-module of the form R/I. where I ≤ R R. 3. where d = GCD(m. Determine all the submodules in: Z6. Show that there exists n n ∈ N and a surjective homomorphism ϕ : R → M. b) HomZ(Zm. 2. Let m. Write all (up to isomorphism) simple Z-modules and all simple K[X]-modules. Under what conditions the Abelian group (Zm. Zn) ≅ Zd. Prove that every R-module homomorphism u : M → N can be written as u = v◦w. n). 4. L ϕ (L ) 8. Let m. where I is a maximal left ideal in R. n ∈ N*. Prove that every factor module of a finitely generated module is finitely generated. Show that M is simple if and only if it is isomorphic to a module of the type R/I. Then there exists a canonical isomorphism M ϕ (M ) ≅ . Generalize the result. Prove the following version of the fundamental isomorphism theorem: Let ϕ : M → N be an R-module homomorphism and let L ≤ R M such that Ker ϕ ⊆ L. where K is a field. +)? 6. Prove that: a) HomZ(Zn. where v is an injective homomorphism and w is a surjective homomorphism. n ∈ N. Zn) ≅ Zn. Z8.

Define the homomorphisms π1 : M1 × M2 → M1 π2 : M1 × M2 → M2. by: π1(x1. X2 in T. Exact sequences The methods of constructing new structures (from a given collection of structures) are extremely important. The Cartesian product M1 × M2 = {(x1. The Abelian group M1 × M2 becomes a left R-module by defining: r(x1. Let ξ1. x2) ∈ M1 × M2. M1 × M2 is in fact the direct product of the groups (M1. rx2). y2) := (x1 + y1. x2) | x1 ∈ M1. x2) := x1 and π2(x1. x2 + y2). (y1. The homomorphisms π1 and π2 are called the canonical projections of the direct product .3 Direct sums and products. Then Tξ1∩Tξ2 is not a finitely generated T-module. x2) := (rx1. and The R-module M1 × M2. It is trivial to check the R-module axioms.3 Direct sums and products. ∀(x1. +). ∀(x1. x2) ∈ M1 × M2. x2). ∀ (x1. where I is the ideal of S generated by {(X1 − X2)Xi | i ≥ 3}. We begin with the case of two R-modules M1 and M2. together with the homomorphisms π1 and π2. ξ2 be the images of X1. Exact sequences 81 9. S = K[Xn. y2) ∈ M1 × M2. +) and (M2. ∀r ∈ R. x2) + (y1.II. x2) := x2. x2 ∈ M2} is made into an Abelian group by defining: (x1. is called the direct product of M1 and M2. n ∈ N] and T = S/I. Let K be a field. Is it true that any intersection of finitely generated submodules is a finitely generated submodule? (Hint.) II. rings…) some of the most used constructions are the direct product and the direct sum. In the case of modules (and also groups.

Let I be a set (seen as a set of indices) and let (Mi)i∈I be a family of R-modules indexed by I. for i ∈ {1. We prove first the uniqueness. On the other hand. likewise. ( f + g)(i) := f(i) + g(i). An element f ∈ ∏i∈I Mi is usually written as (xi)i∈I or (xi)I (where we denoted f (i) = xi ∈ Mi. if there exists v as required. Let v : E → M1 × M2 be a homomorphism with the required property.2}): 9 i M 1 × M 2 ⎯⎯→ M i π v E vi Proof. v2(e)).82 II. v1 factorizes through π1 and v2 factorizes through π2. So. ! The construction above can be generalized for an arbitrary family (possibly infinite) of modules. ∀i ∈ I}. g ∈ ∏i∈I Mi. (the universality property of the direct product) Let M1 and M2 be R-modules. Define an operation of addition on ∏i∈I Mi by: ∀f. there exists a unique homomorphism v : E → M1 × M2 such that v1 = π1◦v and v2 = π2◦v (the diagram below is commutative.1 Theorem. for any R-module E and any homomorphisms v1 : E → M1. Recall that the Cartesian product of this family is ∏i∈I Mi = { f : I → ∪i∈I Mi | f (i) ∈ Mi. v2 : E → M2. The direct product satisfies the following universality property: 3. ∀i ∈ I). Let e ∈ E and suppose v(e) = (x1. ∀e ∈ E. Then. We have x1 = π1(v(e)) = (π1◦v)(e) = v1(e) and. ∀i ∈ I. . 9 So. Modules M1 × M2. x2 = v2(e). x2) ∈ M1 × M2. the map v defined in this way is clearly a module homomorphism. then v(e) = (v1(e).

2 Theorem. are called the canonical projec- The R-module ∏i∈I Mi. ∀f ∈ ∏i∈I Mi. ∏i∈I Mi is an Abelian group with respect to this operation. Define the external operation: ∀r ∈ R. (Mi)i∈I. g ∈ ∏i∈I Mi. ∀f ∈ ∏i∈I Mi. vi : E → Mi. using the notation (xi)I : πj((xi)I) := xj. is called the direct product of the family of modules tions. define the homomorphisms πj : ∏i∈I Mi→ Mj by: πj( f ) := f ( j). ∀(xi)I ∈ ∏i∈I Mi. i ∈ I. (yi)I ∈ ∏i∈I Mi. i ∏i∈I M i ⎯⎯→ M i π v E vi . ∀(xi)I ∈ ∏i∈I Mi. which shows that r( f + g) = rf + rg. i ∈ I. (rf )(i) := rf (i). Then. ∀i ∈ I. such that vi = πi◦v. ∀f. The homomorphisms πi.3 Direct sums and products. Exact sequences 83 With the alternate notation: (xi)I + (yi)I := (xi + yi)I. The direct product satisfies a universality property: 3. ∀i ∈ I. ∀(xi)I. r(xi)I := (rxi)I. ∀r ∈ R. we have (r( f + g))(i) = r·( f + g)(i) = r( f (i) + g(i)) = rf (i) + rg(i) = (rf + rg)(i). ∀i ∈ I. Proving that ∏i∈I Mi becomes an R-module with the operations above is routine: for instance. for any R-module E and any family of homomorphisms indexed by I. there exists a unique homomorphism v : E → ∏i∈I Mi. For any j ∈ I.II. together with the family of homomorphisms (πi)i∈I. With the alternate notation: ∀r ∈ R. (the universality property of the direct product) Let (Mi)i∈I be family of R-modules.

then v(e)(i) = vi(e). rect product of the family is denoted by M (or M if I is finite. ∀i ∈ I.3 Theorem. Let e ∈ E and denote v(e) by f ∈ ∏i∈I Mi. Suppose there exists a homomorphism v : E → ∏i∈I Mi with the desired property. If the i =1 n elements). Suppose that an R-module P. Modules Proof. On the other hand. For a finite family M1. set E = Q and vi = qi. i ∈ I. satisfies the property: “For any R-module E and any family of homomorphisms indexed by I. i ∈ I. Proof. there exists a unique homomorphism v : E → P such that vi = pi◦v. …. In (U). Let (Mi)i∈I be a family of R-modules. if such a homomorphism v exists. the di- A remarkable fact is that the universality property of the direct product characterizes the direct product up to a unique isomorphism: 3. pi : P → Mi. v defined this ! n way is a module homomorphism. satisfies also property (U) above. So. the notation is M1 ×…× Mn or I n ∏ M i . ∀i ∈ I. Since Q satisfies also 10 This is exactly the universality property of the direct product. ∀i ∈ I. then there exists a unique isomorphism ϕ : P → Q such that pi = qi ◦ϕ. having modules of the family (Mi)i∈I are all equal to some module M. i ∈ I. Mn. For any i ∈ I. .84 II. vi : E → Mi. and the family of homomorphisms indexed by I.” 10 (U) If the R-module Q. f (i) = πi( f ) = πi(v(e)) = (πi◦v)(e) = vi(e). It is basically the same as for two modules. The direct product of the family (Mi)i∈I is denoted also ×i∈I Mi. ∀e ∈ E. together with the family of homomorphisms qi : Q → Mi. We obtain a (unique) homomorphism ψ : Q → P such that qi = pi◦ψ.

2 says that in R-Mod the direct product of an arbitrary family of objects (Mi)i∈I exists. is called a direct sum of the family (Mi)i∈I if. For this reason. ∀i ∈ I (the diagram below is commutative. (πi)i∈I). ψ ◦ϕ = idP. without involving any elements of the underlying sets of the modules. there exists a unique homomorphism v : S → E such that v◦σi = vi. we get ϕ ◦ψ = idQ. Exact sequences 85 (U). σi : Mi → S. ϕ ◦ψ : Q → Q satisfies qi◦(ϕ ◦ψ) = (qi◦ϕ)◦ψ = pi◦ψ = qi. such that pi = qi◦ϕ.3 Direct sums and products.II. Similarly. The universality property of the direct product is articulated only in terms of objects (modules. for any module E and any family of homomorphisms (vi)i∈I. the universality property above is taken as a definition of the direct product in a category C (replace “module” with “object in C” and “module homomorphism” with “morphism in C”). Indeed.4 Remark. in this case) and homomorphisms. The direct product of a family of modules in C (if it exists) is uniquely determined up to an isomorphism: the proof of the proposition above is valid in any category. An R-module S. By uniqueness. We show that ϕ and ψ are inverse one to each other. apply (U) for P with the homomorphisms (pi)i ∈ I.5 Definition. Dualizing (“reversing the arrows”) in the universality property of the direct product. one obtains the universality property of the direct sum (taken as a definition of this notion): 3. ∀i ∈ I): i M i ⎯⎯→ S σ vi E v . 3. Theorem 3. but the homomorphism idQ : Q → Q has the same property: qi◦idQ = qi. ! guaranteed by (U). ∀i ∈ I. ∀i ∈ I. vi : Mi → E. Let (Mi)i∈I be a family of R-modules. namely (∏i∈I Mi. together with a family of homomorphisms (σi)i∈I. there exists a unique homomorphism ϕ : P → Q.

86 II. The homomorphism v : S → Mj. The direct sum is also called coproduct (a categorical name. so f + g has finite support. ⎩ x. Identify f ∈ (i∈I Mi with σj : Mj → (i∈I Mi ∀x ∈ Mj. ∀j ∈ I. For a fixed j ∈ I. 3. we have (the sum is finite. recall that the support of f is the set Supp f := {i ∈ I | f (i) ≠ 0}. if f. the family (xi)i∈I has finite support. emphasizing the duality with the product). It is easy to check that σj is a homomorphism. if i ≠ j (σj(x))(i) := ⎨ . Note that ∀(xi)I ∈ (i∈I Mi. In other words. Define. g ∈ (i∈I Mi. (σi)i∈I). given by the definition. so σj is injective. then Supp(rf ) ⊆ Supp f. We now construct an object satisfying the definition of the direct sum. in the definition above set E = Mj and vi : Mi → Mj defined as follows: vi = 0 if i ≠ j and vj = idMj. then xj = x. where xi = f (i) ∈ Mi. if i = j and xi = 0. i ≠ j. ∀i ∈ I. by: ⎧0. ∀j ∈ I. If r ∈ R. The canonical injections are indeed injective. if σj(x) = (xi)i∈I. then Supp( f + g) ⊆ Supp f ∪ Supp g. ∀i ∈ I. Indeed. If f ∈ ∏i∈I Mi. i runs only on Supp((xi)I). (i∈I Mi := {f ∈ ∏i∈I Mi | Supp f is finite}. Consider the following subset of ∏i∈I Mi: the “family of elements” (xi)i∈I. We claim that (i∈I Mi is a submodule of ∏i∈I Mi. (xi)I = ∑i∈I σi(xi) . which is finite). satisfies v ◦σj = id.6 Remark. Modules The homomorphisms (σi)i∈I are called the canonical injections of the direct sum (S.

Mn. If the modules in i =1 n the family (Mi)i∈I are all equal to the same module M. As in the case of the direct product.7 Theorem. 3. Exact sequences 87 3.2. Let (Mi)i∈I be a family of R-modules.! The module (i∈I Mi is often denoted ⊕i∈I Mi. the universality property characterizes the direct sum of a family of modules up to an isomorphism. ∀i ∈ I.3 Direct sums and products. ∀i ∈ I. The reader is invited to formulate this precisely and prove it. Nevertheless. ∀i ∈ I. v is indeed a homomorphism: ∀(xi)I. .11). …. together with the homomorphisms (σi)i∈I.II. Proof. the direct sum is a couple (⊕i∈I Mi. For a finite family M1. (σi)i∈I) and is not the same with the direct product {∏i∈I Mi. v(r(xi)I) = v((rxi)I) = ∑i∈I vi(rxi) = ∑i∈I rvi(xi) = r∑i∈I vi(xi) = rv((xi)I). the notation is M1 ⊕…⊕ Mn or ⊕ M i . ⊕i∈I Mi is denoted by M (I) . ∀(xi)I ∈ (i∈I Mi. (πi)i∈I}. we have: v((xi)I) = v{∑i∈I σi(xi)} = ∑i∈I (v◦σi)(xi) = ∑i∈I vi(xi). If we define v in this way. ∀r ∈ R. then. as in the case of the direct product. the module direct sum ⊕i∈I Mi = (i∈I Mi coincides with the module direct product ∏i∈I Mi in 3. Let E be an R-module and let vi : Mi → E. (i∈I Mi is sometimes called external direct sum. In order to avoid the confusion with the notion of inter- nal direct sum of submodules (see 3.8 Remark. (yi)I ∈ (i∈I Mi. is a direct sum of the family (Mi)i∈I. v((xi)I + (yi)I) = v((xi + yi)I) = ∑i∈I vi(xi + yi) = ∑i∈I vi(xi) +vi( yi) = v((xi)I) + v((yi)I). Then the module (i∈I Mi. This shows that the homomorphism v is uniquely determined by the condition v◦σi = vi. If v : (i∈I Mi → E is a homomorphism with v◦σi = vi. If the set of indices I is finite.

with xi ∈ Li. then ϕ((xi)i∈I) = ϕ((yi)i∈I). When are these sums “the same”? 3. For any x ∈ L.88 II. Let L denote their sum. and the (ηi)i∈I are the canonical injections. having finite support. which means that x is the sum of the family (xi)i∈I. such that ∑i∈I xi = 0. x = ϕ((xi)i∈I) = ϕ{∑i∈I σi(xi)} = ∑i∈I (ϕ ◦σi)(xi) = ∑i∈I ηi(xi) = ∑i∈I xi. ⎛ ⎞ c) For any j ∈ I. having finite support. If (yi)i∈I is another family having finite support.9 Example. with xi ∈ Li. ∀i ∈ I. So. having finite support. ∀i ∈ I. (xi)i∈I = (yi)i∈I. Modules 3.7. let ϕ : (i∈I Li → L be the unique isomorphism with ϕ ◦σi = ηi. ⎜ i∈I \{ j} ⎟ ⎝ ⎠ d) For any family (xi)i∈I. L := ∑i∈I Li. there exists a unique family (xi)i∈I. where Li are seen are modules by themselves can be also constructed. b) For any x ∈ L. As an R-module. xi ∈ Li. ∀i ∈ I. there exists (xi)i∈I ∈ (i∈I Li such that ϕ((xi)i∈I) = x. a)⇒b) Let (i∈I Li be the direct sum constructed as in 3. The direct sum of the family (Li)i∈I. we have xi = 0. . with (σi)i∈I the canonical injections. and let ηi : Li → L be the inclusion homomorphisms. Proof. ∀i ∈ I. The following statements are equivalent: a) L is the external direct sum of the modules (Li)i∈I. ∀i ∈ I. We defined the notion of sum of a family (Li)i∈I of submodules of R M. Let R be a commutative ring and let R[X] be the polynomial ring in the indeterminate X. Let M be an R-module and let (Li)i∈I be a family of submodules of M. such that x = ∑i∈I xi. Lj ∩ ⎜ ∑ Li ⎟ = 0. with yi ∈ Li. such that ∑i∈I xi = ∑i∈I yi. ∀i ∈ I. Since ϕ is an isomorphism. R[X] is ( ) isomorphic with the direct sum R N of a countable family of copies of R R.10 Proposition.

3 Direct sums and products. ∀i ≠ j and yj = −xj. then (ϕ ◦ηj)(xj) = ϕ(xj) = vj(xj). there exists a family of finite support (xi)i∈I (xi ∈ Li. Then there exists a family of finite ⎜ ⎟ ⎝ i∈I \{ j} ⎠ support. From d). we obtain that this family is unique: if ∑i∈I xi = ∑i∈I yi. The map ϕ thus defined is an R-module homomorphism: if x. ∀j ∈ I. such that xj = ∑i∈I \ {j} xi. Similarly one sees that ϕ(rx) = rϕ(x). We must show that ϕ is unique with this property. ϕ(x + y) = ∑i∈I vi(xi + yi) = ∑i∈I vi(xi) + ∑i∈I vi(yi) = ϕ(x) + ϕ(y).II. Let j ∈ I. ∀x ∈ L. We obtain that 0 is the sum of the finite support family (yi)i∈I. and x = ∑i∈I xi. then ∑i∈I (xi − yi) = 0. ∀i ∈ I. be such that ∑i∈I xi = 0. with yi ∈ Li. Since 0 has a unique writing as a sum of a finite support family (evidently. ∀i ∈ I). So.5. which shows that ϕ ◦ηj = vj. xj = 0. Let . then x + y = ∑i∈I xi + ∑i∈I yi = ∑i∈I(xi + yi). Exact sequences 89 ⎛ ⎞ b)⇒c) Let xj ∈ Lj ∩ ⎜ ∑ Li ⎟ . (xi)i∈I \ {j}. ∀r ∈ R. so xi = yi. xj = 0. ∀i ∈ I. ∀i ∈ I. we get xj ∈ ⎜ ∑ Li ⎟ . c)⇒d) Let (xi)i∈I. So. ⎛ ⎞ Then xj ∈ Lj. Define the homomorphism ϕ : L → E by: ∀x ∈ L. ∀i ∈ I \ {j}. ∀i ∈ I. If j ∈ I and xj ∈ Lj. and (xi)i∈I. xi ∈ Li. 0 is the sum of the family (0)i∈I). yi ∈ Li. Let RE and let a family of homomorphisms (vi)i∈I. ∀i ∈ I. ∀i ∈ I. If x ∈ L. (yi)i∈I are the unique families of finite support with xi. xi ∈ Li. y = ∑i∈I yi. such that x = ∑i∈I xi. put ϕ(x) := ∑i∈I vi(xi). where (xi)i∈I is the unique family of finite support with xi ∈ Li. where yi = xi. ⎜ ⎟ ⎝ i∈I \{ j} ⎠ d)⇒a) We show that (L. we obtain xi = 0. y ∈ L. ∀i ∈ I and such that x = ∑i∈I xi. where (xi + yi)i∈I has finite support and xi + yi ∈ Li. since xj = ∑i∈I \ {j}(−xi). vi : Mi → E. (ηi)i∈I) satisfies the definition 3. So.

a) The notions of (internal) direct sum of submodules and external direct sum are very close: if S is a direct sum of the modules (Mi)i∈I. there exists a unique pair (a. and Mi ≅ Imσi.11 Definition. Let (Li)i∈I be a family of submodules of an R-module M. a) If V is a K-vector space. xi ∈ Li. Then x = ∑i∈I xi. b) ∈ A × B such that x = a + b. . We say that “the sum of the family (Li)i∈I is a direct (internal) sum” if and only if one of the equivalent statements of the proposition above is satisfied. for any nonzero submodules A and B of R R. one also says that (Li)i∈I is an independent family of submodules. A submodule A of M is called a direct summand of M if there exists 0 ≠ B ≤ R M such that M = A ⊕ B (a submodule B having this property is called a complement of A). b) M = A + B and A ∩ B = 0. 3. then S is the internal direct sum of its submodules (Imσi)i∈I. then any vector subspace U of V is a direct summand in V. Indeed. ! 3. ! 3. ∀i ∈ I. Modules ψ : L → E be a homomorphism with ψ ◦ηi = vi.13 Examples. The following statements are equivalent: a) M = A ⊕ B. and let x ∈ L. B ≤ R M. i ∈ I.14 Remark. Let M be an R-module and let A. It is usual to identify Mi with Imσi and call S the direct sum of the modules Mi. ∀i ∈ I and: ψ(x) = ψ(∑i∈I xi) = ∑i∈Iψ( xi) = ∑i∈Iψ(ηi( xi)) = ∑i∈I vi(xi) = ϕ(x).12 Corollary. with canonical injections (σi)i∈I. b) If R is a domain. 0 ≠ ab ∈ A ∩ B. then R R has no nonzero submodule that is a direct summand.90 II. In this situation. ∀i ∈ I. 3. we have A ∩ B ≠ 0: for 0 ≠ a ∈ A and 0 ≠ b ∈ B. c) ∀x ∈ M.

x = p(y) = p (y) = p(x) = 0 and Im p ∩ Ker p = 0. Furthermore. We have: p(p(x)) = p(a) = a = p(x). then. b ∈ B. Evidently. If x ∈ Im p ∩ Ker p. called the canonical projection on Mj. with b ∈ B. So. σ p The direct summands of an R-module M are closely related with the idempotents of the ring EndR(M).16 Proposition. Exact sequences 91 canonical injections (σi)i∈I. Proof.15 Proposition. ! 3. the mapping pj : M → Mj. Let M and N be R-modules and let u : M → N and v : N → M be homomorphisms such that v◦u = idM.II. then p := iA◦pA (where pA : M → A is the canonical projection and iA : A → M is the canonical injection) is an idempotent of EndR(M) and Im p = A. If M = A ⊕ B. Then p(x) = a.e. Let M be an R-module. 3. (where x = ∑i∈I σi(xi) .3 Direct sums and products. so x − p(x) ∈ Ker p. For any x ∈ M. such that x = a + b. pj is exactly the restriction to M of the canonical projection πj of the direct product. there exist unique a ∈ A. ∀i ∈ I) is a module homo- b) If M = ⊕i∈I Mi is the direct sum of the modules (Mi)i∈I. 2 Suppose now that p ∈ EndR(M) and p = p. then 2 p(x) = 0 and x = p(y). Ker p = B. Therefore. Then . Im p = A. Ker p = B. (An element e of a ring S is 2 called an idempotent if e = e). For any x ∈ M. which means that M = Im p ⊕ Ker p. x & xj. We have p(x) = 0 iff x = 0 + b. M = Im p + Ker p. Conversely. with morphism. let p ∈ EndR(M) be an idempotent. 2 We have p(x) ∈ Im p and p(x − p(x)) = p(x) − p (x) = 0. Considering M as a submodule of the direct product ∏i∈I Mi. with y ∈ M. 2 p◦p = p = p. j j pj◦σj is the identity of Mj : M j ⎯⎯→ ∏ M i ⎯⎯→ M j . i. with xi ∈ Mi. Thus. Then M = Im p ⊕ Ker p. ∀j ∈ I. write x = p(x) + (x − p(x)).

! 3. N = Im p ⊕ Ker p by 3. Modules N = Im u ⊕ Ker v. If I = {1. 2 Proof. (τi)i∈I)) be the direct sum of the family (Mi)i∈I (respectively (Ni)i∈I). the notations are For any x = (xi)i∈I ∈ ∏ Mi. (πi)i∈I) (respectively (∏Ni. n}. τj◦vj : Mj → ⊕Ni is a homomorphism. Let vi : Mi → Ni (i ∈ I) be a family of R-module homomorphisms. In particular.17 Definition. (ρi)i∈I)) be the direct product of the family (Mi)i∈I (respectively (Ni)i∈I). The universality property of the ∏i=1 vi n or v1 ×…× vn. Ker u◦v = {x ∈ N | u(v(x)) = 0} = {x ∈ N | v(x) = 0} = Ker v. Let (∏ Mi. Because v is surjective. Im p = Im u◦v = Im u. For any j ∈ I. …. Let p = u◦v : N → N. Let (⊕ Mi. . Also. (σi)i∈I) (respectively (⊕Ni. Since p = u◦v◦u◦v = u◦id◦v = u◦v = p.2). The universality property of the direct product (3. ∀j ∈ I: The homomorphism v is called the direct product of the family of homomorphisms (vi)i∈I and is usually denoted by ∏i∈I vi or ×i∈I vi. produces a unique homomorphism j ∏ M i ⎯⎯→ M j π v vj j ⎯ ∏ N i ⎯⎯→ N j ρ v : ∏ Mi → ∏ Ni such that ρj◦v = vj◦πj. Similarly one defines the direct sum of the family of homomorphisms (vi)i∈I.92 II.15. vj◦πj : ∏ Mi → Nj is a homomorphism. M is isomorphic to a direct summand of N. We define the direct product and the direct sum of the family of homomorphisms (vi)i∈I. since u is injective. {∏i∈I vi}(x) = (vi(xi))i∈I ∈ ∏ Ni. For any j ∈ I.

then ∏i∈I vi and ⊕i∈I vi are injective (surjective). B) is an Abelian group with respect to homomorphism addition). . b) If ui : Ni → Pi (i ∈ I) are homomorphisms. B). 11 We identify xi with its image through the canonical injection σi(xi). define A the (covariant) functor h : R-Mod → Ab: A ∀ B∈ R-Mod.19 Definition.3 Direct sums and products. ∀j ∈ I. …. n}. w is denoted by ⊕in=1 vi or v1⊕…⊕vn.18 Proposition. 3. Let R-Mod be the category of the left R-modules and Ab the category of Abelian groups. If I = {1. 3. Let vi : Mi → Ni (i ∈ I) be a family of module homomorphisms. (The Hom functors) For any A ∈ R-Mod. Then: a) If vi is injective (surjective).II. where xi ∈ Mi and the family (xi)i∈I has finite support11. If x = ∑i∈I xi ∈ ⊕ Mi. ! (note that HomR(A. Exact sequences 93 direct sum (3. j M j ⎯⎯→ ⊕ M i ⎯ σ vj w j N j ⎯⎯→ ⊕ N i τ The homomorphism w is called the direct sum of the family (vi)i∈I and is denoted by (i∈I vi or ⊕i∈I vi. h (B) := HomR(A. then (⊕i∈I vi)(x) = ∑i∈I vi(xi) ∈ ⊕ Ni. then {∏i∈I ui}◦{∏i∈I vi} = ∏i∈I ui◦vi and {⊕i∈I ui}◦{⊕i∈I vi} = ⊕i∈I ui◦vi.5) yields a unique homomorphism w : ⊕ Mi → ⊕ Ni such that w◦σj = τj◦vj.

The “direct proof” uses the form of the elements in the concrete constructions of the direct sum and the direct product and defines a natural homomorphism from HomR(S. -). -) and HomR(-. In a similar manner one defines the contravariant functor hA : R-Mod → Ab. A). B') It is immediate that h (u) is a morphism in Ab. we then show this is an isomorphism. A) to ∏i∈I HomR(Mi. hA(B) := HomR(B. hA(u) : HomR(B'. Then there exist canonical isomorphisms in Ab: HomR(A. ∀g ∈ HomR(B'. A) with respect to direct products and direct sums. Define α : HomR(S. that h (1B) = 1hA(B) A A A and that h (v◦u) = h (v)◦h (u). For any B ∈ R-Mod. A). A) → HomR(B. A). We give two proofs. it satisfies the universality property of the direct product) and applies then 3. B) → HomR(A. for any u : B → B' in R-Mod. So. Modules A ∀u : B → B' morphism in R-Mod. Proof. B). A) is a direct product in Ab of the family (HomR(Mi.3. a “direct proof” and a “category proof”.20 Proposition. A). The functor hA is denoted also by HomR(-. ∏i∈I Mi) ≅ ∏i∈I HomR(A. A) ≅ ∏i∈I HomR(Mi.94 II. Let A be an R-module and let (Mi)i∈I be a family of R-modules. A) is given by hA(u)(g) := g◦u. 3. A A A HomR(⊕i∈I Mi. A) → ∏i∈I HomR(Mi. for any R-module B and any R-module A homomorphisms u : B → B' and ∀v : B' → B". We study the behavior of HomR(A. h is a functor. also denoted by HomR(A. A))i∈I (more precisely. The “category proof” shows that HomR(S. Direct proof. A). We prove the second isomorphism (the first is proposed as an exercise). A) as fol- is defined as h (u)(g) := u◦g. . ∀g ∈ HomR(A. Let S = ⊕i∈I Mi and let σi : Mi → S be the canonical injections. Mi). h (u) : HomR(A.

But τ(x) is unique with this property. with canonical projections hA(σi).3 Direct sums and products. Exact sequences 95 lows: for any homomorphism g : ⊕i∈I Mi → A. ∀x ∈ X. α is surjective. ! Many facts in Algebra have a convenient form in the language of exact sequences (used intensively Homological Algebra. then. there exists a unique morphism τ : X → hA(S) in Ab such that τi = hA(σi)◦τ. τ = ϕ. τi(x) : Mi → A is a morphism in R-Mod. hA(σi) : hA(S) → hA(Mi) is a morphism in Ab. ∀x. ∀i ∈ I. which is a morphism in Ab. We obtain τ : X → HomR(S. Category proof: For any i ∈ I. So. Since τ(x + y) is the only homomorphism with this property. So. Indeed. A). τ(x + y) = τ(x) + τ(y). We show that hA(S) is a direct product of the Abelian groups {hA(Mi)}i∈I. ∀x ∈ X.e. If (gi)i∈I ∈ ∏i∈I HomR(Mi. ∀x ∈ X. for instance). If ϕ : X → hA(S) is a homomorphism with τi = hA(σi)◦ϕ. A). By the universality property of the direct sum. with xi ∈ Mi. A) = hA(S). the universality property of the direct sum S = ⊕i∈I Mi (definition 3. τ(x) + τ(y) : S → A is an R-homomorphism satisfying (τ(x) + τ(y))◦σi = τ(x)◦σi + τ(y)◦σi = τi(x) + τi(y) = τi(x + y).II. This means that we have to prove that. It is routine to check that α is a homomorphism. i ∈ I. . so g = 0. then g(∑i∈I σi(xi)) = ∑i∈I g(σi(xi)) = 0. If α(g) = (g◦σi)i∈I = (0)i∈I. i.5) supplies a unique homomorphism g : S → A such that g◦σi = gi. ∃! τ(x) : S → A homomorphism such that τi(x) = τ(x)◦σi. α(g) =: (g◦σi)i∈I ∈ ∏i∈I HomR(Mi. ϕ(x) : S → A is a homomorphism with τi(x) = hA(σi(ϕ(x)) = ϕ(x)◦σi. x & τ(x). for any X ∈ Ab and any morphisms τi : X → hA(Mi) in Ab. y ∈ X. for any family having finite support (xi)i∈I. so τ(x) = ϕ(x). This proves that α is injective.

Thus. The sequence (S) is called semiexact (respectively exact) if it is semiexact (respectively exact) at any term.21 Definition. a) If ϕ : E → F is a homomorphism.96 II. b) If A ≤ R B. we have an exact sequence: 0 → Kerϕ ⎯ι E ⎯ϕ F ⎯ ⎯→ ⎯→ ⎯→ Cokerϕ → 0 . then the following sequence is exact: 0 → Kerϕ ⎯ι E ⎯ϕ F ⎯ ⎯→ ⎯→ ⎯→ F Im ϕ → 0 The inclusion homomorphism is denoted by ι and the canonical surjection by π. the homomorphisms that are uniquely determined (or clear from the context) are not marked on the arrows. The module F/Imϕ is called the cokernel of ϕ and is denoted by Cokerϕ. 3. then the sequence 12 π π Called in the sequel sequence of modules. (S): … ⎯ E ⎯u F ⎯v G ⎯ The sequence (S) is called semiexact at F if v◦u = 0.23 Examples. For instance. b) A sequence of the form 0 → E ⎯u F is exact if and only if u ⎯→ is a monomorphism.22 Remarks. A semiexact sequence of modules is called a complex of modules. Modules 3. write 0 → L instead of 0 ⎯0 L . ⎯→ c) A sequence of the form E ⎯u F → 0 is exact if and only if u is an epimorphism. The sequence (S) is called exact at F if Im u = Ker v. a) Often. d) The sequence 0 → E ⎯u F → 0 is exact ⇔ u is an isomor⎯→ phism. 3. Consider a (finite or infinite) sequence of R-modules and module homomorphisms:12 ⎯→ ⎯→ ⎯→ ⎯→ …. Ker u = 0 ⇔ Ker u = Im 0. Indeed. This may be rephrased as Im u ⊆ Ker v. since the only homo⎯→ morphism defined on the module 0 is the zero homomorphism. .

A homomorphism g that makes the diagram commutative is called a homomorphism of extensions.3 Direct sums and products. In other words.25 Example. ιA : A → A ⊕ C is the canonical injection and πC : A ⊕ C → C is the canonical projection (recall that A ⊕ C = A × C). Let A and C be two modules and consider two extensions of A by C: 0 → A ⎯u B ⎯v C → 0 . . Example b) above).II. It is natural in this sense to consider the following definition: 3. If A and C are two R-modules. Exact sequences 97 0 → A ⎯ι B ⎯π B A → 0 ⎯→ ⎯→ is exact. For such a sequence. the diagram below is commutative: ⎯→ B ⎯v ⎯→ C → 0 0 → A ⎯u g ↓ ↓ ↓ ′ ′ 0 → A ⎯u ⎯→ B′ ⎯v ⎯→ C → 0 (the vertical unmarked homomorphisms are identity homomorphisms). 3.24 Definition. An exact sequence of the form 0 → A ⎯u B ⎯v C → 0 ⎯→ ⎯→ is called a short exact sequence. A is isomorphic to the submodule Im u = Ker v of B. 3.26 Definition. and B/Im u = B/Ker v is isomorphic to C. ⎯→ ⎯→ ′ ′ 0 → A ⎯u B′ ⎯v C → 0 . then the sequence πC 0 → A ⎯ι A A ⊕ C ⎯⎯→ C → 0 ⎯→ is short exact. This is the reason B is also called an extension of A by C (cf. ⎯→ ⎯→ We say that the extensions are equivalent if there exists a homomorphism g : B → B' such that g◦u = u' and v'◦g = v. The problem of finding all extensions of A by C is highly nontrivial. So A ⊕ C is an extension of A by C.

gi : Ei → Fi. (gi )i∈Z is also a homomorphism of complexes. ∀i ∈ Z. Let b ∈ B with g(b) = 0.27 Proposition. Modules 3. ∀i ∈ Z.98 II. ⎯ ⎯→ A homomorphism of complexes from E to F is a sequence of module homomorphisms g := (gi)i∈Z. Thus there exists a ∈ A with u(a) = b. which means g(b) − b' ∈ Ker v' = Im u'. since u' is injective. for some b ∈ B. Then v'g(b) = 0 = v(b). since u' = gu. Let b' ∈ B'. a = 0.28 Definition. v'(g(b)) = v'(b'). such that gi + 1◦ui = vi◦gi. 13 . Since v'g(b) = v(b). Consider two complexes: i −1 E: … → Ei −1 ⎯u⎯ Ei ⎯ui Ei +1 → … ⎯→ ⎯→ F: … → Fi −1 ⎯vi −1 → Fi ⎯vi Fi +1 → … .26. The complexes E and F are called isomorphic if there exists an isomorphism from E to F. so b ∈ Ker v = Im u. In the definition 3.26 is an equivalence relation on the class of all extensions of A by C. we obtain 0 = g(b) = gu(a) = u'(a). ! 3.13 Let us show that Ker g = 0. Thus b' = g(b − u(a)) ∈ Im g. then v'(b') ∈ C and the surjectivity of v implies that. g is an isomorphism. the following diagram commutes: i −1 ⎯→ Ei ⎯ui ⎯→ Ei +1 → … … → Ei −1 ⎯u⎯ ↓ gi −1 ↓ gi ↓ gi +1 vi −1 vi … → Fi −1 ⎯⎯→ Fi ⎯ ⎯→ Fi +1 → … The homomorphism g is called an isomorphism if gi is an isomor−1 phism. In other words. The technique used is called “diagram chasing” and it is used extensively in arguments involving diagrams of module homomorphisms. g(b) − b' = u'(a) = gu(a). v(b) = v'(b'). the relation defined at 3. There exists a ∈ A with u'(a) = g(b) − b'. so b = u(a) = 0. In particular. Proof. In this case.

d) There exists a homomorphism v' : C → B such that v◦v' = idC. Exact sequences 99 3.II. The homomorphisms that compose this isomorphism are u0 : A → Im u. ιC are the canonical injections of the direct sum A ⊕ C and πA.29 Examples.. ∀a ∈ A. ⎯→ has the remarkable property that by reversing the arrows we obtain another exact sequence: 0 ← A ←π A A ⊕ C ←ιC C ← 0 . The exten⎯⎯ ⎯⎯ sions of A by C that are isomorphic to the extension direct sum A ⊕ C are characterized in the next proposition: 3. the “direct sum” extension. B is an extension of A by C). Checking the commutativity of the diagram is left to the reader.30. where ⎯→ ιA.3 Direct sums and products. The following statements are equivalent: a) B is isomorphic (as an extension of A by C) with πC 0 → A ⎯ι A A ⊕ C ⎯⎯→ C → 0 . πC are the canonical projections. . c) There exists a homomorphism u' : B → A such that u'◦u = idA. b) Im u = Ker v is a direct summand in B. Proof. a) The exact sequence 0 → A ⎯u B ⎯v C → 0 ⎯→ ⎯→ is isomorphic (in the sense given by the previous definition) to the sequence 0 → Im u ⎯ι B ⎯π B Im u → 0 . in order to define w : C → B/Im u. v0(b + Ker v) = v(b). πC 0 → A ⎯ι A A ⊕ C ⎯⎯→ C → 0 . c)⇒b) and d)⇒b) follow from 3. Proposition. the fundamental isomorphism theo−1 rem). The extension “direct sum” of A by C.e. observe that the surjective homomorphism v induces a canonical isomorphism v0 : B/Ker v → C.16. Let (S): 0 → A ⎯u B ⎯v C → 0 be a short ⎯→ ⎯→ exact sequence of modules (i. Set w = v0 . ∀b ∈ B (cf. idB : B → B. u0(a) = u(a). where ι is the inclusion ⎯→ ⎯→ and π the canonical surjection. b) Any extension of Z2 by Z2 is isomorphic either to Z2⊕Z2 or to Z4.

14). Defining −1 −1 v' : C → B. ⎯→ A monomorphism 0 → A ⎯u B is called split if the short exact u sequence 0 → A ⎯ B → B Im u → 0 splits (⇔ Im u is a direct ⎯→ summand in B). thus. “the sequence ⎯→ ⎯→ splits”) if it satisfies the equivalent conditions of the previous proposition.31 Definition. then b + b' = u(a) + d + u(a') + d' = u(a + a') + d + d'. A short exact sequence u v 0 → A ⎯ B ⎯ C → 0 is called split (or we say. Indeed. d' ∈ D. so d) holds. v' = ϕ ιC. Modules a)⇒c). b)⇒a) Let B = Im u ⊕ D. In the same way is shown that α preserves multiplication with scalars. b' = u(a') + d' ∈ B. Note that α◦u : A → A is idA. Define u' : B → A. Define then α(b) = a ∈ A. This proves c). we have vv' = vϕ ιC = πCιC = idC. ! 3. d. We have u'u = πAϕ u = πAιA = idA (see Remark 3. ∀b ∈ B. We obtain a module homomorphism α : B → A. with a. d) Let ϕ : B → A ⊕ C be an isomorphism such that the diagram ⎯→ B ⎯v ⎯→ C → 0 0 → A ⎯u ϕ ↓ ↓ ↓ πC 0 → A ⎯ι A ⎯→ A ⊕ C ⎯⎯→ C → 0 commutes. ⎯→ An epimorphism B ⎯v C → 0 is called split if the short exact sequence 0 → Ker v → B ⎯v C → 0 splits (⇔ Ker v is a direct sum⎯→ mand in B).100 II. if b = u(a) + d. ϕ is a homomorphism (and isomorphism) of extensions. Consider ϕ : B → A⊕C. We have ϕu = (ιAα + ιCv)◦u = ιAαu + ιCvu = ιA πCϕ = πC (ιAα + ιCv) = πCιAα + πCιCv = v. ∃! a ∈ A and d ∈ D such that b = u(a) + d. u' = πAϕ. so. a' ∈ A. α(b + b') = a + a' = α(b) + α(b'). . ϕ = ιAα + ιCv. So.

M So. ϕ : M → A is a homomorphM ism) such that h (u)(ϕ) = u◦ϕ = 0. Also. Then u'◦ψ : M → B is well defined M and it is a homomorphism. The next proposition says “the functor Hom is left exact”: 3. ∀a ∈ A) is an isomorphism. We have to prove that h (u) is monomorphism and M M M Im h (u) = Ker h (v). then v◦ψ = 0. The M reverse inclusion also holds: if ψ ∈ h (B) (i. so u0 : A → Im u (u0(a) = u(a). Let R M be an R-module and let 0 → A ⎯u B ⎯v C → 0 ⎯→ ⎯→ be a short exact sequence. so Imψ ⊆ Ker v = Im u.e. But u is injective. M M M M M M Since h (v)◦h (u) = h (v◦u) = h (0) = 0. M Proof. consider the exact sequence: 0 → A ⎯u B ⎯ B Im u → 0 ⎯→ ⎯→ M M and write that h is left exact to obtain that h (u) is monomorphism. ψ : M → B) is in M Ker h (v). h (u)(u'◦ψ) = u◦u'◦ψ = ψ. Since u is monomorphism.33 Proposition.! The contravariant functor hM is also left exact.II. In particular. Let R M be a module and let 0 → A ⎯u B ⎯v C → 0 ⎯→ ⎯→ be a short exact sequence. We invite the reader to prove: 3. if u : A → B is an R-module monomorM phism. Exact sequences 101 The behavior with respect to short exact sequences is primordial in the study of functors defined on categories of modules. Ker h (u) = 0. Then the sequence 0 → hM (C ) ⎯hM ⎯→ hM (B ) ⎯hM ⎯→ hM ( A) ⎯(v ) ⎯(u ) M M . let u' : Im u → A be the inverse of u0. If u is monomorphism. then h (u) is an Abelian group monomorphism. Let ϕ ∈ h (A) (i. Im h (u) ⊆ Ker h (v). Then the following sequence 0 → h M ( A) ⎯h ⎯→ h M (B ) ⎯h ⎯→ h M (C ) ⎯(u ) ⎯(v ) is exact in Ab.3 Direct sums and products.32 Proposition. ϕ = 0.e.

c) If A = ∏s∈S As. if and only if for any family (xi)i∈I. then the inclusion is in fact an equality. Prove that he inclusion may be strict. then hM(v) is a monomorphism. the same applies to IAs). Prove that there exists a (canonical) isomorphism A⊕B ≅ B⊕A. 4. Let A and B be R-modules. Generalization. ∀i ∈ I. with canonical projections (πi)i∈I. …. ∀i ∈ I. Modules is exact in Ab. 1. with xi ∈ Mi. if and only if for any x ∈ S. a ∈ A}>. such that x = ∑i∈I σi(xi). 5. In particular. ∀i ∈ I. For any module R A. …. a) Prove that IA = {i1a1 + … + inan | n ∈ N. If I is finitely generated. in ∈ I. prove that the canonical projections are surjective. Prove that the R-module S is a direct sum of the family of R-modules (Mi)i∈I. with canonical injections (σi)i∈I. with xi ∈ Mi. there exists a unique family of elements (xi)i∈I.102 II. a1. 2. Taking the universality property of the direct product as its definition. then IA ⊆ ∏s∈S IAs (we identify As with its image in ∏s∈S As. ! Exercises In the exercises. b) If A = ⊕s∈S As (internal direct sum). then IA = ⊕s∈S IAs. there exists a unique x ∈ P such that πi(x) = xi. . an ∈ A}. having finite support. define the submodule of A: IA := <{ia | i ∈ I. if v : B → C is an R-module epimorphism. Prove that the R-module P is a direct product of family of R-modules (Mi)i∈I. R denotes a fixed ring with identity. Let R be a ring and let I be a left ideal of R. i1. 3.

but v does not split. γ and u' are monomorphisms. any R-monomorphism ψ : R → M splits? 9. then β is an epimorphism. Let R be a PID. v such that uv splits. then α is a monomorphism. 11. that: a) If α. . Give an example of monomorphisms u. then β is monomorphism. Prove that the R-module M is cyclic (generated by one element) ⇔ there exists an exact sequence of the form 0 → R → R → M → 0. then γ is an epimorphism. v are epimorphisms. 10.3 Direct sums and products. b) If α. γ are monomorphisms. c) If β is a monomorphism and α. d) If β is an epimorphism and u'. Then any R-epimorphism ϕ : M → R splits. 8. ⎯→ ⎯→ ⎯→ 7. w such that the following sequence is exact: 0 → Z3 ⎯u Z9 ⎯v Z9 ⎯w Z3 → 0.II. Show that any composition of split monomorphisms (epimorphisms) is a split monomorphism (epimorphism). Exact sequences 103 6. Give an example of an R-module M with a direct summand S in M that has an infinity of complements. Is it true that for any ring R and for any R-module M. Let R be a ring and let M be an R-module. v. using diagram chasing. Find homomorphisms u. Suppose that the following diagram of R-modules commutes and has exact rows: A ⎯u ⎯→ B ⎯v ⎯→ C ↓α ↓β ↓γ u′ v′ ⎯→ B′ ⎯ ⎯→ C ′ A′ ⎯ Prove. γ and v are epimorphisms.

…. On the other hand. xn. 4. xn be a family of elements of M. . …. The existence of a basis is not guaranteed for a module. in studying linear independence we may suppose that x1. ….2. …. xn. R is a ring with identity and all modules are left R-modules. not all zero.4 Free modules Many nice properties of the vector spaces are consequences of the fact that any vector space has a basis. A family of elements in M that is not linearly independent is called linearly dependent. It is natural to define and to study the class of modules that “have a basis”. This is the reason we can speak about a linearly dependent (finite) subset of M. rn ∈ R. rx = 0 implies r = 0. and this increases considerably the difficulty in the study of modules in comparison to the vector spaces. …. r1x1 + … + rnxn = 0 implies r1 = … = rn = 0. A relation of the form r1x1 + … + rnxn = 0. rn ∈ R. In this section. any linear combination of x1. n ∈ N and let x1.104 II. xn are distinct. is called a relation of linear dependence of the family x1. a) If there exists i ≠ j with xi = xj. Let M be an R-module. …. with r1. the notion of linear independence does not depend on the indexing of x1. xn is linearly independent (over R)14 if. Modules II. Thus. Remarks. for any r1. This suggest the following definition: the annihilator of x in R is 14 * The reference to the ring R is often omitted. Definition. In other words.1. …. We say that the family x1. b) The set {x} (containing a single element x ∈ M ) is linear independent if and only if ∀r ∈ R. then the family x1. …. …. xn is linearly dependent: the linear combination xi − xj is 0. 4. xn is 0 if and only if all its coefficients are 0.

Since AnnR(x) is exactly the kernel of the R-module homomorphism ρx : R → M. a) The empty subset of M is linearly independent.3. we say that X is linearly dependent over R. ∑i∈I rixi = 0 implies ri = 0. AnnR(x) is a left ideal in R and (by the isomorphism theorem) the submodule Rx = Imρx is isomorphic as a left R-module to R/AnnR(x). ∀i ∈ I. The family {xi}i∈I is linearly independent if and only if. such that sr = 0).5. More generally. It follows that the sum of the family of submodules {Rxi}i∈I is direct (see 3. Thus. 4. for any family {ri}i∈I of elements of R. 4. . s ≠ 0. ρx(r) = rx. {r} is linearly dependent ⇔ r is a right zero divisor in R (∃s ∈ R. Definition. ∀r ∈ R. Remarks.II. Summarizing: {x} is linearly independent ⇔ AnnR(x) = 0 ⇔ ρx is an isomorphism from R to Rx.∀i ∈ I (prove this!). ∀r ∈ R. If X is not linearly independent. b) Let {xi}i∈I be a family of elements of the R-module M. In fact. c) Every subset of an linearly independent set is linearly independent. {xi}i∈I is linearly independent if and only if the sum of the family of submodules {Rxi}i∈I is direct and AnnR(xi) = 0.4. X is linearly dependent if and only if there exists a finite linearly dependent subset of X. with finite support. a) The set {1}. containing only the unity of the ring R (seen as a left R-module) is linearly independent. Examples. 4. Let X be a subset of R M.4 Free modules 105 AnnR(x) := {r ∈ R | rx = 0} (sometimes denoted by lR(x)). We say that X is linearly independent over R (or free over R) if any finite subset of X is linearly independent over R in the sense of the remark above.11).

…. if any element in M is written uniquely as a linear combination of elements of B. A subset B of an R-module M is called a basis of M if it is simultaneously linearly independent and a system of generators for M. such that x = ∑i∈I riei. 3x = 0 . we obtain that B is linearly independent. as seen at example a). Let B = {ei}i∈I be a basis of R M and let x ∈ M. 4. in the R-module R[X] the set {X | n ∈ N} is linearly independent. If {si}i∈I is another family of elements in R. ∀i ∈ I. 4. Can you generalize this? Of course. with x = ∑i∈I siei. if {fn | n ∈ N} is a family of polynomials such that deg fn ≠ deg fm if m ≠ n. then the family {fn | n ∈ N} is linearly independent. Writing the fact that 0 has a unique writing as a linear combination of elements of B. Since B generates M. Definition. This amounts to saying that a polynomial a0 + n a1X + … + an X (a0. a1. The linear independence of B implies si = ri. an ∈ R) is 0 if and only if a0 = … = an = 0. for x ∈ M. Conversely.8. {r} is a basis of R R if and only if r is right invertible in R. then B is a system of generators of M. ! With the notations above. Proof. Then: B is a basis of M if and only if any x ∈ M is written uniquely as a linear combination of elements of B. 4. a) {∅} is a basis (the only one!) of the module {0}. n c) If R is a domain. the set { 1 } is linearly independent over Z3. More generally. the elements {ri}i∈I in R with x = ∑i∈I riei are called the coordinates of x in the basis B = {ei}i∈I.7 Proposition. Modules b) The Z-module Z3 has no Z-linearly independent subsets: for any ˆ ˆ x ∈ Z3. More generally. then ∑i∈I (si − ri)ei = 0.106 II. b) {1} is a basis of R R. The module M is called a free module if it has a basis. there exists a family {ri}i∈I of elements in R with finite support. . Let M be an R-module and let B be a subset of M. Examples.6.

0. If I = {1. Proof. we have to prove that . (I) with canonical injections σi : R → R ) is free. But. A module homomorphism is determined by its values on a generating set. where M is an R-module and u : B → M is a map. c) v is an isomorphism ⇔ u(B) is a basis. there exists a unique family (ri)i∈I of elements in R such that x = ∑i∈I riei. w ∈ HomR(L. The same thing is obtained for w. (The universality property of the free module of basis B) Let L be a free R-module of basis B and let i : B → L be the inclusion map.II. there exists a unique homomorphism v : L → M with v◦i = u. for an arbitrary generating set. this can be formulated: For any module M and any family (yi)i∈I of elements of M. en = (0. a basis being {ei}i∈I.…. 0. there may be no homomorphism that takes prescribed values on the elements of the generating set. 1. ∀i ∈ I. b) v is surjective ⇔ u(B) is a system of generators for M. u). We have v{∑i∈I riei} = ∑i∈I riv(ei) = ∑i∈I riyi. i) has the following universality property: For any pair (M. for any choice of the values on the elements of a basis. …. we have: a) v is injective ⇔ u(B) is linearly independent. a unique homomorphism takes the respective values on the elements of the basis: 4. the R-module R (the direct sum of | I | copies of R. The pair (L. 0). Let x ∈ L. …. the canonical basis of the free R-module R are e1 = (1.…. This basis is called the canonical basis of R . since (ei)i∈I is a basis. ∀i ∈ I. Uniqueness claim: suppose v. n}. If B = (ei)i∈I. 1).…. 0). R is also called the free R-module on the set I (or the free R-module of ban sis I ). Moreover.9 Proposition. M) satisfy v(ei) = yi = w(ei). there exists a unique homomorphism v : L → M such that v(ei) = yi. In the privileged case of free modules. For the existence claim. (I) (I) where ei = σi(1).4 Free modules (I) 107 c) If I is a set. e2 = (0.

it follows that u is surjective. Let x = ∑i∈I riei and y = ∑i∈I siei be elements in L. for any x ∈ L there exists a unique family (ri)i∈I of elements in R such that x = ∑i∈I riei. b ∈ R. By the result above.11 Proposition. ! The R-modules of the type R (I) = ∑i∈I ariyi + ∑i∈I bsiyi = av(x) + bv(y). Then R L ≅ R R . which generates R . If x = ∑i∈I rixi ∈ Ker u (with (ri)i∈I family of elements in R).10. Let (Mi)i∈I be a family of free R-modules. Since {ei}i∈I is a basis. Since. ∑i∈I riyi = 0}. ∀i ∈ I. M is the (internal) direct sum . Proof. The rest of the proof is left to the reader. v is well defined. Then the (external) direct sum ⊕i∈I Mi is a free R-module. Since Im u includes {ei}i∈I. with (ri)i∈I. By 3. We show that B := ∪i∈I σi(Bi) is a basis of M. so x = 0. It is clear that Ker v = 0 ⇔ (yi)i∈I is linearly independent. then 0 = u(x) = ∑i∈I riei. For any a. having a finite support.108 II. defines a module homomorphism. (I) Proof. ! A direct sum of free modules is also free: 4. (si)i∈I families of elements in R. for any x = ∑i∈I riei with (ri)i∈I family of elements in R. we have: v(ax + by) = v{∑i∈I ariei + ∑i∈I bsiei} = v{∑i∈I (ari + bsi)ei} = ∑i∈I (ari + bsi)yi = We prove a): Ker v = {∑i∈I riei | (ri)i∈I ∈ R . Let M := ⊕i∈I Mi. ri = 0. let σj : Mj → M be the canonical injection and let Bj be a basis of Mj. (I) ∀i ∈ I. For any j ∈ I.10 Proposition. (I) there exists a unique homomorphism u : L → R with u(xi) := ei. (I) are “all” free R-modules: 4. Modules v(x) := ∑i∈I riyi. Let L be a free R-module and let (xi)i∈I be a basis (I) of L. Let (ei)i∈I be the canonical basis in R .

If M is an R-module. if M is an R-module and S is a system of generators of M. ∀k ∈ I. then ∑t∈Tk rktykt = −∑i∈I \{k}∑t∈Ti rityit. Hence M ≅ R /Kerϕ. More precisely. Every module is (isomorphic to) a factor module of a free module.9). we obtain Imϕ = M. ∀t ∈ Tk. We have now the exact sequence .4 Free modules 109 of the family of submodules (Im σi)i∈I. Let (es)s∈S be the canonical basis in R . and S generates M. the structure of an arbitrary module is known. ∀i ∈ I. then there exists an exact sequence: … → Ln+1 → Ln → … → L1 → L0 → M → 0. Define ϕ : R → M by ϕ(es) = s. let L0 be free and let u0 : L0 → M be an epimorphism (as given by the Proposition above). If {rit | i ∈ I. ∑i∈I∑t∈Ti rityit = 0. ∀n ∈ N. ! This simple fact is very important: if one knows the structure of the submodules and of the factor modules of free modules. But (ykt)t∈Tk is a basis in Mk. (S) (S) Proof. t ∈ Ti} is a finite support family elements of R. so ∑t∈Tk rktykt = 0.II. But Mk ∩{∑i∈I \{k} Mi} = 0.12 Proposition. so rkt = 0. ∀s ∈ S (see 4. then there exists an epimorphism ϕ : ( S) (S) R → M. Such a sequence is called a free resolution of M.13 Example. Since σi(Bi) generates Im σi. This method will be used to study the structure of finitely generated modules over a PID. where Ln is free. ∪i∈I σi(Bi) generates M. Since the submodule Imϕ includes S. Then we have an exact sequence: 0 → Ker u0 ⎯i L0 ⎯u0 M → 0 . For the proof. ∀k ∈ I.! with 4. Here is another application of this principle: 4. Apply again the proposi⎯→ ⎯→ tion for Ker u0 and obtain a free module L1 and an epimorphism u1 : L1 → Ker u0. The linear independence of B is easy to show (only the notations are complicated): let Bi = (eit)t∈Ti the basis of in Mi and let yit = σi(eit).

110

II. Modules

)u1 0 → Ker u1 → L1 ⎯i⎯→ L0 ⎯u0 M → 0 ⎯→ since Ker u0 = Im i◦u1. One continues by induction on n such that exists an exact sequence of the form: 0 → Kn → Ln → … → L1 → L0 → M → 0, with Li free, 1 ≤ i ≤ n.

In a vector space, any two bases have the same cardinal. In the case of a free module over an arbitrary ring, this fact is not guaranteed. Nevertheless, the free modules over a commutative ring have this property; in fact, the proof of this result reduces the problem to the case of vector spaces.
4.14 Proposition. Let R be a commutative ring and let M be a free R-module. Then any two bases of M have the same cardinal. Proof. Let us try to connect the R-module M to a vector space over a field. Recall that, if I is a maximal ideal in R, then R/I is a field. A maximal ideal I of R exists (see 1.20). Then R/I is a field (see Appendix, Prime and maximal ideals). The module M has no natural structure of R/I-vector space: the natural “multiplication” given by (r + I)x = rx, for r ∈ R, x ∈ M, is not well defined unless ix = 0, ∀i ∈ I. This is the reason we need to “kill off” the elements x of M for which ix ≠ 0 for some i ∈ I, as follows: Let IM := {a1x1 + … + anxn | n ∈ N, ai ∈ I, xi ∈ M, ∀i = 1, …, n}. A straightforward check shows that IM is an R-submodule of M and that the factor module M/IM is also an R/I-module (i.e. an R/I-vector space!) with respect to the external operation: (r + I)(x + IM) := rx + IM, ∀r ∈ R, ∀x ∈ M. Let π : M → M/IM be the canonical surjection, π(x) = x + IM, ∀x ∈ M. We prove that: if (eα)α ∈ A is a basis of R M, then (π(eα))α ∈ A is a basis in the R/I-vector space M/IM.

II.4 Free modules

111

Indeed, (π(eα))α ∈ A generates the R-module M/IM, since π is surjective and (eα)α ∈ A generates M. It easily seen that it is also generates M/IM as an R/I-module. Let us show that (π(eα))α ∈ A is R/I-linear independent. Let (rα + I)α ∈ A be a family of elements of R/I (where rα ∈ R, ∀α ∈ A), of finite support, such that ∑α∈A (rα + I)π(eα ) = 0 + IM. This means: ∑α∈A rα eα ∈ IM. Note that, for any x ∈ IM, x = a1x1 + … + amxm ∈ IM, for some ai ∈ I, xi ∈ M, ∀i = 1, …, m. Writing every xi as a linear combination of (eα)α ∈ A, it follows that x is of the form ∑α∈A bα eα , where bα ∈ I, ∀α ∈ A (since ai ∈ I and I is an ideal). Thus, ∑α∈A rα eα = ∑α∈A bα eα But (eα)α ∈ A is a basis, so rα = bα ∈ I, ∀α ∈ A. Thus, rα + I = bα + I = 0 + I, ∀α ∈ A. Now, taking two bases (eα)α ∈ A and (vβ)β ∈ B of R M, we obtain that (π(eα))α ∈ A and (π(vβ))β ∈ B are bases in the R/I-vector space M/IM. Since two bases of a vector space have the same cardinal, we obtain |A| = |B|. !
4.15 Definition. Let R be a commutative ring and let L be a free R-module. The cardinal of a basis of R L is called the rank of L, denoted rankR L (or, simply, rank L). The previous result ensures that the definition is independent of a choice of the basis in R L. If R is a field, the rank of a vector space is the same as its dimension.

We study now the homomorphisms between free modules of finite rank. The idea is to mimic the vector space situation: for two fixed bases in the modules, we associate a matrix to a homomorphism. In this manner, the operations with (and the properties of) module homomorphisms translate in matrix language (and conversely). We

112

II. Modules

suppose that the ring R is commutative, although some results hold for any ring with identity.
4.16 Definition. Let E and F be free R-modules and let ϕ : E → F be a homomorphism. Fix e = (e1, …, em) an ordered basis in E and f = (f1, …, fn) an ordered basis15 in F. For any i ∈ {1, …, m}, there exist and are unique aij ∈ R, ( j ∈ {1, …, n}) such that ϕ(ei) = ai1 f1 + … + ain fn. The matrix (aij)1≤i≤m, 1≤j≤n ∈ Mm, n(R) is denoted by Me, f(ϕ) and it is called the matrix of the homomorphism ϕ (in the pair of bases (e, f)). In other words, the i-th row of Me, f(ϕ) is made up of the coordinates ai1, …, ain of ϕ(ei) in the basis f. If E = F, one usually takes e = f and Me, f(ϕ) is denoted simply Me(ϕ). The importance of ordering the bases is quite clear now: a permutation of the basis (e1, …, em) leads to another matrix of ϕ, whose rows are a permutation of the rows of the initial matrix. 4.17 Remark. This is the “algebraic” convention of writing the matrix Me, f(ϕ). Of course, one can agree to the “geometric convention”: write the coordinates of ϕ(ei) in the basis f on the i-th column (which t means considering the transpose matrix Me,f(ϕ)). Choosing this rule t leads to changing property b) in the next sentence into: Me, f(ψ ◦ϕ ) = t t Mf, g(ψ )· Me, f(ϕ ). 4.18 Proposition. a) The mapping: Me, f : HomR(E, F) → Mm, n(R), ϕ & Me, f(ϕ) is an R-module isomorphism.
15

We consider the bases as being totally ordered (the place of the element in the

basis matters).

II.4 Free modules

113

b) If G is a free R-module, g = (g1,…, gp) is a basis in G, and ψ:F→G is an R-module homomorphism, then Me, g(ψ ◦ϕ) = Me, f(ϕ)·Mf, g(ψ). c) The mapping Me : EndR(E) → Mm(R) is an R-algebra antiop isomorphism (i.e. Me : EndR(E) → Mm(R) is an R-algebra isomorphism). Proof. a) Let η : E → F be another homomorphism and let (bij)1≤i≤m, 1≤j≤n = Me, f(η). Then (ϕ +η)(ei) = ϕ(ei) +η(ei) = ∑j aij fj + ∑j bij fj = ∑j (aij + bij) fj So, Me, f(ϕ +η) = (aij + bij)1≤j≤n, 1≤i≤m = Me, f(ϕ) + Me, f(η), so Me, f is an Abelian group homomorphism. If Me, f(ϕ) = 0 (the matrix 0 ∈ Mm, n(R)), then ϕ(ei) = 0, 1≤ i ≤ m. Since (ei)1≤i≤m is a basis in E, ϕ = 0; this shows that Me, f is injective. For the surjectivity, let A = (aij) ∈ Mm, n(R). The universality property of the free module yields a unique homomorphism ϕ ∈ HomR(E, F) such that ϕ(ei) = ai1 f1 + … + ain fn, i.e. Me, f(ϕ) = A. If r ∈ R and ϕ ∈ HomR(E, F), a simple computation shows that Me, f(rϕ) = rMe, f(ϕ). b) Let Mf, g(ψ) = (bjk)1≤j≤n, 1≤k≤p. We have (ψ ◦ϕ)(ei) = ψ{ ∑j aij fj } = ∑j aijψ( fj ) = ∑j aij∑k bjk gk = ∑k { ∑j aijbjk} gk.

So, Me, g(ψ ◦ϕ) = Me, f(ϕ)·Mf, g(ψ). c) The statement follows from a) and b): a) implies that Me is an R-module homomorphism and b) implies that op Me : EndR(E) → Mm(R) is a ring homomorphism. !

In the setting above, if e = (e1,…, en) and f = (f1,…, fn) are ordered bases of the free module E, then every element of the basis f is written uniquely as a linear combination of ei's:
f i = ∑ sij e j , sij ∈ R
j =1 n

We obtain the basis change matrix Te, f = (sij) ∈ Mn(R). If g is another basis in E, then a straightforward computation yields: Te, g = Tf, g·Te, f

114

II. Modules

In particular, In = Te, e = Te, f ·Tf, e, i.e. the basis change matrix is invertible in Mn(R). When is a matrix in Mn(R) invertible? If A ∈ Mn(R), the adjoint matrix of A is A* (the entry (i, j) of A* is i+j (−1) det(Aji), where Aji is the matrix in Mn−1(R) obtained by suppressing the row j and the column i of the matrix A). Then A·A* = A*·A = det(A)·In. This shows that, if det(A) ∈ U(R), then A is −1 −1 invertible in Mn(R) and A = (det A) A*. Conversely, if A is invert−1 −1 ible, A·A = In implies (taking determinants) that det(A·A ) = −1 det(A)det(A ) = 1, so det(A) ∈ U(R). Thus, we have:
4.19 Proposition. Let R be a commutative ring and let n ∈ N*. a) A matrix S ∈ Mn(R) is invertible in Mn(R) if and only if det(S) is invertible in R.

b) Let E be a free R-module of rank n and let e = {e1,…, en}, f, g be bases in E. Then: Te, g = Tf, g·Te, f In particular, any basis change matrix Te, f is invertible. Conversely, if S = (sij) is invertible in Mn(R) and f i = then {f1,…, fn} is a basis in E.

∑ sij e j , 1 ≤ i ≤ n,
j =1

n

!

The facts not already proven in the proposition above have proofs that can be taken word for word from the case of the vector spaces. The multiplicative group of the n×n invertible matrices with entries in R, U(Mn(R)) = {S ∈ Mn(R) | ∃T ∈ Mn(R) such that ST = TS = I} = {S ∈ Mn(R) | det S ∈ U(R)} is also denoted GL(n, R) or GLn(R) and is called the linear general group of degree n over R. The following formalism is useful in many situations involving vectors and matrices. Let R be a commutative ring and let E be an R-module. If m, n, p are positive integers, let Mn, p(E) denote the set of n×p matrices with entries in E. It is evident that Mn, p(E) is an Abelian

II.4 Free modules

115

group with respect to usual matrix addition. For any A = (aij) ∈ Mm, n(R) and any X = (xjk) ∈ Mn, p(E), define the product AX ∈ Mm, p(E) by
AX = (yik) ∈ Mm, p(E), yik = E.

∑ aij x jk , ∀i ∈ {1, …, m}, ∀k ∈ {1, …, p},
j =1

n

Of course, aijxjk is the product given by the R-module structure of We obtain an “external operation” · : Mm, n(R)×Mn, p(E) → Mm, p(E), with the following properties (the proof of which is straightforward): (A + B)X = AX + BX, A(X + Y) = AX + AY, ∀ A, B ∈ Mm, n(R), ∀ X, Y ∈ Mn, p(E). * Moreover, if q ∈ N , then for any A ∈ Mq, m(R), B ∈ Mm, n(R), X ∈ Mn, p(E) we have: (AB)X = A(BX). If I is the identity matrix in Mn(R), then IX = X, ∀X ∈ Mn, p(E). Thus, Mn, p(E) is a left module over the ring Mn(R). As an application, let us see how the matrix of a homomorphism of free modules changes when we change bases in the modules.
4.20 Proposition. Let E and F be free R-modules, rank E = n, rank F = m and let e = (e1,…, en), e' = (e'1,…, e'n) be bases in E, ′ f = { f1,…, fm}, f' = ( f1′,… f m ) bases in F. Let S = (sij) ∈ Mn(R) be the basis change matrix from e to e' and let T = (tij) ∈ Mm(R) be the basis change matrix from f to f'. If ϕ : E → F is a homomorphism, then: −1 Me, f(ϕ) = S·Me', f'(ϕ)·T . Proof. Let Me, f(ϕ) = A = (aij) ∈ Mn, m(R), i.e.: ϕ(ei) = ∑j aij fj, ∀i ∈ {1, …, n}.
t

We view e = (e1, …, en) as a column matrix, e ∈ Mn, 1(E) and, similarly, f ∈ Mm, 1(F). The relations above can be written as ϕ(e) = A·f,

116

II. Modules
t

where ϕ(e) = (ϕ(e1), …, ϕ(en)). In the same way, if B = Me', f'(ϕ), then −1 ϕ(e') = B·f'. Also we have e' = S·e and f' = T·f, or f = T ·f'. Since ϕ is an R-module homomorphism, we have ϕ(S·e) = S·ϕ(e) (prove this!). Thus: −1 −1 ϕ(e') = ϕ(S·e) = S·ϕ(e) = S·(A·f) = (SA)·f = (SA)·(T ·f') = (SAT )·f', which says that the matrix of ϕ in the bases e' and f' is SAT .
−1

!

Exercises
1. Let R be a ring and let L be a free R-module. Then any R-epimorphism ϕ : M → L splits. In particular, if K is a field, any short exact sequence of K-linear spaces 0 → U → V → W → 0, is split. 2. a) Let R be a domain and g1, …, gn ∈ R[X], with deg gi ≠ deg gj if i ≠ j. Then g1, …, gn are linearly independent in the R-module R[X]. b) Let K be a field of characteristic 0 (i.e. n·1 ≠ 0, ∀n ∈ N*, where 2 1 is the identity of K) and let a ∈ K. Then {1, X − a, (X − a) , …, n (X − a) , …} is a basis in the K-vector space K[X]. If p ∈ K[X], compute the coordinates of p in this basis. (Hint. Recall the Taylor series expansion, applied to polynomials.) 3. Prove that in the Z-module Q any subset having at least two elements is linearly dependent and that Z À ZQ. Deduce that ZQ is not free. 4. Let G be a finite Abelian group. Can G be a free Z-module? 5. Characterize the ideals I of the commutative ring R that are free R-modules.

(the converse of Prop. Is it true that any submodule (resp. Let W be a finite dimensional K-vector space and let U. (ii) R/I ≅ R/J (as R-modules). Then M is free and B is a basis of M. prove that dim(U + V) = dim U + dim V − dim(U ∩ V). express |M| as a function of |R| and |B|. 4. Consider the statements: (i) I ≅ J (as R-modules). b) Using the isomorphism theorems. If M is a free R-module of basis B. factor module) of a free module is still a free module? . V ≤ KW. 10. Which are the valid implications between the above statements? 9.9). a) Show that dim(W/U) = dim W − dim U. J be ideals in R. Let M be an R-module with the following property: There exists a subset B of M such that. (iii) R/I ≅ R/J (as rings). 8. 7. for any R N and any function ϕ : B → N.4 Free modules 117 6.II. there exists a unique R-homomorphism ψ : M → N with ψ|B = ϕ. Let R be a commutative ring and let I.

if R is commutative. R denotes a principal ideal domain. In order to build a factor module we need a submodule. Applying these theorems for Z. It is therefore natural to study first the submodules of a free module. a complete description of finitely generated Abelian groups is obtained. any free R-module E has the property that any two bases have the same cardinal (called the rank of E). 118 . Another important application is to the problem of invariant subspaces of an endomorphism of a finitely dimensional vector space: we obtain the existence of a basis in which the given endomorphism has a “simple” matrix (the Jordan canonical form). Recall that. unless otherwise specified.1 The submodules of a free module Any module is a factor module of a free module. In this section. III. Finitely generated modules over principal ideal domains In the case of finitely generated modules over PID's very precise structure theorems are available.III.

of rank ≤ n − 1. Then F is free and rank F ≤ n. By induction. of basis {e1 + L} (easy check) and thus F/G is free of rank 1 (by the case n = 1. B is linearly independent. with a0 f0 + a1 f1 + … + am fm = 0. if a0. Let us show that it is a generating system. fm} being a basis in G). fm} be a basis in G and {f0 + G} a basis in F/G. Then x + G ∈ F/G. Let {f1. so its rank is n − 1. {e1. we may suppose F ≠ 0. If n = 1 let {e} be a basis of E. which implies f1 = … = fm = 0 ({f1. But E/L is free. fm} is a basis in F. Then L is free. x = a0 f0 + a1 f1 + … + am fm. and G = F ∩ L ≤ L. note that F + L ≤ R E. with a ∈ R. f1. but {f0 + G} is a basis in F/G. We get a1 f1 + … + am fm = 0. we are done. for any free R-module H of rank n − 1. for some ai ∈ R.…. Let L := Re2 + … + Ren and G := F ∩ L. Suppose that. so {a} is a basis of Ra. Thus.…. of rank ≤ 1.III. the statement of the theorem becomes: any submodule ( = ideal) of R is free. en}. am ∈ R. this is true : any nonzero ideal of R is of the form Ra. Let x ∈ F. G is free of rank m ≤ n − 1. Since R is a PID. already proven).….…. If G = F. so a0 = 0. en} a basis in E and F ≤ R E. If not. We use an argument by induction on n.1 The submodules of a free module 119 1.1 Theorem. so F = F ≅ F+L ≤ E G L∩F L L Accordingly. Let E be a free R-module of rank n and let F be a submodule of E. We claim that B = {f0.…. free of rank n. Indeed. of basis {e2. Then E = Re ≅ R. This means that x − a0 f0 ∈ G. ! .…. In this case. 0 ≠ F/G is isomorphic to a submodule of E/L. Evidently. Take R E. Proof. so x − a0 f0 = a1 f1 + … + am fm. any submodule of H is free. a1. Therefore. so there is some a0 ∈ R such that x + G = a0 f0 + G. then a0 f0 + G = 0 + G (since a1 f1 + … + am fm ∈ G).

Then there exists a basis e = (e1. dm ∈ R and d1|d2|…|dm. For any basis e' = (e'1. n(R) such that f ' = ∑j aije'j. ∀i ∈ {1. Thus.20) and f' as a matrix in Mm. −1 Comparing with f = De. such that m ≤ n. diagonal) form?1. fm} as in the statement of the theorem. m}.….2 Theorem. en}: fi = ∑j aijej. n(R).…. e'n) in E and any ′ basis f' = ( f1′.…. so −1 f = U (AV)e. E a free R-module of rank n and F a submodule of E. fm) in F. 1(E). with aij ∈ R.…. we have Uf = A(Ve) = (AV)e.…. then these relation are written shortly: f' = Ae'. Then the unique matrix D = (dij) ∈ Mm.… f m ) in F.g. We obtain a matrix A = (aij) ∈ Mm. fi = diei.….120 III. we obtain D = U AV. there exist 1 What about the case of vector spaces? . Can we choose the bases in E and F such that A has a simple (e. If we see e' i as a column matrix. Let R be a principal ideal domain. where d1. e' = Ve and f' = Uf. If {e1. 1(R). en) and f = { f1. (see the discussion preceding 4.…. n(R) with f = De has the properties : dij = 0 if i ≠ j and d11|d22|…|dmm. e' ∈ Mn. Replacing in f' = Ae'. there exists a unique matrix A = (aij) ∈ Mm. m}. Suppose the problem is solved: we have two bases e = (e1. en} is a basis of E and {f1. (Beginning of) Proof.…. The next theorem says the answer is Yes. en) in E and a basis f = ( f1. We see that the existence of the bases e and f with the desired property is equivalent to the following: given A ∈ Mm. 1. Let V ∈ GLn(R) the basis change matrix from basis e to basis e' and U ∈ GLm(R) the basis change matrix from basis f to basis f'. ∀i ∈ {1. any fi is written uniquely as a linear combination of {e1. n(R)..….…. Finitely generated modules over principal ideal domains Let R E and F ≤ R E. as above. fm} is a basis of F. ….

4 Theorem.3 Definitions. ⎡d1 … 0 0⎤ diag(d1. b) We call a diagonal matrix D = diag(d1. Let R be a PID and let m. which is interesting on its own: the proof is in fact an algorithm used to compute the Smith normal form of a matrix. n ∈ N*. If r = min(m. We reduced the statement to a problem concerning matrices with elements in R. n(R) is arithmetically equivalent to a matrix in Smith normal form.….n)) canonically diagonal (or in Smith normal form) if d1|d2|…|dr.2 by proving the next theorem.… . This is an equivalence relation on Mm. a) A matrix D = (dij) ∈ Mm. Let m. n(R) are called arithmetically equivalent (denoted A ∼ B) if there exist invertible matrices U ∈ Mm(R) and V ∈ Mn(R) such that B = UAV. ….…. dr ∈ R. n(R) is a diagonal matrix (i ≠ j implies dij = 0). dr) := ⎢ ! * ! ! ⎥ ⎢ ⎥ ⎢ 0 … d r 0⎥ ⎣ ⎦ denotes the diagonal matrix (dij) ∈ Mm.III. 1. B ∈ Mm. n ∈ N*.… . n(R) with dii = di. n(R) (where r = min(m. The following definitions are helpful: 1.n) and d1. Then any matrix A ∈ Mm. dr) ∈ Mm. We finish the proof of theorem 1.… .1 The submodules of a free module −1 121 invertible matrices U and V such that U AV = D = (dij) ∈ Mm. dr) and D' = diag(d'1. the Smith normal form of A is unique in the following sense: if D = diag(d1. with the additional condition d11|d22|…|dmm. r}. n(R) is called a diagonal matrix if its entries not situated on the main diagonal are zero: i ≠ j implies dij = 0. n(R) (exercise!). ∀i ∈ {1. Moreover. d'r) are in Smith nor- . c) The matrices A.

let us review the transformations that can be performed on the matrix A. We shall see that these are the transformations that arise when computing determinants: swapping rows (or columns).5 Definitions.…. Pij is the matrix obtained from I by swapping row i with row j. m}. where i ∈ {1. addition to a row (column) of another row (column) multiplied by some element. i j 1 i Tij(a) = j ! * ! ! ! 0 + + 1 + a ! * ! + + 0 + 1 * 0 1 . . then d1 ∼ d'1. Tij(a) is obtained from I by adding to the row i the row j multiplied by a. m}. j ∈ {1. Finitely generated modules over principal ideal domains mal form and are arithmetically equivalent. 1. i ≠ j. where i. D is called the Smith normal form of A. where a ∈ R. …. it is uniquely determined up to an association in divisibility of the entries on the diagonal. . Before proving the theorem. ….Type III : Di(u). i ≠ j. i. m} and u ∈ U(R). j ∈ {1.122 III. Let m ∈ N* and let I be the identity matrix of the ring Mm(R). If D is in Smith normal form and is arithmetically equivalent to A. such that the resulting matrix is arithmetically equivalent to A.Type I : Tij(a). An elementary matrix is a square matrix in Mm(R) which is of one of the following types: .Type II : Pij. Di(u) is the matrix obtained from I by multiplying row i by u. dr ∼ d'r (" ∼ " means here “associated in divisibility”). ….

Di(u) are elementary matrices in Mm(R) as above. Pij A is obtained from A by swapping the row i with the row j. Di(u)A is obtained from A by multiplying the row i by u. This follows from the following relations. Pij. respectively III). II. a direct calculation shows that: Tij(a)A is obtained from A by adding to the row i the row j multiplied by a. ∀a. The transformations described above are called elementary transformations of the rows of A (of type I. we obtain the elementary transformations of the columns of A: Type I: ATij(a) is obtained from A by adding to the column i the column j multiplied by a. easy to prove: Tij(a)Tij(b) = Tij(a + b).III. and Tij(a). b ∈ R. If we take elementary matrices in Mn(R) and multiply A on the right with these matrices. so Tij(a)Tij(−a) = Tij(0) = I. All elementary matrices are invertible. Type III: ADi(u) is obtained from A by multiplying the column i by u. .1 The submodules of a free module 123 i j i 1 ! * ! ! ! 0 1 0 0 1 Di(u) = ! ! * ! 0 0 i ++ 0 + 1 Pij = ! * ! j ++ 1 + 0 i +++ u * * 0 0 + 1 0 1 0 If A ∈ Mm. Type II: APij is obtained from A by swapping the column i with the column j. n(R).

We prove P(m) by induction on l(A). In other words. in Z. where: P(m): For any matrix A ∈ Mm. j ∈ {1. where the subscripts indicate the ′ ⎣0m −1.124 III. we prove that P(m) holds for any m. l(a) is the number of prime factors (not necessarily distinct) of a decomposition of a in prime factors. A' ∈ Mm−1. called the length of a. n −1 ⎦ dimensions of the matrix.6 Proposition. a ≠ 0. as follows: if a is a unit. For example. ∀u. l(0) = −∞. there exists d1 ∈ R and a matrix A is arithmetically equivalent to the matrix (written in block 01. Finitely generated modules over principal ideal domains PijPij = I. …. m}. set l(a) = 0. …. Define the number l(a). define the length of A: l(A) := min{l(aij) | i ∈ {1. if a is nonzero and not a unit. we obtain: 1. n(R). b) If A = (aij) ∈ Mm. the inverse of an elementary matrix exists and is also an elementary matrix (of the same type). We can finally prove theorem 1. n(R). l(1) = 0.7 Definition. n−1(R) such that d1 divides all the entries of A' and .1 Am −1. −1 Dij(u)Dij(v) = Dij(uv). any matrix obtained from A by elementary transformations of rows and/or columns is arithmetically equivalent to A. v ∈ U(R). By convention.4: The proof is by induction on m. n(R). l(24) = 4. ! We also need: 1. For any A ∈ Mm. a) Let R be an UFD and a ∈ R. n}}. Looking at the definition of the relation of arithmetic equivalence between matrices. so Dij(u)Dij(u ) = Dij(1) = I. l(8) = 3. n −1 ⎤ ⎡ d1 form): ⎢ ⎥ . More precisely.

As in case 1. For 2 ≤ i ≤ m. we may suppose (swapping some rows and/or some columns) that l(a11) = l(A). as in case 1. the matrix obtained has 0 on the first column. which finishes the proof of P(m) by induction on l(A). ∃ bi ∈ R with ai1 = a11bi. In this case we prove that A is arithmetically equivalent to a matrix C ∈ Mm. This case is essentially Case 1.1).III. Subcase 2.1. since a11|ai1.1 The submodules of a free module 125 If A = 0 (l(A) = −∞). Clearly. l(A) ≥ 1. The condition that a11 divides all entries of A' is fulfilled (a11 is invertible!).2. we may suppose thus from the beginning that a11 is invertible. In the same way we can annihilate the entries of the first row. Let aij be the entry of A for which l(aij) = l(A). A is in Smith normal form. add to the row i −1 the row 1 multiplied by (−a11 ai1) (which amounts to multiply A on −1 the left with Ti1(−a11 ai1)). A is arithmetically equivalent to a matrix of the form B. We may suppose that a11 does not divide an entry of the first row or of the first column. add to the row i the row 1 multiplied by (−bi). Let aij be a unit. Case 1. otherwise a11 divides all entries of the first . 2 ≤ i ≤ m. Thus. A has an entry that is a unit (⇔ l(A) = 0). To simplify notations. Similarly. we make 0's on the first row and obtain a matrix of the form: 0⎤ ⎡a B = ⎢ 11 ⎥ ⎣ 0 A'⎦ with B ∼ A. a11 divides all the entries of A. a11 divides the entries of the new matrix (these entries are linear combinations of the entries of A). then swap the columns j and 1. Subcase 2. n(R) with l(C) < l(A). with aij in the position (1. Indeed. Case 2. except a11. a11 does not divide all the entries of A. Indeed. Swap the rows i and 1. The matrix obtained is arithmetically equivalent to A. We now obtain a matrix that has 0 on the column 1 (except for a11): for each i.

Consider the matrix (written in block form): u v 0 U = -b a 0 0 0 I where I is the (m − 2)×(m − 2) identity matrix (if m = 2. i.entry is d (thus l(C) ≤ l(d) < l(A)). for some a. For any A ∈ Mm. so l(C) ≤ l(d) < l(a11) = l(A). so C := UA ∼ A. suppose that a11 does not divide an element on the first column (the proof in the other case is similar. arithmetically equivalent to A. If a11 divides all the entries of the matrix.1. Since d |a11. b) = 1 and R is a PID.e. This finishes the proof of the existence part. because this implies a11|a21. there exists an entry aij (with i. we obtain a matrix with a11 the only nonzero entry of the first row and of the first column.126 III. Since (a. whose (1.n(R) and 1 ≤ k ≤ min(m.aij. To fix the ideas. v ∈ R such that au + bv = 1. not divisible by a11. the difference being that one multiplies on the right with adequate invertible matrices – see below). Finitely generated modules over principal ideal domains row and of the first column. we obtain an entry on the column 1. Suppose a11-a21 (if not. By adding to the column 1 the column j. a21). But C has as the (1. U is a 2×2 matrix). dau + dbv = a11u + a21v = d. We exhibit a matrix C. Let d = GCD(a11. false. b ∈ R. Let a11 = da. n). we are done! If a11 does not divide all the entries of the matrix. 1)-entry ua11 + va21 = d. a21 = db. We prove now the uniqueness part in theorem 1. l(d) < l(a11) = l(A). Working as in subcase 2.4. So. then I does not appear anymore. swap rows to achieve this). 1) . The matrix U is invertible (det U = ua + vb = 1). j > 1) such that a11 . We cannot have d ∼ a11. let ∆k(A) be the GCD of the minors determinant of a matrix obtained as follows: choose k rows and k columns of A and retain only the entries located at the intersection of the of order k of the matrix A (Recall that a minor of order k of A is the . there exist u.

involving the prime factorizations of the 2 relation holds: det(al1 + bl'1. b ∈ R (similarly for any row li). The existence part of the proof above is in fact an algorithm to find the Smith normal form of a given matrix. …. d1. These relations indicate another method to compute effectively d1. for k ≥ 2.III. …. dr) is in Smith normal form. Note that. R is an Euclidian ring. if V ∈ Mn(R). the rows of UA are linear combinations (with coefficients in R) of the rows of A. ik). ik of UA) are linear combinations of the rows of A (truncated to contain only the entries on the columns i1. …. The claim now follows. then ∆k(A)|∆k(UA). n are not small). …. Indeed. the following . lk) = adet(l1. …. In this case. There are ⎜ ⎟⎜ ⎟ minors of order k in an ⎝ k ⎠⎝ k ⎠ m×n matrix). then ∆k(A)|∆k(AV). ….8 Remark. …. ∀a. if A ∼ B. So. then ∆k(A)|∆k(B) and. ∆k(A) ∼ ∆k(B). So. by symmetry. the rows of a minor of order k of UA (corresponding to the choice of columns i1. dk ∼ ∆k(A)/∆k−1(A) . Similarly. ∆k(B)|∆k(A). lk). lk. if A ∼ D. Thus. …. lk) + bdet(l'1.7 (whose computation is costly. lk) the matrix having the rows l1.e. an easy check show that ∆k(D) ∼ d1…dk. By denoting (l1. dr are determined (up to association in divisibility) by ∆1(A). ∆r(A) and d1 ∼ ∆1(D)∼ ∆1(A). In practice. Applying the fact that the determinant of a matrix is a multilinear function of the rows of the matrix2. ! 1. ….1 The submodules of a free module 127 ⎛ m ⎞⎛ n ⎞ chosen rows and columns. i. if U ∈ Mm(R). …. dr (although the amount of computation is prohibitive if m. the function length defined at 1. it follows that a minor of order k of UA is a linear combination of minors of order k of A. …. If D = diag(d1.

….2. the quotient in the division with remainder of a21 to a11). R denotes a principal ideal domain. m}. 1. The proof above is rewritten word for word (replace everywhere l with ϕ). Finitely generated modules over principal ideal domains elements of the matrix) can be advantageously replaced by the function ϕ in the definition of the Euclidian ring. This places r on the entry (2. 6. Exercises In the exercises. 10. where a21 = a11q + r. if R is Euclidian with respect to ϕ. 3 . 1) of the matrix UA. which has thus ϕ(UA) < ϕ(A). j ∈ {1. and A = (aij) ∈ Mm. subtract from the row 2 of A the row 1 multiplied by q. …. Find the Smith normal form of the matrix: ⎡2 6 9 ⎤ ⎢5 10 12⎥ ∈ M3(Z). 9). define ϕ(A) = min{ϕ(aij) | i ∈ {1. unless otherwise specified.e. 12).. determine a basis of L.n( R). rank L and the factor module 3 Z /L. 6. at the subcase 2. v3 = (0. ⎥ ⎢ ⎢0 6 12⎥ ⎦ ⎣ If L is the submodule of Z generated by v1 = (2.128 III. v2 = (5. n}}. To be precise. with ϕ(r) < ϕ(a11) = min{ϕ(aij)} (i. 12). We invite the reader to check the details and apply the algorithm in the proof in concrete cases (see also the exercises). replace the matrix U with T21(− q).

…. 0). 5. where vi is the column i of the man trix V (vi is seen as an element in R ). Use the invari⎣ ⎦ ants ∆k.…. +). (e1. Let U ∈ GL(m. Show that the Smith normal form of ⎢ ⎥ is ⎣ 0 b⎦ ⎡d 0 ⎤ ⎢ 0 m ⎥ . an) if and only if GCD(a1. Let a. …. n) and d1. n(R) (in the canonical bases). R) such that UAV is in Smith normal form. 0). (Hint. x1. …. n(K). Let K be a field and let A ∈ Mm. 10. an ∈ R. an) ∈ Mn(R). where 1 appears r times (r is the rank of the matrix A). 0. n(R). xn)V = (d. 7.) 3. …. R) such that (x1. Find the Smith normal form of a diagonal matrix diag (a1. n ∈ N and let ϕ : R → R be an R-homomorphism whose matrix is A ∈ Mm. Let n ∈ N and let a1. 8. Show that a basis in Kerϕ is (vr+1. vn). 6. 0. dr nonzero. where d = GCD(a.1.…. xn). …. xn)). ….….…. with r ≤ min(m. …. ….…. b ∈ R. R) and V ∈ GL(n. …. Show that a basis in F = < f1. Show that: there exists V ∈ GL(n. …. Determine all subgroups of (Z×Z. xn ∈ R and d = GCD(x1. namely diag(d1. an) ∈ M1. 0.III. b). 4. en) is a basis in L and {f1. …. m = LCM(a. Find the Smith normal form of a row matrix (a1. Consider the Smith normal form of the row matrix (x1. (Ind. …. Let m. …. fm > can be obtained as follows: * n m * * . Then the Smith normal form of A is diag(1. an) = 1. 9. b). Let n ∈ N . Suppose L is a free R-module of rank n. fm} ⊆ L. 0).1 The submodules of a free module 129 ⎡ a 0⎤ 2. dr. …. R) such that the first row of V is (a1. Show that there exists V ∈ GL(n.

Show that: Ax = b has solun tions x in R if and only if the Smith normal form of A is (D.2 ).) n III. and a detailed description of any submodule of a finitely generated free module is available (Theorem 1. denoted by o(x). b) ∈ Mm. Let m. the order of an element in M is defined up to an association in divisibility. A ∈ Mm. the last column is b).2 Finitely generated modules over a principal ideal domain We are now ready to describe the structure of finitely generated modules over a principal ideal domain. (see also exercise 8. with UAV in Smith normal form. Then a basis in F is {gi | 1 ≤ i ≤ m. M is an R-module and x ∈ M. 1(R). 1(R). Consider the “extended matrix” A = (A. n ∈ N . Consider the equation * Ax = b. n(R) and b ∈ Mm. Let gi := ∑j uij fj ∈ F. n + 1(R) (the first n columns of A are the columns of A. Note that if R is a field. we have .130 III. this is the Kronecker-Capelli theorem. If R is a PID. V ∈ GLn(R). x ∈ R (a linear system of m equations and n unknowns). n(R) such that fi = ∑j aijej (1 ≤ i ≤ n) and let U ∈ GLm(R). Therefore. Thus. We need only some preparations: 2. Since any module is a factor of a free module. Finitely generated modules over principal ideal domains Let A = (aij) ∈ Mm. 0). A generator of the ideal AnnR(x) is called an order of x. then AnnR(x) (= {r ∈ R | rx = 0}) is a principal ideal of R. 11. where D is the Smith normal form of A and 0 is the zero column in Mm. gi ≠ 0}.1 Definition. the task is now easy.

∀i ∈ I}. there exist n. m ∈ N. xn ∈ M such that : M = Rx1⊕ …⊕ Rxm ⊕ Rxm+1…⊕ Rxn. ! We state now the following important theorem. (Invariant factors theorem) Let R be a principal ideal domain and let M be a finitely generated R-module. (D) . Recall that R° is the set of nonzero noninvertible elements of R.III.3 Theorem. and x1. …. ϕ is the direct sum of the family of homomorphisms ηi◦πi : M → Mi/Ni. ∀i ∈ I. Let R be a ring and let M be a left R-module such that M is the direct sum of a family of submodules (Mi)i∈I. 2. Rx ≅ R/Ro(x). Define M ϕ : M → ⊕ i by ϕ(x) = (πi(x) + Ni)i∈I. M = ⊕I Mi. which determines the structure of finitely generated modules over a principal ideal domain. If ϕ : E → F is an R-module isomorphism and x ∈ E. where ηi : Mi → Mi/Ni is the canonical surjection). Then M is a direct sum of a finite number of cyclic submodules. then o(x) = o(ϕ(x)) since AnnR(x) = AnnR(ϕ(x)). This notion generalizes the usual concept of order of an element in an Abelian group.2 Lemma.2 Finitely generated modules over a principal ideal domain 131 AnnR(x) = Ro(x). with m ≤ n. Since x = ∑i∈I πi(x). 2. then the sum of the submodules (Ni)i∈I is direct and we have a canonical isomorphism : ⊕ Mi M i∈I ≅⊕ i . We have Kerϕ = {x ∈ M | πi(x) ∈ Ni. ⊕ N i i∈I N i i∈I Proof. ∀x ∈ M. it follows that Kerϕ = ∑i∈I Ni = ⊕I Mi. One easily checks i∈I N i that ϕ is a surjective homomorphism (in fact. More precisely. Apply now the isomorphism theorem. Let πj : ⊕I Mi → Mj be the canonical surjections. If Ni ≤ R Mi.

Besides. ∀i ∈ {1. yn' ∈ M such that : M = Ry1⊕ …⊕ Rym' ⊕ Rym' +1…⊕ Ryn'.132 III. ∀s ∈ S ((es)s∈S is the canonical basis of R ). The existence part: if S ⊆ M is finite and generates M.….…. …. …. n = n' and di ∼ ei. where F is the kernel of ϕ. m' ∈ N.…. dk ∈ U(R) and dk+1.…. en) in E and a basis f = (f1. dm ∉ U(R). Then Rfi = Rdiei = Rei. m} and d1|d2|…|dm . n}. ∀i ∈ {1. (S) given by ϕ(es) = s.2 provides us with a basis e = (e1. There exists k ∈ N (0 ≤ k ≤ m) such that d1. fk+1. o(xi) =: di ∈ R satisfy the conditions: di ∈ R°. ∀i ∈ {1. so {xi |1 ≤ i ≤ n} generates M. …. (D') and o(yi) =: ei satisfy : ei ∈ R°.…. …. Theorem 1. . dm+1 = … = dn = 0. di ≠ 0 and d1|d2|…|dm. fm. Let (S) n = |S|.…. We have M = Rx1⊕…⊕Rxn because {ei + F |1 ≤ i ≤ n} generates E/F. Changing notations if necessary. with di ∈ R. The numbers n. E = R . The “orders” o(xi) are called the invariant factors of the module M. with m' ≤ n'. i = 1 n with the ’ above properties are uniquely determined. en. the structure of M is determined by dk+1. m'} and e1|e2|…|em' . then m = m'. 1 ≤ i ≤ k. Finitely generated modules over principal ideal domains Moreover.…. Let ψ(ei + F) = xi ∈ M. we may suppose from the beginning that d1. Proof.…. and we can write (applying lemma 2. m ∈ N and the orders o(xi) ∈ R. 1 ≤ i ≤ n. dm are non invertible. in the following sense: if n'. ek+1. and y1. dm. then (S) there is an isomorphism ψ : R /F → M. fm) in F (with m ≤ n) such that fi = diei. em'+1 = … = en' = 0.2): R e1 Re Re Re M≅E/F ≅ ⊕ … ⊕ k ⊕ k +1 ⊕ … ⊕ m ⊕ R em+1 ⊕ … ⊕ R en R e1 R ek Rf k +1 Rf m Re Re = k +1 ⊕ … ⊕ m ⊕ Rem +1 ⊕ … ⊕ Ren Rf k +1 Rf m ≅ Rek +1 ⊕ … ⊕ Rem ⊕ Rem +1 ⊕ … ⊕ Ren Rf k +1 ⊕ … ⊕ Rf m Thus.

4 Definitions. Let R be a domain and let M be an R-module. 1 ≤ i ≤ n. ∀x ∈ M} is called the annihilator of the module M. If M = t(M). . 2. since ∑1 ≤ i ≤ n ri(ei + F) = 0 + F ⇔ ∃si ∈ R such that ∑1 ≤ i ≤ n riei = ∑1 ≤ i ≤ m sidiei ⇔ ri = sidi.5 Proposition. let za(M) := {x ∈ M | ax = 0} (called the annihilator of a in M. then t(⊕i∈I Mi) = ⊕i∈I t(Mi). so r = 0 if m + 1 ≤ i ≤ n and r = sidi if 1 ≤ i ≤ m. 2. rxi = 0 ⇔ rei ∈ F ⇔ ∃si ∈ R such that ∑1 ≤ i ≤ n riei = ∑1 ≤ i ≤ m sidiei. p ||d ⇔ k is the exponent of p in the prime factor decomposition of d). t(M) is called the torsion submodule of M). c) If a ∈ R. if 1 ≤ i ≤ m and o(xi) = 0 if m + 1 ≤ i ≤ n. p ||d means p |d and p -d (if R is a UFD. The elements of t(M) are called torsion elements. also denoted sometimes by AnnM(a) or rM(a)). if t(M) = 0. then t(M) ≅ t(N). Also. Define tp(M) := {x ∈ M | ∃n ∈ N n cu p x = 0} (called the p-torsion submodule of M or the p-submodule of M). if M ≅R N.III. d) AnnR(M) := {r ∈ R | rx = 0. Let R be a domain and let M be an R-module. for any 1 ≤ i ≤ m and ri = 0. M is called a torsion module. o(xi) = o(ei + F) = di. We need to collect the basic proprieties of these invariants. b) Let p ∈ R be a prime element. Indeed.2 Finitely generated modules over a principal ideal domain 133 E/F = R(e1 + F)⊕…⊕R(en + F). for m + 1 ≤ i ≤ n ⇔ ri(ei + F) = 0 + F. k k k+1 k If d ∈ R and k ∈ N. This finishes the proof of existence of a decomposition (D) of M. For the uniqueness part. for any r ∈ R. we need some “invariants”. b) If M = ⊕i∈I Mi. M is called a torsion-free module. a) Define t(M) := {x ∈ M | ∃r ∈ R \{0} with rx = 0}. a) t(M) is a submodule of M and t(M/t(M)) = 0.

rn ∈ R such that x = r1x1 + … + rmxm +… + rnxn and ax = ar1x1 + … + armxm +… + arnxn = 0. Let R be a PID. so Rx1⊕…⊕Rxm ⊆ t(M). Proof.… xm ∈ t(M). b) If M = ⊕i∈I Mi.e. a finitely generated torsion-free module is free. such that ∑i∈I xi := x ∈ t(M). ∀j ∈ J. Then there exist a. It is now clear that. M = t(M)⊕L. then tp(⊕i∈I Mi) = ⊕i∈I tp(Mi). . free. Obviously.…. then t(M) is a direct summand in M and there exists L ≤ R M.134 III. i = 1 n . a) Let x. c) Take a decomposition (D) of M. We show that t(M) = Rx1⊕…⊕Rxm. From M = ⊕i∈I Mi it follows that axi = 0. a) tp(M) is a submodule in M and tp(M/tp(M)) = 0. ! 2. since J is finite). ∀i ∈ I. Because the sum of the submodules Rxi is direct. xi ∈ t(Mi). Let x + t(M) ∈ t(M/t(M)): there exists 0 ≠ a ∈ R with ax ∈ t(M). This shows that ⊕i∈I t(Mi) ⊆ t(M). there is some family (rj)j∈J. x + t(M) = 0 + t(M). x1. Finitely generated modules over principal ideal domains c) Let R be a PID. since dixi = 0 and di ≠ 0. b) Let (xi)i∈I have finite support. We deduce that x ∈ Rx1⊕…⊕Rxm. We have ab ≠ 0 and ab(rx + sy) = 0. arixi = 0. So. Then there exists 0 ≠ b ∈ R with bax = 0. there is some 0 ≠ a ∈ R and r1. such that ax = by = 0. so ri = 0. nonzero. then tp(M) ≅ tp(N). s ∈ R.6 Proposition. In particular. let M be an R-module and p ∈ R a prime element. If x ∈ t(M). t(M) ⊆ ⊕i∈I t(Mi). then ’ ari ∈ AnnR(xi) = 0. Then there exists 0 ≠ a ∈ R such that ax = ∑i∈I axi = 0. so x ∈ t(M) (since ba ≠ 0). b ∈ R. so rx + sy ∈ t(M). Evidently. r ≠ 0 and r∑i∈I xi = 0. with 0 ≠ rj ∈ R and rjxj = 0. i. If i > m. y ∈ t(M) and let r. with xi ∈ Mi. If xi ∈ t(Mi) such that J := supp((xi)i∈I) is finite. if M ≅R N. Thus. If M is finitely generated. such that M = t(M)⊕L. ∀i ∈ I. denoting by L the free module ⊕i > m Rxi. Let r := ∏j∈J rj (well defined.

We have rx = 0 ⇔ ∑i∈I rxi = 0 ⇔ rxi = 0. In k particular. We have Kerϕ = Rp.d and r ∈ R. k c) Let b ∈ R such that d = p b. zp. then zp(Rx) = 0 if p-o(x) and zp(Rx) ≅ R/Rp if p|o(x). then zr(M) = ⊕i∈I zr(Mi). if x ∈ M.3. if p-d. We have p(r + Rd) = 0 + Rd ⇔ pr ∈ Rd ⇔ d | pr. Ins deed.2 Finitely generated modules over a principal ideal domain k k 135 c) Let d ∈ R and k ∈ N such that p ||d. if M ≅R N. Proof. So b| p r and (b. which imply b|r.7 Proposition. Proof. zp(R/Rd) = Ra/Rd. b) Let x = ∑i∈I xi. then tp(Rx) = 0.c)). we get: . For any r ∈ R. a) and b) have similar proofs with 2. p) = 1. In particular. Then tp(R/Rd) ≅ R/Rp . The other inclusion is obvious. a) zr(M) is a submodule in M. ϕ(r) = rb + Rd. with c ∈ R. let r + Rd ∈ tp(R/Rd). tp to the decompositions (D) and (D'). We claim that tp(R/Rd) = Rb/Rd. ϕ(r) = ra + Rd. with xi ∈ Mi. if x ∈ M and o(x) = d. then tp(Rx) ≅ R/Rp . So. then zr(M) ≅R zr(N). ∀r ∈ R. ! 2. c) If R is a PID. and the sum is direct) ⇔ xi ∈ zr(Mi).III. i. ∀i ∈ I. ∀i ∈ I (since rxi ∈ Mi.5 and are proposed as an exercise. b) If M = ⊕i∈I Mi . so r + Rd = 0.5. ∀r ∈ R. Since (d. So. Let M be an R-module and let r ∈ R. Let p|d and a ∈ R with d = pa. a) Exercise. p) = 1. k We have R/Rp ≅ Rb/Rd by the isomorphism theorem applied to ϕ : R → Rb/Rd. c) Let p .e. r ∈ Rb and tp(R/Rd) ⊆ Rb/Rd. pr ∈ Rd = Rpa ⇔ r ∈ Ra. There exist s ∈ N such that p r ∈ Rd. ∀i ∈ I. Since t(M) = Rx1⊕…⊕Rxm = Ry1⊕…⊕Rym' (see 2. so Ra/Rd ≅ R/Rp. Consider the surjective homomorphism ϕ : R → Ra/Rd. then zp(R/Rd) = 0 if p-d and zp(R/Rd) ≅ R/Rp if p|d. s k s p r = dc = p bc. this implies d | r. The idea is to apply the invariants t. p ∈ R is a prime element and d ∈ R. ! We can proceed now to prove the uniqueness part in 2.

1 ≤ i ≤ m. Let us show that di ∼ ei. if M ≅ R/Rd1⊕…⊕R/Rdm. such that p divides ei. By 2. 1 ≤ i ≤ m. By (**). Let p be a prime divisor of d1. 1 ≤ i ≤ m). with 1 ≤ i ≤ m'. with e1|e2|…|em'. 1 ≤ i ≤ m'. (R/Rp) ≅ (R/Rp) (R-module isomorphism). 1 ≤ i ≤ m. It is clear that this is also an R/Rp-module isomorphism. (we used that p|di. It is sufficient to show that αi = βi. k ≤ m'. AnnR(M) = Rem'. we suppose in the sequel that M = t(M) = Rx1⊕…⊕Rxm = Ry1⊕…⊕Rym'. Since d1|d2|…|dm. ei ∈ R°. 1 ≤ i ≤ m. We have : zp(M) ≅ zp(Ry1)⊕…⊕zp(Rym') ≅ R/Rp⊕…⊕R/Rp (k terms). But R/Rp is a field (p is prime and R is a PID). Let p be an arbitrary prime factor of dm and let αi (respectively βi) be the exponent of p in the decomposition of di (respectively ei). …. Finitely generated modules over principal ideal domains M/t(M) ≅ Rxm+1⊕…⊕Rxn ≅ Rym'+1⊕…⊕Ryn'. 1 ≤ i ≤ m. (**) then m = m' and di ∼ ei.7 and (*). by (*). . Ryi ≅ R/Rei. Since Rxi ≅ R/Rdi. AnnR(M) = {r ∈ R| rx = 0. Consider zp(M). so m ≤ m'. m k So. let k be the number of those indices i. so dm ∼ em'. Of course. 1 ≤ i ≤ m. But {xm+1. so m = k. we have to prove that. Indeed. we have α1 ≤ α2 ≤ … ≤ αm (and also β1 ≤ β2 ≤ … ≤ βm). with d1|d2|…|dm. ∀x ∈ M} is an ideal in R. Then p|di. so m = m'. By symmetry. we can write the R-module isomorphisms: zp(M) ≅ zp(Rx1⊕…⊕Rxm) ≅ zp(Rx1)⊕…⊕zp(Rxm) ≅ R/Rp⊕…⊕R/Rp (m terms). ….136 III. To this end. m' ≤ m. It remains to prove that m = m' and di ∼ ei. Use now (**). 1 ≤ i ≤ m. Note first that dm ∼ em'. since two isomorphic R/Rp-vector spaces have the same dimension. AnnR(M) = Rdm (prove!). di ∈ R°. so they have the same number of elements: n − m = n' − m'. yn'} are bases in the free module M/t(M). (*) M ≅ R/Re1⊕…⊕R/Rem'. Suppose that is not so: then there exists j. xn} and {ym'+1.

III.2 Finitely generated modules over a principal ideal domain

137

1 ≤ j ≤ m, minimal with the property that αj ≠ βj (thus αi = βi if i < j). Say, for instance, that αj < βj. Apply tp. By (*) and 2.6, we have: tp(M) ≅ tp(Rx1)⊕…⊕tp(Rxm) ≅ R Rpα1 ⊕ … ⊕ R Rpαm “Multiply” tp(M) by p j (i.e., consider the submodule p j t p (M ) ). We
α

By (**), tp(M) ≅ tp(Ry1)⊕…⊕tp(Rym) ≅ R Rp β1 ⊕ … ⊕ R Rp βm .
α

obtain from (*):
α

p jtp (M ) ≅ p

αj

(R

Rpα1 ⊕ … ⊕ p ⊕ … ⊕ R Rp

)

αj

( R Rp )
αm

≅ R Rp

α j +1 −α j

α m −α j

We have used the proprieties (whose proof is immediate): if α β M = ⊕i∈I Mi and r ∈ R, then rM = ⊕i∈I rMi ; p {R/Rp } ≅ 0 if β ≤ α; α β β −α p {R/Rp } ≅ R/Rp if β > α. From (**), with a similar argument, we have : α α α p j t p ( M ) ≅ p j R Rp β1 ⊕ … ⊕ p j R Rp βm

(

)

(

)

≅ R Rp

β j −α j

⊕ … ⊕ R Rp

β m −α j

Therefore, α −α α −α β −α β −α R Rp j +1 j ⊕ … ⊕ R Rp m j ≅ R Rp j j ⊕ … ⊕ R Rp m j These are decompositions of the type (*), as one easily sees. Let k be the number of indices i for which αi > αj (evidently, 0 ≤ k ≤ m − j). In the left hand side we have k nonzero terms and in the right hand side there are m − j + 1 nonzero terms. The first part of the proof shows that k = m − j + 1, contradicting k ≤ m − j. Thus, we must have αi = βi, 1 ≤ i ≤ m. !
2.8 Remark. An R-module M can have several decompositions of type (D). But the sequence of ideals AnnR(x1) ⊇ AnnR(x2) ⊇ … ⊇ AnnR(xn) is uniquely determined. The generators (uniquely determined up to an association in divisibility) d1, d2, …, dn ∈ R of these ideals are called the invariant factors of the R-module M. Sometimes this name is

138

III. Finitely generated modules over principal ideal domains

given to the sequence of ideals above. If in the principal ideal domain R there is a natural way to choose a generator of an ideal, the invariant factors are the “natural” generators of the annihilator ideals above. For example, in Z the positive generator is chosen, in K[X] (K a field) the monic polynomial that generates the ideal is chosen. ˆ ˆ 2.9 Example. Let Z2 = { 0, 1 }, Z6 = { 0 , 1 , …, 5 } and the Z-module Z2×Z6. We have the following decompositions: Z2×Z6 = Zx1⊕Zx2 = Zy1⊕Zy2, ˆ ˆ ˆ ˆ where x1 = ( 1, 0 ), x2 = ( 0, 1 ), y1 = ( 1, 3 ), y2 = ( 0, 5 ) (check this!). Of course, these are distinct decompositions, but AnnR(x1) = AnnR(y1) = 2Z and AnnR(x2) = AnnR(y2) = 6Z. The sequence of invariant factors of the Z-module Z2×Z6 is: 6, 2.

Exercises
1. Give an example of a finitely generated Z-module such that its torsion submodule has at least two complements. (cf. Prop. 2.5.) 2. Let R be a domain and let u : M → N be an R-module homomorphism. Then u(t(M)) ⊆ t(N). State and prove a similar propriety for tp(M) (R is a PID and p is a prime element in R) and zr(M) (where r ∈ R). 3. Assume R is a domain, M is an R-module and L ≤ R M. Then M/L is torsion-free if and only if t(M) ⊆ L. 4. This exercise gives an example of a Z-module M such that t(M) is not a direct summand in M. (Thus, by 2.5.c), M cannot be finitely generated). Let P = {pn |n ≥ 1} be the set of prime natural numbers and let M := ∏p∈P Zp. Show that:

III.2 Finitely generated modules over a principal ideal domain

139

a) t(M) = ⊕p∈P Zp ( = {(ap)p∈P | ∀p ∈ P, ap ∈ Zp and supp((ap)P) finite}). b) ∀n ∈ Z, n ≠ 0, n·(M/t(M)) = M/t(M). c) ∀x ∈ M, ∃p ∈ P such that x ∉ pM. d) If M = t(M)⊕S, with S ≤ R M, then S ≅ M/t(M). e) t(M) is not a direct summand in M.
5. Let p be a prime natural number and let G = Zp×Zp. Prove that the subgroups of G coincide with the Zp-vector subspaces of G. How many vector subspaces of dimension 1 are there in G? In how many ways can ZG be written as a direct sum of two proper submodules? (Hint: the sum of any two nonzero distinct subspaces is direct, equal to G). 6. State a structure theorem for finitely generated Abelian groups. 7. Give an example of an Abelian group G that is neither cyclic, nor a direct sum of cyclic groups. (Hint: G cannot be finitely generated. An example is Q). 8. Determine all Abelian groups with 100 elements. 9. Let G be a finite Abelian group whose order is squarefree (not divisible with the square of any prime). Then G is cyclic. 10. If G is a finite Abelian group and n is the exponent of G ( = GCD of the orders of the elements of G), then G contains an element of order n. (Hint: use the theorem of invariant factors). Give an example of a group G such that n is a proper divisor of the order of G. 11. Let G be a finite Abelian group. Show that, for any divisor d of |G|, there exists a subgroup of G having d elements. 12. For n ∈ N*, let gn denote the number of isomorphism types of Abelian groups having n elements. For what n ∈ N* we have gn = 1? Determine {n ∈ N | n ≤ 100, gn ≥ 4}. 13. Using the structure theorem of finite Abelian groups, show that * any finite subgroup G of the multiplicative group K (where K is a

140

III. Finitely generated modules over principal ideal domains

field) is cyclic. (Hint. If G has more than one invariant factor and d is the greatest invariant factor, then any element of G is a root of a d X − 1 and d < |G|).
14. Determine the finitely generated Abelian groups with the property that their lattice of subgroups is a chain ( = totally ordered with respect to inclusion).

III.3 Indecomposable finitely generated modules
Theorem 2.3 says that any finitely generated R-module (if R is a PID) is a direct sum of cyclic submodules. Can such a decomposition be refined? In other words, can we decompose further these submodules as direct sums of proper submodules?
3.1 Definition. Let R be a ring (not necessarily commutative). An R-module M is called indecomposable if M ≠ 0 and M has no proper direct summands (if L, N ≤ R M are such that M = L⊕N, then L = 0 or N = 0). An R-module that is not indecomposable is called decomposable. 3.2 Examples. a) The Z-module Z6 is decomposable: Z6 = 2Z6⊕3Z6. b) If R M is isomorphic to a direct product of modules of the type A×B, with A, B nonzero, then M is decomposable. b) If K is a field, a K-vector space V is decomposable if and only if dim V > 1. (Besides, any proper subspace of a vector space is a direct summand). c) The Z-module Z2 is indecomposable (it has no proper submodules altogether).

III.3 Indecomposable finitely generated modules

141

3.3 Proposition. If the lattice of submodules of the R-module M is a chain (it is totally ordered) with respect to inclusion, then M is indecomposable. Proof. Suppose M = A⊕B, with A, B ≤ R M. But the lattice of submodules of M is a chain, so A ⊆ B or B ⊆ A. Thus, A ∩ B = 0 im! plies A = 0 or B = 0. 3.4 Corollary. Let R be a PID, p ∈ R a prime element and k ∈ N. k Then the cyclic module R/Rp is indecomposable. Proof. Suppose k ≥ 1. It is enough to check that the lattice of ideals k k of R that include Rp is a chain (see II.2.3). But the ideal I includes Rp k if and only if I = Ra (I is principal), where a ∈ R, a|p . The divisors of k t t p are p , 0 ≤ t ≤ k, so a ∼ p for some t ≤ k. This says that the ideals k k k−1 2 that include Rp are chained: Rp ⊆ Rp ⊆ …⊆ Rp ⊆ Rp ⊆ R. If k = 0, we have to show that R R is indecomposable. This is true for an arbitrary domain: the intersection of any two nonzero submodules ( = ideals) I, J of R is nonzero: if a ∈ I, b ∈ J are nonzero, then ! 0 ≠ ab ∈ I ∩ J.

If R is a PID, the cyclic modules (isomorphic to) R/Rp , with p prime in R and k ≥ 0, are all finitely generated indecomposable R-modules. This follows from the following result, valid in any commutative ring.
3.5 Theorem. (Chinese remainder theorem) Let R be a commutative ring, n ≥ 2 and I1,…, In ideals of R. a) If Ii + Ij = R for i ≠ j,3 then the product 4 I1·…·In is equal to the intersection I1 ∩…∩ In and there is a natural isomorphism of rings

k

(and of R-modules) η:
The ideals Ii şi Ij are called in this case comaximal. For example, the ideals Za and Zb of Z are comaximal if and only if a and b are coprime.
3

142

III. Finitely generated modules over principal ideal domains

R R R R = ≅ ×… × , I1 ⋅ … ⋅ I n I1 ∩ … ∩ I n I1 In

η(r + I1 ∩…∩ In) = (r + I1, …, r + In), ∀r ∈ R.
b) Conversely, if the homomorphism ϕ : R → R R ×… × , I1 In

ϕ(r) = (r + I1, …, r + In), ∀r ∈ R is surjective (inducing an isomorphism R R R ≅ ×… × as above), then Ii and Ij are comaxiI1 ∩ … ∩ I n I1 In mal ideals, for any i ≠ j. Proof. a) We prove by induction on n that I1·…·In = I1 ∩…∩ In and that η is an isomorphism. For n = 2, I1 + I2 = R implies the existence of x ∈ I1, y ∈ I2 such that x + y = 1. Let z ∈ I1 ∩ I2. Then z = z·1 = zx + zy, with zx, zy ∈ I1·I2, so I1 ∩ I2 ⊆ I1I2. Thus, I1 ∩ I2 = I1I2. R R Let ϕ : R → × , ϕ(r) = (r + I1, r + I2), ∀r ∈ R. ϕ is a ring (and I1 I 2

surjections R → R/Ij). We have:

an R-module) homomorphism (it is the direct product of the canonical Kerϕ = {r ∈ R| (r + I1, r + I2) = (0 + I1, 0 + I2)} = I1 ∩ I2 The isomorphism theorem implies that R/I1 ∩ I2 ≅ Imϕ and the

isomorphism is precisely η. Let us prove that ϕ is surjective (which R R will finish the proof). Let (r1 + I1, r2 + I2) ∈ × . We must exhibit I1 I 2
r ∈ R with r − r1 ∈ I1, r − r2 ∈ I2. Such an element is r = r1 y + r2 x. Indeed, r − r1 = r1y + r2x − r1x − r1y = (r2 − r1)x ∈ I1.
Recall that the product IJ of two ideals I and J is the ideal generated by all the products ij, i ∈ I, j ∈ J. The product of ideals is associative and always IJ ⊆ I ∩ J.
4

III.3 Indecomposable finitely generated modules

143

Similarly, r − r2 ∈ I2. Suppose now that for any k < n and any ideals I1,…, Ik, pairwise comaximal , we have I1·…·Ik = I1 ∩…∩ Ik and η is an isomorphism. Take n pairwise comaximal ideals I1,…, In. Since Ij + In = R, 1 ≤ j ≤ n − 1, there exist aj ∈ Ij, bj ∈ In, such that aj + bj = 1. Multiply these n − 1 equalities:

∏ (a j + b j ) = a1·…·an−1 + b = 1, where b ∈ In, a1·…·an−1 ∈ I1·…·In−1.
j =1

n −1

So, I1·…·In−1 + In = R. Applying the case n = 2 to the comaximal ideals I1·…·In−1 and In, I1·…·In−1·In = (I1·…·In−1) ∩ In = (I1 ∩…∩ In−1) ∩ In We also used the induction hypothesis I1·…·In−1 = I1∩…∩ In−1. Also, case n = 2 says that R R R ≅ × , (I1 ⋅… ⋅ I n−1 ) ⋅ I n I1 ⋅… ⋅ I n−1 I n
r + I1·…·In & (r + I1…In−1, r + In), ∀r ∈ R.

On the other hand, by induction, we have the isomorphism: R R R ≅ ×… × , I1 ⋅ … ⋅ I n −1 I1 I n −1
r + I1·…·In−1 & (r + I1, …r + In−1), ∀r ∈ R.

Combining these isomorphisms, we obtain the result. b) We prove that I1 and I2 are comaximal. Let (1 + I1, 0 + I2, …, R R 0 + In) ∈ × … × . There exists y ∈ R such that (y + I1, y + I2, …, I1 In
y + In) = (1 + I1, 0 + I2, …, 0 + In), i.e. y ∈ I2 and y − 1 =: x ∈ I1. So, ! 1 = −x + y ∈ I1 + I2, which means I1 + I2 = R.
3.6 Corollary. a) Let R be a PID and let a1, …, an ∈ R. If (ai, aj) = 1, ∀i ≠ j, then:

Apply now the Chinese remainder theorem for Ra1. then (a. …. b) is Ra + Rb. there is an isomorphism ϕ : R/Ra1 ×…× R/Ran → M. If d ≠ 0. …. 1 ≤ i ≤ n. we have R/Ra1 ×…× R/Ran ≅ R/Rd. isomorphic to some . …. ∀i ≠ j. 0 + Ran) and xi := ϕ(yi). Also. aj) = 1. (applying ϕ) M = Rx1⊕…⊕Rxn.M is indecomposable if and only if: either M is cyclic. So. b) = 1 if and only if Ra and Rb are comaximal. Indeed. a) If a. c) This follows from a): let M = Rx. …. Ran. let d = p1 1 … ptkt the prime factor decomposition of d (where p1.144 III. the ideal generated by GCD(a. piki and p j j are coprime if i ≠ j. o(xi) = o(yi) = ai. with o(x) = d = a1·…·an ∈ R° and (ai. If k d = 0. b) = 1 ⇔ Ra + Rb = R. Then: . b) We have M ≅ R/Rd. there exist xi ∈ M such that M = Rx1⊕…⊕Rxt. b ∈ R. with o(xi) = piki . 1 + Rai. pt are distinct primes in R).7 Corollary. By a).…. R/Ra1 ×…× R/Ran = Ry1⊕…⊕Ryn. ! 3. isomorphic to R (in this case M is torsion-free) or M is cyclic. Let yi := (0 + Ra1. applying b). so. Clearly. Proof. Let R be a PID and let M be a finitely generated R-module. (a. Finitely generated modules over principal ideal domains R R R ≅ ×… × Ra1 … an Ra1 Ran r + Ra1…an & (r + Ra1. with x ∈ M and let d = o(x). Obviously. So Rxi is k indecomposable. r + Ran). 1 ≤ i ≤ m. then there exist xi ∈ M. then M ≅ R is indecomposable. b) If M is a cyclic R-module M = Rx (x ∈ M). So. such that o(xi) = ai and M = Rx = Rx1⊕…⊕Rxm c) Any cyclic R-module can be written as a direct sum of indecomposable submodules. being isomorphic to R Rpiki . ∀r ∈ R.

if p ∈ P p-di. ! k 3. Indeed. So. q ≠ p}} = 0. There exists a decomposition (D). ax = . so R/Rd ≅ R Rp1 1 × … × R Rptkt . Keeping the notations in 2. then M = Rx. Let M be indecomposable finitely generated.III. as in theorem 2. q ≠ p (only a finite number of yq longs to the intersection. ∀q ≠ p for which yq ≠ 0.3. then tp(M) = 0. p|di}. piki and p j j are coprime if k i ≠ j. If m = 0. then tp(M) ∩ { ∑{tq(M) |q ∈ P. o(xi) = di ∈ R°. n = 1. If m = n = 1. and {p ∈ P | tp(M) ≠ 0} ⊆ {p ∈ P | ∃i. We know that M is a direct sum of cyclic submodules: M = Rx1⊕…⊕Rxm. If M is a finitely generated torsion module over a PID R. for all 1 ≤ i ≤ m. If t > 1. if x be- x = ∑q≠p yq. Let k d = p1 1 … ptkt be the prime decomposition of d (p1.3. More precisely. ∃kq ∈ N such that q q yq = 0. The proof of the “if” part is easy: we saw that the modules k R/Rp are indecomposable. k k are nonzero). From 2. then M is the direct sum of its p-torsion submodules. pt are distinct primes in R). which is clearly decomposable. also. with yq ∈ tq(M). and thus isomorphic to R. which is finite. with o(x) = d. M ≅ R/Rd. So. we see that M is indecomposable only if m = n = 1 or m = 0 and n = 1. then p x = 0 for some k ∈ N. Proof. Let a = ∏ y q ≠0 q kq (finite number of factors!). ….6 we deduce that. We claim that t = 1. if P is a system of representatives of the equivalence classes (with respect to association in divisibility) of the prime elements in R.8 Proposition. If p ∈ P. Evidently. then {p ∈ P | tp(M) ≠ 0} is finite and M is the finite direct sum: M = ⊕p∈P tp(M) Proof. where p ∈ R is prime and k ∈ N* (in this case M is a torsion module). So.3 Indecomposable finitely generated modules k 145 R/Rp .then M is free of rank 1. ∀q ∈ P.

By 3. with xi. The existence part is proven as follows: from theorem 2. p|o(xi)} = ⊕{Ryj | 1 ≤ j ≤ n.3. Let d = o(x) ∈ R and k d = p1 1 … ptkt be its prime piki decomposition. Ai torsion free} (and equal to the cardinal of the set {j | 1 ≤ j ≤ n. Also (p . Thus tp(M) = ⊕{Rxi | 1 ≤ i ≤ m. o(xi). and 3. Let x ∈ M.146 III. Bj torsion free}). We may suppose then that M = t(M) = Rx1⊕…⊕Rxm = Ry1⊕…⊕Ryn.e. i. Then M is a finite direct sum of indecomposable submodules. As in the proof of uniqueness in theorem 2. Proof. v ∈ R such that k k k up + va = 1. So.6 says that every cyclic submodule is a finite direct sum of indecomposable submodules.3. Let R be a PID and let M be a finitely generated R-module. Ryj indecomposable (by 3. The invariant factor theorem and the list of finitely generated indecomposable modules allow us to formulate the following structure theorem: 3. p|o(yj)}.7.b). with o(xi) = . The following property of uniqueness holds: if M = A1⊕…⊕Am = B1⊕…⊕Bn are decompositions of M as direct sums of indecomposable submodules.9 Theorem. o(yj) are prime powers in R). then m = n and there exists a permutation σ ∈ Sn such that Ai ≅ Bσ(i).6. yj ∈ M and Rxi. Finitely generated modules over principal ideal domains k a∑q≠p yq = 0. xi ∈ t pi (M ) . Let p be a prime in R. Then x = (up + va)x = up x + vax = 0.6). 1 ≤ i ≤ m. since p is not associated to any prime q occurring in the decomposition of a. a) = 1. there exist u. We prove that x ∈ ∑ p∈P tp(M). note that the free module M/t(M) has rank equal to the cardinal of the set {i | 1 ≤ i ≤ m. ! Rx = Rx1⊕…⊕Rxt. We have tp(Rxi) = 0 if po(xi) and tp(Rxi) = Rxi if o(xi) is a power of p (see 2. . M is a finite direct sum of cyclic submodules.

1 ≤ i ≤ r. (This definition can be generalized to any type of algebraic . so n = d1·…·dm .11 Examples. d1·…·dm = n and d1|…|dm. o(x1)|… |o(xr). …. Rxi ≅ Ryi). 2 . For 30). dm) of natural example. let {i | 1 ≤ i ≤ m. If M is a torsion module and Rx1⊕…⊕Rxm is o decomposition of M as a direct sum of indecomposables. by 3. the proof is ! finished. 3. 24).10 Remark. a) The elementary divisors of the Z-module 3 Z6⊕Z24 ≅ Z2⊕Z3⊕Z8⊕Z3 are (2.3 says that r = s and o(xi) = o(yi) (so. and d1|…|dm is the sequence of its invariant factors. dm) are (60). …. because in a set all elements are distinct. b) If G is an Abelian group with n elements. 3. The uniqueness part of theorem 2. Since M = ⊕p∈P tp(M) (see 3. numbers (m ≥ 1). Thus there are two types of isomorphism6 of a Abelian groups 5 We avoid the term “set”. The family of elements (o(xi))1≤ i ≤ m is thus uniquely determined (up to an association in divisibility and a permutation) by the module M and is called the family 5 of the elementary divisors of M.III. …. such that o(xi) = p ki and o(yj) = p j . for n = 60. 3). any Abelian group of order n is perfectly determined (up to an isomorphism) by an m-uple (d1.3 Indecomposable finitely generated modules 147 Relabeling if necessary. then G ≅ Z d1 ⊕ … ⊕ Z d m . p|o(yj)} =: {1. then the family of ideals (AnnR(xi))1≤ i ≤ m is independent of the chosen decomposition. r} and l {j | 1 ≤ j ≤ n. o(y1)|… |o(ys). Its invariant factors are (6.7. 3. The elementary divisors are powers of prime elements in R. s}.8). Then tp(M) = Rx1⊕…⊕Rxr = Ry1⊕…⊕Rys. while the elementary divisors can occur more than once. (2. such that: d1 ≥ 2. 6 The class of all groups isomorphic to a given group G is called the type of isomorphism of G. …. Thus. p|o(xi)} =: {1. the possible choices for (d1. with k1 ≤ … ≤ kr and l1 ≤ … ≤ ls.

5). Conversely. the elementary divisors can be obtained by decomposing in a product of powers of primes every di and writing down all the powers that occur. 3 .148 III. Which is this Z-module? structures: Abelian groups. which are the invariant factors of the Z-module. the invariant factors are obtained as follows: write a product containing (only once) all primes in the family of the elementary divisors. …. 3. 2·3. the theorem of invariant factors yields a classification of finitely generated Abelian groups. 2. in general) objective. ordered sets…). modules. the description of all types of isomorphism of the structure is a most important (and hard to attain. if the family of the elementary divisors of a Z-module 2 3 is (2. 5)). The product obtained is dm (the “largest” invariant factor – divisibility-wise). 5)) and Z2⊕Z30 ≅ Z2⊕Z2⊕Z3⊕Z5 (the elementary divisors are (2. . 2. For instance. if the family of the elementary divisors is given. dm) are given. called classification. fields. Finitely generated modules over principal ideal domains 2 with 60 elements: Z60 ≅ Z4⊕Z3⊕Z5 (the elementary divisors are (2 . 2 . c) If the invariant factors (d1. accomplished in the 1980's. Erase from the family of the elementary divisors the powers written in the product and repeat the procedure with what is left. For a given type of algebraic structure. at the greatest power. The classification of the finite simple groups (having no normal proper subgroups) is one of the great successes of group theory. 3. following the procedure above we obtain succes2 3 sively: 2 ·3 ·5. rings. 2. For example. as many times that they arise. Continue until the elementary divisors are exhausted. 3.

Then o(x)| r. where the condition (*) is: ∀s ∈ R. ∀i ≠ j. The idempotents of EndR(M) are 0 and 1M. M an R-module with AnnR(M) = Rr. so it has a maximal element F.III. I = (XY) = XYK[X. 2. (*) C is nonempty. (Hint. then ord(x + y) = ord(x)·ord(y)). and (ord x. AnnR(M/N) ≠ (0). o(xj)) = 1. 5. ∃xi ∈ M with o(xi) = pibi . We want to prove that M is a direct sum of cyclic submodules. then ∃ x0 ∈ C with sx = sx0. with (o(xi). 6. Let b bi = max {b ∈ N | ∃x ∈ M with pi |o(x)} (bi ≤ ai). Prove that: a) For any N ≤ R M. …. Let R be a PID and M an R-module (not necessarily finitely generated) with AnnR(M) ≠ (0). r ≠ 0. Give an example of a ring R and an indecomposable R-module that has a decomposable factor module. Can R be a PID? (Hint. ∀x ∈ M. Y]. r = ∏ pibi . 3. +) an Abelian group. 4. Prove that R M is indecomposable ⇔ the ring EndR(M) contains no idempotents other than 0 and 1M. Let R be a PID. o{∑ xi} = ∏o(xi) by ex. Let R be a PID. XK[X. Give an example of a ring R an indecomposable R-module that has decomposable submodules. M an R-module and x1. so bi = ai) c) Let C = {C ≤ M | C is a direct sum of cyclic submodules and satisfies condition (*)}. . Then o(x1 + … + xn) = o(x1)·…·o(xn). Let K be a field. xn ∈ M. if sx ∈ C. r = p1 1 … pn n . Set y =∑ xi. with pi prime in R and ai ∈ N*. Also. Let a a AnnR(M) = Rr. (G. (This generalizes the known fact: if a. 4. For any i. ∀x ∈ M.3 Indecomposable finitely generated modules 149 Exercises 1. R = K[X. inductively ordered. Y] and M = R/I. b ∈ G. Y] + YK[X. ord y) = 1. Y]}/I ≤ M and is decomposable). b) There exists y ∈ M such that AnnR(y) = AnnR(M).

K[X] is the polynomial ring in the indeterminate X with coefficients in K.) e) F = M. III. (Ind.150 III. we can apply the invariant factors theorem to obtain information on this K[X]-module (therefore on the endomorphism u). We endow the Abelian group (V. V is a finite dimensional K-vector space and u : V → V is a K-endomorphism (also called linear transformation or operator). Because K[X] is a PID. 4. if C ≠ M. . given by (*). n where u = u◦…◦u (n times). Finitely generated modules over principal ideal domains d) ∀C ∈ C. The idea of the theory we develop here is the following: The endomorphism u defines a natural structure of K[X]-module on V. f·v is the image of the vector v by the endomorphn ism f(u) := a0id + a1u + … + an u . then there exists D ∈ C such that C ( D. where id is the identity automorphism of V. the results obtained translate into matrix language. Let u ∈ EndK(V).4 The endomorphisms of a finite dimensional vector space Throughout this section. In other words.1 Definition. K is a field. Furthermore. keeping in mind the connection between endomorphisms and matrices. ∀v ∈ V. +) with a structure of K[X]-module (depending on u): n ∀f = a0 + a1X + … + anX ∈ K[X]. Then set D := C + R(y − y0). define: n f·v := a0v + a1u(v) + … + anu (v) ∈ V. M/C and obtain y∈M such that Apply b) to AnnR(M/C) = AnnR(y) =: µ. So µy ∈ C and let y0 ∈ C with µy = µy0.

It is clear that ϕ is also a K-automorphism. i. is a ring homomorphism (more exactly. as above. a K-algebra homomorphism). Can distinct endomorphisms of V define “the same” K[X]-module structure on V? 4.4 The endomorphisms of a finite dimensional vector space 151 Checking the module axioms is simple. If n f = a0 + a1X + … + anX ∈ K[X]. ∀v ∈ Vw. 4.4) that is the same as the one defined above. ∀f. ϕ◦u = w◦ϕ. We have ϕ(X ·u v) = X ·w ϕ(v). Then the K[X]-modules Vu and Vw are isomorphic if and only if there exists a K-automorphism −1 ϕ of V such that ϕ◦u◦ϕ = w. For any v ∈ Vu. so ϕ(u(v)) = w(ϕ(v)). Let Vu denote the K[X]-module V defined by the endomorphism u. w ∈ EndK(V). Proof. .III. The universality property of the polynomial ring K[X] ensures the existence of a unique K-algebra homomorphism η : K[X] → EndK(V) such that η(X) = u. The homomorphism η defines a K[X]-module structure on V (see remark II. X ·w v = w(v). ∀v ∈ V. are a consequence of the fact that: ( f·g)(u) = f(u)◦g(u) and ( f + g)(u) = f(u) + g(u). Let u ∈ EndK(V). Let ·u denote the external operation of the K[X]-module Vu. It is worth mentioning that the axioms ( f·g)·v = ( f )·(g·v) and ( f + g)·v = f·v + g·v ∀f. Let u. Let ϕ : Vu → Vw be a K[X]-isomorphism.3 Proposition. g ∈ K[X] This means that the “evaluation map in u” from K[X] to EndK(V). ∀f ∈ K[X]. f & f(u).2 Remark. X ·u v = u(v) and. n n η( f ) = f (u) = η(a0 + a1X + … + anX ) = a0idV + a1u + … + anu . g ∈ K[X]. The external operation defined above extends to K[X] × V the external operation (defined on K × V) of the K-vector space V.e. Note that the K[X]-module structure defined on V depends strongly on the endomorphism u.1.

Notation: u ≈ w. Let KV be finite dimensional and let v = (v1.152 III.6 Definition. w ∈ EndK(V) are similar if and only if the matrices Mv(u) and Mv(w) are similar. vn) be a basis of V. . ! for any a0 + a1X + … + anX ∈ K[X]. We say that V is indecomposable relative to u (or u-indecomposable) if V cannot be written as a direct sum of proper u-invariant subspaces. Then: W is an uinvariant subspace ⇔ ∀w ∈ W. n n ϕ((a0 + a1X + … + anX )·u v) = a0·wϕ(v) + a1X·wϕ(v) + … + anX ·wϕ(v). The K-subspace W of V is called invariant relative to u (or u-invariant) if u(W) ⊆ W. B ∈ Mn(K). 4.4 Definition. since ϕ is K-linear.5 Remark. Let u ∈ EndK(V) and W ≤ KV. ϕ◦u = w◦ϕ is written n n ϕ(X ·u v) = X ·w ϕ(v). then ϕ is a K[X]-isomorphism from Vu to Vw. The similarity relation on EndK(V) and the similarity relation on Mn(K) are equivalence relations. 4. A typical example of similar matrices is given by the matrices of an endomorphism in various bases of V. 4. We call the matrices A and B similar (denoted by A ≈ B) if there exists an invertible matrix U ∈ Mn(K) −1 such that B = UAU . X·w ∈ W ⇔ W is a K[X]-submodule of Vu. Indeed. From this we deduce that ϕ(X ·u v) = X ·w ϕ(v). 4. Finitely generated modules over principal ideal domains If ϕ : V → V is a K-isomorphism with ϕ◦u = w◦ϕ. b) Let n ∈ N and A. a) The endomorphisms u. To see that. ∀n ∈ N. …. w ∈ EndK(V) are called −1 similar if there exists ϕ ∈ AutK(V) such that ϕ ◦u◦ϕ = w. Let W be a nonempty set of V.7 Proposition. The endomorphisms u. recall that the rings EndK(V) and Mn(K) are (anti-)isomorphic by the mapping u & Mv(u).

” (In this case.3: . where p is an irreducible polynomial in K[X] and k ∈ N*.. …. then.3. By n induction. u is called a cyclic endomorphism). u(v). since W is a linear subspace in V. Vm”.8 Proposition. …. finitely generated K[X]-module. Vm” ⇔ “K[X]Vu is the direct sum of the K[X]-submodules V1. …. If dimKV = n ∈ N*. The converse is an ! easy exercise. III. ∀v ∈ V. ∀n ∈ N. But Vu is a torsion. “V is the direct sum of the u-invariant subspaces V1. X ·w ∈ W.2 and III.III. u (v) cannot be linearly independent. “V is u-indecomposable” ⇔ “Vu is an indecomposable K[X]-module”. not all zero. b) V is u-indecomposable if and only if Vu is an indecomposable K[X]-module. this n implies (a0 + a1X + … + anX )·w ∈ W. ….7. the vecn tors v. there exist a0. with o(v) = p . i “There exists v ∈ V such that V is generated by {u (v) | i ∈ N}” ⇔ “Vu is a cyclic K[X]-module (generated by v). b) V is u-indecomposable if and only if there exists v ∈ Vu such that k Vu = K[X]v. The following statements (whose proofs are left to the reader) are further examples of translations from vector space language to K[X]-module language. a) Any system of generators KV is also a system of generators for the K[X]-module Vu.4 The endomorphisms of a finite dimensional vector space 153 Proof. ∀ai ∈ K. Consequently. such that a0v + a1u(v) + … + n n anu (v) = 0 ⇔ (a0 + a1X + … + anX )·v = 0. ! Apply now Prop. u(w) ∈ W ⇔ X·w ∈ W. a1. a) The K[X]-module Vu is a finitely generated torsion module. an ∈ K. Proof. 4. u(W) ⊆ W if and only if ∀w ∈ W. We apply to the K[X]-module Vu the structure theorems from III.

o(wt) are the ! elementary divisors of the K[X]-module Vu). ….…. where wi ∈ Vu and o(wi) are powers of irreducible polynomials in K[X]. 4. . If dimK V = n ∈ N*. If dimKV = n and u ∈ EndK(V). Let u.…. The natural number m and the monic polynomials o(v1). Proof. Two endomorphisms (matrices) are similar ⇔ they have the same invariant factors ⇔ they have the same elementary divisors. the invariant factors of a matrix A ∈ Mn(K) are the invariant factors of the unique endomorphism u with Mv(u) = A (similarly for the minimal polynomial. w ∈ EndK(V).…. cu o(v1)|…|o(vm).….154 III. then there exists m ∈ N* and v1.11 Proposition. 1 ≤ i ≤ t. such that Vu is a direct sum of u-invariant subspaces Vu = K[X]v1⊕…⊕K[X]vm. o(wt) ∈ K[X] are uniquely determined (o(w1). We have: Mv(u) ≈ Mv(w) ⇔ u ≈ w ⇔ Vu ≅ K[X]Vw (by 4.3) ⇔ Vu and Vw have the same invariant factors ⇔ Vu ! and Vw have the same elementary divisors. The natural number t and the monic polynomials o(w1). o(vm) ∈ K[X] with the properties above are uniquely determined (o(v1).9 Proposition. The monic polynomials that are the invariant factors (respectively the elementary divisors) of the K[X]-module Vu are called the invariant factors (respectively the elementary divisors) of the endomorphism u. the elementary divisors). vm ∈ Vu. 4.10 Definition. o(vm) are the invariant factors of the K[X]-module Vu). the monic polynomial that generates AnnK[X](Vu) is called the minimal polynomial of the endomorphism u (and is denoted by µu). Finitely generated modules over principal ideal domains 4. Vu is also a direct sum of indecomposable submodules: Vu = K[X]w1⊕…⊕K[X]wt. The same terminology is used for matrices: choosing a basis v in V.

Then v = (e1. ⎢ ⎥ Am ⎦ ⎣0 b) Conversely. …. …. …. then v1∪…∪vm =: v is a basis7 in V. the highest degree invariant factor of u. With the notations in 4. a) It is clear that v is a basis in V (see also II. the following properties are equivalent: a) f = µu. using the definitions.9. in this order. Writing the matrix of u in the basis v. fq). 1 ≤ i ≤ m.4. If f ∈ K[X] is monic. The proof is easy. 7 . 1 ≤ i ≤ m. u(ei) is a linear combination of e1. then the rows of the block Ai correspond to a set of vectors in v that generate an u-invariant subspace Vi (1 ≤ i ≤ m) and V = V1⊕…⊕Vm. …. fq. where V1. in the basis vi. unlike the minimal polynomial of an algebraic element in a field extension. c) f (u) = 0 and ∀g ∈ K[X]. …. b) f (u) = 0 and ∀g ∈ K[X]. µu is o(vm). a) Let V = V1⊕…⊕Vm. implies deg f ≤ deg g. f1. …. Note that. v2 = ( f1. ep. the minimal polynomial of an endomorphism is not necessarily irreducible. p + q = n = dimV. suppose m = 2 and v1 = (e1. If vi is a basis in Vi. g ≠ 0. To keep notations manageable. ….12 Remark. Since V1.4 The endomorphisms of a finite dimensional vector space 155 4. g(u) = 0. V2 are u-invariant. if the matrix of u in a basis v is of the form above. Vm are u-invariant subspaces. 4. and u( fj) is a linear combination of f1. then the matrix of u in the basis v is (written on blocks): 0⎤ ⎡ A1 ⎢ ⎥ Mv(u) = ⎢ * ⎥.III. vm. We totally order the vectors in the basis v. If Ai is the matrix of the restriction of u to Vi. by sequencing the elements of the bases v1. ep. fq). …. Proof. The next result translates in matrix language the fact that V is a direct sum of u-invariant subspaces.13 Proposition. ep). g(u) = 0 implies f |g.11).

o(v) = p . ⎢ 0 A2 ⎥ ⎣ ⎦ b) The task of detailing the proof is left to the reader. p irreducible in K[X]. p = X − ar −1 X −… − a1 X − a0. ! We want to find a basis v of V such that Mv(u) has as “simple” a form as possible. If p ∈ K[X]. Finitely generated modules over principal ideal domains ⎡ A1 0 ⎤ Mv(u) = ⎢ ⎥.156 III.9) the previous result allows us to study the restriction of u to each of the u-invariant subspaces in the direct sum.14 Definition. . k ∈ N*. French mathematician. It is thus natural to study first the case in which Vu is indecomposable: k Vu = K[X]v. Since Vu is a direct sum of indecomposable submodules (theorem 4. N = ⎢… ⎥ ⎢ 0 0 0 … 1 ⎥ ⎢0 0 0 … 0⎥ ⎢ ⎥ ⎢ ⎥ ⎢1 0 0 … 0⎥ ⎢ a0 a1 a2 … ar −1 ⎥ ⎣ ⎦ ⎣ ⎦ The matrix Cp is called the matrix companion of the polynomial p. define the r×r matrices with entries in K: ⎡ 0 1 0 … 0 ⎤ ⎡0 0 0 … 0⎤ ⎢ 0 0 1 … 0 ⎥ ⎢0 0 0 … 0⎥ ⎢ ⎥ ⎢ ⎥ Cp = ⎢ ! * ⎥ . 4. for some v ∈ V. Define the rk×rk matrix (written in block form): Cp N 0 … 0 0 0 Cp N … 0 0 * * k J{ p } = ∈ Mrk(K) * * 0 0 0 … Cp N 0 0 0 … 0 Cp r r −1 J{ p } is called the Jordan cell 8 associated to the polynomial p . k k 8 Camille Jordan (1838-1922).

a matrix of the form k ⎡ J p1 1 ⎢ k J p2 2 ⎢ ⎢ ⎢ ⎣ 0 ( ) ( ) * ⎤ ⎥ ⎥. …. …. ekr−1) such that k Me(u) = J{ p } : In other texts this matrix is called a rational canonical matrix.4 The endomorphisms of a finite dimensional vector space 157 A matrix (written in block form) having on the diagonal Jordan cells (and 0 elsewhere). let p1 1 . Then dimKV = rk and there exists a basis v of KV such that the k k matrix of u in v is the Jordan cell J{ p } associated to p . ptkt be the elementary divisors of u. a) We exhibit a basis e = (e0. with p1. is called a Jordan canonical matrix9 over K.III.e. 4. 9 .15 Proposition. a) Suppose Vu is a indecomposable K[X]-module k and v ∈ Vu is such that Vu = K[X]v. J= ⎢ ⎢ ⎥ * ⎢ ⎥ J ptkt ⎦ ⎣ 0 ( ) ( ) ( ) Proof. i.… . Then there exists a basis of V in which the matrix of u is the Jordan canonical matrix: k ⎡ J p1 1 0 ⎤ ⎢ ⎥ k2 J p2 ⎥. with o(v) = µu = p . k b) In the general case. ⎥ ⎥ J ptkt ⎦ 0 ( ) where p1. pt irreducible and monic in K[X]. where r r−1 p = X − ar−1X −… − a1X − a0 ∈ K[X] is irreducible of degree r and k ∈ N*. pt are monic irreducible polynomials in K[X]. the name Jordan canonical matrix being given only if pi are polynomials of degree 1. ….

… . …. By a). gn−1·v are thus linearly independent in V. we have f |h. hence h = 0. Apply now Prop. Since o(v) = f. gn−1·v form a basis of V. ! k .13. …. …. so also a K-vector space isomorphism). then. u(ekr−1) = a0e(k −1)r + a1e(k −1)r+1 + … + ar−1e(k−1)r+r−1. each such subspace has a basis in which the k k restriction of u has the matrix of the form J{ p }. we have: r−1 i−1 r i−1 r−1 i−1 u(eir−1) = X(X p ·v) = X p ·v = (p + a0 + a1X + … + ar−1X )p ·v i i−1 i−1 r−1 i−1 = p ·v + a0p ·v + a1Xp ·v + … + ar−1X p ·v = eir + a0e(i −1)r + a1e(i −1)r+1 + … + ar−1e(i −1)r+r−1 If i = k. deg f = n. …. with ai ∈ K. 1 ≤ i ≤ n. with p elementary divisor of u. The n elements g0·v. …. ekr−1) is a basis: Lemma. If 1 ≤ i < k. being of distinct degrees. so a0 = … = an−1 = 0. …. hence dimKV = dimK K[X]/(f ) = deg f = n. Finitely generated modules over principal ideal domains r−1 e0 = v. e2r−1 = X p·v = u(e2r−2). the vectors g0·v. gn−1·v are linearly independent: if a0g0·v + … + an−1gn−1·v = 0. e(k −1)r+1 = Xp·v = u(e(k −1)r). … k −1 r−1 k −1 e(k −1)r = p ·v. for any g0. and deg h < n. These equalities.9). ekr−1 = X p ·v = u(ekr−2). r−1 er = p·v. together with the relations (0). The vectors g0·v. We get back to proving that Me(u) = J{ p }. 4. Proof of the lemma. … . gn−1 ∈ K[X].158 III. e1 = X·v = u(e0). …. with deg gi = i. since k p ·v = 0. The next lemma proves that e = (e0. prove the claim. …. er−1 = X ·v = u(er−2). (k − 1). then h·v = 0. Vu ≅ K[X]/(f ) (K[X]-module isomorphism. which means they are a basis. whose dimension is n. But the polynomials g0. and f = o(v). If u ∈ EndK(V) is an endomorphism such that Vu = K[X]v for some v ∈ V. gn−1 are linearly independent in the K-vector space K[X]. where h = a0g0 + … + an−1gn−1. er+1 = Xp·v = u(er). b) Decompose Vu as a direct sum of u-indecomposable u-invariant subspaces (see 4.

…. Then the diagonally canonical matrix D ∈ Mn(K[X]). If the elementary divisors of A are p1 1 .3): since (v1. Let v = (v1. …. such that φi = diεi. ptkt . ϕ(ei) = vi. en) and the K[X]-homomorphism ϕ : E → Vu. d2. (see the proof of 2. di ∈ K[X]. rank F = n and there exist two bases ε = (ε1. Any matrix A ∈ Mn(K) is similar to a Jordan k canonical matrix. The invariant factors of Vu are the noninvertible polynomials among d1. …. vn) be a basis in V such that Mv(u) =: A = (aij) ∈ Mn(K).… .…. How do we find the elementary divisors of an endomorphism u (of a matrix A)? This amounts to finding its invariant factors. there exists a matrix B ∈ Mn(K[X]) such that f = Be. we need such a matrix B. Since Vu is torsion. The homomorphism ϕ is surjective. XI − A = ⎢ … ⎥ ⎢ −a − an 2 … X − ann ⎥ ⎣ ⎦ n1 Proof. Then the invariant factors of u (of the matrix A) are the polynomials of degree > 0 on the diagonal of the diagonally canonic matrix D ∈ Mn(K[X]). . 4. Recall the proof of the invariant factors theorem (2. take a K[X]-module E. vn) generates K[X]Vu. φn) in F. dn? For any basis f in F. Let fi =: Xei − (ai1e1 + ai2e2 + … + ainen) ∈ E.17 Theorem. arithmetically equivalent to B.….b). d2. dn). dn. then ! A ≈ J. εn) in E and φ = (φ1.…. The non constant polynomials on the diagonal are then the invariant factors of A. d2. The next theorem says: take the matrix XI − A and find its Smith normal form. 1 ≤ i ≤ n. d1|d2|…|dn.III. is exactly D = diag (d1. Kerϕ =: F is a free submodule in E and E/F ≅ Vu. 1 ≤ i ≤ n. where J is the matrix at 4.16 Corollary. Thus. …. …. free of basis e = (e1.4 The endomorphisms of a finite dimensional vector space 159 4. arithmetically equivalent to the matrix ⎡ X − a11 − a12 … − a1n ⎤ ⎢ −a X − a22 … − a2 n ⎥ 21 ⎢ ⎥ ∈ Mn(K[X]).3) How do we find d1.15.

hence r = 0. then these relations can be written f = (XI − A)e. X ei is expressed as: m X ei = ∑j qj fj + ∑i rjej.…. u(vi) = ai1v1 + ai2v2 + … + ainvn Thus. Finitely generated modules over principal ideal domains If f = (f1. deg gi ≤ deg g1. then y = ∑i giei = ∑i qi fi + r. gei can be written in the same form. So. one easily sees that. 1 ≤ j ≤ n. so ϕ(r) = ϕ(∑i ciei) = ∑i civi = 0. (*) So. By inducm tion. ϕ(fi) = Xvi − (ai1v1 + ai2v2 + … + ainvn) = u(vi) − (ai1v1 + ai2v2 + … + ainvn) = 0. we obtain y = ∑i qi fi. X ei = Xfi + ai1 Xe1 + ai2 Xe2 + … + ain Xen. gi X = ∑j aji gj. Using (*). rj ∈ K. 1 ≤ i ≤ n. Let us prove the linear independence. fn) is linearly independent. Lemma. en) is a basis. 1 ≤ i ≤ n. 1 ≤ i ≤ n. Let g1 be the polynomial of maximal degree among the gi (relabel if necessary). This shows that ( f1. (**) 2 From (**) we deduce that. f2. fn) is a basis in F = Ker ϕ. f =: ( f1.160 III. Proof of the lemma. fn). for some qj ∈ K[X]. for some qj ∈ K[X]. 1 ≤ j ≤ n. f2. …. Let us prove that f is a system of generators for F. contradiction. if y = ∑i giei ∈ F. rj ∈ K. For any 1 ≤ i ≤ n. ∀m ∈ N*. This shows that fi ∈ Kerϕ = F. with gi ∈ K[X]. with ci ∈ K. for some qj ∈ K[X]. ci = 0. and r ∈ E is of the form ∑i ciei. 1 ≤ i ≤ n. Finally. . f2. 1 ≤ i ≤ n. Since (e1.…. 1 ≤ i ≤ n. we obtain: 2 X ei = ∑j qj fj + ∑i rjej. with gi ∈ K[X]. Using (*). If g1 ≠ 0. So. Since v is basis in V. g1 X = ∑j aj1 gj implies deg{∑j aj1 gj} ≤ max j(deg(aj1 gj)) ≤ max j(deg gj) < 1 + deg g1 = deg g1X. it follows that ∑i gi Xei = ∑i gi { ∑j aij ej} = ∑j{ ∑i aijgi }ej.…. But r = y − ∑i qi fi ∈ F. Note that: Xei = fi + ai1e1 + ai2e2 + … + ainen. Suppose ∑i gi fi = 0. ∀g ∈ K[X].

If u is an endomorphism of the K-vector space V and A is the matrix of u (in some basis). …. the minors obtained by selecting k rows {i1. More generally. 2. There are ⎜ ⎟ such minors. …. dn. for some S ∈ GLn(K). cn are: c1 = a11 + a22 + … + ann =: Tr(A) (called the trace of A). This definition is correct: Two similar matrices A and B have the −1 same characteristic polynomial: if B = SAS . then the characteristic polynomial fu of u is by definition fA.. dk + 1. .e. …. ! 4.III. The relations fi = Xei − (ai1e1 + ai2e2 + … + ainen) show that f = (XI − A)e.4 The endomorphisms of a finite dimensional vector space 161 We continue the proof of the theorem. Let A = (aij) ∈ Mn(K) and let n n−1 n−2 n fA = det(XI − A) = X − c1 X + c2 X + … + ( − 1) cn. where k = min{i | di noninvertible}. Let A ∈ Mn(K). n}. cn = det(A). ik} of the ⎛n⎞ matrix A). …. ik} and the columns with the same indices {i1. 4. ….19 Remark. one for each choice of a subset ⎝k ⎠ of k indices from {1. The polynomial fA := det(XI − A) ∈ K[X] is called the characteristic polynomial of the matrix A.18 Definition. cn ∈ K.…. We used the fact that the matrix XI commutes with any matrix in Mn(K[X]). then the invariant factors of Vu (of the endomorphism u) are dk. with c1. −1 −1 then: fB = det(XI − SAS ) = det(S·(XI − A)S ) = −1 = det(S)·det(XI − A)·det(S ) = fA. If XI − A ∼ D = diag(d1. d2. with d1|d2|…|dn. dn). Writing the definition of det(XI − A) and arranging the terms by the like powers of X. …. ck (1 ≤ k ≤ n) is the sum of the minors of order k of A on the main diagonal (i. the coefficients c1.

1. Its characteristic polynomial is fu = fA.15. since obviously S ∈ GL(n. c1 =: Tr(u) = Tr(A) is called the trace of u and cn =: det(u) = det(A) is called the determinant of u. detS. The following results are about matrices in Mn(K). The matrix XI − A is arithmetically equivalent to the matrix in Smith normal form D = diag(1. By 4. A and B have the same invariant factors: the polynomials of degree > 0 on the diagonal of D. Thus. −1 Proof. The characteristic polynomial of a matrix A is the product of the invariant factors of A (and it is equal to the product of the elementary divisors of A). 4. Let A. dm are the invariant factors of A. d1. then there exists S ∈ GL(n. whose matrix is A (in some basis of V). cn are uniquely defined by u. ! Let n ∈ N*. Suppose now that XI − A ∼ XI − B and let D ∈ Mn(K[X]) be the canonical matrix with D ∼ (XI − A) ∼ (XI − B). dm) ∈ Mn(K[X]). There exist S.e. according to 4.17. and the coefficients c1. Then A and B are similar matrices if and only if XI − A and XI − B are arithmetically equivalent matrices in Mn(K[X]). detT ∈ K*) such that XI − A = SDT. A and B have the same elementary divisors. B ∈ Mn(K). …. …. 4. as above. So. T ∈ U(Mn(K[X])) (i. dimV = n. Proof. K[X]). A and B are similar with the same Jordan canonical matrix.162 III. …. If A ≈ B. Finitely generated modules over principal ideal domains The same terminology applies for an endomorphism u ∈ EndK(V).20 Proposition. so XI − A ∼ XI − B. We have d1·…·dm = detD = det(S(XI − A)T) = detS·fA·detT . …. where d1. but they can be translated in statements on endomorphisms of a K-vector space of dimension n. −1 Then XI − B = S (XI − A)S.21 Proposition. K) such that B = S AS.

1 ≤ i ≤ t. On the other hand. Then the elementary divisors of A are piki . The next proposition shows that these are all the Jordan canonical matrices similar to A.III. Sir William Rowan Hamilton (1805-1865).23 Proposition. diverse Jordan canonical matrices are obtained. a) Let p ∈ K[X] be monic and irreducible and let k k ∈ N*. by reordering the elementary divisors. namely k p . b) Let A be a Jordan canonical matrix whose diagonal is made up by the Jordan cells J piki . b) (The Frobenius theorem)11 The characteristic polynomial and the minimal polynomial of a matrix A ∈ Mn(K) have the same irreducible factors in K[X]. The Jordan cell J{ p } has only one elementary divisor. 1 ≤ i ≤ t. b) This is a consequence of d1·…·dm = fA and d1|…|dm. it is clear that the product of the elementary divisors equals the product of the invariant factors. Let d1. ( ) Arthur Cayley (1821-1895). a) (The Cayley-Hamilton theorem) 10 Any matrix A ∈ Mn(K) is a root of its characteristic polynomial: fA(A) = 0. dm(A) = 0 and dm| fA. a) The minimal polynomial of A is dm (the invariant factor of highest degree). dm ∈ K[X] be the invariant factors of A. So. 10 . is the Jordan canonical matrix that is similar to A uniquely determined? Of course. ! Given a matrix A. ! 4. German mathematician.4 The endomorphisms of a finite dimensional vector space 163 This means that the monic polynomials d1·…·dm and fA differ by the factor detS·detT ∈ K*. where pi ∈ K[X] are monic and irreducible. Proof. British mathematicians.22 Corollary. thus fA(A) = 0. which shows that they are equal. 4. 11 Ferdinand Georg Frobenius (1849-1917). ….

perhaps in different order. then J is called the Jordan canonical form of A. b) There exists KV and u ∈ EndK(V).164 III. So. ∀y ∈ V (the dot denotes the external K[X]-module operation of V). 1 ≤ i ≤ t. then A and B have the same elementary divisors (4.11). V is a K[X]-module (being a factor module of K[X]) and a K-vector space of dimension k·deg p =: n. So J{ p } has the k same elementary divisors as Vu. The classical notions of eigenvector and eigenvalue of an endomorphism are closely connected to its invariant subspaces of dimension one. The matrix of ui is exactly J piki and the first part of the proof shows that piki is the only elementary divisor of ui. a) Let V := K[X]/{ p }. V = Vu is an indecomposable k K[X]-module. namely piki . 1 ≤ i ≤ t.1. The endomorphism u defines on V a structure of a K[X]-module. Finitely generated modules over principal ideal domains c) Let A. If A ∈ Mn(K) and J is a Jordan canonical matrix such that J ≈ A.24 Definition. c) If A ≈ B. Let ui be the restriction of u to Vi. u(y) = X·y. and its only elementary divisor is o(v) = p . Then A and B have the same Jordan cells. Then V is written as a direct sum of u-invariant subspaces: Vu = V1⊕…⊕Vt. 1 ≤ i ≤ t. B ∈ Mn(K) be Jordan canonical matrices such that A ≈ B. k Proof. k where v := 1 + { p } ∈ V. The Jordan canonical form of A is uniquely determined up to an order of the Jordan cells on the diagonal. namely only p . Let u ∈ EndK(V). The elementary divisors of u are obtained writing down all the elementary divisors of the restrictions ui. Vu. There exists k k a basis (as in 4. It is easy to see that these two structures of K[X]-module coincide and that V = K[X]v. ! ( ) ( ) 4. A and B have the same Jordan cells. . By b).15) in which u has the matrix J{ p }. as in definition 4. Vi being the u-invariant subspace corresponding to the Jordan cell J piki . whose matrix is A (in some basis).

then X − λ| fu. b)⇔c) This follows from dimK K[X]v = deg o(v). So. If u ∈ EndK(V). i. thus (X − λ)·v = 0. Proof. . this means X·v = λv.III. ! The reader is invited to give an alternate proof. Suppose v ∈ V is an eigenvector of u for the eigenvalue λ. Let v ∈ V. a)⇒b) We have u(v) = λv. Then o(v) = X − λ.4 The endomorphisms of a finite dimensional vector space 165 4. Each such vector v is called an eigenvector of u for the eigenvalue λ.9. λ ∈ K is an eigenvalue of u if and only if λ is a root of fu. Because fu and µu have the same irreducible factors. (o(v) = 1 is impossible. Also µu | fu. hence o(v) = X − λ. The following statements are equivalent: a) v is an eigenvector of u for the eigenvalue λ. u ∈ EndK(V) and λ ∈ K. the characteristic polynomial of u. is in AnnK[X](v) = (X − λ). µu. In K[X]-module language for Vu. there exists v ∈ V such that o(v) = µu. X − λ|µu. if fu(λ) = 0. so X − λ| fu. ! 4. since the minimal polynomial of u. o(v) = X − λ ∈ K[X]. o(v)| X − λ. which means that λ is a root of fu.e.26 Proposition. v ≠ 0. Then o(g·v) = X − λ. Conversely. 4. such that u(v) = λv. which is irreducible. using facts from the theory of systems linear equations. b) Considering v ∈ K[X]Vu. so µu = (X − λ)g for some g ∈ K[X]. since it implies v = 0). c) dimK (K[X]v) = 1 (the submodule of Vu generated by v has K-dimension 1).27 Proposition. b)⇒a) o(v) = X − λ implies u(v) = λv.25 Definition. g·v ∈ V is an eigenvector for the eigenvalueλ. an element λ ∈ K is called an eigenvalue of u if there exists a vector v ∈ V. By 4. we have X − λ|µu. Proof.

Finitely generated modules over principal ideal domains Note that if fu has no roots in K. c ∈ R and b + 4c < 0. If K is an algebraically closed field (in particular. or X − bX − c. 2 with b. the monic irreducible polynomials in R[X] are X − a. then the irreducible monic polynomials in K[X] are of the form X − a. k Thus the Jordan cell J{(X − a) } is: a 1 0 … 0 0 0 a 1 … 0 0 ∈ Mk(K) . k 2 a ∈ R (and the Jordan cell J{(X − a) } is the one above). 1 J{(X − a) } = 0 0 0 0 k * * * * 0 0 … … a 0 a If K = R. then u has no eigenvalues and no eigenvectors. We describe now the Jordan cells in the important cases K = C and K = R. K = C). a ∈ K . * * * * ⎢ ⎥ ⎢ 0 1 0 0⎥ ⎢ ⎥ c b 1 0⎥ ⎢ ⎢ 0 0 1⎥ ⎢ ⎥ c b⎦ ⎣ . in which case the Jordan cell 2 k J{(X − bX − c) } is: ⎡0 1 0 0 ⎤ ⎢c b 1 0 0 ⎥ ⎢ ⎥ 0 1 0 0 ⎢ ⎥ ⎢ ⎥ c b 1 0 ⎢ ⎥ * * * * ⎢ ⎥ 2 k J{(X − bX − c) } = ⎢ ⎥ ∈ M2k(R).166 III.

p(λn). symmetric. there exists t −1 U ∈ GL(n. Moreover. Then A is similar to A. The diagonal of T is λ1. V is a finite dimensional K-vector space and u is an endomorphism of KV. λn. In this case. Prove that the eigenvectors corresponding to distinct eigenvalues of u are linearly independent. Can you exhibit four such matrices? Generalization. such that A = U AU. (Ind.III. Let A ∈ Mn(K) such that the characteristic polynomial of A splits in factors of degree 1 in K[X] (one says that A has all its eigenvalues in K). Determine the endomorphisms u ∈ EndK(V) whose minimal polynomials are of degree 1. ⎢ − 21 9 7⎥. …. 3. Give an example of two matrices having the same minimal polynomial and the same characteristic polynomial. 2. then p(A) has the eigenvalues p(λ1). . K is a field. ….4 The endomorphisms of a finite dimensional vector space 167 Exercises In the exercises. Compute the characteristic polynomial and the minimal polynomial of the following matrices: ⎡ ⎢ ⎢ ⎢ ⎢ ⎣ 1 0 0 1 1 −1 0 0 1 − 2⎤ ⎡1 ⎥ ⎡ − 7 3 3⎤ ⎢0 0 0 ⎥. …. Compute p(T). λn in K. Is the converse true? 6. 5. A is called trigonable. with T upper triangular. 4. K). Then A is similar to an upper triangular matrix T = (tij) ∈ Mn(K) (tij = 0 if i > j). 1. Give an example of three matrices in M3(Q) whose only eigenvalue is 2 and any two matrices are not similar. Let A ∈ Mn(K) and let p ∈ K[X]. ⎢ ⎥ ⎢0 0 0⎥ ⎢ ⎢ − 6 2 4⎥ ⎢ ⎣ ⎦ 0 2⎥ ⎦ ⎣0 t 0 1 1 0 1 0 1 0 0⎤ 1⎥ ⎥ 0⎥ 1⎥ ⎦ 8. Let A ∈ Mn(K). but are not similar. If A has all its eigenvalues λ1. Let A ≈ T.) 7.

Finitely generated modules over principal ideal domains 9. A. 13. So. Then A is a R-linear combination of I. u has an invariant subspace of dimension dim V − 1. For any fixed k.) 14. Let V = U ⊕ W (direct sum of subspaces). Conversely. Then any v ∈ V can be uniquely written as v = u + w. Let R be a commutative ring with identity and let A ∈ U(Mn(R)) be an invertible matrix. ∀w ∈ W. Use the Cayley-Hamilton theorem. π(u + w) = u (π is called the projection on U along W) and ρ(u + w) = u − w. multiply relation i with Bik and sum after i to obtain fu(X)·ek = 0 = fu(u)(ek). ∀k ≥ 1.168 III. Let u ∈ EndK(V) be an endomorphism having the eigenvalue 0. Define π. …. (Ind. Let u ∈ EndK(V) be a nilpotent endomorphism (∃r ≥ 1 such that r k u = 0). . Then V = U ⊕ W. Let R be a commutative ring with identity and let n ∈ N*. W u-invariant subspaces and dim U = 1. 12. (ρ is called the symmetry with respect to U along W). Generalize the relevant notions and prove the Cayley-Hamilton theorem: if E is a free R-module of rank n and u ∈ EndR(E). Let Bik ∈ R[X] be the algebraic complement of bik in the matrix B.) 11. with U. n−1 A . en) and XI − A = B = (bij) ∈ Mn(R[X]). (Hint: let A the matrix of u in a basis (e1. ∀k ≥ 1. Let f be the characteristic polynomial of u. In the R[X]-module Eu the relations ∑j bijej = 0 hold. …. Then Tr u = 0. then u is nilpotent. ρ : V → V by: ∀u ∈ U. w ∈ W. fu = det B. with u ∈ U. (Hint. then u is a root of the characteristic polynomial: fu(u) = 0.) 10. Let u ∈ EndK(V). Show that π and ρ are K-endomorphisms of V and find their minimal polynomials. ∀i. if char K = 0 and k Tr u = 0. in the relation f(u) = 0 apply Tr and deduce that 0 is an eigenvalue of u. Then V has no u-invariant proper subspaces ⇔ the K[X]-module Vu is simple ⇔ the characteristic polynomial of u is irreducible in K[X].

169 . Standard facts about rings and vector spaces are a prerequisite: polynomial rings. y ∈ K. bases and dimension in vector spaces. A field K has no zero divisors (in other words. x ≠ 0 and y ≠ 0 implies xy ≠ 0. ·) with identity 1 (with 1 ≠ 0). Any field has at least two elements: 0 and 1.1. factor rings. IV. with the property that every nonzero element is invertible with respect to multiplication. Some elementary properties of cardinals and Zorn's Lemma are used in the proof of existence of the algebraic closure of a field.IV. a more detailed treatment is found in most Abstract (Modern) Algebra introductory texts. prime and maximal ideals. K is a domain): for any x. Knowledge of polynomial ring arithmetic is recommended (as provided in the chapter “Arithmetic in integral domains”). Most of these facts can be found in the Appendices. ring isomorphism theorems. + . Field extensions This chapter contains the basic concepts and results from the theory of field extensions. Algebraic extensions Recall that a field is a commutative ring (K.

The subfield K of L is called proper if K ≠ L. the map σ : K → L is a homomorphism if and only if. we call the triple (K. but R can be identified with the set of the couples (a. it is natural to consider that C is an extension of R. L are rings (with identity). This notion of field extension is too restrictive. ϕ(a) = (a. b ∈ R. Recall that the nonempty set K of the field L is a subfield in L if it is closed under addition and multiplication and becomes a field with the induced operations. if σ : K → L is a field homomorphism. The homomorphism σ : K → L being injective. ∀x. In this setting.170 IV. In fact. we −1 have x − y and xy ∈ K. L be fields. 0) and then identifies R with its image ϕ(R). σ) a field extension of K. x ≠ 0 implies ∃x ∈ K. we also say that “L is an extension of K”. L. then σ is −1 injective. which is a subfield of C. For instance. a ∈ R. It is thus natural to consider the following definition: 1. The objects we study are field extensions: if L is a field and K is a subfield in L. we may identify the field K with its image σ(K). which is a subfield in L.1 Definition. R is not a subset of C. We ex- . y ∈ K with y ≠ 0. Field extensions All rings and all ring homomorphisms considered are supposed to be unitary. if K. Thus. ∀x ∈ K. If σ : K → L is a field homomorphism. A widely used characterization is: K is a subfield of L if and only if for any x. b) with a. 0). More generally. so −1 −1 σ(x)σ(x ) = σ(xx ) = σ(1) = 1. a) Let K. the rigorous interpretation is the following: one defines the field homomorphism ϕ : R → C. But C is usually constructed as the set of all couples (a. Indeed. endowed with an addition and a multiplication that make it a field. so σ(x) is nonzero and thus Kerσ is {0}. y ∈ K: σ(x + y) = σ(x) + σ(y) σ(xy) = σ(x)σ(y) σ(1) = 1.

b ∈ . b) The field C of complex numbers is an extension of R. a) The real numbers field R is an extension of Q. e) If n is a natural number.24.1. let Fp denote the field of integers modulo p. This identification allows to consider K as a subfield of L. and it is an } extension of Q.IV. For any prime p. For example. if a ∈ K and x ∈ L. The extension K ⊆ L is called proper if the inclusion is strict (i. “L/K is a field extension” or “L is an extension of K”. σ(K) ⊂ L). the ring Zn of the integers modulo n is a field if and only if n is a prime. Let HomK(L. a mapping ϕ : L → E is called a K-homomorphism if ϕ is a ring homomorphism and ϕ|K = idK (the identity function of K). the field of rational fractions with coefficients in K (K(X) is the quotient field of the polynomial ring K[X]) is an extension of K. 1. ϕ is a K-algebra homomorphism). [ 2 ] = a + b 2 ∈ - { a. E) denote the set of all K-homomorphism from L to E. Algebraic extensions 171 press this by writing “K ⊆ L is a field extension”.e. we call ϕ : L → E a K-homomorphism if ϕ is a field homomorphism with ϕ ◦σ = τ (in other words. τ : K → E. is a field. b) If K ⊆ L and K ⊆ E are extensions of K. In this context. then K(X). d) If K is a field. i. If one considers the general definition above. we call an intermediate field (or subextension) of the extension K ⊆ L any subfield E of L that includes K. Note that L is an extension of K if and only if L is a field also structured as a K-algebra (the structural homomorphism is precisely σ). c) The set . the field of rational numbers. we write a·x instead of σ(a)·x.e. there exist field homomorphisms σ : K → L. The intermediate field E is called proper if E ≠ K and E ≠ L. We shall construct an extension of F2 at 1.2 Examples. we identify σ(a) ∈ L with a ∈ K. For any element a ∈ K. .

1.172 IV. i} is a basis of C over R. Field extensions f) A method to construct subfields. Indeed. y ∈ K and σ ∈ End(K) then H −1 H σ(x − y) = σ(x) − σ(y) = x − y. so [C : R] = 2. One can define the degree of an extension of skew fields (division rings): if L Evariste Galois (1811-1832). then L is canonically a K-vector space: the multiplication of a “scalar” in K with a “vector” in L is their multiplication in L. n ·1 ≠ 0. is defined as follows: . The dimension of L as a K-vector space is called the degree 40 of the extension K ⊆ L and is denoted by [L : K] or (L : K). is the following: let K be a field and let H be a set of field endomorphisms of K (homomorphisms defined on K with values in H K). French mathematician. ∀σ ∈ H} is a subfield of K (called H the fixed subfield of H). We determine now all prime fields. 39 40 is a skew field and K is a subfield (skew) of L. Let 1 be the identity element of R and let n ∈ N. The set K := {x ∈ K|σ(x) = x.3 Definition. In example b) above. A basic tool is the concept of degree of an extension. so x − y ∈ K . then L is naturally a K-vector space and one defines the degree [L : K] as dimKL.4 Definition. If K ⊆ L is an extension. {1. . for any n ∈ N*. xy ∈ K if y ≠ 0. if x. Recall the notion of characteristic of a ring R with identity. then char R = 0. denoted char R. very important in Galois theory39. Likewise. An extension is called finite if its degree is finite.if. n ·1 denotes the multiple 1 + … + 1 (n terms). What are the degrees of the other extensions? 1. the characteristic of R. The fields that have no proper subfields are called prime fields.

Since ψ(Q) is a subfield in K. For example. we have ψ(Q) = K. ϕ(n) = n·1.6 Proposition. then 0 = n ·1 = (a·1)·(b·1). Proof. n = ab. Suppose n = char R ≠ 0. If char K = 0. Algebraic extensions 173 . so ψ is an isomorphism. ! . char F2 = 2. n·1 ≠ 0. which is a subfield in K. Suppose char K = 0. contradicting the minimality of n. Define the ring homomorphism ϕ : Z → K. In particular. Kerϕ = pZ. by contradiction. 1. then there exists a unique isomorphism K ≅ Fp. b ≠ 0. if R is a unitary subring of S. Since for any n ∈ Z. with a. applying the isomorphism theorem. the characteristic of a field is 0 or a prime. b ∈ N*. Then char R is the natural generator of the ideal Kerϕ.IV. 1. char Z = char Q = 0. then char R = char S. then there exists a unique isomorphism K ≅ Q. If. If char K = p > 0. so a·1 = 0 or b·1 = 0. Let K be a prime field. b < n. where 1 is the identity of K. then char R is the smallest n ∈ N* such that n ·1 = 0.7 Proposition. The proof of the equivalence of these definitions is left to the reader. a < n. Imϕ = K.if there exists n ∈ N* such that n ·1 = 0. Define the homomorphism ψ : Q → K that extends ϕ. n·1 is invertible in K. ∀a. Then the characteristic of R is 0 or a prime. If char K = p. Z/pZ = Fp ≅ Imϕ. A related notion is the characteristic exponent of a field: the characteristic exponent of the field K is 1 if char K = 0 and is p if char K = p > 0. b ∈ Z. The characteristic of R can be defined also as follows: there exists a unique ring homomorphism ϕ : Z → R (prove!). −1 ψ(a/b) = ϕ(a)ϕ(b) .5 Remark. Let R be a domain. ∀n ∈ Z.1. Proof. The reader is invited to prove the uniqueness part. Since K has no proper subfields. 1. But R has no zero divi! sors.

x − y ∈ ∩Li. ! 1. n is finite and F is n isomorphic (as an Fp-vector space) to (Fp) . and denoted by . so it is equal to both of them. then P ∩ Q is a subfield both in P and in Q. If (Li)i∈I is a family of subfields of the field L and x. y ∈ K with y ≠ 0. Let K ⊆ L be a field extension and let S be a subset of L. called the field generated by K and S. so x − y ∈ Li. The intersection of all subfields of K is a subfield P. this shows that P is the unique prime subfield in K.10 Application.174 IV. which is infinite!). ∀i ∈ I. The rest of the statement follows from the list of prime subfields. which cannot have proper subfields (any subfield F of P would be a subfield in K. for instance. then x. absurd!).11 Definition. The same argument works for −1 ! xy if y ≠ 0. y ∈ Li.if char K = p > 0. the cardinal of a fin nite field is p .9 Theorem. If n is the degree [F : Fp]. for some prime p and some n ∈ N*. This means there do not exist fields with 10 elements. 1.e. Proof. . If Q is a prime subfield of K. In particular: . The intersection of all subfields of L that include K ∪ S is a subfield of L. then K is an extension of Q. 1.if char K = 0. we have x − y and xy ∈ K. y ∈ ∩Li. If F is a field with a finite number of elements. Use the characterization: K is a subfield of L if and only if −1 for any x. then K is an extension of Fp. i. Any field K contains a unique prime subfield. Thus. Field extensions 1. so F includes P. Proof. so char F = p > 0 (for some prime p) and F is an extension of the field Fp. ∀i ∈ I. The intersection of any family of subfields of a field is a subfield. The following definition describes a basic construction in field extensions. Finite fields. then char F cannot be 0 (otherwise F is an extension of Q.8 Lemma.

then K(S) is denoted by K(x1. xn}. the ring K[S] is a domain (it is a subring of the field L) and K[S] ⊆ K(S). R[S] is the subring generated by R ∪ S in A (R[S] coincides with the R-subalgebra of A generated by S). 41 .…. and S is a subset of A. included in L. K(∪i∈I Ei). 1. Obviously. for a family (Ei)i∈I of extensions of K. included in L.IV. The composite of the fields E and F (denoted by EF) is the subfield of L generated by E and F: EF = K(E∪F) = E(F) = F(E). Algebraic extensions 175 K(S).….1. We denote by K[S] the subring of L generated by K ∪ S (the intersection of all subring of L that include K ∪ S). and α is called a primitive element. xn) and K[S] by K[x1. Let K ⊆ L be a field extension and let E. In general. the composite of the fields (Ei)i∈I is the subfield of L generated by ∪i∈I Ei. xn].41 It is easy to see that K(S) (respectively K[S]) is the smallest subfield (respectively subring) of L that includes K ∪ S. The field of fractions of the domain K[S] is canonically isomorphic to K(S) (see the universality property of the field of fractions).…. If S = {x1. the extension K ⊆ L is called a simple extension. An extension K ⊆ L with the property there exists a finite subset S of L such that L = K(S) is called a finitely generated extension. This notation is used also for ring extensions: if R is a subring of the ring A. A primitive element need not be unique: for instance.12 Definition. K(α) = K(α + 1). One also says that K(S) is obtained by adjoining to K the elements in S. F be extensions of K. Do not confuse with the concept of finite extension (which means that its degree is finite)! If there exists α ∈ L such that L = K(α).

in ) ∈ ' n ⎬ 1 n ⎫ ⎪ where ∑ means the sums are finite.i )∈' ⎩ 1 n 1 n 1 n n n ∈ '* . T is the smallest subring of L that includes K and S. T is the image in L via the evaluation homomorphism of polynomial ring in S indeterminates).e. −1 b) K(S) = {αβ |α. S1. a) Let T be the set of all polynomial expression in the elements of S.…. ' ⎪ ⎭ set 42 with two elements (resp. i. F} = EF. inf{E.176 IV. (Alternatively. ∀ ( i1 . β ≠ 0}. ai … i ∈ K . x1 . The set IF(L/K) of all intermediate fields of a given extension L/K is ordered by inclusion and IF(L/K) is a (complete) lattice 42: for any E. S2 be subsets of L.… . ≤) is called a lattice (resp. … . a complete lattice) if any subset .13 Remark.…. namely: K[S] = ⎧ ⎪ i i ' ⎨ ∑ ai … i x1 … xn ⎪( i . sup{E. in ) ∈N n ai1…in ≠ 0} is finite. β ∈ K[S]. Let L/K be an extension and let S. that includes K and S. Then: a) K[S] is the set of all polynomial expression in the elements of S. the {(i1. On the other hand. F ∈ IF(L/K). F} = E ∩ F. c) K[S1∪ S2] = K[S1][S2] = K[S2][S1] and K(S1∪ S2) = K(S1)(S2) = K(S2)(S1). Proof. with coefficients in K. Thus. An ordered set (S. d) K(S) = ∪{ K(T) | T ⊆ S.14 Theorem. T finite}. any subring of L that includes K and S must also include T. with coefficients in K. any subset) of S has a least upper bound and a greatest lower bound in L. 1. One checks directly that T is a subring of L. xn ∈ S . Field extensions 1.

g ( x1 . is in K(S).… xn ) ≠ 0⎬ . eva is called the “homomorphism of evaluation in a”. X n ]. The usual notation for eva( f ) is f (a). If f (a) = 0.… xn ) f ∈ K [ X 1 . …. x is transcendental over K if and only if evx induces a K-isomorphism between the polynomial ring K[X] and the subring K[x] generated by K and x in L.1. x is algebraic over K if and only if the evaluation homomorphism evx : K[X] → L is not injective. with α. xn ∈ L.16 Examples. xn}. xn] = { f ( x1 . If the element x is not algebraic over K (i.… . But the set of all these elements is a subfield in L (standard check).IV. any element of the form αβ −1. g(a) ≠ 0}.… xn ) ⎭ If S = {a} ∈ L: K[a] = {f (a) | f ∈ K[X]} and K(a) = {f (a)/g(a) | f. x is called transcendental over K. β ∈ K[S]. Thus. 2 is alge- . g ∈ K[X]. as a root of X − 2 ∈ Q[X]. The following notion is central in all the theory we describe. we say „a is a root of f”. : K[x1. g ∈ K [ X 1 . ….15 Definition. 1. with x1.… . xn) = ⎨ ⎩ g ( x1 . We say that is algebraic over K if there exists a nonzero polynomial f ∈ K[X] such that f (x) = 0. In the particular case S = {x1. …. the element 2 braic over Q. Thus. 1. called the “value of f in a”.e.… xn ) ⎫ f . X n ]} ⎧ f ( x1 . Algebraic extensions 177 b) Since K[S] ⊆ K(S) and K(S) is a field. then eva( f ) = b0 + b1a + … + bna ∈ L. Let K ⊆ L be a field extension and let x ∈ L. If f ∈ K[X]. …. In other words. a) In the extension Q ⊆ R. where eva : K[X] → L is the unique K-algebra homomorphism with the property that eva(X) = a. β ≠ 0. K[a] = Im eva. d) Exercise. f = b0 + b1 X + … + n n bn X . ! c). evx is injective). K(x1.

a)⇒b) If f were reducible. and is injective. without any reference to Q. d) A complex number that is algebraic (respectively transcendental) over Q is called by tradition algebraic number (respectively transcendental number). d) f (x) = 0 and. Field extensions b) Any element of K is algebraic over K. x is algebraic over K if and only if Ker evx ≠ {0}. Let f be a monic polynomial with coefficients in K. The element X in the extension K ⊆ K(X) is transcendental over K. h ∈ K[X]. the evaluation homomorphism evX : K[X] → K(X) is the canonical inclusion. With the notations above.17 Theorem. Indeed. g(x) = 0. g(x) = 0 implies f |g. then f = gh. This underscores the importance of identifying the field over which the element is algebraic. algebraic over K. b) f (x) = 0 and f is irreducible. e) The real numbers e (the sum of the series ∑n≥0 (n!)−1 ) and π (the length of a circle of diameter 1) are transcendent (over Q). These facts were proven by Hermite in 1873 and Lindemann in 1882.178 IV. c) Let K(X) be the field of rational fractions in the indeterminate X with coefficients in the field K. deg g < deg f. Since g(x)h(x) = f (x) = 0. Let K ⊆ L be a field extension and let x ∈ L. for any g ∈ K[X]. x is a root of g . for some g. g ≠ 0}. which is fundamental in describing the extension K ⊆ K(x). However. c) f is a generator of the ideal Ker evx = {g ∈ K[X] | g(x) = 0}. The following statements are equivalent: a) f (x) = 0 and deg f is minimal among the nonzero polynomials that vanish in x: deg f= min{deg g | g ∈ K[X]. X is obviously algebraic over K(X). 1. with 1≤ deg h. The generator of the ideal Ker evx (K[X] is a PID!) is a nonzero polynomial. Proof.

The minimal polynomial of x over K is denoted by Irr(x. More generally. Q) = X – 2 since X – 2 ∈ Q[X] is monic. whose degrees are less than deg f. R) = X − 4 2 . By the hypothesis. using the form of the elements in Q( 2 )). The fact X − 2 that has minimum degree among the polynomials with coefficients in Q( 2 ) that vanish in 4 2 is equivalent to 4 2 ∉ Q( 2 ). (prove this. 1. f |g. because f is irreducible. b) Irr( 4 2 . so { f } = Ker evx. So. b)⇒c) Clearly. 4 4 .19 Examples. contradicting the hypothesis. b) K[x] is a field. K) = X − a.IV.18 Definition. K). Ker evx = { f } or Ker evx = K[X]. c)⇔d) is evident. The following statements are equivalent: a) x is algebraic over K. Let L/K be a field extension and let x ∈ L be algebraic over K.1. Let K ⊆ L be a field extension and let x ∈ L. the ideal generated by f is maximal. 1. 1. g ≠ 0. so deg f ≤ deg g. 4 2 is a root and it is irreducible in Q[X] (by Eisenstein's criterion). a) Irr( 4 2 . In the PID K[X]. ( f ) ⊆ Ker evx. 2 2 c) Irr( 4 2 . But obviously 1 ∉ Ker evx (⇔ Ker evx ≠ K[X]).20 Theorem (characterization of algebraic elements in an extension). d)⇒a) Let g ∈ K[X] with g(x) = 0. c) K[x] = K(x). Algebraic extensions 179 or of h. K) or min(x. Irr(a. for any field K and any a ∈ K. Q( 2 )) = X − 2 . We call minimal polynomial of x over K the monic polynomial in K[X] that satisfies one of the equivalent properties in the previous proposition.

then g = a0 + a1X + … + an − 1X ∈ { f }. K) n−1 If n = deg f.. this is also a isomorphism of K-vector spaces. x }.. a1. X. a)⇒d) At a)⇒b) we saw that we have a K-isomorphism of fields K[X]/{ f } ≅ K(x). On the other hand. K)): [K(x) : K] = deg Irr(x. there exists a n linear dependence relation. This means that h + { f } is a linear combination n −1 with coefficients in K of the classes of 1. we have g = 0. It is evident the b)⇔c). This shows that x is algebraic over K. not all zero.e. we get 2 n+1 a0x + a1x + … + anx − 1 = 0. Besides. i. a0{1 + { f }} + a1{ X + { f }} + … + an − 1{ X n −1 where a0. isomorphic to K[X]/{ f }. The ideal Ker evx of K[X] is generated by f. x.180 IV. X n −1 + { f }} = 0 + { f }. Evidently. then the degree of the extension K ⊆ K(x) is the degree of the minimal polynomial Irr(x. So. Since deg f = n. n−1 form a basis. for some n ∈ N and a0. an ∈ K.. an − 1 are 0. if x is algebraic over K. If X. . an − 1 ∈ K. …. …. the classes of 1.K) ∈ K[X] and let evx : K[X] → L be the evaluation homomorphism in x. a1. so x is a root of a nonzero polynomial with coefficients in K.. is also a field. . In the K-vector space K[X]/{ f }. Proof. −1 c)⇒a) If x = 0. …. a)⇒b) Let f = Irr(x. Let n = deg f. Multiplying by x. Suppose x ≠ 0 and let x = a0 + n a1x + … + anx ∈ K[x] be the inverse of x. Then K[x]. a0. Field extensions d) The extension K ⊆ K(x) is finite. i d)⇒a) The infinite family {x | i ∈ N} of elements of the finite dimensional K-vector space K(x) is linearly dependent. a0·1 + a1x + … + anx = 0. so f |g. a1. a K-basis of the extension K(x) is {1. The isomorphism theorem implies K[X]/{ f } ≅ Im vx = K[x]. the ideal { f } is maximal and K[X]/{ f } is a field. Since f is irreducible in K[X]. the division with remainder theorem shows that any class modulo f of a polynomial h ∈ K[X] has a un representative of degree less than n. then all is evident.. …. . X ..

1.IV. n −1 n−1 + { f } in the basis 1. Taking classes modulo ( f ). …. x. g) = 1. there exist (and can be effectively computed) u. the inverse of 3 y = 1 + 3 2 + 3 4 . Take g ∈ K[X]. The proof of a)⇒b) above uses essentially the fact that if f is irreducible in K[X]. Using the extended Euclid algo3 2 rithm for X −2 and 1+X+X . We have Irr(x. in the extension Q ⊆ Q (3 2 ) . We must show that any nonzero element in K[X]/{ f } has a multiplicative inverse. . Let K be a field and let x be algebraic over K. Keep the conventions above. X 1. This also suggests a way to compute the inverse of an arbitrary elen−1 ment in K[x] expressed in the basis {1. Let x = 3 2 . x }: for any element n −1 =: g(x). The factor ring K[X]/{ f } appears also in the following proof. By the extended Euclid algorithm. dimK K(x) = dimK K[X]/{ f } = n. where g ∈ K[X]. The isomorphism K[X]/{ f } ≅ K(x) takes the basis 1 + { f }. we obtain 1 = ( 3 2 − 1)(1 + 3 2 + 3 4 ). this relation shows that v + { f } is the inverse of g + { f } in K[X]/{ f }. where v. as follows. K) = f = X − 2 and 2 y = g( 3 2 ).g. x. This implies f . This result can be proven in a more “elementary” way. The degree of the extension K ⊆ K[x] (equal to deg Irr(x. 1. K)) is called the degree of x over K. v ∈ K[X] such that uf + vg = 1. ! …. u ∈ K[X] are such that uf + vg = 1. where g = 1 + X + X . x . then K[X]/{ f } is a field. X + { f }. Let us find. so GCD( f. then y is v(x). Algebraic extensions 181 So. Evaluating in 2 .21 Remark. we find that 3 2 3 1 = ( − 1)(X − 2) + (X − 1)(1 + X + X ). y = a0 + a1 X + … + an − 1 X −1 deg g < n.22 Definition. g ≠ 0. …. g + { f } ≠ 0 + { f }.

that can be complex numbers or “other similar numbers”. ∀a ∈ K. only polynomials with numerical coefficients were considered). where a. Concrete fields can be constructed following the procedure in the proof. The polynomial f is irreducible in R[X]: it 2 has degree 2 and has no roots in R. ϕ(a) = a + { f }.24 Examples. In 1792. which is a field. denoted by i. Here are two examples. Then there exists an extension of K in which f has a root 43. The importance of this result is not just theoretical. Albert Girard. was known intuitively for a long time and often tacitly accepted in 17th to 19th centuries’ mathematical arguments (of course. Any element of this field is written uniquely as the class of a polyno2 mial of degree at most 1 modulo (X + 1). α = X + { f } is a root of f. In the UFD K[X]. is a root of f. 1. it is sufficient to prove the claim for the case when f is irreducible. 2 a) Let f = X + 1 ∈ R[X]. extension of R. f with one of its irreducible factors. Proof. Consider the factor ring K[X]/{ f } =: L. if needed. is of the form a + bX = a + b ⋅ X = a + b ⋅ i . 43 . states –without proof– that an equation of degree n has n roots. Any root of a factor of f is also a root of f. since f is irreducible. whose proof is simple for a modern mathematician. In L. the polynomial f is a product of irreducible polynomials. So. because the canonical mapping ϕ : K → K[X]/{ f }. is a field homomorphism. Let K be a field and let f ∈ K[X].23 Proposition. Let g be the class of the polynomial g ∈ R[X] modulo 2 2 (X + 1). deg f ≥ 1. b ∈ R. Pierre Simon de Laplace gives an elegant proof of the “Fundamental Theorem of Algebra” – any nonconstant polynomial with complex coefficients has a complex root – admiting though that the roots exist “somewhere”. Around 1629.182 IV.e. suppose that n f = a0 + a1X + … + anX n Then f (α) = a0{1 + { f }} + a1{X + { f }} + … + an {X + { f }} = {a0 + a1X n ! + … + anX } + { f } = f + { f } = 0 + { f }. Moreover. Field extensions 1. Identifying the real This result. Replacing. i. R[X]/(X + 1) is a field. Indeed. L is an extension of K. The class of X modulo (X + 1).

Let p be a prime number and let n ∈ N*. Let f ∈ K[X]. deg f = n ≥ 1. b ∈ R. ∀a. Let K be a field and let f ∈ K[X]. b ∈ F2. K := F2[X]/{ f } is a field with 4 elements. c. Thus. 2 For instance. Fp) is an irreducible polynomial of degree n in Fp[X]. n b) Construction of a field with p elements.4). for any a. b. take L = K. deg f = n. So. There ex- . with a. Proof. for any field F and any g ∈ F[X] with deg g < n. then Fp[X]/{ f } is a field with p elements. we have 2 α = α + 1. If α denotes the class of X modulo f. 2 In this field we have i = −1. If f ∈ Fp[X] is an irreducible polynomial of degree n n. In fact.1. the addition and multiplication rules in K are given by the following rules: (a + bα) + (c + dα) = (a + c) + (b + d)α. Suppose that. Then there exists an extension L of K such that f is written as a product of polynomials of degree 1 in L[X] (one also says “f splits over L” or “f has n roots in L”). Induction by the degree of f. 2 (a + bα)·(c + dα) = ac + (ad + bc)α + bdα = ac + (ad + bc)α + bd(α + 1) = (ac + bd) + (ad + bc + bd)α. with a.IV. If deg f = 1. ·) (see 3.25 Corollary. 1. the polynomial f = X + X + 1 is irreducible in F2[X] (it has degree 2 and has no roots in F2). then Irr(α. this construction can be taken as a definition of C. n Conversely. we can uniquely write a generic element of 2 R[X]/(X + 1) in the form a + bi. b ∈ F2 (we identify any a ∈ F2 with its image a + { f } in F2[X]/{ f }). d ∈ R This field is isomorphic to the complex number field C. Since α is a root of f. if L is a field with p elements and α ∈ L is a generator of (L*. Algebraic extensions 183 number a with its class a . any element in K is written uniquely as a + bα. Thus. there exists an extension of F in which g splits. the multiplication of two elements is given by: (a + bi)·(c + di) = (ac − bd) + (ad + bc)i.

yn} is an L-basis. Field extensions ists an extension E of K in which f has a root a. Let us show that {xiyj | 1≤ i ≤ m. Obviously. if {x1. bn ∈ L such that z = b1y1 + … + bnyn. xm} is a K-basis of L and {y1. then {x1 y1. Using this in the expression for z. 1. Each bj (1 ≤ j ≤ n) is of the form bj = aj1x1 + … + ajmxm. for some aji ∈ K (1≤ i ≤ m). f splits also in L. because {y1. …. xm y1.26 Theorem (The transitivity of finite extensions) a) Let K ⊆ L and L ⊆ M be a tower of finite field extensions. …. concerning the dimension of a vector space. x1yn. We will study this concept in IV. So X − a divides f in E[X]. 1 ≤ j ≤ n} with coefficients in K.….2. hence f = (X − a)g. for some g ∈ E[X]. then K ⊆ L and L ⊆ M are finite extensions. …. Let us prove that {xiyj | 1≤ i ≤ m. 1 ≤ j ≤ n} generates the K-vector space L. The following theorem and its consequences illustrate the strength of this translation. For a given f ∈ K[X]. We saw that “x is algebraic over K” is equivalent to “K(x) is a finite dimensional K-vector space”. there exists b1. 1 ≤ j ≤ n} is K-linearly independent.184 IV.14. Apply now the induction hypothesis for the field E and g ∈ L[X] to obtain an extension field L ! of E in which g splits. If z ∈ M. . we obtain that z is a linear combination of {xiyj | 1≤ i ≤ m. xmyn } is a K-basis of M. This translates the property of an element of being algebraic into a Linear Algebra property. Then K ⊆ M is a finite extension and the degree is “multiplicative”: [M : K] = [M : L] [L : K] Moreover. …. Proof. b) If K ⊆ M is a finite extension of fields and L is an intermediate field. yn} is a L-basis of M. …. …. the “minimal” extension of K over which f splits is called the splitting field of f over K. a) It is sufficient to prove the claim on bases.

b ∈ Q (again a contradiction). 3 ) . uniquely as a + b3 2 + c3 4 . with a. then ∑ 1≤ j ≤ n that 1≤ j ≤ n ∑ aij xi = 0.Q( 2 )) 2 = X − 3. Q) = X − 2) and a basis is {1. Then (xiyj)(i. if the extensions are not necessarily finite. b ∈ Q. the degree of the extension is 3 and a Q-basis of Q (3 2 ) is Eisenstein's criterion (or observing that X − 2 has no rational roots). then 3 = a or 3 = 2b . yn} implies If ∑ aij xi y j = 0 . c ∈ Q. ⎜ ⎟ 1≤i ≤m ⎝ 1≤ j ≤n 1≤i ≤ m . with a. b. If ab ≠ 0. …. then 2 ∈ Q. n 2n −1 . with aij ∈ K. 3 2 . 6 }. The first extension has de2 gree 2 (since Irr ( 2 .…. The same argument { } . using 3 Thus. X 2 − 3 is irreducible over Q ( 2 ) . the extension Q ⊆ Q (n 2 ) has degree n and a Q-basis is 1. Irr ( 3 2 . b) The extension Q ⊆ Q( 2 . xm} are K-linearly independent.IV.1. by absurd. 3 ) has degree 4 and a Q-basis is {1. because {x1. because 3 ∉ Q ( 2 ) : if. If 2 2 ab = 0. with a.28 Examples. 3 = a + b 2 . consider the tower of extensions Q ⊆ Q ( 2 ) ⊆ Q( 2 )( 3 ) = Q( 2 . for the proof. for any i. b) More generally. n 2 .Q) = X − 2. suppose that (xi)i∈I is a K-basis of L and (yj)j∈J is an L-basis of M. then 2 2 3 = a + 2b + 2ab 2 . because Irr ( 3.j)∈I×J is a K-basis of M. The L-linear independence of {y1. contradiction. a) Consider Q (3 2 ) /Q. The extension Q ( 2 ) ⊆ Q( 2 )( 3 ) has also degree 2. This means that any element of Q (3 2 ) is written * shows that for any n ∈ N . Hence aij = 0 for any i and j.27 Remarks. since the proof does not use the commutativity of multiplication. 3 4 } . 1. 2 . 3 {1. Algebraic extensions 185 ⎛ ⎞ ⎜ ∑ aij xi ⎟ y j = 0 . 2 }. Indeed. a) The theorem is also true for division rings. 3. 1. 1≤ j ≤ n ⎠ where ∑ aij xi ∈ L. ! ….

[K(T) : K(u)] = deg Irr(T. Let us find Irr(α. We have α = 5 + 2 6 ⇒ (2 6 ) = (α 2 − 5) ⇒ (α 2 − 5) − 24 = 0 . Q( 3 ) . 3 ) is evident. Note that this extension has the proper intermediate fields Q ( 2 ) . 3 ∈ Q(α). 3 ∈ Q(α). ⎨ ⎩ 2 + 3 =α where α. α is the root of (X 2 − 5) − 24 . K(u)) is associated to h = g(X)u − f (X) ∈ K(u)[X]. 6 }. We have α = 5 + 2 6 . deg g). K(u)). At this stage. The element 2 + 3 =: α is a primitive element: Q( 2 . Let 2 us show that 2 . Q). Q) has degree equal to [Q(α) : Q] = [Q(α) : Q ( 2 ) ]·[ Q ( 2 ) : Q] = 2·2 = 4 2 2 2 2 2 Thus. Let K be a field. so it must be equal to Irr(α. which is monic. 3 }. so 6 ∈ Q(α). 3 ) = Q( 2 + 3 ) . Q ( 6 ) . Thus 6α = 3 2 + 2 3 =: β ∈ Q(α). g) = 1. The polynomial Irr(α. a basis of the extension Q ⊆ Q( 2 . K(T) the field of rational fractions with coefficients in K. u ∈ K(T). shows that 2 . Since K(T) = K(u)(T). 2 . but it is an immediate consequence of Galois theory. so K(u) ⊆ K(T) is finite. Field extensions Thus. . The system ⎧3 2 + 2 3 = β . β ∈ Q(α). in Q[X] and has degree 4. g ∈ K[T] with u = f /g and ( f. which shows that T is algebraic over K(u). Q). The inclusion Q( 2 + 3 ) ⊆ Q( 2 . u ∉ K and f.186 IV. a basis of the extension Q ( 2 ) ⊆ Q( 2 )( 3 ) is {1. Applying the method in the theorem of transitivity of finite extensions. Then the extension K(u) ⊆ K(T) is finite and [K(T) : K(u)] = max (deg f. this is difficult to prove. These are all the proper intermediate fields. Let h = g(X)u − f (X) ∈ K(u)[X] (where g(X) is the polynomial in X obtained by replacing T with X in the polynomial g). We claim that Irr(T. 3. 3 ) is {1. c) Here is a nontrivial example of a finite extension. We have h(T) = 0.

29 Definition. X]. deg g). It is interesting to remark that any intermediate field K ⊆ L ⊆ K(T) with K ≠ L is of the form L = K(u). seen as a polynomial in Y with coefficients in K[X]. there is a K-isomorphism K[Y] ≅ K[u] (where Y is an indeterminate). The polynomial h is irreducible and vanishes in T. g) = 1.IV. Thus. considering the tower of extensions K ⊆ K(u) ⊆ K(T). we obtain that q|g and q| f in K[X]. q ∈ K[Y. We have to prove that h = g(X)u − f (X) ∈ K(u)[X] is irreducible. the theorem of transitivity of finite extensions would imply that K ⊆ K(T) is finite). So. since ( f. We have 1 = degY r = degY p + degY q. note that u is transcendental over K. K(u)). Suppose r = pq. u ∉ K.1. We −1 have ub − a ≠ 0. Proposition 1. with p. respectively g. respectively b. so we may suppose degY p = 1 and degY q = 0. This fact is known as “Lüroth’s Theorem” and is of significance in Algebraic Geometry (see MORANDI [1996]. K ⊆ K(T) is infinite (T is transcendental over K) so K ⊆ K(u) must also be infinite (otherwise. then the leading coefficient of h is ub − a. Thus. The irreducibility of g(X)u − f (X) ∈ K(u)[X] is thus equivalent to the irreducibility of r = g(X)Y − f(X) ∈ K(Y)[X]. WALKER [1950]). Algebraic extensions 187 We have deg h = max (deg f. X]. implying that q ∈ K*. are the leading coefficients of f. so it is associated to Irr(T. This is clear if deg f ≠ deg g. because otherwise u = ab ∈ K. q ∈ K[X] and p = cY + d. The extension K ⊆ L is . h is irreducible in K[Y][X]. for some c. where a.20 says that [K(T) : K(u)] = deg Irr(T. contradicting the assumption u ∉ K. deg g). this is equivalent to r being irreducible in K[Y][X] ≅ K[Y. K(u)) = deg h = max (deg f. d ∈ K[X] and r = g(X)Y − f(X) = (cY + d )q. First. if deg f = deg g. Identifying the coefficients of the powers of Y. Let degY r be the degree of r. for some u ∈ K(T). A field extension K ⊆ L is called algebraic extension if each element in L is algebraic over K. Since K[Y] is a UFD and K(Y) is its field of quotients. 1. Indeed.

the induction hypothesis says that K ⊆ K(x1. 12. …. Prop.20. …. …. xm). xm} is a K-basis of L. (transitivity of algebraic extensions) If K ⊆ L and L ⊆ M are algebraic extensions. …. xn is also algebraic over K(x1. …. 1. any algebraic finitely generated extension is a finite extension. * Proof. Let K ⊆ L be finite and let x ∈ L. then K[x1. An easy argument by induction on n ∈ N* shows that. xn − 1.17).20 shows that x is algebraic over K. xn) is a finite extension. p.30 Proposition. If n > 1.35. so L is a finitely generated extension of K.188 IV. For n = 1. …. …. Any finite extension is algebraic and finitely generated. ! 1. xn). then K ⊆ M is an algebraic extension. In particular. 1. xn − 1) ⊆ ! K(x1. 1. the extension K(x1. Then K ⊆ K(x1. Field extensions called transcendental if it is not algebraic (it contains at least one transcendental element over K). …. xn − 1)(xn) is thus finite. Then K ⊆ K(x) is finite (as an intermediate field of K ⊆ L) and Prop. xn) (see exercise 1. xn ∈ L are algebraic over K. The conclusion follows by applying the transitivity of finite extensions to the tower K ⊆ K(x1. xn − 1) is finite.32 Theorem. 232). the conclusion follows from Prop. The converse of this result is nontrivial and it is known as "Zariski’s Lemma” (see for instance SPINDLER [1994].31 Proposition. xn − 1) ⊆ K(x1. If {x1. then certainly L = K(x1. …. Suppose K ⊆ L is a field extension.…. Since xn is algebraic over K. xn − 1). xn] = K(x1. xn ∈ L are algebraic over K. if K ⊆ L is an extension and x1. …. …. . n ∈ N* and x1. …. We prove the claim by induction on n ∈ N . 1. Proof.

with b ≠ 0. we show that x is contained in a finite extension of K and apply then Prop. Let K ⊆ L be a field extension. ab are algebraic over K. The extension Q ⊆ L = Q n 2 n ∈ N* is algebraic and infinite. …. the product and the quotient (if it exists) of two algebraic elements over K are also algebraic elements over K. by contradiction. the sum. m ∈ N*.35 Proposition. and each of these is algebraic (they are even finite). Q ⊆ L is algebraic because L is the union of its intermediate fields of the form Q ({n 2 n ≤ m}) . But Q ⊆ L is not finite: if. b). ! 1. By transitivity. Why? ({ }) .33 Example. then K ⊆ K(a. an)(x) is also finite. …. Since x is algebraic over L. Thus a + b. as elements of K(a. b ∈ L. If K ⊆ L is a field extension. Proof. then each intermediate field would have the degree a divisor of n. K ⊆ K(a0. If a.36 Example. Let a0. 1. Let x ∈ M. an)(x).1. …. 1. Then x is algebraic over K(a0. an ∈ L be the coefficients of f. This shows that ! K'L is a subfield of L. …. K ⊆ K'L. a − b. b) is finite. …. the set of all elements in L that are algebraic over K is denoted by K'L and is called the algebraic closure of K in L. −1 ab. …. 20. 1. are algebraic over K. so it is an algebraic extension. Q'C is a subfield in C. But Q ⊆ Q (n +1 2 ) has degree n + 1. the difference. Then K'L is a subfield of L and an extension of K. 1.IV. Algebraic extensions 189 Proof. Clearly. an infinite algebraic extension of Q. there exists 0 ≠ f ∈ L[X] such that f (x) = 0.31. by Prop. In order to prove that x is algebraic over K. We have now the tower of finite extensions K ⊆ K(a0. [L : Q] = n ∈ N*. an) ⊆ K(a0. an) is finite. In particular. 1. an) and K ⊆ K(a0.34 Definition.

3. Field extensions Exercises 1. The same problem for Q ⊆ Q(α) and 3 the element 1 − α. Find a basis of the extension Q ⊆ Q[ 3 ]. bi ∈ K: Prove that: the system (S) has a solution in E ⇔ (S) has a solution in K. 7. 1 ≤ i ≤ n + 1. Let K be a field and let (Ki)i ∈ I be a chain of subfields of K (∀i. then the field of fractions of the domain K[S] is canonically isomorphic to K(S). K) = n. Let K ⊆ L be an extension and let p ∈ L[X] \ K[X]. then deg Irr(β. Let K ⊆ L be an extension. j ∈ I. α·x = αx (the multiplication in L). Show that L is a K-vector space with respect to: the addition of L and the “scalar multiplication” given by: ∀α ∈ K. 8. algebraic over K. Prove that ∪i ∈ I Ki is a subfield in K. K) divides n.190 IV. Prove that [L : K] = 1 if and only if K = L. Q). Find Irr(1 + 3 . Express (1 + 3 ) in this basis. 2. If S is a subset of the field L and K is a subfield in L. deg p = n. Interpret this as a system of n + 1 equations. Can you prove other similar properties of (S)? 4. … ⎪ ⎩am1x1 + … + amnxn = bm . where α is a root of X + X + 1. If deg Irr(α. Ki ⊆ Kj or Kj ⊆ Ki). the unknowns being the coefficients of the polynomial. Let K ⊆ L be an extension and let α ∈ L. βi ∈ K. Suppose p(αi) = βi. with αi. Let K ⊆ L be an extension and let x ∈ L. ∀x ∈ L. for any β ∈ K(α).) 5. (Hint. Consider the linear system with coefficients aij. 6. Prove that x is transcendental over K if and only if K[x] is K-isomorphic to the K-algebra K[X] of polynomials in the indeterminate X. −1 ⎧a11x1 + … + a1nxn = b1 ⎪a21x1 + … + a2nxn = b2 (S): ⎨ . Prove that | p(K) ∩ K | ≤ n. Let K ⊆ L be an extension.

x > 0 implies ϕ(x) > 0. x = a . d ≠ 1 and d is not the multiple of any square of a prime). 11. y) = K(xy). xn). n ∈ N* and let x1. K). 13. xn] = K(x1. 15. Let ψ : K[X] → L[X] be the ring isomorphism that extends ϕ and with ψ(X) = X. (Hint. Suppose Irr(α. with x. where deg f = n. Show that m ∉ Q( n ). y ∈ L. Let K ⊆ L be an extension. y ∈ K and (m. where α ∈ L is a root of a polynomial of the form 2 2 2 X − b (if char K ≠ 2). respectively X − b or X − X − b (if char K = 2). For 0 ≠ β ∈ K(α) given as a K-linear n−1 −1 combination of 1. Show that ϕ|Q = id. (Hint. Let K and L be fields and let ϕ : K → L be a field isomorphism. More generally. Let K ⊆ L be an extension such that L = K(x. describe a method to find β and Irr(β. Prove that p ∈ K[X] has a root in K ⇔ ψ(p) has a root in L. 16. Prove that Q( d ) ≅ Q( e ) ⇔ d = e. then K(x. If m n there exists m. Let K ⊆ K(α) be an algebraic extension. y). Prove a similar property for extensions of Fp (where p is a prime). 10. Let K and L be extensions of Q and let ϕ : K → L be a field homomorphism. Let ϕ : R → R be a field homomorphism. Let K ⊆ L be an extension of degree 2 (”quadratic extension”). (Hint. Let d.1. n) = 1. for some a ∈ R. . α . for some b ∈ K.IV. with d ∈ Z squarefree (d ≠ 0. …. Let m. xn ∈ L.) 17. e ∈ Z. squarefree. algebraic over K. n ∈ Z be distinct and squarefree. Use that ϕ|Q = id and that between any two real numbers there is a rational number. α. Prove that any quadratic extensions of Q is of the form Q( d ).) b) Prove that ϕ = id. if S is a subset of L consisting of algebraic elements over K. a) Prove that ∀x ∈ R. …. Algebraic extensions 191 9.) 12. Show that L = K(α). …. …. K) = f is known. then K[S] = K(S). Then Q( m . n ∈ N* such that x ∈ K. Deduce ϕ is an 2 increasing map. Then K[x1. n ) has degree 4 over Q. 14.

Let K ⊆ L ⊆ E be a tower of extensions and let x ∈ E be algebraic over K. y ≠ 0}. q = Irr(y. (1 − i )5 2 }. Let x ∈ C be integral over Z. deg p = m. Field extensions 18. Let L and E be intermediate fields of the extension K ⊆ F. 26. y) : K] = mn. c) Formulate a sufficient condition for [K(x. Let R ⊆ S be an extension of domains (R is a subring of the domain S and 1 ∈ R). y) : K] = mn. Show that IF(L/K) = {E | E is a subfield in L. K) ≤ mn. a) Prove that deg Irr(x + y. give an example of extension of degree n of Q.192 IV. the extension Q ⊆ Q(x). K). K ⊆ E} is a complete lattice with respect to inclusion (any family of subfields (Ei)i∈I has sup and inf in IF(L/K)). . deg q = n. Let K ⊆ L be an algebraic extension. Let L/K be an extension. 22. K) ≤ mn and deg Irr(xy. An x ∈ S is called integral over R if x is the root of a monic polynomial in R[X]. What is the form of an arbitrary element in sup{Ei | i∈I}? 20. y linear combinations of elements in F with coefficients in E. Q) ∈ Z[X]. Under what assumptions is the composite EF equal to the set of all linear combinations of elements in F with coefficients in E? 19. 25. b) deg Irr(x. K). Prove that Irr(x. y ∈ L be algebraic over K. Let K ⊆ L be a field extension and let x. Find Irr(x. where d ∈ Z is squarefree. Show that any subring of L that includes K (any K-subalgebra of L) is a field. Is the converse true? 21. For any n ∈ N*. Prove −1 that E(F) ( = the composite EF) is equal to the set {xy | x. K(y)) = m ⇔ deg Irr(y. are roots of a monic polynomial of degree 2 with integer coefficients. ∀ 24. Then [L(x) : L] ≤ [K(x) : K]. K(x)) = n ⇔ [K(x. Let p = Irr(x. 23. Let E and F be intermediate fields of the extension K ⊆ L. Deduce that the integral elements in Q ( d ) . Q) and a basis of x ∈ { 3 − i. 1 + 3 3 .

n) = 1. then g is irreducible in L[X]. It is thus natural to consider the following definition: . deg g = p. with p prime. Reduce to the case when S is finite. d) If the degrees [L : K] and [E : K] are finite and coprime. Show that X − 2 ω = cos(2π/5) + i·sin(2π/5).IV. If g is irreducible in K[X]. where IV. f) If [E : K] = 2 and K = L ∩ E. Let K ⊆ L be an extension of degree n and let g ∈ K[X].2 Roots of polynomials. K = L ∩ E. 28. Bézout's theorem says that the polynomial f ∈ R[X] has the root a ∈ R if and only if X − a divides f in R[X]. then [EL : E] ≤ [L : K]. then E ⊆ EL is algebraic. 27. e) If [EL : K] = [E : K]·[L : K]. 5 E is irreducible in Q(ω)[X]. then [EL : K] = [E : K]·[L : K]. but [EL : K] < 9. Algebraically closed fields Let R be a domain. then K = L ∩ E. (p. then E < S > = EL. Algebraically closed fields 193 EL L L∩E K a) If K ⊆ L is algebraic.2 Roots of polynomials. then [EL : E] = [L : K] and [EL : K] = [E : K]·[L : K]. (Hint.) c) If [L : K] is finite. g) Give an example such that [E : K] = [L : K] = 3. b) If S is a subset of L and K < S > = L (S generates the K-vector space L).

By the induction hypothesis. If n = 1.1 Definition. ( X − a1 ) 1 . Suppose the claim is true for n − 1 and let f be as in the statement. we obtain that ( X − an )mn divides g. an ∈ R are distinct roots of f. ( X − an )mn is coprime with ( X − a1 )m1 …( X − an −1 )mn−1 .f. Proof. By Prop. of multiplicities m1. Let R be a domain and let f ∈ R[X]. If a1. for some g ∈ R[X]. . The natural number n is called the multiplicity of the root a. deg f = n. If n > 1. this is the definition. a is called a multiple root of f (double if n = 2. Let K be the field of fractions of R. f ∈ R[X] is a nonzero polynomial. each root counts with its multiplicity (unless otherwise specified). Suppose R is a domain. When counting the roots of a polynomial. I. We now seek a criterion to decide if a polynomial has multiple roots. ….5. are irreducible and m pairwise not associated in divisibility. …. Of course. a ∈ R and n ∈ N. Proof. ( X − an )mn are pairwise relatively prime. Then f has at most n roots in R. If n = 1. We see f as a polynomial in K[X] and ap! ply the preceding proposition. Since ( X − an )mn divides ( X − a1 )m1 … ( X − an −1 )mn−1 g . It is useful to introduce the notion of formal derivative of a polynomial with coefficients in an arbitrary ring. …. a is called a simple root of f. triple if n = 3 etc).2 Proposition.6.3 Corollary. We say that a is a multiple root of f with n n+1 multiplicity n if (X − a) | f and (X − a) . Field extensions 2. Let R be a UFD and let f ∈ R[X] be nonzero. Use induction on n. Since the polynomials X − ai. 2. f = ( X 1 − a1 )m1 …( X n −1 − an −1 )mn−1 g . then ( X 1 − a1 )m1 …( X n − an )mn divides f in R[X].194 IV. mn respectively. each root is counted with its multiplicity. ! 2. 1 ≤ i ≤ n.

∀f. a) Use induction on deg f. we obtain the result. Direct calculations show that the formal derivative has the usual properties of the derivative of a function as known from calculus: { f + g}' = f ' + g'. . deg df ≤ deg f − 1. The uniqueness part amounts to the R-linear independence of i {(X − α) | i ∈ N} in R[X]. Let R be a commutative unitary ring and let n f = a0 + a1X + … + anX ∈ R[X]. 2. ….2). bn ∈ R such that f = ∑ bi ( X − α )i .6. If f = a0 + a1X. m − 1}. { fg}' = f 'g + fg'. Note that d : R[X] → R[X] is the unique R-module homomorphism n n−1 with the property that d1 = 0 and dX = nX . let f ∈ R[X] of degree n > 0 and let α ∈ R. ∀f ∈ R[X]. then f (α) = 0. Evidently.IV.4 Definition. d : R[X] → R[X].2 Roots of polynomials. for If α any i ∈ {0. Composing the homomorphism d with itself n times (n ∈ N*) is den n n n−1 0 noted by d . ∀n ∈ N* (apply the universality property of a free module). Thus. Writing g in the form given by the induction hypothesis and replacing in the previous equality. n (n) d f is often denoted by f . with b0 ∈ R and g ∈ R[X]. easy to prove. a) There exist and are unique b0. (i) b)0≤i ≤n is a root of multiplicity m (m ∈ N*) of f. then f = (a0+ a1α) + a1(X − α). The (formal) derivative of f is the polynomial n −1 df := a1 + 2a2 X + … + nan X . deg g = n − 1. Let R be a domain. ∀n ∈ N*. ∀a ∈ R.g ∈ R[X]. If deg f = n > 1. {af }' = af '. Proof.5 Proposition. then α is a multiple root of f (of multiplicity at least 2). c) If f (α) = f '(α) = 0. by the integer division with remainder theorem applied to f and X − α (I. Algebraically closed fields 195 2. …. (1) Other notations: df = f ' or df = f . we obtain f = (X − α)g + b0. d = id. d = d◦d .

an ≠ 0. (Viète's relations) 44 Suppose R is a domain.7 Proposition. m − 1}. bm−1 are all 0. (X − α) divides f if and only if b0. (X − α) | f. Applying the result above. 1 ≤ k ≤ n. we saw that f and f ' are not relatively prime. 2. b1. having the roots x1. 1540-1603. ∀i ∈ {0. the criterion above allows to decide if it has multiple roots. we have: 44 In honor of François Viète. n f = a0 + a1X +…+ anX is a polynomial in R[X]. ! Without knowing the roots of a polynomial. We obtain f (α) = b0 = 0 and 2 f '(α) = b1 = 0. then ( f. …. Thus: 2. g)Ω[X]. But the GCD of two polynomials can be obtained using Euclid's algorithm and does not depend on the field considered: if f. ∀i ∈ {0. (i) On the other hand. n} (proof by induction on (i) n).196 IV. Then f has multiple roots (in some extension of K) if and only if f and f ' are not relatively prime. if g = GCD(f. …. …. . g)K[X] = ( f. If f has multiple roots. we obtain that α is a multiple root of f if and only if is simultaneously a root of f and of its derivative: (X − α)| f and (X − α)| f ' in Ω[X]. f ') has degree ≥ 1.1. Let Ω be an extension of K in which f splits (IV. Then: f = an(X − x1)…(X − xn) For any k.6 Proposition. ! Suppose K is a field. Ω is an extension of K. Conversely. g ∈ K[X]. Field extensions m b) By a).…. So. c) Let f be written as in a). α ∈ Ω and f ∈ K[X]. then g has a root α in Ω (the roots of g are among the roots of f ) and α is a multiple root of f since f(α) = f '(α) = 0. xn ∈ R.25). This implies that f (α) = 0. French mathematician. f (α) = i!bi. This implies that the GCD of f and f ' in Ω[X] has degree ≥ 1. Proof. Let K be a field and let f ∈ K[X].

…. so they are associated in divisibility in K[X] (K is the field of fractions of R).…. . Xn. 2.n} xi1 … xik = (− 1)k an −k .2. ! The fields that have no proper algebraic extensions are highly interesting.…. 1 k In particular. 2.…. If deg f = n and f has the roots x1. xn)). which is an element of R.….…. where s1.….…. there exists h ∈ R[X1. Algebraically closed fields 197 an ∑{i . Thus: g(x1. The polynomials g and f have the same degree and g | f. sn(x1. xn) ∈ R.…. Since g and f have the same leading coefficient.IV. Let R be a subring of the domain S and let f ∈ R[X] be monic. xn ∈ S.…. by prop. Let K be a field. then g(x1. g = f. The following statements are equivalent: a) There exist no proper algebraic extensions of K. for any 1 ≤ i ≤ n.….…. b) There exist no proper finite extensions of K. xn) ∈ R. any root in R of f divides a0. Consequently.2. 2. an(x1 + … + xn) = – an − 1 an(x1 x2 + x1 x3 + … + xn −1 xn) = an − 2 … n an x1… xn = (–1) a0. Xn]. They are characterized by the next theorem.i }⊆{1.2 Roots of polynomials. xn) = h(s1(x1. The polynomial g = an(X − x1)…(X − xn) divides f.8 Corollary. for any symmetric polynomial g ∈ R[X1. sn). Proof. Proof. The relations between roots and coefficients show that si(x1. The other ! equalities result by identifying the coefficients of g and f. ….…. Xn] such that g = h(s1. By the fundamental theorem of the symmetric polynomials.9 Theorem. sn are the fundamental symmetric polynomials in the indeterminates X1. xn).

Proof. x ∈ K. 2.10 Definition. Prop. c)⇒d) Let f ∈ K[X]. d) Any polynomial of degree at least 1 with coefficients in K has a root in K. so g splits over K. so K = K(x) and x ∈ K. 2. since any finite extension is algebraic. f )⇒a) Let L be an algebraic extension of K and let a ∈ L.11 Examples. if F is a field with n elements (n ≥ 2). such that f does not split in K[X]. e)⇒f ) Evident.198 IV. In- deed. so f = (X − a)g. ∀a ∈ F. Choose f of minimum degree with this property. d)⇒e) Suppose by contradiction that there exists f ∈ K[X]. hence it has degree 1. An extension L of the field K is called an algebraic closure of K if L is algebraically closed and K ⊆ L is an algebraic extension. deg f ≥ 1. contradiction. deg f ≥ 1. ! so a ∈ K. a) A finite field is never algebraically closed. a)⇒b) Clear. By hypothesis. A field satisfying the equivalent properties above is called an algebraically closed field. with g ∈ K[X]. (“K is algebraically closed in L”). f) The irreducible polynomials in K[X] are the polynomials of degree 1. the polynomial f = 1 + ∏ a∈F ( X − a ) ∈ F[X] has degree n and f (a) = 1. Since f = (X − a)g. Then Irr(a. the algebraic closure of K in L coincides with K. so f has no roots in F. Then K ⊆ K(x) is finite. . b)⇒c) Let x ∈ K'L. Field extensions c) For any extension L of K. 1. But deg g < deg f. f also splits over K. f has a root a ∈ K. K) is an irreducible polynomial in K[X].23 ensures the existence of an extension L of K in which f has a root x. Since x is in K'L. e) Any polynomial f of degree n ≥ 1 with coefficients in K splits over K ( f has n roots in K).

r). Then |L| ≤ max(|K|. First. K) and a = ar. because C is uncountable and any algebraic extension of Q is countable (see next lemma). The function ϕ is injective. although complex numbers continue to play an important role in mathematics. Any f ∈ P has a finite number n{ f } of roots in L. also interesting in its own right. |N|). …. so |L| ≤ |P| × |N| = max(|P|. then L is algebraically closed. Define ϕ : L → P × N as follows: if a ∈ L. n n so: |P| = |∪n≥1 Pn| = ∑n≥1 |Pn|. n −1 n ψ(a0 + a1X + … + an − 1X + X ) = (a0. This fact is known as “The Fundamental Theorem of Algebra” 45 and will be proven later. |Pn | ≤ |K| . let us prove a technical lemma. |N|). 45 . …. On the n other hand. let ϕ(a) = ( f. then K'L is an algebraic closure of K (prove this!). if K ⊆ L is an extension and L is algebraically closed. We have to show |P| ≤ max(|K|. But Q'C. so |P| = ∑n≥1 |Pn| ≤ ∑n≥1 |K| = This name (kept for historic reasons) expresses the concept that Algebra studies mainly complex numbers. Note that P is the union of the disjoint sets (Pn)n ≥1 . Proof. More generally. Algebraically closed fields 199 b) The complex field C is algebraically closed. is injective. Of course. 2. C is not an algebraic closure of Q. {a1.2 Roots of polynomials. where Pn is the set of polynomials in P of degree n. an − 1). We establish an indexing of these. Let P be the set of all irreducible monic polynomials in K[X] and let a ∈ L. since ψ : Pn → K . |K| = |K| if K is infinite.12 Lemma. an{ f }}. c) If K and L are isomorphic fields and K is algebraically closed.IV. Let K ⊆ L be an algebraic extension. is an algebraic closure of Q. where f = Irr(a. the algebraic closure of Q in C. |N|). An important problem is the existence (and uniqueness) of an algebraic closure of a given field K. This is no longer true since the 19th century.

The problem is that the algebraic extensions of K do not form a set! Fortunately. the union of the elements in B is in A (prove!) and is an upper bound for B. 46 . 48 In fact.a) suggests looking for the algebraic closure as a maximal element46 in the “set” of all algebraic extensions of K. 47 This choice for | M | is suggested by the previous lemma. then |K| is finite and |P| ≤ ∑n≥1 |K| = |N| (as a countable union of finite sets). The set A is ordered: for any F. 2. we can include K in a “sufficiently large” set M (that contains “all” algebraic extensions of K) and confine our search to algebraic extensions of K included in M. Th. Then an algebraic closure of K exists. Let K be a field. This means that there exist real transcendental numbers (even uncountably many!). Zorn's Lemma assures the existence of a maximal element F of A.9. Field extensions n n |K × N| = |K|. We want to be sure that M includes a "copy" of any algebraic extension of K. 2. E is an algebraic extension of K. Let M be a set such that47 | M | > max(|K|. Zorn's Lemma will be used. F ≤ L if and only if F is a subfield of L. we need just an injective function α : K → M. The previous lemma says that You guessed. Let E be an algebraic extension of F.13 Theorem. All proofs of the existence of the algebraic closure use some form of the Axiom of Choice. thus. Proof. the extension Q ⊆ R is not algebraic. We prove that F is an algebraic closure of K. If K is finite. but we assume K ⊆ M to simplify notations. |N|) and K ⊆ M. The set A is nonempty (K ∈ A) and is inductively ordered.48 Let A be the set of algebraic extensions of K included in M. L ∈ A.200 IV. Indeed. because R is not countable. if B is a chain in A. By the transitivity property of algebraic extensions. ! An immediate consequence is that any algebraic extension of Q is countable.

…. Then ϕ : E → ϕ(E) is an F-isomorphism of fields and thus ϕ (E) is an algebraic extension of K. the splitting field of a family F over K depends on the choice of an algebraic closure Ω of K. every polynomial in the family F splits in the splitting field of F over K. any polynomial in the family) splits. 2. ϕ(E) ∈ A and F ⊆ ϕ(E). the addition in ϕ(E) is defined by ϕ(x) + ϕ(y) := ϕ(x + y). By transport of structure. …. such that ϕ|F = idF. F = ϕ(E). injective. 2 . b) The algebraic closure of K is the splitting field over K of the family of all nonconstant polynomials in K[X]. let Ω be an algebraic closure of K and let F be a family of polynomials in K[X]. 2.2 Roots of polynomials. Because F is maximal in A. xn ∈ Ω. ∀x. Of course. S := K({x ∈ Ω | ∃f ∈ F such that f(x) = 0}). it is interesting to study the “smallest” field over which the polynomial (respectively. so F = E. a) The splitting field of X − 2 over Q is Q( 2 ).IV.15 Remarks. If f ∈ K[X] has the roots x1. For a field K and a polynomial (or a family of polynomials). If F = { f }. S is called the splitting field of f over K. The splitting field of the family F over K is the field S obtained by adjoining to K all the roots in Ω of the polynomials in F. then the splitting field of f over K is K(x1. a) At first glance. We can then define ϕ : E → M. 2. …. This shows that F is algebraically ! closed. We shall prove though that any two splitting fields of F over K are K-isomorphic. Let K be a field. 2 b) The splitting field of X − 2 over R is R. xn) = K[x1. y ∈ E). This is the reason we say the (and not a) splitting field of F over K.16 Examples. xn]. ϕ(E) becomes a field (for instance. So.14 Definition. Algebraically closed fields 201 | E | < | M |.

the roots of X − 2 are 3 2 . then ϕ(α) is also a root of f. Since deg g = n − 1. ω 2 3 2 ) . K)) ≤ deg f = n. Field extensions 3 c) The splitting field of X − 2 over Q is Q( 3 2 . Proving the uniqueness (up to a K-isomorphism) of the splitting field of a family of polynomials over K requires some results on the extension of field homomorphisms. 2.ω ). ω ) = Q (3 2 . Apply the K-homomorphism ϕ to this equality: .ω 3 2 .17 Proposition. Indeed. if x1.ω 2 3 2 and Q (3 2 . then [K(x1) : K] = deg(Irr(x1. n Proof.202 IV. then [L : K] ≤ n!. the extension to K[X] of the homomorphism σ is denoted also by σ. then σ has a unique extension to a ring n homomorphism τ : K[X] → L[X]. The existence and uniqueness of τ are a consequence of the universality property of the polynomial ring K[X]. This is the unique ring homomorphism τ : K[X] → L[X] satisfying τ|K = σ and τ(X) = X. and L is its splitting field over K. d) If f ∈ K[X] is a polynomial of degree n. …. in particular. α ∈ E is a root of f ∈ K[X]. We shall use frequently the following elementary fact: if σ : K → L is a field homomorphism. By a harmless abuse of notation. and ϕ : E → L is a K-homomorphism. where ω ∈ C is 2 3 a root of X + X + 1. but has deep implications. The following property is very simple. If E and L are extensions of K. Let f = a0 + a1 X + … + an X ∈ K[X]. namely τ(a0 + a1X + … + anX ) n = σ(a0) + σ(a1)X + … + σ(an)X . xn ∈ L are the roots of f. Indeed. Note that L is a splitting field over K(x1) of g := f/(X − x1) ∈ K(x1)[X]. apply an induction to obtain that [L : K(x1)] ≤ (n − 1)! and so [L : K] ≤ n!. These results have also other important applications. it is instrumental in the determination of the Galois group of an extension. ω 3 2 . Then f(α) = a0 + a1α + n … + anα = 0.

for an irreducible polynomial f ∈ K[X]. Suppose σ : K → K' is a field isomorphism. ϕ : F → Ω . L ⊆ F. then K(α) is isomorphic to K'(α') by an isomorphism σ' that extends σ and σ'(α) = α'. there exists a K'-isomorphism γ : K'[X]/{ f '} → K'(α'). Proof. If E is an algebraic extension of L. E = {(F.2 Roots of polynomials. If α' is a root of f ' := σ { f } ∈ K[X] in some extension E' of K'. such that γ {X + { f '}} = α'. 2.! The proposition below uses the isomorphism K[X]/{ f } ≅ K(α). Algebraically closed fields n n 203 0 = ϕ(a0 + a1α + … + anα ) = a0 + a1ϕ(α) + … + anϕ(α) = f(ϕ(α)). ϕ) | F is a subfield of E. In other words. ϕ is a homomorphism. let Ω be an algebraic closure of K and let σ : L → Ω be a K-homomorphism. where α is algebraic over K and f = Irr(α. ϕ{h + { f }} = -1 σ (h) + { f '}. We saw that there exists a K-isomorphism η : K[X]/{ f }→ K(α) such that η{X + { f }} = α. We also have an isomorphism ϕ : K[X]/{ f }→ K'[X]/{ f '}. . Similarly. regardless of the choice of an extension E in which f has a root and regardless of the choice of a root (of f ) α ∈ E. Let K be a field.IV.19 Theorem. Then γ ◦ϕ ◦η is the isomorphism σ' we need: K (α ) ⎯η ⎯→ K [ X ] ( f ) ⎯ϕ K ′[ X ] ( f ′) ⎯γ K (α ′) ⎯→ ⎯→ −1 ! The seemingly pointless degree of generality in the previous proposition (taking two isomorphic fields K and K' instead of just K) is in fact useful in the next Theorem. f ∈ K[X] is irreducible and α is a root of f in some extension E of K. K). K(α) is the same (up to a K-isomorphism). for any h + { f } ∈ K [ X ] ( f ) . 2. Let E be the set of “extensions” of the homomorphism σ. then there exists a K-homomorphism τ : E → Ω that extends σ. ϕ|L = σ}. let L be an algebraic extension of K.18 Proposition. Proof.

x' is a root of ϕ{ f } (in Ω).20 Corollary. Thus. a maximal element (F. then any two splitting fields of the family F over K are K-isomorphic.18 to the following situation: ϕ : F → ϕ(F) field isomorphism. E is inductively ordered and has. which finishes the proof. d) Let Ω be an algebraic closure of K. The canonical inclusion ι : K → Ω has an extension to a K-homomorphism σ : Ω → Ω' (by the previous theorem). isomorphic to Ω (by σ). Then: a) Any two algebraic closures of K are K-isomorphic. by Zorn's Lemma. let f = Irr(x. If F ≠ E. We obtain an isomorphism ϕ' : F(x) → ϕ(F)(x') that extends ϕ. The element x is algebraic over F. Then any algebraic extension of K is K-isomorphic to a subfield of Ω that includes K. c) If F is a family of polynomials in K[X]. ϕ). a) Let Ω and Ω' be algebraic closures of K. Since Ω' is algebraic over σ(Ω). Let us prove that F = E. Let σ : Ω → Ω' be a K-isomorphism. f ∈ F[X].! 2. ϕ' is an extension of ϕ to F(x). it is bounded above by { ∪i∈I Fi. σ(Ω) = Ω'. Moreover. x is a root of f (in E). Since ϕ(F)(x') ⊆ Ω. σ is an isomorphism. pick x ∈ E \ F. ϕ) ≤ (F'. ϕ}. where ϕ : ∪i∈I Fi → Ω is defined by ϕ(x) = ϕi(x) if x ∈ Fi (this definition is independent of the choice of i ∈ I such that x ∈ Fi). F). If f ∈ F. We have to prove that K(R) and K(R') are K-isomorphic. if {(Fi. contradicting the maximality of (F. Proof. then σ(α) is a root of f in Ω'. ϕ') if and only if F ⊆ F' and ϕ'|F = ϕ. defined by: (F. ϕi)}i∈I is a chain in E. The image of σ is a subfield σ(Ω) of Ω'. as in a). 2. so it is algebraically closed.204 IV. Apply Prop. Let K be a field. and α ∈ Ω is a root of f. Let R (respectively R') the set of all roots in Ω (respectively Ω') of the polynomials in F. This shows that σ (R) ⊆ R'. Field extensions E is ordered by " ≤ ". A straightforward proof shows that " ≤ " is indeed an order relation. Considering . b) If f ∈ K[X]. then any two splitting fields of f over K are K-isomorphic. c) Let Ω and Ω' be algebraic closures of K. ϕ). Thus.

21 Theorem. Note that all proofs use some analysis.F. 49 . a subfield of K. Jean le Rond d’Alembert proposes an incomplete proof in 1746. then f has a real root. so σ establishes a bijection between R and R'. Let deg f = n ∈ N and let s{ f } be the exponent of 2 in the prime s{ f } s{ f } + 1 . (the Fundamental Theorem of Algebra)49 The field C is algebraically closed. the first one in 1797. b) is a particular case of c).IV. C. Augustin Louis Cauchy (1820). If s{ f } = 0. If we remember the form of the elements in K(R). Algebraically closed fields 205 : Ω' → Ω. we obtain that σ (K(R)) = K(σ (R)) = K(R'). The “theorem of Liouville” (which is due in fact to Cauchy. Proof. The present proof belongs to Pierre Samuel and has the advantage (?) of being more “algebraic”. 2. then deg f is odd and f has a real root. Thus. L is K-isomorphic to ϕ(L). Also known as the d’Alembert-Gauss theorem. Other proofs are due to Jean Argand (1814). Gauss gives four correct proofs of this theorem. The essential role is played in fact by the order properties of R. We prove the claim factor decomposition of n (2 | n and 2 by induction on s{ f }. we obtain analogously that σ (R') ⊆ R. Let us prove first that any polynomial f of degree ≥ 1 with real coefficients has a complex root. d) Let K ⊆ L be algebraic.n).2 Roots of polynomials. because fundamental (topological) properties of R do not possess purely algebraic descriptions. Then the canonical inclusion ι : K → Ω extends to a K-homomorphism ϕ : L → Ω by 2. ! σ −1 −1 Part d) above says that a given algebraic closure Ω of K includes “all” the algebraic extensions of K.19. 1844) –“any holomorphic bounded function on C is constant”– proves the theorem in one line. This follows from the well known fact from analysis: the polynomial function ϕ : R → R associated to f has both positive and negative values (since the limits of ϕ at + ∞ and − ∞ are infinite and have opposite signs). If deg f is odd. so the restriction of σ to K(R) is a K-isomorphism between K(R) and K(R'). the continuity of ϕ implies the existence of c ∈ R with ϕ(c) = f (c) = 0.

deg f = n = 2 m. ga has a complex root. Prop. Consider the following element in K: uij(a) := titj + a(ti + tj) = uji(a). the polynomial above coincides with ga. Let ga := ∏{i .σ( j)} | {i. a polynomial in K[X]. 2. tn. such that uij(a) ∈ C and uij(b) ∈ C. If σ is a permutation of the set {1. for any a ∈ R. …. On the other hand. We claim ga has coefficients in R. the roots of ga are uij(a). with s s(g) < s. …. The above proves that. Because R is infinite and P is finite. . the polynomial obtained from ga by the action of σ on the roots (i. Suppose that any polynomial g ∈ R[X]. n}. s > 0. This means that ga has coefficients symmetric polynomials in t1. j} ∈ P} = P. …. with a ≠ b. tj ∈ C. j} ∈ P. To show that these polynomial expressions are real. b ∈ R. …. j} ∈ P. tn with real coefficients. s{ f } = s. ∀i) is ∏{i. …. there exist a. has a complex root. so there exists {i. j}∈P (X − uσ (i )σ ( j ) (a )) Since {{σ(i). Field extensions Let s ∈ N. K = C(t1.e. j} ∈ P. j}∈P ( X − uij (a )) . It follows immediately that s := ti + tj and p := titj are complex numbers and so ti. the coefficients of ga are polynomial expressions of t1. The polynomial ga has degree n(n − 1)/2 and (a priori) coefficients in K. fix a ∈ R and let {i. Let P be the family of subsets of {1. Let f ∈ R[X]. n} having two elements. tn). there exists {i. Thus. 2. tn with real coefficients.8 implies that ga ∈ R[X]. …. 2. tn be the roots of f in a splitting field of f over C. Let t1. let us study how they change under a permutation of the roots t1. …. so s(ga) = s − 1 and the induc2 tion hypothesis applies to ga ∈ R[X]. Indeed. as 2 roots of X − sX + p ∈ C[X]. ti & tσ(i). such that uij(a) ∈ C.206 IV. n (n − 1) We have = 2 s −1 m (2 s m − 1) . with uij(a) ∈ C.

let g be the conjugate of g (the polynomial obtained by the complex conjugation of the coefficients of g).2 Roots of polynomials. ⎩⎝ − v u ⎠ ⎭ 50 a) Show that H is a skew field (division ring) with respect to addition and multiplication of matrices.IV. then g has ! the root α . The following statements are equivalent: a) R is a domain. An elementary computation shows that g ⋅ g ∈ R[X] and the first part applies. c) Any polynomial of degree1 in R[X] has at most one root in R. If g(α) = 0. we are done. 50 The skew field of quaternions was discovered (invented?) by W. Exercises Throughout the exercises. Let R be a commutative ring.R. Therefore. Hamilton. If g (α) = 0. Algebraically closed fields 207 In the general case of a polynomial g ∈ C[X]. b) Any polynomial p ∈ R[X] of degree n ≥ 1 has at most n roots in R (the roots being counted with their multiplicity). 2. (The skew field of quaternions) Let ⎧⎛ u v ⎞ ⎫ H = ⎨⎜ ⎟ ∈ M 2 (C ) u. v ∈ C ⎬ . . 1. K is a field and Ω is an algebraic closure of K. there exists α ∈ C with g (α)⋅g(α) = 0.

Show ⎝0 1⎠ ⎝0 − i⎠ ⎝ −1 0⎠ ⎝ i 0⎠ that any element of H is written uniquely as a1 + bi + cj + dk. irreducible. xn −1) is a splitting field of f over K. (Q is called the quaternion group). xn be the roots of f in some extension L of K. Field extensions ⎛1 0⎞ ⎛i 0 ⎞ ⎛ 0 1⎞ ⎛0 i ⎞ b) Let 1 := ⎜ ⎟ . xn). 7. Does this contradict 2. b. Show that E is a splitting field of f over K if and only if E a splitting field of f over L. Prove that Q is not commutative and has 8 elements. deg g ≥ 2. xn are roots of f (not necessarily all of them). If E is a splitting field of f over K. β ∈ Ω are roots of g.3? e) Prove that H is an extension of C. d ∈ R. where L is algebraically closed and L is not algebraic over K. 5. where x1. Suppose L = K(x1. Suppose α − β ∈ K and let β = α + b. Write down the multiplication table of Q.208 IV. where a. Deduce that H is a skew field. Let K ⊆ L ⊆ E be a tower of extensions and let f ∈ K[X]. …. c. then E is a splitting field of f over L. If α. Give an example of an extension K ⊆ L. Then K(α) = K(α + b). i := ⎜ ⎟ . k := ⎜ ⎟ . (Hint. c) Let Q be the multiplicative subgroup of H* generated by i and j. 8. Let K ⊆ L ⊆ E be field extensions and let f ∈ K[X]. …. f) Give an example of a countable skew field. 9. 4. b ∈ K. there exists a K-isomorphism ϕ : K(α) → K(β) = K(α) that takes α to . Let char K = 0 and let g ∈ K[X]. 3. Because g is irreducible. Show that K(x1. …. then α − β ∉ K. then L = Ω. …. j := ⎜ ⎟ . If the extension K ⊆ L ⊆ Ω is such that any nonconstant polynomial in K[X] splits over L. 2 d) Prove that X + 1 ∈ H[X] has an infinity of solutions in H. Show that (Q'R)(i) = Q'C. 6. Let f ∈ K[X] be a polynomial of degree n and let x1.

ϕ(ϕ(α)) = ϕ (α). Also: p . ϕ(x) = x .… are roots of ϕ.3 Finite fields We now apply the results obtained so far (notably the existence of the algebraic closure of a field) to determine all finite fields. then f(g(α)) = 0 and: f(g(X)) is irreducible in K[X] ⇔ [K(α) : K] = deg f(g(X)) = deg f ·deg g. Let f = a∏(X − βi). where β = β1. We already know (1. K(β)). p a prime. If α is a root of g(X) − β. n . Then: f(g(X)) is irreducible in K[X] ⇔ f is irreducible in K[X] and g(X) − β is irreducible in K(β)[X]. these are all distinct. char K = p > 0.char F is a prime p > 0. Then α. the field of integers modulo p). (Hint. Let f. Since ϕ (α) = α + nb and char K = 0. These facts raise the following questions: Given a prime p and n n ∈ N*. then: .10) that. In the tower of extensions K ⊆ K(β) ⊆ K(α). ∀x. βn ∈ Ω. [K(β) : K] = deg Irr(β.3 Finite fields 2 n 209 β = α + b. nonconstant and let β be a root of f in Ω. Let K be a commutative ring. is it unique (up to isomorphism)? How can one construct it? 3. Of course. where n = [F : Fp].…. does there exist a finite field with p elements? If it exists. g ∈ K[X]. ∀x ∈ K. ϕ(α). if F is a finite field.|F| = p . ϕ (α).F is a finite extension of Fp ( = Zp. ϕ(xy) = ϕ(x)ϕ(y).IV. K) and [K(α) : K(β)] = deg Irr(α. ….) IV. n for any n. .) 10.1 Definition. y ∈ K. Let: ϕ : K → K.

2 Theorem. der of an element in a finite group. Conversely. 3. German mathematician. for example in K[X]. We use this endomorphism especially if K is a field (but also for rings. for any x ∈ Fp. a subring in K. n a) Let n ∈ N* and let q := p . If char K = 0. where m divides n. for any divisor m of n there m exists a unique subfield of F with p =: r elements. p p The last equality holds because p divides the binomial coefficients ( )= p i p ( p −1)…( p −i +1) i! . for any 0 < i < p. Consequently. the Frobenius endomorphism is taken to be the identity function. considered with coefficients in 51 Ferdinand Georg Frobenius (1849-1917). see also the proof below). The Frobenius endomorphism is the identity in the p case of the field Fp (by the little Fermat theorem. then m |K| = p . Field extensions ϕ(x + y) = (x + y) = p 0 ≤i ≤ p ∑ ( p) x i p −i yi = x + y . Let p be a prime and let Ω be an algebraic closure of the field Fp. ϕ is a ring homomorphism. namely the splitting field of X − X over Fp. The image of ϕ is K : = {x | x ∈ K}. Applying the Lagrange theorem for the orq −1 q = 1. so x = x. . a) Suppose F ⊆ Ω is a field with q elements. called the Frobenius endomorphp p ism51 of K. Proof. Thus. In parn ticular. x = x. ϕ = id : K → K. n b) If F is a field with p elements and K is a subfield of F. we obtain x ∀x ∈ F*. ·) has q − 1 elements. namely r K = {x ∈ F| x = x}. The group * (F . where K is a field of characteristic p). any element of q F is a root of the polynomial X − X.210 IV. The prime subfield of F is Fp. There exists a unique subfield F of Ω q having q elements. there exists a finite field with p elements and any two fields n with p elements are isomorphic.

Let us prove that there exists a field with q elements F ⊆ Ω. then F is the subfield of all elements fixed by this homomorphism.6).IV. This q means that F is the splitting field of X − X over Fp and F is the unique subfield with q elements of Ω. Let L := {x ∈ Ω | x = x} = {x ∈ Ω |ψ(x) = x}. Let F be the set of all roots of g in Ω. q So. We deduce that F. the elements of F are exactly the roots of X − X in Ω. This polynomial can have at most q distinct roots in Ω. honors Evariste Galois. since q−1 its derivative is qX − 1 = −1 (see 2. ! The finite field with p elements (unique up to an isomorphism) is denoted by GF(p ) or F pn . b) Let K be a subfield of F (we suppose that F ⊆ Ω). is a field homomorphism (ψ = ϕ where ϕ is the Frobenius). thus. The notation GF(p ). We claim F is a subfield of Ω. We have p = |F| = s . if m divides n. Let Ω be an alger braic closure of F and let ψ : Ω → Ω. who determined the structure of the finite fields in 1830. We have 2 t r t (ψ ◦ψ) (x) = ψ(x ) = x r and. since q |F| = q. We show that L ⊆ F. The q argument above says F must be the splitting field of X − X over Fp. this is m mt n possible only if s is of the form p and p = p . ∀x ∈ Ω. t t t n x ∈ L ⇒ ψ(x) = x ⇒ ψ (x) = x ⇒ x r = x . ∀x ∈ Ω.3 Finite fields 211 Fp. On the other hand. ψ (x) = x r . Indeed. ψ(x) = x . Let s := |K| n t and let t := [F : K]. the splitting field of g over Fp. . q F = {x ∈ Ω | x = x}. let g = X − X. since p is a prime. But r = p . the number of roots of g in Ω is exactly q: g has at most q roots and has no multiple roots. q n ψ : Ω → Ω. Any subfield E of F with r elements is equal to L because all elements of E satisfy the equation r x = x. an acronym for n n n “Galois Field”. L is a subfield with r elements of Ω. then n = mt for some t. which implies m|n. So. x ∈ L ⇒ x ∈ F. has q elements. by induction. ∀x ∈ Ω. where m r r = p . ψ(x) = x . and F = n {x ∈ Ω | x p = x}. Conversely.

·). which has at most m roots in R.212 IV. contradiction. where n q = p . where n = d1·…·dm > dm. a field with q elements. see exercise 3.5 Definition. then G is cyclic. Lagrange's theorem implies that m | n. 3. Field extensions 3. which have also numerous other applications in algebra. |Cm| = dm and d1 | d2 |…| dm. If R is a domain and G is a finite subgroup of U(R) (the multiplicative group of units of R).2). so G is cyclic. Thus. Let F be the splitting field of g = X − X over Fp. Let K be a field and let n be a positive integer.3 Proposition. Fp[X]/{ f } ≅ Fp(α). n Fp[X]/{ f } is a field with p elements. Use the invariant factors theorem. any finite subgroup of K* (where K is a field) is cyclic. ! We prove now that any finite division ring is commutative (it is a field). |C1| = d1. Proof. ….4 Lemma. q Proof. G ≤ (U(R). there exists at least an irreducible polynomial of degree n in Fp[X]. any element m in G is a root of X − 1. The proof of this result needs some facts on roots of unity and cyclotomic fields. We have that F = Fp(α) and deg f = [Fp(α) : Fp] = [F : Fp] = n. Let m be the exponent of G. ⋅) is a cyclic group. …. generated by x. Cm are cyclic. The next lemma says that (F*. for any such polynomial f. Also. Let G be a finite group with n elements. Another proof. Suppose G is not cyclic. On the other hand. Fp). Then G ≅ C1×…×Cm. where m ≥ 2 and C1. number theory etc. In particular. Then X d m − 1 has n roots in R. For any n ∈ N*. there exists an element x having order m in G (for an elementary proof of this. So. applied to the finite Abelian group G. Since G is Abelian. Let Un(K) . Let α be a generator of F* and let f = Irr(α. The n element ζ ∈ K is called an nth root of unity in K if ζ = 1. n ≤ m and m = n. f is irreducible of degree n in Fp[X]. ! 3. (the GCD of the orders of the elements of G).

3 Finite fields 213 (or simply Un if K cannot be confused) be the set of the nth roots of n unity in K. Let Ω be an algebraic closure of the field K and let n ∈ N*. a) Un(Q) = { −1. We denote by Pn(K) (or simply Pn) the set of all primitive nth roots of unity in K: Pn(K) := {x ∈ Un(K) | ord x = n}. 1} if n is even and Un(Q) = {1} if n is odd. 3. U4(F7) = {1. 1 ≤ k < n. n Since X − 1 has at most n roots. 6}. then Un(K) = Um(K). We have ϕ(n) = |U(Zn)|. the number of invertible elements in the ring Zn. Un(K) is cyclic (as a finite subgroup of K*). cu (k. (k. 4}. 2kπi n where e / = cos(2kπ/n) + isin(2kπ/n). P4(F5) = {2. The function ϕ : N* → N* is called the Euler phi function. There are ϕ(n) complex primi2kπi n tive nth roots of unity. n) = 1. (⇔ n m ζ = 1.6 Remark. the extension K ⊆ K(ζ ) is called a cyclotomic extension.7 Examples.IV. 3. 3. If ζ is a root of unity in some extension of K.8 Proposition. with (p. ∀m < n ⇔ ζ generates Un). Un is a subgroup with at most n elements of the multiplicative group (K*. Here ϕ(n) is the number of all integers k. m) = 1. The element ζ ∈ Un is called a primitive nth root of unity in K if the order of ζ in Un is n. namely e / . t b) If char K = p > 0 and n = p m. …. 3}. 2kπi n b) Un(C) = {e / | k ∈ {1. c) U4(F5) = F5* = {1. n}}. there exists a primitive nth root of unity in K if and only if Un(K) has n elements. Un(K) := {x ∈ K | x = 1}. but ζ ≠ 1. . P4(F7) = ∅. 1 ≤ k < n. 2. 3. Thus. ·). n) = 1. a) Un(Ω) has n elements if and only if char K does not divide n.

of Greek origin and meaning approximately “circle dividing”. ∀y ∈ K.214 IV. 3. m | n implies Um ⊆ Un. Let n ∈ N* and let K be a field containing a primitive nth root of unity. so x ∈ Um. so deg Φn = ϕ(n). In other words. f. Then h ∈ K[X].( k . 3. In the complex plane. if g divides f in L[X]. a) Un(Ω) has n elements if and only if f = X − 1 has no n −1 n multiple roots ⇔ { f '. Let p =: q. g ∈ K[X] and h ∈ L[X] such that f = gh. the complex nth roots of unity are the vertices of a regular polygon with n sides inscribed in the unit circle. f } = 1 ⇔ {nX . The polynomial (in K[X]) Φn := ∏ζ ∈P ( K ) ( X − ζ ) n is called the nth cyclotomic polynomial over K. then Φn := ∏ (X − e 2 kπi n ) k <n . If x ∈ Un. it is sufficient to study Un(K) when char K does not divide n. a) Suppose K ⊆ L is an extension. Field extensions n Proof. x ∈ Un implies x = 1. is a field endomorphism (necessarily injective) of K (it is a power of the Frobenius endomorphq m ism).9 Definition. then q n q x = (x m ) = 1. 3.n )=1 is called the nth cyclotomic polynomial (dropping any reference to the field). This justifies the name cyclotomic. X − 1} = 1⇔ n·1 ≠ 0 ⇔ char K does not divide n.! According to the above. The roots of Φn are distinct and they are exactly the primitive nth roots of unity in K. But y & y . t b) In general.10 Remark. If char K = 0 (one usually takes K = C to ensure a primitive nth root of unity exists). . so y = 1 implies y = 1.11 Lemma. there exists a primitive nth root of unity in some extension of K. Thus. then g divides f in K[X]. In this case.

then Φn ∈ Z[X]. with deg r < deg g. Then af. there exists a primitive dth root of unity in K. then ζ is a primitive dth root of unity. primitive. we obtain f = ± f1 ∈ Z[X]. The division with remainder in L[X] of f to g is f = gh + 0. since c(u) = 1. d) If char K = 0 (for example K = C). r ∈ K[X] such that f = gq + r. c(af )·c(bg) = ab. g1 ∈ Z[X]. where g ∈ Z[X] and g is monic. Then: a) For any d dividing n. 3. . a) If ζ is a primitive nth root of unity and m = n/d. Let a (respectively b) be the LCM of denominators of the coefficients of f (respectively g). b) X n − 1 = ∏d n Φ d holds. since u is monic. Likewise. c) We prove by induction by m (1 ≤ m ≤ n) the statement: "for any d with 1 ≤ d ≤ m. c) The coefficients of Φn are in the prime subfield P of K. g1 ∈ Z[X].12 Proposition. with g ∈ Q[X]. u = fg.IV. Simplify by ab = c(af )·c(bg) to obtain f1·g1 = u. The uniqueness of the quotient and of the remainder in L[X] implies that r = 0 and h = q ∈ K[X]. Since f (which is monic) is associated in divisibility with f1. d | n implies Φd ∈ P[X]". and f1. namely the nth roots of unity in K. bg ∈ Z[X] and af·bg = abu.3 Finite fields 215 b) If f ∈ Q[X] is monic and divides (in C[X]) a monic polynomial u ∈ Z[X]. then f ∈ Z[X] and u = fg. So af = c(af )·f1. g1 have leading coefficients 1 or − 1. m Proof. f1. Let n ∈ N* and let K be a field containing a primitive nth root of unity. a) The division with remainder theorem in K[X] guarantees the existence of q. with f1. b) By a). b) The polynomials in the equality are monic and have the same roots. Thus c(af )·f1·c(bg )·g1 = abu. bg = c(bg )·g1. ! g ∈ Z[X]. Proof. Taking the contents of the polynomials.

d ∀n ∈ N*. Φm ∈ Q[X]. F2(α) is the field with 8 elements F8. The Möbius inversion In characteristic 0 the behavior above is not present: the cyclotomic polynomials over C are irreducible (in Z[X] and in Q[X]): . If α is a primitive 7th root of unity. SPINDLER [1994]) gives an explicit expression for Φn. d < n). If m | n. d m . One can use the formula X n − 1 = ∏d n Φ d to recursively compute formula (see e. d) Repeat word for word the proof of c) (replace P with Z). where the right hand side is the irreducible factorization of Φ7 in F2[X] (the polynomials have degree 3 and no roots in F2). then X − 1 = Φm·g. ∀d < m. in terms of products of polynomials X − 1 or their inverses. ! Use lemma 3. where g = ∏ Φ d ∈ P[X]. to obm tain that X − 1 = Φm·g. So. From Φ1Φ7 = X − 1 we deduce 6 5 Φ7 = X + X +…+ 1. for d | n. where g ∈ Z[X]. Φ7 is reducible in F2[X].11. If m . By trial and error (or using a computer) we obtain 3 2 3 Φ7 = (X + X + 1)(X + X + 1). d <m since by the induction hypothesis Φd ∈ P[X].b) to deduce Φm ∈ Z[X]. there is nothing m to prove. Why? Φn (for this.216 IV. monic. then Φ1 = X − 1 ∈ P[X]. Apply now the lemma 3.11.13 Example.n. Consider the 7th cyclotomic polynomial Φ7 over an 7 algebraic closure of F2. Thus. all Φd are needed. 3. F2) = X + X + 1 or X + X + 1. d | m. We suppose the claim true for any q < m and we prove for m. Any nonzero element of F8 is a primitive 7th root of unity.g.a) to obtain that Φm is also in P[X]. Field extensions If m = 1. Let 1 < m ≤ n. 3 2 3 then Irr(α.

then K(ζ) and K(β) are K-isomorphic via an isomorphism that m takes ζ to β. Then Φn is irreducible in Z[X] and it is the minimal polynomial over Q of any complex primitive nth root of unity. Indeed. we obtain that all the primitive nth roots of unity are roots of f. Let g = a0 + a1X + … + X . 3. π : Z[X] → Zp[X] is the unique extension to Z[X] of the canonical n homomorphism Z → Zp). Let ζ be a primitive complex nth root of unity and let f = Irr(ζ. Φn = fg. we show that f and Φn have the same roots. Because f = Irr(α.11). Thus.IV. p . We will show that: p If α ∈ Pn and f (α) = 0. so f | Φn. if β is a root of f. By lemma p 3. ζ = 1 if and only if m β = 1. with g ∈ Z[X]. monic. In what follows. Q).n. Proof. so ord ζ = ord β. In particular. an inductive argument based on (*) shows that m f (ζ ) = 0. f | g1 in Z[X] (by 3. Any root of f is a primitive nth root of unity. for any prime p not dividing n. Since α ∈ Pn. we have. so p p α is a root of g. so g1 = fh. We show that f = Φn. α is a root for g1 := g(X ) ∈ Z[X]. Let n ∈ N*. for any q ∈ Z[X].14 Theorem. Suppose by contradiction that f (α ) ≠ 0. This shows that any root of f is a root of Φn. Since any primitive nth root m of unity is of the form ζ for some m coprime with n. in other words.b).3 Finite fields 217 3. let α ∈ Pn and let p be prime. ∀m ∈ N*. then f (α ) = 0.(*) Since f (ζ ) = 0. Q). We need now to prove that any primitive nth root of unity is a root of f.11. let π(q) denote the polynomial q reduced modulo p (π(q) is the polynomial in Zp[X] whose coefficients are the images in Zp of the coefficients of q. more precisely. In order to prove (*). with h ∈ Z[X]. p p p p f (α )g(α ) = Φn(α ) = 0. Then: .11. for any m relatively prime to n. monic. n Since f divides X − 1 in C[X] and f is monic.b) implies that f ∈ Z[X].

If x ~ y. is an equivalence relation on G. . there exists an irreducible common factor h of π{ f } and π(g). b). Thus π{ f } | π(g) in Zp[X]. This is absurd. defined by: ∀x. d) Let Ca := {x ∈ G | x ~ a } (the conjugacy class of the element a). where the sum runs on a system S of representatives of the conjugacy classes of the elements not contained in the center of G. b) The set C(a) := {x ∈ G | xa = ax} (called the centralizer of a in G) is a subgroup of G. so all irreducible divisors of π{ f } divide also π(g). which will be used here to prove that any finite division ring is commutative. so π(Φn) divides n n −1 π(X − 1). p p Since g1 = fh. ∀y ∈ G} (called the center of G) is a normal Abelian subgroup of G. ! We will encounter cyclotomic extensions again in the study of Galois Theory. 3. Proof. Let G be a group and let a ∈ G.218 p IV. Then |Ca| = [G : C(a)] (the index of the subgroup C(a) in G).n implies n invertible in Zp) and π(X − 1) is coprime to n −1 . 2 deg h ≥ 1. But then π(Φn) = π{ f }π(g) is divisible with h . y ∈ G. a). a) The set C(G) := {x ∈ G | xy = yx.15 Proposition. We need now some elementary facts on the conjugacy classes of a finite group. e) (conjugacy classes formula) We have: |G| = |C(G)| + ∑a∈S [G : C (a )] . so it has n multiple roots. we say that x is conjugate with y (or y is a conjugate of x). π(g1) = π(g) = π{ f }π(h). Field extensions p n p p p p nX π(g ) = π(g) = (π(a0) + π(a1)X + … + X ) = π(a0) + π(a1) X + … + pn p pn X = π(a0) + π(a1)X + … + X = π(g1). who has no multiple roots: its derivative is nX ≠ 0 in n Zp[X] (p . x ~ y if −1 and only if exists z ∈ G such that y = z xz. since Φn | X − 1. Because deg π{ f } = deg f. c) The “conjugacy relation” in G. c) have standard proofs and are left to the reader.

Let m = [C : Zp] and let n = [K : C]. Then m n |C| = p := q and |K| = q . since −1 −1 x ax = y ay implies C(a)x = C(a)y. (Wedderburn 52 . C := {x ∈ K | xy = yx. ·) is C*. Proof. Because ϕ is obviously surjective. Clearly. e) G is the disjoint union of the conjugacy classes. Let C be the center of the ring K. taking cardinals and using ! |Ca| = [G : C(a)]. We can prove now the following celebrated result: 3. Suppose that n > 1. Then S ∪ C(G) is a system of representatives for the conjugacy classes. Let S be a system of representatives as in the statement. ∀x ∈ G. ∀y ∈ K}. We have thus the tower of extensions of division rings Zp ⊆ C ⊆ K. The injectivity of ϕ also follows. Let p = char K > 0. The unions are disjoint. for any x. Scottish mathematician. Define −1 ϕ : G/C(a) → Ca by ϕ(C(a)x) = x ax. 1909) Any finite division ring is commutative. Let K be a finite division ring. so.3 Finite fields −1 219 d) It is clear that Ca = {z az | z ∈ G}. The centralizer of a in the group K* is 52 Joseph Henry Maclagen Wedderburn (1882-1948). so G = ∪{Ca | a ∈ S ∪C(G)} = {∪{Ca | a ∈ S }}∪ C(G). ϕ is correctly defined −1 (x ax is independent on the representative x of the class C(a)x). consider {x ∈ K | xa = ax} =: Z(a). For any a ∈ K.16 Theorem. a ∈ C(G) ⇔ C(a) = {a} ⇔ |Ca| = 1. y ∈ G: C(a)x = C(a)y ⇔ xy ∈ C(a) ⇔ axy = xy a −1 −1 ⇔ x ax = y ay. . One easily verifies that C is a commutative subring of K. which is easily seen to be a subfield of K that includes C. Let G/C(a) := {C(a)x | x ∈ G} be the set of right cosets of the subgroup C(a) in G. |Ca| = |G/C(a)| = [G : C(a)].IV. we obtain the formula. The center of the group (K*. and C is a field (of characteristic p). In−1 −1 −1 deed. It is sufficient to prove that n = 1.

from (1) −1= we obtain that Φn(q) | (q − 1).27. so a ∈ C*. ∀a ∈ S. * Z(a) = K. So |Φn(q)| > q − 1. In the formula above. Let d(a) = [Z(a) : C]. the claim follows. contradiction with Φn(q) | q − 1. By 1. |K | = q − 1. contradicting the choice of S). d (a ) −1 a∈S q The cyclotomic polynomial (over C) Φn divides (X − 1)/(X − 1) n in Z[X]. * so Zp ⊆ C ⊆ Z(a) ⊆ K implies da | n. Φn(q) n n d(a) divides q − 1 and also divides (q − 1)/(q − 1). d ≠n da | n. so |Z(a)| = q . Note that a ∈ S implies da ≠ n (otherwise qn − 1 . |Φn(q)| = d(a) ∏ζ ∈P n q − ζ . da ≠ n and X ∏d d (a ) Φ d . ! in a plane). 3. We have |q − ζ | > q − 1.b) implies X − 1 = Φn· ∏ Φ d . the multiplicativity of degrees holds also for extensions of division rings. So. Indeed. ·) apply the conjugacy classes formula: |K | = |C | + * * ∑a∈S [K * : Z (a ) ]. represent the complex nth roots of unity . In the group (K . [K : Z(a)*] = q −1=q−1+∑ n * n * * S being a system of representatives of the conjugacy classes of elements not contained in C*. ∀n ≥ 2. since d n .12. |C | = q − 1. On the other hand.220 IV. ∀a ∈ S. so: qd (a ) − 1 (1) n d(a) qn − 1 . ∀ζ ∈ Pn (to see this. Field extensions d(a) Z(a)* := Z(a)\{0}.

Show they have the same roots. (Hint.3 Finite fields 221 Exercises 1. b ∈ G. q) = q n c) Compute N(p. If α generates the group K*. the product of all irreducible monic n polynomials in K[X] with degree dividing n is X q − X . Let K be a finite field with q elements. Let K be a finite field of characteristic p and let α ∈ K*. a) Show that. ∀n ∈ N*. What happens if K is infinite? 5. show that any function ϕ : K → K is polynomial (there is a polynomial g ∈ K[X1.…. xn) = g(x1. More generally. . Let K ⊆ L be an extension of degree n of finite fields. Let G be a finite group and let exp(G) := LCM{ord a | a ∈ G} (the exponent of G). b ∈ G. ord b) = 1. 3.) b) Let N(n.….….IV. Prove that: a) If a. q) for p a prime. xn). Then: ∑d|n d·N(d. b) For any a. Let K be a finite field. then ord ab = (ord a)(ord b). Show that the lattice of all intermediate fields of the extension is isomorphic to the lattice of the divisors of n (ordered by divisibility). 2. ab = ba implies that there exists c ∈ G with ordc = [ord a. Xn] such that n ϕ(x1. q) be the number of irreducible monic polynomials in K[X] having degree n. ab = ba and (ord a. ord b]. then K = Fp(α). ∀(x1.…. c) If G is Abelian. n ∀x ∈ K). there exists an element g of G such that ord(g) = exp(G). Show that any function ϕ : K → K is polynomial (there exists a polynomial g ∈ K[X] such that ϕ(x) = g(x). Is the converse true? 4. xn) ∈ K ).

) The following exercises are about the solutions in finite fields of polynomial equations in several unknowns. If K = Fp. Let Ω be an algebraic closure of the field Fp. Find the number of irreducible polynomials of degree 2 in F5[X]. Determine the number of equivalence classes in Mn(K) with respect to similarity. q = 1 if |Ur(F)| = 1. of (total) degree < n(q − 1). q = ∑{x | x ∈ F}. such that. (Hint. then αSr = Sr). SHAFAREVICH [1985]. g = X − X − a. deg g ≥ 2. where d = (q − 1. then g(X + c) = g(X). for any n ≤ 3. q = −1 if (q − 1) | m .…. Then: . d) Let f ∈ F[X1. q = ∑{x | x ∈ Ur(F)}. For any subset S of the set of all prime numbers P. ∀c ∈ Fp. Show that: r a) Ur(F) = {x ∈ F* | x = 1} is a cyclic subgroup of F*. consider the subfield CS of Ω. Field extensions 6. 12. 11. with d elements. q) for r ≤ n. Let K ⊆ L be an extension of finite fields. Give an example of a finite field K and of an irreducible polynomial g ∈ K[X].9). (Hint. then p α − β ∈ K (cf. CS is the composite of the fields {Fpn | n is a product of primes from S}. Let F be a finite field with q elements and let r ∈ N*. Let p be a prime and let k ∈ N.) 7. Then Sr. then any polynomial of degree 2 in F5[X] has a root in F5[α]. q = 0 if |Ur(F)| > 1. Then any irreducible polynomial of degree n in K[X] splits in L[X]. For some developments. if α. exercise 2. r). Let K be a finite field with q elements. b) Let Sr = Sr.) 9. 8. Make a connection to the N(r. with a ∈ Fp. In the particular case Zp. If 2 α is o root of X + 2 ∈ F5[X]. (Hint. (Hint. see BOREVICH. Sr. Then Φpk = Φp( X p k −1 ). Use that two matrices are similar ⇔ they have the same elementary divisors. Then Tm.m. q = 0 if (q − 1) . 10. Tm.222 IV. Xn]. [L : K] = n. If α is a generator of Ur(F). β are roots of g in Ω. Then Ω has uncountable many subfields. m c) Let m ∈ N* and Tm = Tm. these yield results on “congruences mod p”.

and at least one m is < q − 1. …. deg g < n.4 Transcendental extensions 223 (Hint. I) is the field of rational fractions in the indeterminates (Xi)i∈I with coefficients in K. then K ⊆ K(X. xn) = 0 is a nonzero multiple of p. also g(x1. Then the sum above is a product of n sums of type Tm. We call {x1. xn) | (x1. …. 0).4 Transcendental extensions Recall that the extension K ⊆ L is called transcendental if it is not algebraic: L contains a transcendental element over K. xn) ∈ F n} = 0. 4. Xn]. ….…. ….) 13. xn) = 1. xn) = 0 ⇔ f (x1. ∑{ f (x1. It is sufficient to suppose f is a monomial. …. …. xn} be a finite subset of L. The field K(X) of rational fractions with coefficients in K is a transcendental extension of K.IV.…. I) is a transcendental extension. the equation has solutions other than (0. xn} an algebraically independent set over K (or say that x1. char F = p and n g ∈ F[X1. More generally. xn are algebraically independent over . R is a transcendental extension of Q. xn) = 0. xn) ∈ F n} and use the previous exercise.…. Xn] is a homogeneous polynomial. (Hint. …. xn) = 0 is a multiple of p.1 Definition. Let q−1 . (Warning) Let F be a field with q elements. …. deg g < n. …. char F = p and g ∈ Zp[X1. Then the n number of solutions in F of the equation g(x1. a) Let K ⊆ L be a field extension and let {x1. xn) | (x1. For example. if (Xi)i∈I is a family of indeterminates and K(X. ….…. xn) ≠ 0 ⇔ f := 1 − g f (x1. IV. Calculate ∑{f (x1. (Chevalley) Suppose F is a field with q elements. Then deg f < n(q − 1) and g(x1. ….) 14. Then the number N of solutions in F of the equation g(x1. In particular. ….

Thus. another way to say this is: the purely transcendental extensions of K are (up to K-isomorphism) the extensions K ⊆ K(X. S] → K[S] v(Xs) = s. b) The extension K ⊆ L is called a purely transcendental extension if L is generated over K by a set of algebraically independent elements over K: there exists a subset S of L. e } is not algebraically independ- . ∀s ∈ S. Likewise. Let (Xs)s∈S be a family of indeterminates indexed by S. is injective. ….224 IV. The set {e. distinct and f ∈ K[X1. This isomorphism between the domains K[X. ….2 Remark. …. a) The singleton {x} is algebraically independent over K if and only if x is transcendental over K. The subset S is algebraically independent over K if and only if the unique homomorphism of K-algebras v : K[X. S) and K(S). …. An arbitrary subset S of L is called algebraically independent over K if any finite subset of S is algebraically independent over K. the subring generated by K and S in L (the K-subalgebra of L generated by S). xn) = 0. xn) = 0 an algebraic dependence relation over K for S. I). we call the equality f(x1. with f ≠ 0 and f(x1.3 Examples. xn) = 0 implies f = 0. This means that there exist x1. The image of this homomorphism is K[S]. ∀s ∈ S. is an isomorphism. xn ∈ S. S] and K[S] induces an isomorphism between the fields of fractions K(X. for any polynomial f ∈ K[X1. S] → L such that v(Xs) = s. {π} and {e } are algebraically 5 independent sets over Q. Xn]. …. an algebraically independent set over K behaves the same as a set of indeterminates in the field of fractions of a polynomial ring with coefficients in K. S is algebraically independent over K if and only if v : K[X. algebraically independent over K. Xn]. …. where (Xi)i∈I is a family of indeterminates. The set S is called algebraically dependent over K if S is not algebraically independent over K. f(x1. 4. Thus. Field extensions K) if. {e} is algebrai3 cally independent over Q. 4. In other words. a subset of L. such that L = K(S).

where f = Y − X ∈ Q[X. tn) = 0. T be disjoint subsets of L. tn ∈ T and a polynomial f in n indeterminates with coefficients in K[S] such that f (t1. Prove this. …. Suppose S ∪ T is algebraically independent over K. b) The empty set is algebraically independent over any field. The symmetric elementary polynomials s1. ….4 Proposition. Y). distinct. c) Let K be a field and let n ∈ N*. Then S ∪ T is algebraically independent over K if and only if T is algebraically independent over K(S). Proof. the elements X and Y are algebraically independent over K. with g nonzero polynomial in n indetermi! nates with coefficients in K(S). …. sn are algebraically independent over K in K(X1.…. sn. Write down this equality. to obtain an algebraic dependence relation over K for S ∪ T. e ) = 0. Let K ⊆ L be an extension and let S. Then there exist s1. Y] ≅ K[X . although K[X . let T be algebraically independent over K(S) and suppose S ∪ T is algebraically dependent over K. Y] ( K[X.Y].IV. Group the terms by the monomials in t1. tn ∈ T. then there exist t1. 5 5 K[X. …. It is not known if {e.4 Transcendental extensions 5 5 225 ent over Q because f (e. π} is algebraically independent over Q.…. tn to obtain a relation of the form g(t1.…. Xn] is written uniquely as a polynomial with coefficients in K of s1. with S algebraically independent over K. Y]. sm ∈ S. If T is algebraically dependent over K(S). Xn). …. …. taking into account the form of the elements in K[S]. Can you generalize? So.” 5 d) In K ⊆ K(X. sm. …. 4. …. tn) = 0. …. and a nonzero polynomial f in m + n indeterminates with coefficients in K such that f (s1. . Y] (K-isomorphism). This statement is equivalent to the uniqueness part in the fundamental theorem of symmetric polynomials: “Any symmetric polynomial in K[X1. t1. Conversely. t1. tn) = 0.

c) S is a minimal set of algebraic generators of L over K. …. Xm]. ….6 Examples. a) The extension K ⊆ L is algebraic if and only if the empty set ∅ is a transcendence basis of L over K. 4. Let K ⊆ L be an extension and let S be a subset of L. xm )α k = 0 . I) over K. n}. b) S is called a transcendence basis for the extension K ⊆ L (or a transcendence basis of L over K) if S is algebraically independent over K and S algebraically generates L over K. The following statements are equivalent: a) S is a transcendence basis of L over K. Let K ⊆ L be an extension and let S be a subset of L. xm ∈ S and fk ∈ K[X1. c) {e} is a transcendence basis of Q ⊆ Q(e. for some x1. with fn(x1. Proof. …. xm ∈ S and fk ∈ K[X1. b) S is a maximal algebraically independent set over K. then (Xi)i∈I is a transcendence basis of the field of rational fractions K(X.5 Definition. b) If (Xi)i∈I is a set of indeterminates. Since S algebraically generates L over K. a) S is called a set of algebraic generators of L over K (or say that S algebraically generates L over K) if K(S) ⊆ L is an algebraic extension. …. ….226 IV.… . there exist x1. so S ∪{α} is not algebraically independent over K. xm )α k = 0 . xm) ≠ 0. 2 ). Xm] such that α satisfies an equation of the form: 0≤ k ≤ n ∑ f k ( x1. ∀k ∈ {0. . This is equivalent to the fact that any element α ∈ L satisfies an equation of the form: 0≤ k ≤ n ∑ f k ( x1. ….… . 4. a)⇒b) Let α ∈ L \ S. Field extensions 4.7 Proposition.

Then: a) There exists a transcendence basis B of L over K such that S ⊆ B ⊆ T. so S \{s1} algebraically ! generates L over K.IV. b) The extension K ⊆ L has a transcendence basis. B is nonempty. 4. If it is not so. To this end. transcendence basis) in field extensions. There is a similarity between the concepts of linear independence (respectively generating set. …. contradicting the minimality of S. Proof.8 Theorem. Indeed. with S ⊆ B ⊆ T. i. contradicting the maximality of S. c)⇒a) It is sufficient to see that S is algebraically independent over K. Then the previous result shows that S ∪{α} is algebraically independent over K. α is transcendental over K(S \{α}) (otherwise S would be algebraically dependent over K). S \{α} is not a set of algebraic generators of L over K. since S ∈ B. a) We look for a transcendence basis as a maximal algebraically independent set B over K. …. …. define B := {C | S ⊆ C ⊆ T. sn ∈ S and f ∈ K[X1.…. S is algebraically independent over K and T is a set of algebraic generators of L over K. Order B by inclusion. C is algebraically independent over K}. sn) = 0. then there exist s1. Relabeling if necessary.4 Transcendental extensions 227 b)⇒a) We must prove that K(S) ⊆ L is algebraic.e. with f ≠ 0 and f (s1. a)⇒c) We must prove that. sn). Let K ⊆ L be a field extension and let S. we may suppose that there exists a monomial in f that contains a power of s1. basis) in vector spaces and algebraic independence (respectively set of algebraic generators. This means that s1 is algebraic over K(s2. T be subsets of L such that S ⊆ T. The same method used at vector spaces (Zorn's Lemma) is used to prove the existence of a transcendence basis of an extension. for any α ∈ S. Suppose α ∈ L is not algebraic over K(S). Xn]. The set B is inductively ordered (straightforward proof) and has thus a maximal . L is a transcendental extension of K(S \{α}).

But this contradicts the fact that S is a maximal algebraically independent set over K. So. (use 4. We claim that B is a transcendence basis. because L is algebraic over K(T) and the transitivity of algebraic extension applies. Thus. Thus. if not. T be transcendence bases of K ⊆ L. Suppose any t ∈ T is algebraic over K(S \{s}). any element in the first base can be replaced with some element in the second base. Sst is algebraically independent over K. Proof. K ⊆ K(S) is purely transcendental and K(S) ⊆ L is algebraic. transcendental over K(S \{s}). Let s ∈ S. for two transcendence bases. s is transcendental over K(Sst). with S a transcendence basis of L over K. Then. There is left to show that L is algebraic over K(B). there exists t ∈ T such that (S \{s}) ∪{t} is a transcendence basis.9 Lemma. But this contradicts the fact that S is a minimal set of algebraic generators of L over K (see 4. any . which says that. for any s ∈ S. Then L is algebraic over K(S \{s}). Since L is algebraic over K(S). so Sst ∪{s} = S ∪{t} is algebraically independent over K. For any s ∈ S and t ∈ T. If α ∈ S is transcendental over K(B). B is algebraically independent over K since B ∈ B. Field extensions element B. we use the following lemma. any field extension K ⊆ L can be “decomposed” in a tower of extensions K ⊆ K(S) ⊆ L. Indeed. there exists t ∈ T. then B ∪{α} is algebraically independent over K (by 4. let Sst := (S \{s}) ∪{t}. We remark that s is algebraic over K(Sst). Let K ⊆ L be a field extension and let S. ! b) Apply a) with S = ∅ and T = L. 4.4).4.7).4). As the bases in a vector space. to obtain a transcendence basis. contradicting the maximality of B. Let us show that Sst is also a set of algebraic generators of L over K. any two transcendence bases of an extension have the same cardinal. it is sufficient to prove that any α in S is algebraic over K(B). By 4.228 IV. For the proof.

ti+1.11 Definition. …. si+1. We have: |S| = |S'| = |∪t∈T St| ≤ |T × N| = |T|. t2. Proof. For any t ∈ T. so |T| = |Sn| = n = |S|. tn ∈ T such that Sn := {t1. we deduce Sn = T. …. Let S' = ∪t∈T St.10 Theorem. so |T × N| = |T|. si+2. s2. …. Since Sn ⊆ T and T is also a transcendence basis. L is algebraic over K(S'). ti. since S is a minimal set of algebraic generators of L over K. Thus. ! 4. Since L is algebraic over K(S).IV. ! By symmetry. Suppose first that K ⊆ L has a finite transcendence basis S = {s1. let S and T be two such bases. sn} is a transcendence basis. Let K ⊆ L be a field extension. 4. Let 1 ≤ i < n and suppose we have found t1. So S' = S. since t is algebraic over K(S). the previous lemma yields t1 ∈ T such that S1 := {t1. T and si+1 to obtain ti+1 ∈ T such that Si+1 := {t1. sn}. In the case when any transcendence basis of L over K is infinite. there exists a finite subset St of S such that t is algebraic over K(St). Since T is algebraic over K(S') and L is algebraic over K(T). |S| ≤ |T|. Apply again the previous lemma for Si. …. Thus. sn} is a transcendence basis.4 Transcendental extensions 229 element in S is algebraic over K(Sst). |T| ≤ |S|. This theorem shows that the following definition is correct. tn} is a transcendence basis. The cardinal of a transcendence basis of L over K is called the transcendence degree of the extension L/K and is denoted trdeg (L/K). Any two transcendence bases of a field extension K ⊆ L have the same cardinal. sn} is a transcendence basis. …. ti ∈ T such that Si := {t1. there exist t1. .…. We used the fact that T is infinite. For s1 ∈ S. …. Let T be another transcendence basis of L over K. …. so |S| = |T|. …. t2. the transitivity property implies L is algebraic over K(Sst).

The transcendence degree is “additive” (compare with the property 1. …. Field extensions 4.4. the . …. In order to prove that S ∪ T algebraically generates M over K. Any transcendental extension has an infinity of intermediate fields. …. Xn) over K. sn} is a transcendence basis of K(X1.230 IV. It is clear that S ∩ T = ∅. Xn).3 says that the symmetric fundamental polynomials s1. sn}. Proof. so trdeg (L/K) = n. S ∪ T is algebraically independent over K. This means that {s1. Example 4. Let S. …. 4.26 of the degree of an extension). remark that M is algebraic over L(T). respectively L ⊆ M.12 Example. Then: trdeg (M/K) = trdeg (M/L) + trdeg (L/K). so L(T) is algebraic over K(S)(T). Since T is algebraically independent over L and K(S) ⊆ L. …. sn are algebraically independent over K. T be transcendence bases for K ⊆ L. Xn} is clearly a transcendence basis of K ⊆ L. so it is enough to prove that S ∪ T is a transcendence basis for K ⊆ M. if B is a transcendence basis. …. ! Exercises 1. Prove that an extension K ⊆ L is finitely generated if and only if its transcendence degree is finite and. {X1. But L is algebraic over K(S). On the other hand. so there exists a transcendence basis of K ⊆ L that includes {s1. Let K ⊆ L ⊆ M be a tower of field extensions. T is algebraically independent over K(S). By 4. Let K be a field and let L := K(X1. 2. The transitivity of algebraic extensions implies that M is algebraic over K(S)(T) = K(S ∪ T).13 Proposition.

4. ….) .over K(B). (Hint. there exists a transcendence basis B included in S. Prove that. xn). K(x1. for any x1. 7. Show that any K-endomorphism of Ω is an automorphism. Let K ⊆ L be an extension and let α ∈ L be algebraic over K. Let K ⊆ C be a field extension such that C is algebraically closed. Let Ω be an algebraically closed field and let K be a subfield in Ω such that trdeg (Ω/K) is finite. Then B is a transcendence basis of L ⊆ L(x1. (Hint: If L = K(S). (Hint. Then L is algebraic and finitely generated -by S. xn) = K(x1. then Irr(α. 10. xn) ∩ Ω =: L is a finite extension of K. 8. …. xn).) 3. Then L is not an algebraically closed field.) 5. xn ∈ C. Prove that the following statement is false: “If E and F are subfields of the field L such that E ⊆ L and F ⊆ L are algebraic. Show that there exists a K-endomorphism of Ω that is not an automorphism. Show that there exists a field isomorphism C ≅ Ω. with r ≤ n. then E ∩ F ⊆ L 2 2 is algebraic”. (Hint. xr} be a transcendence basis of K ⊆ K(x1. K(t)). Let Ω = K'C. Prove that for any field K. …. If t ∈ L is transcendental over K.) 6. K(X) is not algebraically closed. with S finite. K) = Irr(α. Let B = {x1. 9. Let K ⊆ L be a purely transcendental extension.IV.4 Transcendental extensions 231 degree [L : K(B)] is finite. Then there exists a K-isomorphism K(α) ≅ K(β) that takes α in β if and only if α and β are either both transcendental. …. K(X + X) ∩ K(X ) = K. Let Ω be an algebraically closed field and let K be a subfield in Ω such that trdeg (Ω/K) is infinite. Let Ω be an algebraic closure of C(X). …. either both algebraic and have the same minimal polynomial over K. Let K ⊆ L be a field extension and let α. β ∈ L. ….

although a lot of partial results were proven. A major part of modern number theory and Algebra owes its existence to the efforts of proving FLT. the English mathematician Andrew Wiles proved the “Shimura-Taniyama conjecture”. “Fermat’s Last n n n Theorem” claims that if n ≥ 3 is an integer. Representation Theory. Galois Theory The idea behind modern Galois Theory is the following: for a given field extension. All attempts at proving this assertion failed until 1995. Various properties of the extension can then be deduced by investigating its Galois group. then the equation x + y = z has no positive integer solutions. Taylor finally gave a proof which is now accepted as correct. The recent proof of Fermat's Last Theorem 1 uses Galois Theory as a basic tool. 1 232 .V. and the list is far from complete. French mathematician. Wiles and R. The idea of studying a certain structure (in our case. Pierre de Fermat (1601-1665). a group) has been very fertile in 20th century mathematics. It appeared soon that some parts of the proof were wrong. In 1995. but A. It can be found in many areas: Algebraic Topology. one associates to it a group (the Galois group of the extension). a field extension) by associating to it another structure (in our case. Algebraic Geometry. a statement which implies FLT (a fact proven by Gerhard Frey and Ken Ribet in 1986). Class Field Theory (recently generalized in the form of the “Langlands Correspondence”).

one can easily verify that L is a subfield of L containing K and that S <S> L = L . Let K ⊆ L be a field extension.1 Automorphisms 233 V. Thus. a K-automorphism of L is a bijective field homomorphism σ : L → L with σ |K = id. ∀σ ∈ H} and denoted by L . which we now define. Checking that G(L/E) is a subgroup in G(L/K) is trivial. denoted Gf /K.V. The Galois group of the extension K ⊆ L is the set of K-automorphisms of L (where the group operation is the composition of maps).1 Definition. defined as H {x ∈ L | σ(x) = x.e. in this −1 case. if K is understood. called the Galois Connections. the set of those σ ∈ G(L/K) for which σ|E = <id. defining L = {x ∈ L | σ(x) = x. To any intermediate field E ∈ IF(L/K) we associate the subgroup G(L/E). to any subgroup H of G(L/K) associate the fixed field of H. σ is also a K-automorphism of L. If f is a polynomial in K[X] and F is the splitting field of f over K. denoted by Gal(L/K) or G(L/K). for any subset S of G(L/K). where < S > is the subgroup generated by S. 2 S . i. or Gf. We denote by IF(L/K) the set of intermediate fields of the extension K ⊆ L and by Subg(G(L/K)) the set of subgroups of the group G(L/K). S ∀σ ∈ S}. 1. The proof of the fact H that L is an intermediate field of K ⊆ L is immediate2. given a field extension K ⊆ L. we define: More generally.1 Automorphisms Recall that. G(F/K) is called the Galois group of the polynomial f over K. The transfer of properties from intermediate fields of the extension K ⊆ L to subgroups of the Galois group and vice versa is accomplished by means of two natural maps. The other way around.

Φ(E) = G(L/E). Since ϕ is injective (being a field homomorphism). so they are of the form ( i1 . where y ≠ 0. First. by the Fundamental Theorem of Galois Theory. For such an element. Ψ(H) := L . K ⊆ E be field extensions and ϕ : L → E a K-homomorphism. and x. let us review some simple but farreaching properties of K-homomorphisms. b) For any f ∈ K[X] and any root α of f in L. a) Viewing L as a K-vector space. −1 c) Any element of L is of the type xy . then σ = ϕ. c) If S is system of generators of the extension K ⊆ L (that is. Galois Theory Φ : IF(L/K) → Subg(G(L/K)). for algebraic extensions. in ) ∈' n ∑ ' i ai1… in x1i1 … xnn .2 Proposition. Thus. n n b) Let f = a0 + a1X + … + anX . Let K ⊆ L.234 V. a) If L = E and K ⊆ L is finite. ∀E ∈ IF(L/K). we have: . ϕ(α) is a root of f. Applying ϕ. since ϕ is a K-homomorphism and the ai's are in K. L = K(S)). then ϕ is an automorphism. A natural question arises: Under what conditions the maps Φ and Ψ are bijective and inverse to each other? The answer is given. ϕ is a K-vector space endomorphism of L. ∀H ∈ Subg(G(L/K)). Ψ : Subg(G(L/K)) → IF(L/K). 1. we have dim Imϕ = dim L. Then a0 + a1α + … + anα = 0. ϕ(α) is a root of f in E. and σ : L → E is a K-homomorphism such that σ|S = ϕ|S. it follows that n a0 + a1ϕ(α) + … + anϕ(α) = ϕ(0) = 0. So Imϕ = L. Proof. H y ∈ K[S]. where ai1… in ∈ K and ∑' indicates the fact that the sum has a finite number of terms. ….

4 Definition. The map ϕ : G(L/K) → Sn. Indeed. Then: a) σ permutes the roots of f in L.…. c) View Sn as the group of permutations of the set R := {x1.1 Automorphisms i i ϕ (∑ ' ai1…in x11 … xnn ) = ∑ ' ai1…inϕ ( x1 )i1 …ϕ ( xn )in = 235 ∑ ' ai …i σ ( x1 ) 1 n i1 i i …σ ( xn ) n = σ (∑ ' ai1…in x11 … xnn ) .3 Corollary. K) = Irr(β. 3 . ….3 Proof. One also says “β is a conjugate of α”. xn) is the splitting field of f over K. Let K ⊆ L be a field extension. σ & σ|R. K). b) If L = K(S) (S generates L over K). We shall exploit this point of view later in order to obtain data on the Galois group of a polynomial. by b)). It is useful to introduce the following concept: 1. xn}. β elements of L. This definition agrees with the terminology “conjugate numbers”. ! 1. i which shows that ϕ = σ. α and β have the same minimal polynomial This is in fact the original view that Evariste Galois had on the notion of group associated to a polynomial. We say that α and β are conjugate over K if they have the same minimal polynomial over K: Irr(α. …. is ! a homomorphism (injective. ∀σ ∈ G(L/K). used to designate the complex numbers α = a + ib and β = a − ib (where a. xn are the distinct roots of f in an algebraic closure of K and L = K(x1. then σ is determined by its action on S. let σ ∈ G(L/K) and let f ∈ K[X]. then G(L/K) = Gf is isomorphic to a subgroup of the symmetric group Sn. In other words. algebraic over K. α and β are roots of the same irreducible polynomial f with coefficients in K. These simple facts are fundamental in determining the Galois group of an extension. Let K ⊆ L be a field extension and let α. c) If x1.V. b ∈ R).

An algebraic element α over K can have at most n conjugates. ∀σ ∈ G(L/K).5 Corollary. there are at most n possibilities to ! choose σ(α) from. a) Consider the extension R ⊆ C and let G = Gal(C/R). The following examples illustrate this technique. but Q ( Q{ 3 2 }. For any σ ∈ G. 1. ∀a. Thus. So. where n is the degree of α over K (n = deg Irr(α. in the sense of the above definition. Since { 3 2 } generates the extension. −i}. Then its Galois group G has at most n elements. The corollary 3. If σ (i) = i. the elements of the Galois group can be found by inspecting their action on a generating set.6 Examples. Moreover. Galois Theory over R (find it!). So. b ∈ Q. Because C = R[i]. Since σ(α) is a conjugate of α. σ(i) ∈ {i. where a. γ and δ are conjugate over Q. The same remark can be made for the “conjugates” γ = a + b d and δ = a − b d . Who is 2 .15. A K-automorphism σ ∈ G is determined by its value in α. one must remember that they transport any element in one of its conjugates. K)). Let K ⊆ K(α) be a simple field extension. then σ = id. if σ (i) = −i. where n = [K(α) : K] is the degree of α. it is sufficient to look for the action of the automorphisms in G on i. Because the mini2 mal polynomial of i is X + 1. but 3 2 is the only root of X − 2 in Q{ 3 2 }. b ∈ R. then σ is the “complex conjugation”: σ (a + bi) = a + bσ (i) = a − bi. G(C/R) consists of two elements: the identity map and the conjugation map.a) says that σ(α) is a conjugate of α. b) Let us find G = G (Q(3 2 ) Q ) . where i = − 1 . the subfield generated by F2 and X . σ (3 2 ) is a root of 3 3 X − 2.236 V. G = {id}. c) Let K = F2(X) (the rational function field with coefficients in F2) and let F := F2 ( X 2 ) . σ = id. ∀σ ∈ G. so σ (3 2 ) = 3 2 . 1. Proof.

these are necessary and sufficient conditions for the bijectivity of the Galois connections. if X = f /g. Note that the given extensions are simple (each is generated by a single element α). It turns out that for finite extensions. where X is algebraic over F as a root of h = Y − X ∈ F[Y]. The fact that X generates K over F implies σ = id. in c). so X is a double root of h. then Xg = f.1 Automorphisms 237 G(K/F)? Obviously. σ (α) must be a conjugate of α. In the cases b) and c) above a bijective correspondence between the subgroups of the Galois group and the intermediate fields cannot exist: the Galois group is trivial (thus it has only one subgroup). “Separability”: For any α ∈ L. the minimal polynomial would be of degree 1. g ∈ F2 [X ]. the left hand side has odd degree and the right hand side has even degree. in the splitting field of 3 X − 2 over Q). α has no conjugates distinct from itself because its minimal polynomial has only one root (which is a double root). But in both cases α has no conjugates in the extension: in b). Thus. which 2 means that X ∈ F. The 2 2 . K) has all its roots in L. K = F(X). So G(K/F) = {id}. so any automorphism σ in the Galois group is perfectly known by its action on α . the conjugates exist. K) has no multiple roots. This is even the minimal polynomial of X over F: otherwise. In this equality of polynomials in X. But X ∉ F: indeed. with f. but F ( K. Irr(α. Let us look closer at the reasons behind this phenomenon. then σ (X) can only be X (the unique root of h in K). Take now σ ∈ G(K/F). it appears natural to consider the following two properties that an algebraic extension K ⊆ L may have: “Normality”: For any α ∈ L. 2 Note that h can be written (in K[Y]) h = (Y − X) .V. moreover. but the extensions are not trivial (they have at least two intermediate fields). but they are in a larger extension (for instance. Irr(α. contradiction.

1 Definition. 2. An easy example of normal extension is the extension K ⊆ Ω.238 V. Remark that the algebraic extension K ⊆ L is normal iff any irreducible polynomial in K[X]. d) There exists a family of polynomials F ⊆ K[X] such that L is the splitting field of F over K. let us give some necessary and sufficient conditions for an algebraic extension to be normal. one says “L is normal over K”. Proof. K) has all its roots in L. checking the above definition is rarely practical (for any α.2 Proposition. K). has all its roots in L. The following statements are equivalent: a) K ⊆ L is normal. A field extension K ⊆ L is called a normal extension if it is algebraic and for any α ∈ L. Galois Theory following two sections are devoted to the study of the conditions above. V. In this case. one must find Irr(α. so α ∈ L. we have ϕ(L) = L. c) For any K-homomorphism ϕ : Ω → Ω. where Ω is an algebraic closure of K. a)⇒b) Let ϕ : Ω → Ω be a K-homomorphism and take α ∈ L. In concrete situations. then ϕ(α) is a root of f. which has a root in L. Let K ⊆ L be an algebraic extension and let Ω be an algebraic closure of L. we have ϕ(L) ⊆ L. Before giving other examples. Irr(α. b) For any K-homomorphism ϕ : Ω → Ω. K) and its roots). .2 Normal extensions 2. If f is Irr(α.

so ϕ(L) ⊆ L. so the splitting field of F over K is included in L. Since f has degree 2 (because K(α) = L) and f is divisible by X − α in L[X]. 3 deed. the roots of any polynomial in F are in L. Because τ(L) ⊆ L and τ(α) = σ(α) = β. ! 2.ω ) Q . then ϕ(S) ⊆ S (by 1.V.ω ) = . But ϕ is injective (it is a field homomorphism) and S is finite. let β ∈ Ω be another root of f. we must adjoin the other roots of X − 2. If ϕ : Ω → Ω is a K-homomorphism. K).2 Normal extensions 239 b)⇒a) Fix an α ∈ L and denote by f its minimal polynomial over K. We obtain the extension 2 Q(3 2 . So. All roots of f lie in Ω. c) Any extension K ⊆ L of degree 2 is normal. a) The extension Q(3 2 ) Q is not normal. 3 2ω . we obtain β ∈ L. d)⇒b) Let S be the set of the roots in Ω of the polynomials in F. There exists a K-isomorphism σ : K(α) → K(β) which takes α to β. S = {α ∈ Ω | ∃f ∈ F such that f(α) = 0}. K) | x ∈ L}. where ω ∈ C is a root of the polynomial X + X + 1. the roots of X − 2 are 3 3 2 . a)⇒c) By b). L = K(S). 3 2ω 2 and so Q(3 2 . Extend σ (see theorem IV. a)⇒d) Let F = {Irr(x. We saw that S ⊆ L and that ϕ(S) ⊆ S. pick α ∈ L \ K and let f = Irr(α. In- 3 Q(3 2 . because 3 Irr( 3 2 . The fact that L is included in the splitting field of F over K is obvious. K) and let S be the set of all roots of f in Ω. 3 2ω 2 ) is the splitting field of X − 2 over Q. Indeed. It remains to show that ϕ is surjective. we already know that ϕ(L) ⊆ L. Take α ∈ L.2.19) to a K-homomorphism τ : Ω → Ω. f decomposes in linear factors in L[X]. By hypothesis. Because K ⊆ L is normal. all roots of f are in L.b)).2.3 Examples. b) If we want a normal extension of Q that contains 3 2 . so ϕ|S : S → S is a bijection. Thus there is a β ∈ S with ϕ(β) = α. let f = Irr(α. 3 2ω . Q) = X − 2 has also complex non real roots.

6 Definition. “K-homomorphism” can be replaced by “K-automorphism”. If K ⊆ L is normal. More precisely. Let K ⊆ L be a finite extension. b) above can be generalized: if K ⊆ L is not normal. in the characterizations above. So. every K-homomorphism ϕ : Ω → Ω is an automorphism. 2. Since Ω is a normal extension of K. xn} ⊆ L be such that L = K(S). then ϕ|L is a K-automorphism of L (an element of G(L/K)). K). Galois Theory In order to prove that a given finite extension K ⊆ L is normal. 2. The situation in Examples a). let S = {x1. the most frequent approach is to prove that L is a splitting field for some polynomial in K[X]. If K ⊆ L is algebraic.240 V. Then there exists a unique normal extension N of K such that L ⊆ N ⊆ Ω. Ω an algebraic closure of K and L an extension of K. we have N ⊆ F. K)⋅ …⋅ Irr(xn. 2. N is the composite of the conjugates of L over K in Ω: N = K(∪{σ(L) | σ ∈ AutK(Ω)}) . for any normal extension K ⊆ F with L ⊆ F ⊆ Ω. Ω is an algebraic closure of L and ϕ : Ω → Ω is a K-homomorphism. which is the smallest with this property. Proof. If we take f = Irr(x1. …. Suppose K ⊆ L is normal. Also.5 Corollary. Ω is an algebraic closure of L and σ is a K-automorphism of Ω. then σ(L) is called a conjugate extension of L over K in Ω.7 Proposition. one gets a normal extension. The converse is already proven. Let K be a field. by adjoining to L the roots of the minimal polynomials of a set of generators of L over K.4 Remark. then L is the split! ting field of f over K. L ⊆ Ω. 2. Then the extension K ⊆ L is normal iff there exists f ∈ K[X] such that L is the splitting field of f over K.

if K ⊆ L is finite. we can take S finite.2 Normal extensions 241 Moreover. L ⊆ M. We have however the following: . We consider I = {Irr(x. the splitting field of I over K (that is. of L over K). Obviously. has the property that σ(E) = E. M ⊆ N and N ⊆ M. so the splitting field of Irr(x. K) over K is included in F. ( 2 ) and Q( 2 ) ⊆ Q(4 2 ) are normal. Since N is the splitting field over K of a family of polynomials. Let S ⊆ L such that L = K(S). Thus. M satisfies the same conditions as N. then {σ(L) | σ ∈ AutK(Ω)} = {τσ(L) | σ ∈ AutK(Ω)} So. τ(M) = K({τσ(L) | σ ∈ AutK(Ω)}) = K({σ(L) | σ ∈ AutK(Ω)}) = M. K) | x ∈ S}.V. that includes L. for any x ∈ S. then K ⊆ N is finite. From the uniqueness of N it follows that M = N. The uniqueness of N results from the minimal condition proved above: if M is another extension with the same properties. N is a finite extension of K. Let M be the composite of the conjugates of L over K in Ω. which means that M ⊆ E. it follows that the normal closure does not depend (up to a K-isomorphism) on the algebraic closure Ω that we choose. being of degree 2. So. Proof. take N to be the splitting field of the family I over K. ∀σ ∈ AutK(Ω). Normal extensions are not transitive: . So. ⊆ . so I is also finite. but Q ⊆ Q(4 2 ) is not normal (why?). so M = N. If τ is a K-automorphism of Ω. On the other hand. then S ⊆ F. If F is a normal extension of K with L ⊆ F ⊆ Ω. If L is finite over K. This shows that M is a normal extension of K. N) is included in F. any normal extension E of K. so σ(L) ⊆ σ(E) = E. ! The extension N constructed above is “the smallest” (in the sense of inclusion) normal extension of K that includes L and is called the normal closure (in Ω) of the extension L/K (or. It is clear that N includes L and N is normal over K.

then F ⊆ FE is normal. Exercises In the exercises. ! 2. So. Proof. b) Let Ω be an algebraic closure of L and ϕ : Ω → Ω be K-automorphism.242 V.2 (using the automorphisms of the algebraic closure or using splitting fields). so ϕ(EF) ⊆ EF and ϕ(E ∩ F) ⊆ E ∩ F. We propose to the reader to give alternate proofs for the results on normal extensions. K is a field and Ω is an algebraic closure of K. b) If K ⊆ E and K ⊆ F are normal. Then: a) If K ⊆ E is normal.9 Proposition. so E = K(S). Then E ⊆ L is normal. Proof.8 Proposition. Let F be a family of polynomials over K such that L is the splitting field of F over K. One can see that the results on normal extensions can be proven using either one of the characterizations given at 2. K ⊆ EF and K ⊆ E ∩ F are ! normal. Let K ⊆ L be a normal extension and E an intermediate field. included in L. Galois Theory 2. then K ⊆ EF and K ⊆ E ∩ F are normal. Then FE = F(K(S)) = F(S). Then ϕ(E) ⊆ E. F algebraic extensions of K. Then L is the splitting field of F (viewed as included in E[X]) over E. . ϕ(F) ⊆ F (as normal extensions). so FE is a splitting field over F for I (viewed as a subset of F[X]). a) E is a splitting field over K for some family I ⊆ K[X]. where S is the set of all roots in Ω of the polynomials in I. Let K ⊆ L be an extension and E.

K ⊆ E ⊆ L an intermediate field and F an algebraically closed field. Q ⊆ Q( 4 2 ). then E ⊆ F. Q ⊆ Q( 3 5 ). Let g ∈ K[X] such that g(p(x)/q(x)) = 0. 8. Give an example to show that Cα = Cβ does not necessarily imply K(α) = K(β). a) If q(x) ≠ 0 then q(x') ≠ 0. Find the normal closures for: Q ⊆ Q( 3 ). 9. K(α) ⊆ K(β) ⇒ Cα ⊆ Cβ and that K(α) = K(β) ⇒ Cα = Cβ. Find the degree and a basis of the extension Q ⊆ L. 2. Prove that: x and y are conjugate over K ⇔ there exists a K-homomorphism ϕ : Ω → Ω such that ϕ(x) = y ⇔ for any normal extension K ⊆ L such that L ⊆ Ω and x. q ∈ K[X]. and F is a normal extension of K.V. For any α ∈ Ω. 5. Show that F := {x ∈ Ω | L contains all roots of Irr(x. 3. Give an example showing that the assumption that K ⊆ L is normal cannot be omitted in the previous statement. K)} is a subfield of L. Let L be the splitting field over Q of the polynomial X − 9. Let x ∈ Ω and p. there exists a K-homomorphism ϕ : L → L such that ϕ(x) = y. for any conjugate x' of x over K. y ∈ L. then there exists ϕ ∈ AutK(L) such that h = ϕ ( g ) (where ϕ : L[X] → L[X] is the unique K-algebra homomorphism that extends ϕ and takes X to X). denote by Cα the splitting field over K of the polynomial Irr(α. for any conjugate x' of x over K. Then any K-homomorphism ϕ : E → F can be extended to a K-homomorphism ψ : L → F. β ∈ Ω. Let L be an extension of K. 4 . Let K ⊆ L be a normal extension and f ∈ K[X] a irreducible monic polynomial.2 Normal extensions 243 1. Prove that ∀α. If K ⊆ E ⊆ L is such that K ⊆ E is normal. h ∈ L[X] are irreducible divisors of f (in L[X]). K). b) Suppose q(x) ≠ 0. Let K ⊆ L be a normal extension. g(p(x')/q(x')) = 0. Let x. 7. Then. Q ⊆ Q( 3. y ∈ Ω. 6. If g. 4. included in Ω. 3 5 ).

11.1 Definition. L. a fixed algebraic closure of K. 12. b) An irreducible polynomial f ∈ K[X] is called a separable polynomial if f has no multiple roots (in Ω). For any n ∈ N*. E be two distinct extensions of degree 3 of K. It is convenient to suppose (without any loss of generality) that all algebraic extensions of the field K that we consider are included in Ω. Let L. 13. 6. Galois Theory 10. 4 . Let K ⊆ L be a field extension.244 V. d) An element of L is called inseparable over K if it is algebraic and not separable over K. van der Waerden. included in Ω. This terminology. introduced by B. Note that f ∈ K[X] is separable if and only if every root of f (in Ω) is separable over K. Show that any extension of finite fields is normal. we fix a field K. Then [LE : K] ∈ {6. We have [LE : K] = 6 ⇔ L and E are conjugate over K. 3. 9}. c) An arbitrary polynomial in K[X] is called separable if all its irreducible factors are separable. expresses the idea that the roots of f are "separated" (distinct). a) An element α of L is called separable 4 over K if it is algebraic over L and Irr(α.3 Separability In this section. V. give an example of a normal extension of degree n. K) has no multiple roots (in Ω). Give examples of normal extensions of Q of degrees: 4. and 8.

g) If K ⊆ L is algebraic and not separable (there exists at least one element of L. inseparable over K). 3. Irr(α.3 Separability 245 e) An irreducible polynomial f ∈ K[X] is called an inseparable polynomial if it has repeated roots f) The extension K ⊆ L is called a separable extension if it is algebraic and every element of L is separable over K (one also says “L is separable over K”). L) divides Irr(α. Note that the irreducible polynomial f ∈ K[X] is separable iff the number of distinct roots of f in Ω equals its degree. The characterization of the polynomials having repeated roots using the formal derivative (IV. If K ⊆ M is separable. with an ≠ 0. 2 The following simple result will be often used. then K ⊆ L and L ⊆ M are separable. Let f = a0 + a1 X + … + an X . then f is separable (f has no repeated roots). K) in L[X]. K ⊆ L is separable. K ⊆ L is called an inseparable extension (or we say that “L is inseparable over K”).4 Proposition. Clearly. 3. Let us show that any α in M is separable over L. We know that Irr(α. We remark that the converse is also true (to be proven in 3. Proof. its roots i and − i are thus separable over Q. a) f has repeated roots (is inseparable) iff f ' = 0 ( f' is the formal derivative of f ). be an irreducible polynomial with coefficients in K. a) The polynomial X + 1 ∈ Q[X] is separable over Q.6) is essential in the proof. Let K ⊆ L ⊆ M be algebraic extensions. K) has no multiple roots.2 Example. ! The next result characterizes the irreducible separable polynomials. 3. n . L) cannot have multiple roots.14).2.3 Proposition.V. On the other hand. b) If char K = 0. so Irr(α.

n−1 b) If char K ≠ 0. Conversely.4. So. Let d = GCD{ f.i. and g ∉ K[X ] (if g ∈ K[X ]. Furthermore. then d = f. 3. Proof. then α is algebraic over K (by transitivity) and . Galois Theory p c) If char K = p > 0. we have i·1 = 0 or ai = 0. ∀i ∈ {1. f ' = a1 + … + nan X ≠ 0 because nan ≠ 0. then i·1 ≠ 0. then f is inseparable iff f ∈ K[X ]. we must have d = 1. then g is irreducible (otherwise f would be reducible!) maximality of e). Proposition 3. so ai = 0. i ∈ {1. If α is an algebraic element over L. a field K is perfect iff any irreducible polynomial in K[X] is separable. Let g ∈ K[X] with f = g X p . c) Let char K = p > 0 and f ' = 0. ( ) e Let e be the greatest integer such that f ∈ K[ X p ].i.6 Proposition. Suppose K ⊆ L is algebraic and K is perfect. g is separable. Proof.b) says that any field of characteristic 0 is perfect. If f ' ≠ 0. so f has no repeated roots. Rephrasing. Since iai = 0. …. A field K is called a perfect field if any algebraic extension of K is separable. Every algebraic extension of a perfect field is a perfect field. if f ' = 0. If p . Since f is irreducible. In p other words. f ∈ K[X ].5 Definition. …. such that f = g X p . The coefficients of f ' are iai = (i·1)·ai. a) Suppose f is inseparable. n}. so deg d ≥ 1 and f has multiple roots. there exists e ∈ N* and an irreducible separable polynomial g ∈ K[X]. So. f '}. n} with p . d is a divisor of f ' and deg d ≤ deg f ' < deg f. the study of separability is meaningful only in characteristic p > 0. then f ∈ K[ X p p p e +1 e ( ) e ]. Thus ai = 0. contradicting the ! The fields that pose no problems of separability are called perfect: 3.246 V.

of characteristic p). then p p β = α = a. n}. If. ⎠ ⎝ i =0 i =0 p (we used the fact that g & g is an endomorphism of the ring K[X]. K) > 1. L) has the same property. The injectivity of the Frobenius endomorphism of Ω implies that β = α. if ϕ is not surjective. bip X pi 3. so α is inseparable over K. L) divides Irr(α. …. where K = {x | x ∈ K}). If char K = 0. there exist bi ∈ K such that ai = bip . If β is another root of f.3 Separability 247 Irr(α. We have α ∉ K (because otherp wise a ∈ K ). f has only one root. so deg Irr(α. then there exists an inseparable algebraic elep ment over K. ! 3. The finite fields. So. with multiplicity p. so ⎞ ⎛ n f= ∑ = ⎜ ∑ bi X i ⎟ . It will be enough to show that.7 Proposition.V. ∀i ∈ {0. n Since ϕ is surjective. Proof. Let a ∈ K \ K . But this shows that f is reducible! Conversely. n p . with ! multiplicity equal to its degree. Irr(α. p then f ∈ K[X ]. K) in L[X]. Therefore. so Irr(α. α is separable over L. the Frobenius is the identity map and all is evident. the fields of characteristic 0 and ! the algebraically closed fields are perfect. this means p p p K = K. ∀x ∈ K) is an automorphism. Consider the polynomial f = p X − a ∈ K[X] and a root α of f in Ω. The field K is perfect iff the Frobenius endomorphism is an automorphism (if char K = p > 0. an irreducible inseparable polynomial f ∈ K[X] would exist. Suppose that the Frobenius endomorphp ism (ϕ : K → K. ϕ(x) = x .8 Corollary. The minimal polynomial of α over K divides f and has the root α. Let char K = p > 0. let K be a perfect field. Let f= i =0 ∑ ai X pi . bi ∈ K. ai ∈ K. K) has only simple roots. on the contrary.

. h is not separable. σ|K = ι}. Ω) is in a bijective correspondence with the set of homomorphisms from L to Ω' that extend ϕ. This leads to the idea of using K-homomorphisms as a tool in the study of separability. η homomorphism}.10 Definition. Consequently. Ω')| = |HomK(L. a) Let K ⊆ L be an algebraic extension. Ω) (the K-homomorphisms defined on L with values in Ω) is denoted by [L : K]s and is called the separable degree of the extension K ⊆ L. Ω)| = [L : K]s does not depend on ϕ and Ω'. [L : K]s is the cardinal of the set {σ : L → Ω | σ homomorphism. another way to characterize the separable degree is given at c) below.9 Example. A K-homomorphism of extensions of K carries a root of a polynomial f ∈ K[X] in another root of f. Let K ⊆ L be an algebraic extension. where ι : K → Ω is the canonic inclusion. P(L/K. Indeed. 5 Recall that Ω is a fixed algebraic closure of K that contains L. Ω' another algebraic closure of L and ϕ : K → Ω' a field homomorphism5. The corollary above shows that.11 Proposition. For simple extensions. 3.248 V. A “natural” example of such a field is Fp(X) =: K. then [M : K]s = [M : L]s [L : K]s. We have that: h is irreducible (by Eisenstein's criterion applied for the prime p element X ∈ Fp[X]). we must look among infinite fields of characteristic p > 0. In other words. in order to find a field that is not perfect. because h ∈ K[Y ]. b) The separable degree is multiplicative: if K ⊆ L ⊆ M are algebraic extensions. consider h = Y − X ∈ K[Y]. 3. The cardinal of the set HomK(L. Then HomK(L. ϕ. the extension K ⊆ K(α) is inseparable. |P(L/K. ϕ. Ω') := {η : L → Ω' | η|K = ϕ. Galois Theory 3. Denoting by α a root of h. p The choice is correct.

v(ϕ) = ϕ(α) is injective. ∀σ ∈ P(L/K. The previous part yields the desired ! inequality. where ι : K → Ω is the canonic inclusion. this shows that P(M/K. a) We must find a bijection between P(L/K. b) For the sake of simplifying the notation. ι. 3. ι. σ) when σ runs over P(L/K. Ω') we associate ψ ◦σ. K) and let R be the set of all conjugates of α (the roots of f in Ω). It is easy to see that these maps are inverse one to each other.V. the map v : HomK(K(α). ι). b) α is separable over K. d) If α is algebraic over K. . ι)| = |P(L/K. For any η ∈ P(M/K. These sets are mutually disjoint and |P(M/L. denote P(L/K. then [K(α) : K]s is equal to the number of conjugates of α in Ω. To every −1 σ ∈ P(L/K. ι). c) Let f = Irr(α.12 Proposition. In other words. Ω'). ϕ. d) Irr(α. ι)|⋅[M : L]s = [L : K]s⋅[M : L]s. ϕ. ι).2. Ω) we associate ψ ◦η ∈ P(L/K. K) = [K(α) : K]. then the following conditions are equivalent: a) [K(α) : K]s = [K(α) : K]. Ω') and P(L/K. Thus. ϕ. Ω). but deg Irr(α. Proof. then [K(α) : K]s ≤ [K(α) : K]. ι)|. then ϕ(α) is a root of f. ι). there exists a K-isomorphism between K(α) and K(β). as shown before. The other way round. on the other hand.3 Separability 249 c) If L = K(α) is a simple algebraic extension of K.19). to every η ∈ P(L/K. K) has at most deg Irr(α. σ)| = [M : L]s. ϕ. ι. ι) is the union of the sets P(M/L. Ω) → R. then ψ (σ(x)) = ψ (ϕ(x)) = x = ι(x). K) roots. Therefore. If L = K(α) is a simple algebraic extension of K. Ω) by P(L/K. This map is also surjective: for any β ∈ R. [M : K]s = |P(M/K. ϕ). There exists an isomorphism ψ : Ω → Ω' which extends ϕ (by IV. Ω): −1 −1 if x ∈ K. η|L ∈ P(L/K. we can write: [M : K]s = |P(M/K. which belongs to P(L/K. any K-homomorphism defined on K(α) is uniquely determined by its action on α. If ϕ : K(α) → Ω is a K-homomorphism.

holds in the general case of finite extensions: 3. xn) : K(x1)]s[K(x1) : K]s ≤ [K(x1)(x2. and c) ⇒ b) is obvious. if we look at d) above. so it is also over K(β). comparing the last two ! relations. But [K(α) : K] = [K(α) : K(β)][K(β) : K]. Proof. …. So: [K(x1)(x2. Since the (separable) degree is multiplicative. Suppose that the inequality holds for any finite extension that has a set of generators with at most n − 1 elements and let us prove it for the extension K ⊆ K(x1. which amounts to [K(β) : K]s = [K(β) : K]. xn). which means that [K(α) : K(β)] = [K(α) : K(β)]s and [K(α) : K]s = [K(α) : K]. xn) : K(x1)]s ≤ [K(x1)(x2. …. a) ⇔ b) is clear. Proof. …. We have: [K(x1) : K]s ≤ [K(x1) : K] and [K(x1)(x2.250 V. the inequality it is already proven. xn) : K(x1)]. Then: a) [L : K]s ≤ [L : K]. a) If K ⊆ L is simple. Rewriting (*). (*) α is separable over K. we get [K(α) : K] = [K(α) : K(β)][K(β) : K]s. Let K ⊆ L be a finite extension. we get the conclusion. The multiplicativity of the separable degree shows that [K(α) : K]s = [K(α) : K(β)]s[K(β) : K]s.13 Proposition. …. . b) K ⊆ L is separable iff [L : K]s = [L : K]. Galois Theory c) K(α) is a separable extension of K. …. given above. The characterization of separable simple extensions using the separable degree. it follows that [K(β) : K]s = [K(β) : K]. a)⇒c) Let β ∈ K(α). We must show that β is separable over K. xn) : K(x1)][K(x1) : K].

By induction. We prove by induction on n the next Proposition: “For any n ∈ N*. Since ble over [L : K(α)]s ≤ [L : K(α)]. If K ⊆ L and L ⊆ M are separable. If n > 1. with degrees strictly smaller than [L : K] = n. Multiplying these equalities gets us to the conclusion. …. Irr(α. an). Then K ⊆ K(A) is a separable extension. and any separable extension K ⊆ L of degree n. …. The case n = 1 is trivial. so α is separable over L'. we get [L : K(α)] = [L : K(α)]s and [K(α) : K] = [K(α) : K]s. . so [L' : K]s = [L' : K]. Proof. Suppose now [L : K]s = [L : K] and yet there exists α ∈ L. So. contradiction. 3. We shall use the characterization of separability with the separable degree in the finite case. If Κ(α) ≠ L. If L = K(α). inseparaK. 3. α is separable and the previous proposition applies. ! Notice the resemblance with the proof of the transitivity of algebraic extensions. L'). multiplying these relations yields ! [L : K]s < [L : K].14 Theorem (transitivity of separable extensions) Let K ⊆ L ⊆ M be algebraic extensions. which means that K ⊆ L'(α) is separable. Evidently. [L : K]s = [L : K]”. Let a0. an ∈ L be the coefficients of Irr(α. then the extensions K ⊆ K(α) and K(α) ⊆ L are separable. Then [K(α) : K]s < [K(α) : K]. then K ⊆ M is separable. Let K ⊆ L be any extension and let A be a set of elements in L. let α ∈ L \ K.15 Corollary. Thus [L'(α) : L']s = [L'(α) : L']. because K ⊆ L is separable. L) = Irr(α.V. L) and let L' = K(a0.3 Separability 251 b) Suppose K ⊆ L is separable. Take α ∈ M and let us show that α is separable over K. separable over K. The multiplication of the last two equalities yields [L'(α) : K]s = [L'(α) : K]. The finite extension K ⊆ L' is separable. α is separable over K.

β ∈ L. …. β) is evident. …. β) = K(γ). xn}. By transitivity. n}. f = Irr(α. vides h (and g.(i) Indeed.12).252 V. Galois Theory Proof. Proof. Every element in K(A) is a polynomial expression with coefficients in K in a finite set of elements of A. ∀j ∈ {1. 3. β1. n = deg g. 2 ≤ j ≤ n}. reduces ! to the already proven case. αm the roots of f in Ω. β = β1. With c chosen this way. Let h(X) = f (γ − cX). L an extension of K and α. the condition on c amounts to c ∉ {(αi − α)(β − βj) | 1 ≤ i ≤ m. K). Since β is separable. βn the roots of g in Ω. too) in Ω[X]. Suppose A = {x1. The following lemma is the essential step in proving this. x2) is separable. x2}. m}. algebraic over K. let γ = α + cβ. is that any finite separable extension is simple. the fact that d = X − β ∈ K(γ)[X] implies then β ∈ K(γ). α = α1. We claim that there exists c ∈ K* such that: ∀i ∈ {1. a polynomial with coefficients in K(γ). n > 2. i. so K(x1) ⊆ K(x1)(x2) is separable. The idea is to show that the GCD of h and g is X − β .16 Lemma.e. which means that X − β di- −1 . m = deg f. K ⊆ K(x1. since K is infinite. we can suppose that A is finite. K(α. Then K ⊆ K(α. Thus. The case A = {x1. …. Let K(α. g = Irr(β. Since x1 is separable over K. and the set above is finite. K). it is enough to show that β ∈ K(γ) (this will imply also α = γ − cβ ∈ K(γ)). …. Let K be an infinite field. Because K(γ) ⊆ K(α. α + cβ = αi + cβj ⇔ i = 1 and j = 1. …. there exists c ∈ K satisfying (i). β) = E. β) is a simple extension (it has a primitive element). where β is separable over K. The polynomials h and g have no other We have h(β) = f (γ − cβ) = f (α) = 0. We show that γ is a primitive element. by induction. A remarkable fact. very important in Galois Theory. x2. K ⊆ K(x1) is separable (by 3. As x2 is separable over K. …. βn are distinct. it is separable over K(x1).

…. Recall that every algebraic extension of a field of characteristic 0 is separable. Clearly. then L is also finite. so βj = β1 = β. thus. GCD(h. which means it is among α1. then γ − cβj is a root sures that j = 1. K L is “the largest” separable extension of K in- . The remaining results (until the end of the section) are a somewhat deeper study of (in)separability. the preceding lemma applies. …. xn) = L. ! 3. lemma 3.V. in Ω[X]. g) = X − β ∈ K(γ)[X]. Proof. Suppose now that K is infinite.19 Definition. But h. For the extensions of the type K ⊆ K(α. ! 3.16 gives also a practical procedure to find a primitive element (or. The reader interested primarily in the Fundamental Theorem of Galois theory may skip directly to the next section. g) = X − β. g ∈ K(γ)[X]. Thus. αm. xn ∈ L such that K(x1. if the condition (i) is hard to verify). 3. that is.3 Separability 253 common roots in Ω but β: if h(βj) = f (γ − cβj) = 0. But then condition (i) en- β ∈ K(γ). If K is a finite field. and any generator of L* can be taken as a primitive element. The multiplicative group L* is cyclic. all that follows is relevant only in characteristic p > 0. (Primitive element theorem) Any separable finite extension K ⊆ L is simple (it has a primitive element). If K ⊆ L is a field extension.17 Theorem.15. at least. of f. The general case follows by an easy induction argument on n ∈ N* with the property that exist x1.18 Remark. x2). β). x2 ∈ L such that L = K(x1. the separable closure of K in L is a subfield of L and a separable extens sion of K. the separable closure s of K in L is the set K L := {x ∈ L | x separable over K}. with K infinite. …. so GCD(h. By 3. a class of good candidates for it. If there exist x1.

then Irr(α. An algebraic element α of the extension K ⊆ L is called purely inseparable over K if Irr(α. K) = ( X − α ) p . where e e is minimal with α p ∈ K. e) The separable degree of the extension K ⊆ K(α) is 1. K) has only one root in Ω (α itself. The following result gives a clearer picture of purely inseparable elements over a field K with char K = p: roughly put. it follows that: an element is simultaneously purely inseparable and separable over K iff it belongs to K. the opposite of separability: 3.21 Proposition. The following statements are equivalent: a) α is purely inseparable over K.254 V. . The separable closure of K in Ω (an algebraic closure of K) is called s s the separable closure of K. Notice that K is the splitting s field over K of the family of all separable polynomials in K[X] (thus K is unique up to a K-isomorphism). e c) There exists e ∈ N such that α p ∈ K. In other words. K) is of the form (X − α) . The following concept is. b) The only conjugate of α in Ω is α. n Irr(α. An extension K ⊆ L is called purely inseparable if any element in L is purely inseparable over K. any purely inseparable element is inseparable. 3. Directly from the definitions. K)). e If α is purely inseparable over K. d) The minimal polynomial of α over K is of the form e e e ( X − α )p = X p − α p . Evidently. (K L includes any separable extension K ⊆ E with E ⊆ L). which has order of multiplicity deg Irr(α. Let char K = p > 0 and let α ∈ Ω.20 Definition. denoted K . Galois Theory s cluded in L. in a certain sense. they are “roots e of order p of some element in K”.

We have f = ( X − α )p so the coefficient of X p f ∈ K[X ]. a)⇒c) and a)⇒d). Then deg g ≥ deg f. Let K ⊆ L be a field extension.m).c): [K(α) : K]s is the number of conjugates of α in Ω. 3. The polynomial f is irreducible in K[X]. If α p ∈ K. K). m −1 is (m ⋅ 1)α p . Let n ∈ N* be such that f = (X − α) . K). m) = 1. c)⇒d). we get α p ∈ K. Let K ⊆ L be an algebraic extension. But α is not purely inseparable 6 6 over K. because g' = 0. Since α is inseparable over K. 6 but not purely inseparable.11. g is divisible by f = Irr(α. a) K ⊆ L is purely inseparable iff [L : K]s = 1. . The set := {x ∈ L | x is purely inseparable over K} is called the purely inseparable closure of K in L. Since m·1 ≠ 0 in K e d (because p .V. that is.22 Example. thus f = ( X − α ) p . a)⇔b) b) just rephrase the definition of pure inseparability. e Proof. Take K = F2(X) and g = Y − X ∈ K[Y]. b)⇔e) It follows from 3. K ⊆ L is normal and Gal(L/K) = {id}. since g = Y − α and 6 is not a power of 2 = char K. Let f := Irr(α. i KL g := X p − α p = ( X − α )p ∈ K[X] and g(α) = 0. In this case. p n ( e ( ) e e d ) = (X m pe −α p e e ) m .24 Proposition. Then there exist e. ! 3. then So. There exist algebraic elements that are inseparable. d ≥ e. d)⇒ a) Evident. d d 3. for some d ∈ N. so m = 1.3 Separability 255 m ∈ N with n = p m and GCD(p.23 Definition. irreducible (by Eisenstein: X is prime in the UFD F2[X]). char K = p > 0. A root α of g is inseparable over K.

let e e K ⊆ L and L ⊆ M be purely inseparable and let α ∈ M. the degree of Ki ⊆ Ki + 1 is a power of p. Since β is purely inseparable over K. K). We must show −1 that α − β. If i E is a purely inseparable extension of K. then [L : K] = deg Irr(α. Thus. β be purely inseparable over K. d) Let α. K). that is. …. being the product of these degrees. Galois Theory b) Let K ⊆ L ⊆ M be a tower of extensions. K) has all its roots (namely α itself) in L. i d) The purely inseparable closure KL of K in L is a subfield of L. αn in L. so [L : K]s = 1 and G(L/K) = {id}. c) If L = K(α) is a simple extension. so K ⊆ L is normal. Then: e d . K ⊆ L is purely inseparable. α is purely inseparable over L. a) Let K ⊆ L be purely inseparable. Let Ki = K(α1. αi). c) If K ⊆ L is purely inseparable and finite. L = K(α1. Let e. d ∈ N with β p ∈ K for some d. Conversely. ∀i ≤ n. E ⊆ L. then σ(α) is a root of Irr(α. For any α ∈ L. If σ : L → Ω is a K-homomorphism. then [K(α) : K]s ≤ [L : K]s = 1. then E ⊆ KL. ∀α ∈ L. KL ∩ K L = K. so α is purely inseparable over K. Irr(α. with β ≠ 0. Ω). αβ. we have a tower of simple extensions (which are d e+d α p ∈ K.…. so [L : K] is still a power of p. then [L : K] is a power of p. …. which is a power of p. Then: K ⊆ M is purely inseparable iff K ⊆ L and L ⊆ M are purely inseparable. β are purely inseparable. For every i. Morei s over. Proof. if [L : K]s = 1 and α ∈ L. Any α ∈ M has the property that α p belongs to K (and to L) for some e. αn) for some finite set of purely inseparable elements α1.256 V. b) Let K ⊆ M be purely inseparable. purely inseparable. clearly. the canonic inclusion is the only element in HomK(L. so σ(α) = α. β p ∈ K. So α p ∈ K. In the general case. Then there exists e ∈ N with β = α p ∈ L. Conversely. Then. by b)) K = K0 ⊆ K1 ⊆ … ⊆ Kn = L.

s Then L is separable over C. i s i s Suppose now that L = KL K L = KL(K L). This means that L is generi ated over KL by separable elements over K (which are also separable i i ! over KL).V. Let K ⊆ L be an algebraic extension. so K L ∩ KL = K. a) Let α ∈ L and let g ∈ K[X] be the irreducible separable polynomial such that Irr(α. Finally.b). that is. Since K L ⊆ L is purely s s s inseparable. Let a = α p . It follows that L is separable over KL. [L : K L]s = 1.4. −1 i A similar argument works for αβ and β . Since KL contains all i purely inseparable elements over K in L. K). thus. So e e K ⊆ K L is separable). a is separable over K. This shows that L = C.3 Separability 257 =α − β ∈ K.25 Proposition. . Suppose L is separable over KL. Then: s a) K L ⊆ L is purely inseparable. α is purely inseparable over K L. so it is purely inseparable over C. i s i c) KL ⊆ L is separable iff L = K L KL.11. s s e s s s c) Let C be the composite K L KL. purely inseparable L separable ⇔ L = K LKL s i KL separable s KL purely inseparable i K Proof. so [L : K]s = [K L : K]s = [K L : K] (because s s i i α p = a ∈ K L. K)(X) = g{ X p } (see 3. KL includes any purely inseparable extension K ⊆ E with E ⊆ L. It follows that g(a) = 0. then α ∈ K. b) By 3. so g = Irr(a. [L : K]s = [L : K L]s·[K L : K]s. if α is separable and s i ! purely inseparable over K. (α − β ) pe+d ( ) ( ) pe pd pd pe 3. s b) [L : K]s = [K L : K].c)). L is also purely inseparable over K L.

then h = g.27 Remark. then KL ⊆ L is separas i ble (so L = K LKL). Consequently. so α = β (by the injectivity of the field homomorphism q q x & x ). Let g ∈ K[X] be a monic irreducible factor of f.14 shows that if K ⊆ L is normal. for every algebraic extension s K ⊆ L. Conversely. as K L ⊆ L is purely inseparable. If K ⊆ L is finite and char K = p > 0.28 Proposition. Galois Theory The proposition above shows that. For an algebraic extension K ⊆ L. m Proof. the separable closure K L is important in the sense that we can decompose K ⊆ L in a tower of extensions: the separable extension s s K ⊆ K L followed by the purely inseparable extension K L ⊆ L.258 V. If h ∈ K[X] is irreducible monic and h| f. then the polynomial f = X p − a is irreducible in K[X]. but K L ⊆ L is not always separable (see exercise 13). The results proven until now show that the following properties hold: a) K ⊆ L is finite ⇒ [L : K] = [L : K]s·[L : K]i. ∀m ≥ 1. then [L : K]i is s a power of p. g and h have α as a common root m p m . i c) K ⊆ L is purely inseparable ⇔ L = KL ⇔ [L : K]s = 1. then a ∉ K . Let q := p and take α. since i i K ⊆ K L is purely inseparable. β to be roots of f in Ω. But [L : K]i is not necessarily equal to the greatest power of p that divides [L : K] and i neither equal to [KL : K]. 3. If a ∉ K . i Exercise 1. s b) K ⊆ L is separable ⇔ L = K L ⇔ [L : K]i = 1. Let K have characteristic p > 0 and let a ∈ K. if X p − a is irreducible in K[X] for some p m ≥ 1. define the s inseparable degree of L over K to be [L : K]i := [L : K L]. α is the only root of f in Ω : f = (X − α) . Then q q α = a = β . Indeed.26 Definition. The i situation is not symmetric for the purely inseparable closure KL. 3. 3.

which is irreducible. Since K is a subfield. they t are equal. …. In addition. being irreducible and monic. If L = K(a1.21). Conversely.3 Separability 259 and hence are not mutually prime. b) Suppose S is finite and L = K(a. We have b = f(0) = −a. where t ∈ N*. an). Exercises In the exercises. where C is the set of coefficients of Irr(a. E) is a monic divisor (in Ω[X]) of f and that E = K(C). Let g = X + 3X + 1 ∈ K[X] and α a root of g in Ω. if a = b p for some b ∈ K. K). contradiction. c) Is the statement still true if L/K is infinite? 3 . 3. E) is injective and that S is finite. a ∈ K . Let L/K be a finite extension. ! This proposition gives a method to exhibit purely inseparable elements.V. Let f = Irr(a. any purely inseparable element can be constructed by this method (cf. If t > 1. so f = g. t is a power of p. c ∈ K. 1. for any E ∈ S. K denotes a field and Ω an algebraic closure of K. b). E). p p p −a ∈ K . contradicting that f is irreducible. If K is infinite. It follows that t = 1. then f = X p ( m −1 −b ) p . use an induction on n to prove that L/K is simple. So f = g . because m t t·grad g = p . a) Suppose L = K(a). show that L is simple considering the set of intermediate fields K(a + bc). Show that. with a ∈ L. We want to prove that L/K is simple iff it has a finite number of intermediate fields. Irr(a. Deduce that E & Irr(a. Denote by S the set of all intermediate fields of L/K. Under what conditions is K ⊆ K(α) separable? 2. Let b := g(0) ∈ K.

Let L = S(v). Show that α ∈ K. Suppose F is a field of characteristic 2 and put K = F(X. of multiplicity > (deg f )/2. Let char K = p > 0 and f ∈ K[X]. 9. Then y = g(x') ⇔ x' is conjugate to x over K(y). if p -deg f. Let α be a separable element over K and g = Irr(α. (Hint: Using formal derivatives. if α is a multiple root (having multiplicity n) of some f ∈ K[X]. then x ∈ K. (Hint. Suppose char K = p > 0 and K ⊆ L is an algebraic extension. i p Prove that KL = K is equivalent to: ∀x ∈ L. Galois Theory 3. of multiplicity ≥ n. then f is reducible in K[X]. S = K(u) and 2 v = uY (v is a root of T − uY ∈ S[T]). Let u 2 be a root of the polynomial T + T + X ∈ K[T]. K). [K(x) : K] ≤ n. Let K ⊆ L be a normal extension. 11. then t ∈ K. 7. Suppose f has a repeated root (in Ω). 2 b) If t ∈ L and t ∈ K. Deduce that K L = S. Y ). Let K ⊆ L be algebraic and let x. Suppose K ⊆ L is an algebraic extension and S is an intermediate s field with K ⊆ S separable and S ⊆ L purely inseparable. 8. Then every irreducible polynomial in K[X] decomposes in L[X] in a product of irreducible polynomials that have the same degree. Let n ∈ N and let K ⊆ L be an algebraic separable extension such that ∀x ∈ L. 12. then n g | f. Find a base in L/K and use the form of the elements in L). Then K ⊆ K L is normal. s . Then [L : K] ≤ n. Prove that. 10. x' ∈ L be conjugate over K. 4.) 6. Then S = K L. if x ∈ K. Then K ⊆ L has at n−1 most 2 intermediate fields. 5. Let K ⊆ L be a simple extension of degree n. show that any conjugate of α is a root of f. Prove that: a) K ⊆ S is separable of degree 2 and S ⊆ L is purely inseparable of s degree 2. Let K ⊆ L be a normal extension. Prove that. 13. Let y ∈ K(x) and g ∈ K[X] such that y = g(x).260 V. Suppose K is perfect and f ∈ K[X] has a repeated root α (in Ω).

d) F ⊆ L is not simple. xn) = L. Can an extension of degree p have an infinity of intermediate fields? 17. Y} is a p-basis of F ⊆ L. Y). …. n Show that [L : K] = p (a set {x1. f) {X. with L ⊆ E normal and K ⊆ L purely inseparable. Suppose K ⊆ L is a finite extension of characteristic p > 0. F(X + βY) = F(X + γY) iff β = γ.V. xn} with these properties is called a p-basis of K ⊆ L). p − 1} is a i { } linearly independent set over K ⇔ ∀i ∈ {1. d) The extension K ⊆ L has a p-basis. i i e) {x1. …. α ∈ F. Prove that ∀β. Take p = 2 and K = F2 in the previous problem. Let F = K(X . e) Deduce that F ⊆ L has an infinity of intermediate fields. such p that L ⊆ K. the field of rational p p functions in two indeterminates X and Y over K.3 Separability 261 c) KL = K. a) Prove that K ⊆ L is purely inseparable. Then K ⊆ E is normal. xi ∉ K({x1. in ∈ {0. b) Suppose {x1. …. Let char K = p > 0 and consider L = K(X. x2) ( … ( K(x1. xn} is a p-basis ⇔ x11 … xnn i1 .… . Let K ⊆ L ⊆ E be a tower of algebraic extensions.….… . 2 b) [L : F] = p . n}. . …. c) Any two p-bases of K ⊆ L have the same cardinal (called the p-dimension of the extension K ⊆ L). xn} \ {xi}). 16. γ ∈ F. Show that: a) F ⊆ L is purely inseparable. 15. xn} ⊆ L is such that K ( K(x1) ( K(x1. Y ). p c) For any α ∈ L. 14. ….

b) Let F ( E ( L be an intermediate field. German mathematician. Notice the linear algebra methods. one of the creators of algebraic number theory. An essential step in the proof is the Dedekind lemma6. (u.262 V.4 The Fundamental Theorem of Galois Theory We gathered sufficient data on normal extensions and on separable extensions to deal with the problem of bijectivity of the Galois connections we stated at the beginning of the chapter. ·). Show that there exists a unique polynomial PE ∈ F2[X. n} are distinct semigroup homomorphisms. Galois Theory a) Show that {1. but such that there is no g ∈ K[X]. 2 2 with u. V. such that E = F(PE).5. we generalize Corollary 1. of the form PE = Xu + Yv + XYw. 6 . v. such that Julius Wihelm Richard Dedekind (1831-1916). ∀i ∈ {1. Show that E & PE establishes a bijection between the set {E | F ( E ( L is an intermediate field} and the set of the polynomials PE of the form above. v. concerning the order of the Galois group of a finite extension. irreducible. Y ]. Give an example of a normal extension K ⊆ L such that L is a splitting field for a polynomial f ∈ K[X].1 Proposition. X. n ∈ N* and σi : G → (K*. First.…. If α1. Y ]. 18. w) = 1. 4. αn ∈ K. ·) is a semigroup. K is a field. w ∈ F2[X . …. a) (Dedekind's Lemma) Suppose (G. XY} is a basis of the extension F ⊆ L and write the general form of an element of L. Y. with L the splitting field of g over K.

…. so we obtained a linear dependence relation with less than m terms. we have: α1σ1(xy) + … + αmσm(xy) = α1σ1(x)σ1(y) + … + αmσm(x)σm(y) = 0 (2) we get α2(σ2(y) − σ1(y))σ2(x) + … + αm(σm(y) − σ1(y))σm(x) = 0. b) If K ⊆ L is a finite extension. ….4 The Fundamental Theorem of Galois Theory 263 α1σ1(x) + … + αnσn(x) = 0 for any x ∈ G. σm} and suppose. ∀x ∈ K. Consider the matrix: ⎡ σ 1 ( x1 ) σ 1 ( x2 ) + σ 1 ( xn ) ⎤ ⎢σ ( x ) σ ( x ) + σ ( x ) ⎥ 2 2 2 n ⎥ ∈ Mm. Since σ1 ≠ σ2. (1) We may even suppose that m is the smallest having this property. …. αm ∈ K. and the number of conjugates of x is finite. ∀x ∈ L. …. by contradiction. Then any σ ∈ G(L/K) is determined by its values in x1. we notice that G(L/K) is finite. In this equality. σn are linearly independent. Proof. …. all nonzero. there exists y ∈ G such that σ1(y) ≠ σ2(y). σn has at least m terms. Subtracting (2) from (3) . xn. σ1. …. Suppose the statement is false. So let G(L/K) = {σ1. xn} a K-basis of L. ∀x ∈ K. n(L). Replacing x with xy in (1). then |G(L/K)| ≤ [L : K]. let [L : K] = n and take {x1. α2(σ2(y) − σ1(y)) is nonzero. Indeed. Relabelling if necessary. then α1 = … = αn = 0. (3) (3) is obtained by multiplying (1) with σ1(y). there exists m ≤ n and α1. contradiction with the minimality of m.V. that m > n. b) First. But σ(x) is a conjugate of x. a) The statement can be rephrased as follows: in the G K-vector space K of functions defined on G with values in K. in the sense that any linear dependence relation between σ1. such that α1σ1(x) + … + αmσm(x) = 0. A =⎢ 2 1 ! * ! ⎥ ⎢ ! ⎢σ ( x ) σ ( x ) + σ ( x )⎥ ⎣ m 1 m 2 m n ⎦ α1σ1(x)σ1(y) + … + αmσm(x)σ1(y) = 0.

αm ∈ L. contradiction. we exhibited a K-isomorphism σβ : L → L. Let α1. 4. so it has a primitive element α. …. Proof. which is evidently an element of G(L/K). H H [L : L ] = |H| and G(L/L ) = H. We get that α1 = … = αm = 0.3 Proposition. The extension is finite and separable. But K(α) and K(β) are K-isomorphic by an isomorphism σβ that transports α to β. Summarizing. …. Artin7) Suppose L is a field. The extension is normal. normal. It is enough to show that |G(L/K)| ≥ [L : K] =: n. for every conjugate β of α. (4) holds for any x ∈ L. K) has n roots in L. Dedekind's lemma ! applies. Then |G(L/K)| = [L : K]. xn. n}. so its rows are linearly dependent. (E. H is a finite subH group of the group Aut(L) of the automorphisms of L and L is the H fixed field of H. ∀i ∈ {1. …. Galois Theory The rank of A is at most n.2 Corollary. [K(β) : K] = n. so Irr(α. For any root β of Irr(α. (4) Since any x ∈ L is a K-linear combination of x1. K). separable. thus K(β) = L. Then the extension L ⊆ L is finite. Let K ⊆ L be a finite. Emil Artin (1898-1962).264 V. 4. normal and separable extension. So deg Irr(α.! The next proposition is an important step in proving the fundamental theorem and shows that the converse of the preceding result also holds. and σi are K-homomorphisms. not all zero. K) = n. Austrian mathematician (he lived most in Germany and the USA). 7 . Viewing σi as homomorphisms from L* to L*. such that α1σ1(xi) + … + αmσm(xi) = 0.

Let G K0 = L . xn) ( H L (x1. hence every element in L is algebraic over L .! . ∀x ∈ L. Let x ∈ L. H But fx(x) = 0. it is simple by the primitive element theoH H rem. the isomorphism σ': L[X] → L[X] H that extends σ leaves fx invariant. and these are distinct. The set Cx = {σ(x) | σ ∈ H} is finite and fx := ∏ y∈C ( X − y ) . ….4 Corollary. ∀n ∈ N*. H In conclusion. xn) is of degree at most |H|. Then. H H H ! |G(L/L )| = [L : L ] = |H| and H = G(L/L ). Proof. x For any automorphism σ ∈ H. Note H that fx has all roots in L. 4. we have K = K0. As H H H H ⊆ G(L/L ) (and |G(L/L )| = [L : L ] from the preceding Corollary). then g(σ(x)) = 0.…. the simple extension H H H L ⊆ L (x) is finite and its degree is at most |H|. Moreover. xn) : L ] > |H|. L ) = fx. Indeed. If L ⊆ E is a finite extension (where E ⊆ L). separable and [L : K0] = |G| = [L : K]. H H L (x) ≠ L. Pick n ∈ N* such that H H [L (x1. L ⊆ L is finite and its degree is at most |H|.b).1. Because K ⊆ K0 ⊆ L. so [E : L ] ≤ |H|. since the H H finite extension L ⊆ L (x1. We have reached a contradiction. If σ ∈ H. Then K ⊆ L is normal and separable and |G(L/K)| = [L : K]. Thus. H H Irr(x.V. Summarizing. so the coefficients of fx are in L . normal and separable.…. Consider the polynomial Proof. The previous arguments show that. if g ∈ L [X] is such that g(x) = 0. ∀σ ∈ H. so any root of fx is a root of g. The proposition above ensures that K0 ⊆ L is normal. Suppose that L ⊆ L is infinite. xn+1). …. We have |G| = [L : K] from 4. xn) : L ] is strictly increasing.….4 The Fundamental Theorem of Galois Theory 265 |Cx| ≤ |H|. then σ(Cx) = Cx. L ⊆ L is algebraic. for any x ∈ L. fx | g. Let K ⊆ L be a finite extension with |G(L/K)| ≥ [L : K]. This implies that there exists a H sequence (xn)n≥1 of elements of L such that L (x1. Let us show that it is finite. Let G = G(L/K). This in turn means that the sequence H H [L (x1. as we saw. take then some y ∈ L \ L (x).

∀E ∈ IF(L/K). E1 ⊆ E2 ⇒ G(L/E1) ⊇ G(L/E2).266 V. G(L/L ) ⊇ H.6 Theorem. normal and separable extension of fields. G(L E) Proof. H . Its proof is a mere application of the definitions: 4. Then the Galois connections are inclusion-reversing maps that are bijective and inverse to each other. H b) For any H ∈ Subg(G(L/K)). ∀H ∈ Subg(G). L / ⊇ E. ∀σ ∈ H}:= L . G(L/L ) = H. Let K ⊆ L be an extension. c) Φ is inclusion reversing: ∀E1. d) Ψ is inclusion reversing: ∀H1. Galois Theory Recall that the Galois connections are defined as follows: for an extension K ⊆ L. Φ(E) = G(L/E). the intermediate fields E with K ⊆ E normal correspond to normal subgroups of G(L/K). Ψ(H) = {x ∈ L|σ(x) = x. L / = E and [L : E] = |G(L/E)|. c). whose Galois group is G. b) For any subgroup H ≤ G(L/K). H1 ⊆ H2 ⇒ H1 L ⊇ LH 2 . ∀σ ∈ G(L/E)} ⊇ E. Then we define: Φ : IF(L/K) → Subg(G). Then: G(L E) a) For any E ∈ IF(L/K). (The fundamental theorem of Galois Theory) Let K ⊆ L be a finite. Via these maps. E2 ∈ IF(L/K). a) L / = {x ∈ L|σ(x) = x. b).5 Proposition. More precisely: G(L E) a) For any intermediate field K ⊆ E ⊆ L. ! 4. H Ψ : Subg(G) → IF(L/K). H2 ∈ Subg(G(L/K)). The result that follows holds for any extension and collects some general properties of Galois connections. d) are proposed as exercises. denote by IF(L/K) the set of its intermediate fields and by Subg(G) the set of subgroups of G.

3 ensures that [L : L / ] =|G(L/E)|. As L / ⊇ E G(L E) (by 4. Applying the isomorphism theorem. by IV. where we de−1 H noted η τη by σ ∈ H and used the fact that σ(x) = x. normal and separable. But τ'(L) = L. 4. so we can consider η as belonging to −1 G(L/K). so τ' ∈ G(L/K) and res(τ') = τ. The fundamental theorem of Galois Theory says that. we get G(L/K)/G(L/E) ≅ G(E/K). H d) If H is a normal subgroup in G(L/K). G(L/E) is a normal subgroup in G(L/K).2. ∀τ ∈ H. and G(L /K) is canonically isomorphic to the factor group G(L/K)/H. d) Let Ω be an algebraic closure of L and η a K-automorphism of H H Ω. then L is a normal extenH sion of K. H ∀x ∈ L . The H ! isomorphism follows by c). since every τ ∈ G(E/K) extends to a K-homomorphism τ' : L → Ω (where Ω is an algebraic closure of L). Then E ⊆ L is finite. for a finite Galois extension. the Galois connections are bijective. ∀x ∈ L . Prop. G(E/K) is canonically isomorphic to the factor group G(L/K)/G(L/E). Ker res = {σ ∈ G(L/K) | σ|E = id} = G(L/E). c) Consider the “restriction” homomorphism res : G(L/K) → G(E/K). On the G(L E) G(L E) other hand. observe that η(L) = L (because K ⊆ L is normal). We have σ|E ∈ G(E/K) because K ⊆ E is normal.5). Proof. The kernel of res is a normal subgroup in G(L/K).3.2. This amounts to show that. we have E = L / . 4. First.19. b) Proved at 4. keeping in mind that H = G(L/L ).V. since K ⊆ L is normal.4 The Fundamental Theorem of Galois Theory 267 c) For any intermediate field K ⊆ E ⊆ L with K ⊆ E normal. a) Let E be an intermediate field of K ⊆ L. A natural question . for any σ ∈ G(L/K). A normal and separable extension is called a Galois extension. Besides. We must prove that η(L ) ⊆ L . we have τη(x) = η(x). The homomorphism res is surjective. We have τη(x) = η(η τη)(x) = ησ(x) = η(x). so [L : E] = |G(L/E)| by Cor. res(σ) = σ|E.

Φ(F)) = Φ(E)∩Φ(F) ∈ Subg(L/K).− 2 } (these are the roots of 8 Also known as a poet.1. Mat.3. Cerc. F) = E∩F (see problem 7). A similar statement holds for inf(E. and normality fol2 2 4. In order to find this group. K)) are exactly {σx | σ ∈ G(L/K)} =: {x1.b). His result states that: Any field extension for which the Galois connections are bijective and inverse to each other is a finite Galois extension. they are anti-isomorphic as lattices: sup(E. . Acad. since the Galois connections are inclusionreversing bijections.P. elements. 195-259). by the fundamental theorem. we look on the action of the automorphisms in G on the generators 2 and 3 . Irr(x. If σ ∈ G.7 Example. under the pen name Ion Barbu (1895-1964). IF(L/K) and Subg(G(L/K)) are anti-isomorphic as ordered sets. if we suppose that the extension is finite.28. Moreover. So.268 V. R. Why? We have m = n iff L = K(x)). Stud.R. Galois Theory arises: “Which are all field extensions for which the Galois connections are bijective?” The answer to this problem was given in 1951 by the Romanian mathematician Dan Barbilian 8 (in the paper Soluţia exhaustivă a problemei lui Steinitz (The exhaustive solution of the Steinitz problem). if K ⊆ L is Galois of degree n and x ∈ L. K) = (X − x1)…(X − xm). separability is automatic (the characteristic is 0). If K ⊆ L is finite Galois. so the Galois group G has 4 lows from the fact that L is a splitting field of (X − 2)(X − 3) over Q. …. the conjugates of x over K (the roots of Irr(x. F) = EF in IF(L/K) corresponds to inf(Φ(E). that is: G(L/EF) = G(L/E)∩G(L/F). The extension Q ⊆ Q( 2 . Notice that. xm} (where necessarily m | n. We remark that. 2 (1951).. then σ ( 2 ) ∈ { 2 . the result is a consequence of 4. 3 ) =: L is Galois: The degree is 4 (see example IV.

corresponding to the proper subgroups <σ >. Q( 3 ) . <η > = {id. Since G has 4 elements. The subgroups of G are {id}. So. so L ⊇ Q( 3 ) . τ}. The following result is often used in arguments on field extensions: 4. But [L< σ > : Q] = [L : Q]/[L : L< σ >] = 4/2 = 2 = [Q( 3 ) : Q] . We remark that G ≅ Z2×Z2 (the 4 Klein group).4 The Fundamental Theorem of Galois Theory 2 269 Irr( 2 . We have σ( 3 ) = 3 . σ ( 3 ) ∈ { 3. η} and G. Let K ⊆ L be finite Galois and let K ⊆ M be an extension (suppose L and M are subfields of a larger field F). Here is a sample of the applications of Galois Theory. τ( 2 ) = 2 and τ( 3 ) = − 3 . one sees immediately that Q( 2 ) . the equality holds since “the degrees match”: [Q( 3 ) : Q] = [L< σ > : Q]. so <σ> these are all intermediate fields. <σ > = {id. <τ > = {id.− 3}. according to the following table: σ τ η id 2 2 − 2 2 − 2 − 3 3 3 3 − 3 For instance. Q) = X − 2). Then M ⊆ ML is finite Galois and G(ML/M) is isomorphic to a subgroup of G(L/K) (namely. . Q( 6 ) are proper distinct intermediate fields. likewise. with G(L/M ∩ L)). the automorphisms in G are determined by their action on generators. σ}. any element of G being of order 2. the extension has 3 proper intermediate fields. On the other hand. and the number of all possible choices is also 4.8 Proposition.V. <τ >. <η >. Based on the table above one can compile the multiplication table of G. The connections between the remaining subgroups and the remaining intermediate fields are established similarly.

normal and separable. where I is the image of the res homomorphism. But {x ∈ ML| σ(x) = x. the fundamental theorem reads in this case: 9 Wolfgang Adolf Ludwig Helmuth Krull (1899-1971). finite and separable. so ML is the splitting field over M of f (considered in M[X]). We have I L = {x ∈ L|σ(x) = x. but not necessarily finite. Galois Theory ML L M L∩M K Proof. German mathematician. By the fundamental theorem. ! The classical Galois theory that we presented here using a “linearized” approach. It is injective: ∀σ ∈ G with σ|L = id. So. this is tantamount to the fact that the fixed fields of I and G(L/M∩L) are equal. due to Dedekind and Artin. I ∀σ ∈ G(ML/M)} = M. Then σ|L ∈ G(L/K). I is a subgroup of G(L/K). ∀σ ∈ L. Thus G ≅ I. they agree on ML). . L is the splitting field over K of a separable polynomial f ∈ K[X]. M ⊆ ML is normal. we have σ = id (since σ and id agree on M and on L. ∀σ ∈ G(ML/M)}. Let σ be an automorphism in G := G(ML/M). has generalizations and counterparts in multiple directions: Infinite Galois Theory treats the case of an extension that is algebraic.270 V. Let us show that I = G(L/M ∩ L). The idea is to make the Galois group of the extension a topological group (by means of the Krull 9 topology) . res(σ) = σ|L. so L = M ∩ L. Consider the group homomorphism res : G → G(L/K).

(Hint: Look for the action of the automorphisms on 2 the generators 3 2 and ω. A Galois theory for commutative rings was developed by CHASE. HARRISON. Differential Galois theory looks into the problem of “explicit” solutions of differential equations. 4. where they are order reversing). where L = Q( 8 2 ). ROSENBERG. Nicolae devised a theory of this type that generalizes. The name reflects the fact that these theories establish inclusion preserving bijections between the lattice of intermediate fields and the lattice of subgroups of a certain group associated to the extension (as opposed to Galois theory. 3. Find G(L/Q) and all the subfields of L. Find G and the fixed field of G. among others. 2003). in a paper published in 1965. Kummer theory (T. ALBU. Cogalois Theory. New York. For an introduction and references. Marcel Dekker. 2. Exercises 1. where ω + ω + 1 = 0). the Romanian mathematicians T.V. Let K be a field and let K(X) be the rational function field over K.4 The Fundamental Theorem of Galois Theory 271 The Galois connections establish bijections from the set of all intermediate fields of the extension to the set of all closed subgroups of the Galois group. Recently. The same problem. Co-Galois theories. see for instance GOZARD [1997]. Let L be the splitting field over Q of X − 2. Albu and F. Let G = G(L/Q). for the polynomial X − 2 ∈ Q[X]. 4 3 . An example of such a theory is the Kummer theory that we present in the next chapter.

the purely inseparable closure of K in L (so I ⊆ L is Galois). . 6. aX + b . 7. The group GL(2. b) Prove that ϕ ∈ AutK(K(X)) = Gal(K(X)/K) iff there exist a. Galois Theory a) Let ψ : K(X) → K(X) be the unique K-homomorphism with ψ(X) = X + 1. denoted PGL(2. Use IV. b) G(L/E∩F) = < G(L/E) ∪ G(L/F) >. b. the subgroup generated by G(L/E) and G(L/F).1. d ∈ K. Show that: a) G(L/EF) = G(L/E) ∩ G(L/F). (Hint. K). F be intermediate fields. K)/S is called the 5.272 V. and I is the identity matrix). res(σ) = σ |S is an isomorphism. Suppose L/K is a normal finite extension and S is the separable closure of K in L. Then: G(LE/K) ≅ G(L/K) × G(E/K). K). K ⊆ E be Galois extensions such that L ∩ E = K. with ad − bc ≠ 0 such that ϕ(X) = cX + d H ≠ Gal(K(X)/K). K) → AutK(K(X)). whose kernel is cX + d the subgroup S of scalar matrices (matrices of the type aI. b) The fixed field of G(L/K) is I. c) L = SI. is a surjective group homomorphism. with ⎝c d ⎠ ϕ(X) = aX + b . Φ ⎜ ⎟ = ϕ. Let L/K be a Galois extension and let E. Let K ⊆ L. with a ∈ K projective linear group of degree 2 over K. Show that: a) K ⊆ S is normal and res : G(L/K) → G(S/K). Determine the subgroup H generated by ψ . So AutK(K(X)) ≅ PGL(2. c. In particular.28c)) ⎛a b ⎞ c) Show that Φ : GL(2. find the fixed field of H. but these subgroups have the same fixed field.

respectively of < A ∪ B > ? 8.1 Ruler and compass constructions 273 c) If A. B ≤ G(L/K). Such problems were one of the focal points of the antique Greek geometry. Determine all subfields of Q(ω). where ω is a primitive root of unity of order 9. The idea is to take an Abelian extension of degree 6 of Q. Prove that the Galois group of the extension is isomorphic to U(Z9) and the fixed field of the complex conjugation automorphism is Q(cos(2π/9)). which are the fixed fields of A ∩ B. This exercise shows how to obtain a normal extension of degree 3 of Q using Galois theory10. a normal extension of degree 3 of Q.VI. Applications of Galois Theory VI. VI. the fixed field of H is normal and has degree 3 over Q. A number of celebrated problems stayed un10 A more general way to construct such extensions uses discriminants. taking a subgroup H of order 2 of the Galois group.1 Ruler and compass constructions Geometric ruler and compass constructions have a substantial historical interest. Consider the cyclotomic extension Q ⊆ Q(ω). .

The main difficulties seem to have been the transfer of the problem from Geometry to Algebra and a precise formulation of the concept of “constructible by ruler and compass”.A set of initial points is given (usually two points).The compass can draw the circle centered in a constructed (or given) point and passing through another constructed (or given) point.For any two points (given. It is remarkable that the proof of impossibility of famous constructions requires only elementary concepts of field extensions theory. with the ruler one can draw the straight line passing through these points. which will answer the following famous problems: “Angle trisection”: Construct an angle whose measure is 1/3 of the measure of a given angle. Construct a regular heptagon. giving yet another demonstration of the power of the algebraic methods. In this section. The solution (often in the negative) was given only in th the 19 century. . for what natural numbers n is it possible to construct a regular polygon with n sides? The rules of construction are simple and well-known: .274 VI. despite the efforts of some of the best mathematicians. at the dawn of the theory of field extensions. . or already constructed as intersections of lines or circles). More generally. We want to obtain general constructibility criteria. Applications of Galois Theory solved for millennia. “construction” means exclusively “ruler and compass construction”. “Square the circle”: Construct a square whose area is equal to the area of a given circle. “Cube duplication”: Construct a cube whose volume is the double of the volume of a given cube. .

in the sense that they lead to the same set of constructible points C(S). The reader is invited to formulate in this form the problems in the list above. If A and B are distinct points in P. Sometimes CS is defined as the set of circles centered in a point in S with radius equal to the distance between two arbitrary points in S.1 Ruler and compass constructions 275 These are the only constructions allowed. Note that drawing a line (a circle) does not mean that all points belonging to it are constructed. The two definitions are in fact equivalent. and |AB| is the length of the segment [AB]. The definition we have adopted for CS corresponds to a “collapsible compass”: one cannot “transport” with the compass the distance between two points. 1. AB denotes the line determined by the points A and B. In what follows.11 The point M ∈ P is called constructible12 in one step from S if it satisfies one of the following conditions: P belongs to the intersection of two distinct lines in DS. P denotes an Euclidian plane. construct a set T of points (satisfying some property)”. Let S be a set of points in P (called initially constructible points). Prove this! 12 We omit saying “by the ruler and compass” in what follows. 11 . the construction of a line reduces to the construction of two distinct points on that line. A point is constructed only if it is identified as an intersection between lines (or circles. one realizes that all are equivalent to a problem of the following type: “For a given set S of points in the plane. or lines and circles). [AB] is the (closed) segment determined by A and B. Indeed. Let DS be the set of the lines determined by two distinct points in S and let CS be the set of circles centered in a point in S and passing through a point in S.1 Definitions and notations.VI. Analyzing the various problems of construction. an angle is determined by three points (the vertex of the angle and a point on each side) etc.

We suppose from now on that S is a subset of P with at least two points. Indeed.276 VI. an angle. Let C0(S) := S and define by recurrence Cn(S) := C1(Cn − 1(S)). Cn(S) ⊆ Cn + 1(S). Proof. let A. Exercise in elementary geometry.2 Remark. b) S ⊆ C1(S). The point A is the intersection of the line AB with the circle of center B passing through A. 1. * ∀n ∈ N . Applications of Galois Theory P belongs to the intersection of two distinct circles in CS. A point in P is called constructible from S if the point is in C(S) := ∪n∈N Cn(S).…) is determined by points in C(S). Let S ⊆ P with |S| ≥ 2 and let A. a) The symmetric of B with respect to A is constructible from S. 1. ! Assumptions and notations. if |S| ≥ 2. Let C1(S) denote the set of points that are constructible in one step from S. So.0). . the angle. then the parallel through C to AB is constructible. If one takes only one initially constructible point. The segment OI has thus length 1.…) is said to be constructible from S if the line (the circle. B ∈ S. P belongs to the intersection of a circle in CS with a line in DS. C ∈ C(S) be distinct. Fix two distinct points O and I in S. of coordinates (0. a) |S| = 1 ⇔ C1(S) = S. b) The midpoint D of the segment [AB] and the perpendicular on [AB] in D are constructible. B. If the set S is understood. d) If C ∉ AB. A line (a circle.0). There exists a unique system of Cartesian coordinates in the plane P such that O is the origin. and I has coordinates (1.3 Lemma. ∀n ∈ N. Let κ : P → R × R the mapping that associates to a point in P its coordinates in the system above. c) The perpendicular from C on AB is constructible. we say constructible instead of constructible from S. nothing else can be constructed from it.

define TR as the set of real numbers that occur as coordinates for the points in T. and say “the point (x. b) The point (0. By 1. construct the symmetric M' of M with respect to O. In the hypotheses above. In our hypotheses. a)⇒b) Let M be the (constructible) point (x. then the midperpendicular of the segment [MM'] (i. TR := {x ∈ R | ∃ y ∈ R such that (x.e. As a consequence of the constructions in 1. 1. The point (0. x) belongs to the intersection of the circle centered in O of radius [OM] with Oy.4 Definition. If the set S is clear from the context. y) of its coordinates. Proof.5 Proposition. Thus. The following assertions are equivalent: a) x is constructible from S. y) is constructible. d) There exists y ∈ R such that (y.VI. SR contains at least 0 and 1. x) is constructible. 0) is constructible from S. the coordinate axes Ox and Oy are constructible from O and I. y) ∈ T or (y. S can be seen as a subset of R × R. y)”. ! The concept of constructible complex number is also natural (and useful. If T is a set of points in P. we say simply “x is constructible”. as we will see shortly): . the Oy axis). x) ∈ T}. The real number x is called constructible from S if the point of coordinates (x. The other implications are equally easy and are left to the reader.1 Ruler and compass constructions 277 We often identify a point M ∈ P with the couple (x.3. 0). Let x ∈ R.3. c) There exists y ∈ R such that (x. 1. 0 and 1 are always constructible. Therefore. x) is constructible.

Omitting S from the notation K simplifies the notation. Let K be the set of constructible reals13. c) The real numbers a and b are constructible.5). Then the perpendiculars from (a. y) & x + iy from R × R to C). y) ∈ K × K. 1} ⊆ SC . The following statements are equivalent: a) The complex number a + ib is constructible.e. y ∈ K}. The set of constructible complex numbers is K[i] = {x + iy | x. 0) and (0. The complex number z = x + iy (where x. we mean the real numbers constructible from S. so (a. the affixes of the points in S): SC = {x + iy ∈ C | (x. We have {0. then the points A = (a. C(S) = K × K.8 Remark. b ∈ R. 13 . b)⇒c) Let (a. b) be a constructible point.. y) is constructible if and only if (x. Of course. ! 1. Proof.6 Definition. Let a. The previous proposition says that the point (x. The set C is in one-to-one correspondence with the plane P (composing the bijection κ above with (x. It is natural to show that the following result holds: 1. Applications of Galois Theory 1. b) are constructible. b) is the intersection of the perpendicular in A on Ox with the perpendicular in B on Oy. Let SC be the set of complex numbers that correspond to the points in S (i.7 Proposition. 0) and B = (0. Thus. b) The point (a. b) are constructible (use prop. y) ∈ S}. (a. b) on the axes Ox and Oy are constructible. b) is constructible. a)⇒b) If a + ib is constructible. y ∈ R) is called constructible from S if the real numbers x and y are constructible from S. In other words.278 VI. 1.

ab are constructible. a) and (− a. We use the constructions from 1.3. K[i] is a subfield in C. a + b ∈ K. 0) intersects Oy in −1 (0. of coordinates (a. Suppose now that a > 0. is constructible. we use the trigonometric form of complex numbers. ab ).9 Proposition. a + b. 0) passing through (0. So. b) Since K is a subfield in R. then its conjugate a − bi ∈ K[i]. It is enough to prove that − a. b) K[i] is a subfield of C. Let −1 a. b ≠ 0. K is the smallest subfield of R that includes SR. 0) and I(0. 0) in two points. Proof. a ∈ K. z and z ∈ K. if L is a subfield of R. so (− a. b ∈ K. 0) is constructible. Then z = ± r (cos(α 2 ) + i sin(α 2 )) . 1) are in S. −1 For the constructibility of ab . the coordinates of the points in S. and is closed under square roots. its symmetric with respect to O. r·sinα ∈ K. Moreover. 0) determine a line d. Let z = r(cosα + isinα). whence − a ∈ K. For proving that K[i] is closed under square roots.1 Ruler and compass constructions 279 1. − a ). with r > 0. The parallel to this line through (a. x ∈ K. We have .0) and (0. 0) is constructible. closed under square roots: for any x ∈ K with x > 0. a) K is a subfield of R. So. L contains SR and L is closed under square roots. a + b). a) 0 and 1 belong to K. Since K is a field. with a. The next lemma implies that. Moreover. b ∈ K. a. a ) and (a. The constructible points (0. 0) intersects the perpendicular on Ox in (a. 0) is constructible and intersects d in (b. then K ⊆ L. r·cosα. closed under square roots and under complex conjugation: for any z ∈ K. The perpendicular on Ox in (b. ((a + 1)/2. If a + bi.VI. The point (a. It is clear that K contains the coordinates of any point in S. since O(0. The circle centered in ((a + 1)/2. 0). 1). consider the line determined by (b. K[i] is the smallest subfield of C that includes the set SC of the affixes of points in S and is closed under square roots and conjugation. is an element in K[i].

We have cos(α 2 ) = ± (1 + cosα ) 2 ∈ K. such that A1A2 and A3A4 are not parallel and M is their intersection. sin(α/2) ∈ K. Use the notations in Definition 1. then C(L × L) = L × L: the set of points constructible from L × L coincides with L × L. such that x. this equation can be written as ax + by = c. yi) ∈ L × L. In the same way. M(x. The equation of the line A1A2 is (x − x1)(y2 − y1) = (y − y1)(x2 − x1). Then M(x.1. y ∈ L or there exists u ∈ L. 4}. u > 0. ∀i ∈ {1. for some d. If (x. a) Let L be a subfield of R and let (x. . y) is the intersection of two lines in DL × L. Applications of Galois Theory 2 2 2 2 r cos α + r sin α = r .280 2 2 VI. Likewise. 2. then x. y). then K ⊆ L. Since L is a field. r and r ∈ K (which is closed under square roots). b. y) is constructible in one step from L × L. with a. c) Let E be a subfield of C such that E is closed under conjugation and square roots and E ⊇ SC. (otherwise the lines would be parallel). Cramer's formulas show that the solution (x. We get also cosα. Case I. sinα ∈ K. closed under square roots. The next lemma shows that any subfield of C containing the affixes of points in S and is closed under conjugation and square ! roots must include K[i]. b) If L is a subfield of R. y) of the system is in L × L.10 Lemma. constructible in one step from L × L. 3. Let Ai(xi. y) is the solution of the system ⎧ax + by = c ⎨ ⎩dx + ey = f The determinant of the system is nonzero. e. Then K[i] ⊆ E. y ∈ L( u ) . In particular. 1. the equation of A3A4 is dx + ey = f. c ∈ L. so r ∈ K. u > 0. f ∈ L. is in L × L or in L( u ) × L( u ) for some u ∈ L. so z ∈ K[i]. Proof. if SR ⊆ L. a) It is enough to prove that any point M(x. y) ∈ R × R.

y) is the solution of the system: ⎧( x − a )2 + ( y − b )2 = r 2 ⎨ ⎩dx + ey = f ( ) where a. closed under square roots. then e ≠ 0 and swap d with e). E = L[i] = {x + iy | x. d. the result follows from a). b. M belongs to the intersection of two circles in CL × L. L is a subfield of R. y ∈ L. Subtracting the equations. t ∈ L. Since x. If x ∈ L. By a). ⎨ ⎪( x − c )2 + ( y − d )2 = t 2 ⎩ Case II. 2 where u = q − 2sp ∈ L. y ∈ L( u ) . We must show that the solutions (if any) of a system of the form ⎧( x − a )2 + ( y − b )2 = r 2 ⎪ . for some p. So. so x. together with one of the initial equations. then x − iy ∈ E (E is closed under conjugation). On the other hand. f ∈ L. if x. C(L × L) = L × L. y ∈ L.e. −1 −1 2 x = − d ey − d f. for some d. b) The subfield L being closed under square roots. r. Thus. y ∈ E. c. . Case III.VI. So. if x. The equation of a circle in CL × L is (x − a) + (y − b) = r . are in L( u ) × L( u ) for some u ∈ L. y ∈ L}. Thus. K ⊆ L. x. q. x. we obtain a system of the type studied at case II. Then M(x. M belongs to the intersection of a line in DL × L with a cir2 2 2 cle in CL × L. This equations must have real solutions (the circle and the line intersect). r ∈ L. y ∈ R and x + iy ∈ E. so x ∈ L. s ∈ L. e. y = 2 p −1 ⋅ − q ± q 2 − 2 sp . The first equation becomes py + qy + s = 0. then C(S) ⊆ C(L × L). We saw that the equation of a line in DL × L is dx + ey = f.1 Ruler and compass constructions 281 Suppose d ≠ 0 (if d = 0. So. u > 0. c) Let L = E ∩ R. If SR ⊆ L (i. Conversely. x > 0. because i = − 1 ∈ E. then x + iy ∈ E. then x ∈ E (E is closed under square roots) and x ∈ R. b. for some a. y ∈ R. the quadratic terms cancel out and we obtain linear equation in x and y. S ⊆ L × L).

…. meaning that “x belongs to the smallest subfield of R that includes Q(U) and is closed under square roots”. there is a tower K0 ⊆ K1 ⊆ … ⊆ Kn such that UR ⊆ Kn and [Ki : Ki − 1] ≤ 2. n}. T ⊆ S. n}. This means that there exists a finite set U ⊆ Cm−1(S) such that any point in T is constructible in one step from U. so K[i] ⊆ L[i] = E. lemma 1. TR ⊆ K n ( u1 . …. 14 . then there exists a tower of extensions K0 ⊆ K1 ⊆ … ⊆ Kn such that TR ⊆ Kn and [Ki : Ki − 1] ≤ 2. y) is constructible from the set of points S” has a purely algebraic form: “x and y belong to the smallest subfield of R that includes Q(SR) and is closed under square roots”. so TR ⊆ SR ⊆ Q(SR) = K0.10. we have K ⊆ L.… . Proof. It seems though that Gauss knew as early as 1796 this criterion of constructibility. for any finite set of points T. if T ⊆ Cm(S). Let m > 0 and let T = {A1. Ar} be a finite subset of Cm(S). From the induction hypothesis. So. …. 1.L. who proved it and published it in 1837 (when he was “elève-ingénieur des Ponts-et-Chaussées”). If m = 0. |U| ≥ 2). For any point As ∈ T (s ∈ {1. r}). Wantzel. …. ! We have proven that the geometric statement “the point M(x. us > 0.a) says that As has its coordinates in Kn (then set us = 0) or in a quadratic extension of Kn of the form K n ( us ) . 1 ≤ i ≤ n. K0 = Q(SR) and [Ki : Ki −1] ≤ 2 for any i.282 VI. ∀i ∈ {1. “⇒” It is enough to prove by induction on m that. This algebraic translation allows us to say about some real number x: “x is constructible from U” (where U is a set of real numbers). Note also that any rational number is constructible (from any U ⊆ R. for some us ∈ Kn. by a). such that x ∈ Kn. ∀i ∈ {1. ur ) and we have a tower of extensions The result belongs to M.11Theorem. Applications of Galois Theory The hypothesis SC ⊆ E is thus equivalent to SR ⊆ L.14 A real number x is constructible from S if and only if there exists a tower K0 ⊆ K1 ⊆ … ⊆ Kn of subfields of R.

we know that Kn−1 ⊆ K. If n > 0. ur ) .a). ! We can easily prove now that some classic ruler and compass constructions problems are impossible: 1. we have Kn ⊆ K (recall K is the field of the real numbers constructible from S). If n = 0. such that Kn = Kn−1( u ).12 Corollary. sin(θ/3)) from {O. we deduce Kn ⊆ K. x is algebraic over K0 and K0 ⊆ K0(x) ⊆ Kn. since K is ! closed under square roots. . In this case there exists u ∈ Kn−1. (Necessary condition for constructibility) Let x be a real number constructible over S and let K0 = Q(SR). [Kn : Kn−1] ≤ 2 implies either that Kn = Kn−1 (and the induction hypothesis shows that we are done) or that Kn is an extension of degree 2 of Kn−1. Using 1. Proof.… . for any tower Q(SR) = K0 ⊆ … ⊆ Kn ⊆ R.VI. 1 ≤ i ≤ n . The extension Kn in the previous theorem is such that x ∈ Kn and [Kn : K0] is a power of 2. [K0(x) : K0] divides [Kn : K0].9. with [Ki : Ki−1] ≤ 2. so. Thus. Then x is algebraic over K0 and its degree over K0 is a power of 2 (there exists e e ∈ N such that [K0(x) : K0] = 2 ). a) The trisection of an angle of measure θ is equivalent to the construction of the point (cos(θ/3). satisfying the condition that each extension has degree at most 2.13 Proposition. we deduce that K includes the subfield of R generated by SR. which means that [K0(x) : K0] is a power of 2. u > 0. then K0 = Q(SR). By induction.1 Ruler and compass constructions 283 K0 ⊆ … ⊆ Kn ⊆ K n ( u1 ) ⊆ … ⊆ K n ( u1 . which is exactly Q(SR). 1. “⇐” We prove by induction on n that.

the intersection of the other side with the unit circle is (cos α. trisecting the angle of measure θ is tantamount to the constructibility of cos(θ/3) from 3 {cosθ}. Indeed. (cosθ. Proof. Conversely. We have Irr(u. 15 can be trisected). a) Constructing an angle of measure α is equivalent to constructing the point (cos α. so [Q(u) : Q] = 3. sin α) or (cos α. a 90° angle . since [Q( 3 2 ) : Q] = 3 is not a power of 2. The formula cos(3x) = 4cos x − 3cosx implies that u := cos(θ/3) satisfies the equation: 3 4u − 3u = cosθ. if cos α is constructible. since [Q(u) : Q] is not a power of 2. This result does not say that no angle can be trisected (for instance. On the other hand. there exists no ruler and compass construction of the trisection of an arbitrary angle. − sin α). This shows that u is not constructible.15 b) Cube duplication is impossible (it is impossible to construct a cube whose volume is the double of the volume of a given cube). The real number 3 2 is not constructible. but that some angles (the 60° angle) cannot be trisected. Thus. sin α) is given. c) Squaring the circle is impossible (it is impossible to construct a square whose area is equal to the area of a given circle). Applications of Galois Theory I. Thus. The cube having double the volume has the side 3 2 . sin α). A 60° angle cannot be trisected by ruler and compass (a 20° angle is not constructible). with coefficients in Q(cosθ). b) Choose the unit length to be the length of the side of the cube. For θ = 60°. if the angle has its vertex in O one of the sides is Ox. the initially constructible points are O and I. then the angle between OP and Ox has measure α. we obtain that u := cos20° is a 3 root of g := 8X − 6X − 1 ∈ Q[X]. Q) = g (since g has no rational roots). 3 We are led to the study of the polynomial 4X − 3X − cosθ . then sin α = ± 1 − cos2 α is constructible. if P(cos α. Thus.284 VI. sinθ)}.

! 1. then z is algebraic over L0 = Q(SR)(i) and its degree over L0 is a power of 2 (there exists e e ∈ N such that [L0(x) : L0] = 2 ). the initially constructible points are O and I. b) If z is a complex number constructible over S. The chain of subfields L0 ⊆ L1 ⊆ … ⊆ Ln satisfies the required conditions. we have Ln ⊆ K(i). a) The complex number z is constructible from S (i.15 Theorem.11 is used. …. z ∈ K(i)) if and only if there exists a chain of subfields of C. ∀t ∈ {1. so it is not constructible. L0 ⊆ L1 ⊆ … ⊆ Ln. there exists a chain K0 ⊆ K1 ⊆ … ⊆ Kn of subfields of R. with [Lt : Lt−1] ≤ 2. we show that for any chain Q(SR)(i) = L0 ⊆ L1 ⊆ … ⊆ Ln of subfields of C. for any t ∈ {1. . n}. But π is transcendental over Q and π is also transcendental. Let Lt := Kt(i). so the side of the square with area π is π . “⇐” The same technique as in Theorem 1.9. By 1. Proof. with x. (Characterization of constructible real numbers) Suppose U ⊆ R. and Ln−1 ⊆ K(i) by hypothesis. for any t ∈ {1. ! For the formulation of a necessary and sufficient criterion of constructibility.11 is useful: 1. the following complex version of Theorem 1. For n > 0. n}. y ∈ K.11. By induction on n. x ∈ R is algebraic over Q(U) and N is the normal closure of Q(U) ⊆ Q(U)(x). ….e. a) “⇒” Let z = x + yi. …. If n = 0. x. n}. y ∈ Kn and [Kt : Kt − 1] ≤ 2. [Ln : Ln−1] ≤ 2 implies Ln = Ln−1 or Ln is a quadratic extension of Ln−1. L0 = Q(SR)(i) and [Lt : Lt −1] ≤ 2.14 Theorem. b) Exercise. The area of the circle of radius 1 is π. apply 1. such that z ∈ Ln.VI. Since K(i) is closed under square roots.1 Ruler and compass constructions 285 c) Choosing the unit length to be the radius of the circle. Then x is constructible from U if and only if [N : Q(U)] is a power of 2. Ln ⊆ K(i). with K0 = Q(SR).

Theorem 1. Keeping the same notations.14 implies that σ(Kn) ⊆ K(i).16 Remark. Q(U)) ∈ Q(U)[X]. ! 1. ∀t ∈ {1.14. Because α ∈ K(i). e}. then the normal closure of Q(U) ⊆ Q(U)(x) is included in L and has its degree over Q(U) a power of 2. if x belongs to such an extension L.7). ∀t ∈ {1. The proof of theorem 1. ∀t ∈ {1. …. implies that N ⊆ K(i). e} (so |Gt| = 2 ). G = G0 ⊇ G1 ⊇ … ⊇ Ge = {id}. Proposition 11) there exists a chain of subgroups of G. …. From (Appendix 6. the chain of extensions K0 = σ(K0) ⊆ σ(K1) ⊆ … ⊆ σ(Kn) has the property that [σ(Kt) : σ(Kt−1)] ≤ 2. On the other hand.11 shows there exists a tower K0 ⊆ K1 ⊆ … ⊆ Kn of subfields of R. so N ⊆ E. we have K0 ⊆ K1 ⊆ … ⊆ Ke = N (we applied the Galois correspondence). n} and x ∈ Kn. So. Since x ∈ E and E is normal over K0. Let α be a primitive element of the separable finite extension K0 ⊆ N. We must show that x ∈ K (the numbers constructible from U) if and only if [N : K0] is a power of 2. n}. …. Let E be the normal closure of K0 ⊆ Kn. We have [Kt : Kt−1] = [Gt−1 : Gt] = 2. We now solve the classical problem of the constructibility of regular polygons. Therefore x ∈ R ∩ K(i) = K. so e its Galois group G has 2 elements.14. g splits over E. e “⇐” The Galois extension K0 ⊆ N has the degree 2 for some e. Indeed. ∀t ∈ {1. Then N is the splitting field of g over Q(U) := K0. Denoting by Kt the fixed field of Gt. where H is the set of the K0-homomorphisms from Kn to C (see 2. with [L : Q(U)] a power of 2 and x ∈ L. n}. “⇒” Theorem 1. …. e−t with [Gt − 1 : Gt] = 2. such that [Kt : Kt − 1] ≤ 2. E ⊆ K(i) and thus N ⊆ K(i). E = K0{∪{σ(Kn)|σ ∈ H}}. ….286 VI. we also have: x is constructible from U if and only if there exists a normal extension L of Q(U). Applications of Galois Theory Proof. Let g := Irr(x.b) guarantees that [K0(α) : K] = [N : K0] is a power of 2. theorem 1. . ∀t ∈ {1. For any σ ∈ H.

t ∈ N and p1.17 Theorem. b ∈ N*.16 Proof. At 1. we saw that the angle α is constructible if and only if cosα is constructible. Using computers. we get u·2π/b + v·2π/a = 2π/ab and we apply the fact that if the angles α and β are constructible. the construc- tion of the angle 2π/n is equivalent to the construction of every angle 2π piαi . 65537.1 Ruler and compass constructions 287 1. …. No other Fermat primes have been found yet. 16 . e > 1. We claim that the angle 2π/ab is constructible if and only if the angles 2π/a and 2π/b are con- structible. any multiple of it can be constructed (so the angles 2π/a. 257. 2π/b are constructible). …. 2. if n = p1 1 … pk k . e If n = 2 . 4. Here. with (a. where e. b) = 1. for any u. For m = 0. This allows us to suppose from now on that n is a power of a prime. which are indeed prime. Let a. 5.13. let u. “constructible” means “constructible from two points (O and I )”. …. 1801) Let n ≥ 3. with p1.VI. then the angle 2π/n is constructible (induction on e: e e−1 the angle 2π/2 is obtained by bisecting the angle 2π/2 ). 17. v ∈ Z such that ua + vb = 1. To construct a regular n-gon is equivalent to construct the angle17 2π/n. then. for any i ∈ {1. 3. The direct implication is simple: for a given angle (for in- stance 2π/ab). (Gauss. pk distinct primes. 1. Such a prime number is called a Fermat prime. it has been proven that the above are all the 40000 Fermat primes less than 10 . The regular n-gon is constructible by ruler and compass if and only if the prime factor e decomposition of n is n = 2 ·p1·…·pt. 17 We say “the angle u” instead of “the angle of measure u”. pt are m distinct primes of the form 22 + 1 . one obtains 3. Multiplying by 2π/ab. for some m ∈ N. v ∈ Z. Conversely. k}. the angle uα + vβ is constructible. α α Thus.

for b b some a. a odd. which means that p is a Fermat prime. the degree of the cyclotomic extension Q ⊆ Q(ζ ) is ϕ(n) = α α α−1 α α−1 ϕ(p ) = p − p . p = 22 a + 1.12. . 15. Proof. 17}. as is a regular 9-gon. of degree ϕ(p) = p − 1. Suppose e is not a power of 2. so [Q(ζ ) : Q(cos(2π/n))] ≤ 2. Let ζ := cos(2π/n) + isin(2π/n). 6. so p is not a prime. 5. In fact. a ≥ 3. a regular n-gon is constructible ⇔ n ∈ {3. Let d := 22 . then p | 2 . [Q(cos(2π/n )) : Q] is a power of 2. let p be a Fermat prime. so [Q(ζ ) : Q(cos(2π/n))] ≥ 2. Let e ∈ N with α−1 e e e p (p − 1) = 2 . The extension Q ⊆ Q(ζ ) is normal. 16. Hence e must be a power of 2. for n ≤ 20. b ∈ N. So α = 1 and p − 1 = 2 . Let n ∈ N. ! 1.19 Example. But i·sin(2π/n) is a 2 2 root of (cos(2π/n) − X = 1. The next lemma says that [Q(ζ ) : Q] = 2·[Q(cos(2π/n)) : Q]. We α prove that cos(2π/p ) is constructible ⇔ α = 1 and p is a Fermat prime.15.288 VI. using 1. We may suppose that ζ = cos(2π/n) + i·sin(2π/n). so p − p is a power of 2. d a + 1 is divisible by d + 1. Since a is odd . e b whence p = 2 + 1. which is a power of 2. On the other hand. n ≥ 2 and let ζ be a primitive complex nth root of unity. p a prime. ζ ∉ Q(cos(2π/n)). ! 1. A regular heptagon is impossible to construct by ruler and compass. Let us prove that cos(2π/p) is constructible. Applications of Galois Theory α Suppose from now on that n = p . Then [Q(ζ) : Q(cos(2π/n ))] = 2. Then. So. p ≠ 2. cos(2π/p) belongs to Q(ζ ). Conversely. 4. so p = d a + 1. Suppose cos(2π/n) is constructible. contradiction. 8. The remark following Theorem 1. Since ζ = cos(2π/p) + isin(2π/p) −1 and ζ = cos(2π/p) − isin(2π/p) are in Q(ζ ). Clearly.18 Lemma. 10. applies and we obtain the constructibility of cos(2π/p). 12. If α ≥ 2. α ∈ N*. absurd. Then e = a ·2 .

“module homomorphism” means “linear mapping” etc. “free module” translates into “vector space”.…. Let e = (e1. the definitions and the proofs are the same as in the more general case of an extension of commutative rings R ⊆ S such that S is a free R-module of finite rank. These are particular cases of concepts that we study now. The reader less familiar with these concepts may assume in what follows that R ⊆ S is a finite extension of fields (so. For instance. where aij ∈ R are defined by the relations In other words. the conjugates of x over K are the roots of the minimal polynomial of x over K. Although we use the notions of trace and norm mainly in the case of a finite extension of fields. (EndR(S). and the product of all its conjugates is a − db . +. This is what we called “the norm of a + b d ” and appeared in the study or arithmetic properties of the ring Z[ d ]. The sum of all conjugates of 2 2 a + b d is thus 2a. °) is a ring with identity. such that S.VI. S is a commutative ring with identity and R is a subring of S containing the identity.2 Trace and norm If K ⊆ L is a finite extension and x ∈ L. if x = a + b d . "+" denotes homomorphism addition and "°" the usual map composition. then a − b d is the only conjugate of x over Q.en) be a basis of the free R-module S. Let R ⊆ S be an extension of commutative rings with identity 18 . We denote EndR(S) := {ϕ : S → S | ϕ is an R-module homomorphism}. with d a square free integer and a. with its canonical R-module structure. In this case. are elements in K. as the Viète relations show. each occurring with its multiplicity.4. b ∈ Q. We recall some facts on free modules from II. is free and has rank n. the matrix of ϕ in the basis e is the n×n matrix Me(ϕ) = (aij)1 ≤ i. 18 . The sum (and the product) of these conjugates. If ϕ ∈ EndR(S).2 Trace and norm 289 VI. j ≤ n. S is an R-vector space of finite dimension).

there exists −1 U ∈ U(Mn(R)) such that Mf(ϕ) = U·Me(ϕ)·U . −1 Mf(ϕ) = U·Me(ϕ)·U . called the characteristic polynomial of ϕ. then there exists a unique matrix U = (uij)1 ≤ i. If f = (f1. ∀i ∈ {1. The characteristic polynomial of a vector space endomorphism inspires the following definition (cf.): 2. in this case. …. Applications of Galois Theory n ϕ(ei) = ∑ aij e j . ϕ & Me(ϕ). j ≤ n ∈ Mn(R) such that: fi = ∑ uij e j . j =1 n f = (f1. The characteristic polynomial of ϕ is correctly defined (it is independent on the choice of a basis): if f is another basis. Thus. III. Define P(ϕ) := det(XI − Me(ϕ)) ∈ R[X]. The matrix: ⎡ X − a11 − a12 … − a1n ⎤ ⎢ −a X − a22 … − a2 n ⎥ 21 ⎥ ∈ Mn(R[X]) XI − A := ⎢ ⎥ ⎢ … ⎢ −a − an 2 … X − ann ⎥ ⎦ ⎣ n1 is called the characteristic matrix of A. which is a ring anti-isomorphism. −1 det(XI − Mf(ϕ)) = detU·det(XI − Me(ϕ))·det(U ) = det(XI − Me(ϕ)).…. fn) is a basis in R S ⇔ U is invertible in Mn(R) ⇔ det U ∈ U(R) (the group of units of R). Let ϕ ∈ EndR(S) and let e be an arbitrary basis in R S. so XI − Mf(ϕ) = −1 U·(XI − Me(ϕ))·U . . Let I ∈ Mn(R) the identity matrix and let A = (aij) ∈ Mn(R). n}.4.18.290 VI. j =1 We obtain a function Me : EndR(S) → Mn(R). ….1 Definition. fn) is an n-uple of elements in S.

3 Proposition. we have. The notations we use take this into account. but not the terminology.2 Trace and norm n n−1 291 + … + a1 X + a0. +. det(θx) =: NS/R(x) ∈ R is called the norm of x. For any x ∈ S. . det(ϕ) = det(A). NS/R(a) = a . is an R-module homomorphism and also a ring homomorphism (in other words. TrS/R(a) = na. It is clear that these notions depend not only on x.VI. θ : (S. NS/R(xy) = NS/R(x)·NS/R(y). the coefficients ai ∈ R Let P(ϕ) := X + an − 1X depend only on ϕ (not on the choice of the basis). Prove this! Using this remark. and norm NS/R : S → R. x & θx. so the following definitions are correct: Tr(ϕ) := − an − 1 (called the trace of ϕ) n det(ϕ) := (−1) a0 (called the determinant of ϕ). n n ! So. θx(y) = xy. S/R) ∈ R[X] is called the characteristic polynomial of x. Tr(θx) =: TrS/R(x) ∈ R is called the trace of x.2 Remark. ·) → (EndR(S). TrS/R : S → R is an R-module homomorphism. ◦). is an R-module homomorphism. ∀y ∈ S. but on the extension S of R. We apply the definitions above to θx ∈ EndR(S). NS/R(ax) = a NS/R(x). ∀a ∈ R: TrS/R(x + y) = TrS/R(x) + TrS/R(y). We defined therefore the mappings trace TrS/R : S → R. so some caution is recommended. the next properties are easy to prove: 2. TrS/R(ax) = aTrS/R(x). then Tr(ϕ) = Tr(A) = a11 + a22 + … + ann. In the conditions above. the function θx : S → S. ∀x. +. If A = (aij) ∈ Mn(R) is the matrix of ϕ in some basis. y ∈ S. The following notations and terms are used: P(θx) =: P(x. 2. θ is an R-algebra homomorphism).

L/K). let x ∈ L and let P(x. K) (1) d Irr(x. P = f . Let K ⊆ L be a finite extension of fields.292 VI. K). which means f |g. So. As for P(x. The Frobenius theorem implies that P and f have the same irreducible factors. if g ∈ K[X] is such that g(θx) = 0. Since f is irreducible. Applications of Galois Theory The case we are interested in is when R and S are fields. written as P(x. we obtain f = Q. NL/K(x) = (NK(x)/K(x)) [L : K(x)] = {∏σ ∈H σx} . L/K) ∈ K[X]. (2) where d = [L : K]i is the inseparability degree of L over K. so g(x) = 0. θP(x) = 0. Let H := HomK(L. we have Q(x) = 0. Let f := Irr(x. We suppose that all algebraic extensions of K are subfields of Ω. then θg(x) = 0. which means P(x) = θP(x)(1) = 0. This proves (1). L/K) be its characteristic polynomial. Identifying the coefficients of X n−1 Let Q := P(x. K) = P(x. so f |Q. We have f | P. The Cayley-Hamilton theorem for θx ∈ EndK(L) says that P(θx) = 0. f is the t only irreducible factor of P. in (1). Also. Let P := P(x. 2. Taking the degrees. g(θx) = θg(x) (use 2. so t = [L : K(x)]. L/K) = P(x. K(x)/K) [L : K(x)] . we obtain n := deg P = [L : K] = t·deg f = t·[K(x) : K]. Ω) be the set of K-homomorphisms from L to Ω. . Then: [L : K(x)] P(x. K(x)/K). L/K) = Irr(x. This shows that f is the minimal polynomial of the endomorphism θx. d (4) f (θx) = θf(x) = θ0 = 0. Note that. Thus. for some t ∈ N*. a fixed algebraic closure of K.4 Theorem. Fix a finite extension of fields K ⊆ L. since P(x) = 0. L/K) = {∏σ ∈H (X − σx)} . since deg Q = [K(x) : K] = deg f. for any g ∈ K[X].2 for the proof). K(x)/K) and P(x. In particular: TrL/K(x) = [L : K(x)]·TrK(x)/K(x) = d·{∑σ ∈H σx} (3) Proof.

Then. K(x)/K) and reduces the computation of P(x. and σ0 extends to a unique K-algebra isomorphism σ' : Ω[X] → Ω[X] with σ'(X) = X) and obtain q q f = (X − σx) σ'(g). This proves (2). m f ={∏σ ∈ H (X − σx)} . for any q ∈ N. TrL/K(x) = ∑σ ∈G σx = {∏σ ∈Hom(L. This implies that: for any q ∈ N. In the general case. we obtain the first equality in (4. Setting X = 0 in (1). namely {σx |σ ∈ H}. . which is the first equality in (3). Assume for the moment that L = K(x). L/K) = f [L : K(x)] where m[L : K(x)]/r = [K(x) : K]i·[L : K(x)]/[L : K(x)]s = [K(x) : K]i·[L : K(x)]i = [L : K]i. we have: P(x. So. Apply σ' (σ : L → Ω extends to a K-isomorphism σ0 : Ω → Ω.3. Let n := |H| = [L : K]s (the separable degree of L over K). ! 2. So. Ω) (X −σx)} = {∏τ ∈Hom(K(x). K) = P(x. the inseparability degree of L over K. In conclusion. Indeed. Ω) has r := [L : K(x)]s extensions to L (see V. L/K) to the case L = K(x).11. m = [L : K]i = [K(x) : K]i. which proves the claim. then the root σx ∈ Ω q has also multiplicity m. for any x ∈ L. we obtain [L : K] = mn = m[L : K]s. Ω) (X − σx) = ∏τ ∈ Hom(K(x).2 Trace and norm 293 we get TrL/K(x) = [L : K(x)]·TrK(x)/K(x). NL/K(x) = ∏σ ∈G σx. Ω) (X −τx) } m[L : K(x)]/r m [L : K(x)] . Ω) (X − τx)r. Comparing the degrees. (X − x) | f ⇔ q (X − σx) | f. Let K ⊆ L be a Galois extension and let G be its Galois group.).5 Corollary.) Formula (1) says that Irr(x. The second part of (3) and (4) follows from (2). (X − x) | f in Ω[X] ⇔ q ∃g ∈ Ω[X] with f = (X − x) g. each σ ∈ HomK(K(x). ∏σ ∈Hom(L. The polynomial f has n distinct roots in Ω. By (1). If m is the multiplicity of the root x of f and σ ∈ H.VI.

2 An important property of the trace and the norm is the transitivity property: If R ⊆ S ⊆ T are extensions of commutative rings such that S . From (3) and (4) we have. L) = G(L/K). σn and TrL/K(x) = σ1x +…+ σnx. g is separable and G(L/K) = {id.if char K = 2: 2 2 2 NL/K(a + be) = (a + be) = a + b d. (4). Proof. Conversely. (3). for any a. Applications of Galois Theory In particular. then σ1 +…+ σn = 0. ….6 Proposition. . If g = X − d is irreducible in K[X] (⇔ d is not a square in K) and e = d is a root of g in Ω. …. where σ(e) = −e. Let K be a field and let d ∈ K. suppose K ⊆ L is separable. the extension K ⊆ K(e) =: L has degree 2 and it is normal. for any σ ∈ G. σn are linearly dependent. Ω) = HomK(L. 2. then g is inseparable and G(L/K) = {id}. Formula (3) implies TrL/K = 0. Then HomK(L. d = 1 in (2). Ω) has n = [L : K] elements σ1. TrL/K(a + be) = 2a. ! contradicting Dedekind's lemma. b ∈ K: . if char K ≠ 2. In fact the inseparable extensions are exactly the extensions for which Tr = 0: 2. then char K = p > 0 and the inseparability degree d = [L : K]i is a power of p (3. Let K ⊆ L be a finite extension.if char K ≠ 2: 2 2 NL/K(a + be) = (a + be)(a − be) = a − b d. so σ1. If K ⊆ L is inseparable. e}. If TrL/K = 0. A basis is {1. So.294 VI. If char K = 2. HomK(L. TrL/K(a + be) = 2(a + be) = 0. Then TrL/K : L → K is the 0 mapping if and only if K ⊆ L is inseparable. σ}.27). NL/K(σx) = NL/K(x) and ! In the case of separable extensions.7 Example. TrL/K(σx) = TrL/K(x).

2.4.11). let σ := η|L ∈ HomK(L. fix a K-automorphism σ : Ω → Ω with σ |L = σ. Ω). Ω)×HomL(E. τ). ∀x ∈ E: TrL/K(TrE/L(x)) = [L : K]i·{∑σ ∈H σ {[E : L]i·∑τ ∈Gτx}} = [E : K]i·∑η∈Jηx = TrE/K(x). Moreover. J := HomK(E. Proof. τ is an L-homomorphism : ∀x ∈ L. η = Ψ(σ. Ω)×HomL(E. for any x ∈ T: NT/R(x) = NS/R(NT/S(x)).τ)∈H×G σ τx } ! This fact is proven exactly as the theorem of transitivity of finite extensions. Then: TrE/K = TrL/K◦TrE/L and NE/K = NL/K◦NE/L. There exists a bijection Ψ : HomK(L. then T is a free R-module of finite rank19 and. σ τ ( y ) = σ σ −1η ( y ) = η(y). We use 2. the proof for the norm being similar. Ψ is also injective. We have. We shall prove this for field extensions. . Ω)×HomL(E. Ω). We show that Ψ is surjective. since. Ω). defined as follows: for any σ ∈ HomK(L. Ω) are finite and have the same cardinal (see V. G := HomL(E. define Ψ(σ. Ω).2 Trace and norm 295 is a free R-module and T is a free S-module (of finite ranks).(4). τ) ∈ HomK(L. Ω). (The transitivity of the trace and norm) Let K ⊆ L ⊆ E be finite extensions of fields. τ) := σ ◦τ (which is a K-homomorphism from E to Ω since σ and τ are K-homomorphisms).VI. ∀y ∈ E. Ω). a proof for the general case uses linear algebra methods. Let H := HomK(L. Ω) and HomK(E.4. 2. Ω).3.(3). Since HomK(L. TrT/R(x) = TrS/R(TrT/S(x)).8 Theorem. We prove for the trace Tr. If η : E → Ω is a K-homomorphism. For any (σ. τ(x) = σ −1 (η ( x )) = σ −1 (σ ( x )) = x. 19 = [E : L]i[L : K]i·{∑(σ. Ω) → HomK(E. Let τ := σ −1 ◦η : E → Ω.

K) = X + an − 1 X + … + a0. where a ∈ Q. The same problem for Q ⊆ Q (3 2 ) . If K ⊆ F ⊆ L is an intermediate field. | K | = q. Determine the fixed field 3 of ϕ. ϕ = id. Is it true in general that F = K ({NL/F(x) | x ∈ L})? 7. Deduce that the norm N : L → K is surjective. Use the trace to find an element in the fixed field. Let K ⊆ L be a finite extension and let x ∈ L such that L = K(x) and n n−1 Irr(x. where L = Q(ω). where L = Q(ω). e) Calculate the cardinal of the set {x ∈ L | NL/K(x) = 1}. Let K ⊆ L be an extension of finite fields. Let p > 2 be a prime and let ω be a primitive complex pth root of unity. | L | = q . a) Write explicit formulas for NL/K(x) and TrL/K(x). 2. ∀x ∈ L. Prove that ϕ ∈ AutK(K(X)). (Hint. 4. Show that NL/Q(ω) = 1. Can the hypothesis “L is a finite field” be weakened? c) Prove that the norm NL/K is a homomorphism of multiplicative groups N : L* → K*. Write down formulas for the trace and the norm of an arbitrary element of a quadratic extension Q ⊆ Q ( d ) . (Ind. Show that NL/Q(1 − ω) = p. Let n > 1 and let ω be a primitive complex nth root of unity. 6.) . Find Irr(1 − ω.296 VI. Compute n NL/Q(a − ω). b) Prove that Tr : L → K is surjective. n Then NL/K(x) = (−1) a0 and TrL/K(x) = − an − 1. with d ∈ Z squarefree. 3. Let L/K be finite Galois. d) If α is a generator of the cyclic group L*. Q)). then N(α) is a generator of K*. Applications of Galois Theory Exercises 1. Let K be a field and let K(X) be the field of fractions of K[X]. 5. then F = K({TrL/F(x) | x ∈ L}). Let ϕ : K(X) → K(X) be the unique K-homomorphism with ϕ(X) = 1/(1 − X).

K ⊆ L is a separable extension. n) = 1}. Then L = K(ω). σω ∈ Un(L) (σ permutes the roots of n k k k X − 1). Recall some notations and facts from n IV.n and L n is a splitting field of X − 1 over K. K ⊆ L is a Galois extension and its Galois group G is isomorphic to a subgroup of (U(Z/nZ). G is Abelian and [L : K] is a divisor of ϕ(n). For any k ∈ N.3 Cyclic extensions and Kummer extensions We fix a field K. then Un(L) = {ω | 0 ≤ i < n} and clearly L = K(ω). ·). for ϕ(X) = 1/X. an algebraic closure Ω of K and we suppose that all the algebraic extensions of K are subfields of Ω. Let ω be a primitive nth root of unity in L. so |Pn(L)| = ϕ(n). if |Un(L)| = n. Let L be an extension of K. respectively ψ(X) = 1 − X. the . If L contains a primitive nth root of unity ω. VI. Determine the subfield of K(X) fixed by ϕ and ψ.1 Proposition. since (X − 1)' = nX ≠ 0 is n relatively prime to X − 1. being i a splitting field.n. A primitive nth root of unity exists in Ω ⇔ |Un(L)| = n ⇔ char K . ·). Pn(L) := {ω ∈ Un(L) | ordω = n} is the set of primitive nth roots of unity in L. The same problem. the group of invertible elements of Z/nZ. In particular.VI. If ω generates Un(L). 3. The splitting field n of X − 1 over K is also called the n-th cyclotomic field over K. The extension K ⊆ L is also normal. We study first the n splitting field of the polynomial X − 1 over K.3 Cyclic extensions and Kummer extensions 297 8.3: Un(L) := {x ∈ L | x = 1} is a finite subgroup (necessarily cyclic) of (L*. then any generator of Un(L) is called a primitive nth root of unity. Thus. (i. ω = 1 ⇔ σ(ω ) = 1 ⇔ (σω ) = 1. Suppose n ∈ N*. K is a field with char K . For any σ ∈ G. then i Pn(L) = {ω | 1 ≤ i < n. n n−1 Proof.

Irr(ω. such i that σω = ω . (i. then we can have [K(ω) : K] < ϕ(n). The extensions of the form K ⊆ K(a). ·)) m c) Irr(a. a) If K already contains some nth roots of unity. . where a has the property that a = b ∈ K for some n ∈ N. ∃! i < n. Indeed. are very important in the problem of solvability by radicals of polynomial equations. for some c ∈ K. with m | n . We obtain a function ψ : G → U(Z/nZ). Let b ∈ K* and let L = K(a).3 Proposition. if n = 8 and K = Q(i) (i = −1). n) = 1}. Since U(Z/nZ) = {i + nZ | 0 ≤ i < n. so ω = ±i. ·). and ω is a primitive nth root of unity. A suggestive writing 1n for this situation is a = n b or a = b / and an extension as above is written K ⊆ K( n b ). ψ is injective. K) = Φn (the nth cyclotomic polynomial) whose degree is ϕ(n). n) = 1} is Abelian and has ϕ(n) ele! ments. So. 2 Thus Irr(ω. (i. These extensions have a simple description if K contains a primitive nth root of unity.298 VI. K) = X − i. n n b) [L : K] = m is the order of bK* in the group (K*/K* . n) = 1. n n (K* := {x | x ∈ K*} is a subgroup in (K*. [K(ω) : K] = 2 < ϕ(8) = 4. Let K be a field containing a primitive nth root of n unity ω. 3. n 3. Applications of Galois Theory order of σω in Un(L) is equal to ordω = n. Define then ψ(σ ) = i + nZ. whence σω ∈ Pn(L). So. then G(Q(ω)/Q) ≅ U(Z/nZ). Then: a) K ⊆ L is a Galois extension and G(L/K) is cyclic of order m. K) = X − c. b) If K = Q. then (ω ) = 1. ∀σ ∈ G. |G(Q(ω)/Q)| = [Q(ω) : Q] = ϕ(n) and ψ in the proof must be an isomorphism. since ψ(σ ) = 1 + nZ implies σω = ω ⇒ σ = id. where a is a root of X − b (in Ω). the other claims are evident.2 Remark. (i. An easy check shows that ψ is a homomorphism. So. 2 2 4 2 For instance. But i Pn(L) = {ω | 0 ≤ i < n.

define ϕ : G → Z/nZ. for some c ∈ K. a subgroup of Z/nZ. Then m = |G| = ordη. then σ(a) ∈ {aω | 0 ≤ i < n}. with i < n. then K ⊆ K( n b ) is cyclic. Since b ∈ K* . A Galois extension (normal and separable) whose Galois group is cyclic (respectively Abelian) is called a cyclic extension (respectively Abelian extension). q n q n q i q q (a ) = b = c . a ∈ K. m m c) deg Irr(a. |G| divides n. thus K ⊆ L is normal. Thus. ϕ(σ ) = i + nZ. We have si m t NL/K(a) = ∏σ ∈G σa = ∏0 ≤ i < m aω = a ω ∈ K. whence a = cω . From m | q and q | m we deduce that m = q. We have q | m. so a ∈ K. Let us q n q n show that m | q. It is also separable (because char K does not divide n). So. If σ ∈ G := G(L/K). The previous proposition says: if K contains a primitive nth root of unity and b ∈ K*.VI.3 Cyclic extensions and Kummer extensions n i 299 Proof. For any i σ ∈ G. there exists a unique i < n such that σ(a) = aω . any finite cyclic extension of K is of the form K ⊆ K( n b ) (if K contains a primitive nth root of unity). We want to prove that. b = c . ! 3. G is isomorphic to Imϕ. b) Let η be a generator of G. conversely. so it is i Galois. Let q den n note the order of bK* in grupul (K*/K* . There exists s s < n such that η(a) = aω . so it is cyclic. by the theorem of Lagrange. This step is crucial in the proof of the characterization of solvability by radicals of an equation.4 Definition. Therefore. We use the following result (very important by itself): . since ϕ(σ ) = 0 + nZ implies σa = a ⇒ σ = id. m m n m m n n with t ∈ N. b = (a ) = (a ) ∈ K* . a) The roots of X − b are {aω | 0 ≤ i < n}. This shows that ordη = m divides q. We obtain ω = 1 ⇒ η = id. Thus. The function ϕ is a injective homomorphism. so a is q q q sq q sq q fixed by η: η(a ) = η(a) = a ω = a . K) = [L : K] = m. ·). and a is a root of X − a ∈ K[X].

Let G := G(L/K) and let σ be a generator of G. n−1 xn−1 := xη(xn−2) = xη(x)…η (x). …. with a ∈ K. {ω a | 0 ≤ i ≤ n − 1} are n distinct conjugates of a over K. DedeThe homomorphisms id. Consider u : L → L as an endomorphism of the K-vector space L. Setting y := σ (t) finishes the proof. there exists a ∈ L with i i i σ(a) = ω a. so 2 n η(t) = η(x0)η(z ) + η(x1)η (z) + … + η(xn−1)η (z) 2 = x1η(z ) + x2η (z) + … + z (we used that xn−1 = NL/K(x) = 1 and n η = id) We remark that η(t) = t/x0. 0 ≤ i ≤ n − 1. Let K be a field containing a primitive nth root of unity ω and let K ⊆ L be a cyclic extension of degree n. One implication is easy: if x = σ(y)/y. Then: NL/K(x) = 1 ⇔ there exists y ∈ L* such that x = σ(y)/y. since x0 ≠ 0. n−1 from L* to L* are distinct. η. Proof.5 Lemma. Proof. x1 := xη(x0) = xη(x). Applications of Galois Theory 3. ! 3. Thus. …. Then there exn ists a ∈ L*. n NL/K(ω) = ω = 1.6 Proposition. we have u ≠ 0. K) has degree ≥ n. xn−1 ∈ L. n−1 ( “the Lagrange-Hilbert resolLet u := x0·id + x1·η + … + xn−1η vent” ). By Hilbert's Satz 90. so σ (a) = ω a. published in 1897). Thus. let σ be a generator of G(L/K) and let x ∈ L*. Because This result is included as "Satz 90" in the Hilbert's monumental Bericht über die Theorie der algebraischen Zahlkörper (Zahlbericht for short. using NL/K(y) = NL/K(σy) we obtain that NL/K(x) = NL/K(σy)/NL/K(y) = 1. …. (Hilbert Satz 90)20 Let K ⊆ L be a cyclic extension of degree n. t = u(z) = x0z + x1η(z) + … + xn−1η (z). defined as follows: x0 := x.300 VI. such that L = K(a). so there exists z ∈ L such that n−1 u(z) =: t ≠ 0. η kind's lemma says they are linearly independent in the L-vector space L* L . so Irr(a. n−1 −1 Suppose now that NL/K(x) = xσ(x)…σ (x) = 1. Let η := σ and let x0. 20 . Since ω ∈ K. so x = x0 = t/η(t).

where y = −u(t). id.7 Lemma.3 Cyclic extensions and Kummer extensions n n n n n n 301 The polynomial X − a ∈ K[X] has the root a and its degree is n. so a ∈ K. defined as follows: x0 := x.6 determine all cyclic extensions of degree n of a field K of characteristic exponent p.5).3 and 3. xn−1 ∈ L. (Hilbert Satz 90. so it ! is equal to Irr(a. additive version) Let K ⊆ L be a cyclic extension of degree n. K). σ that TrL/K(z) ≠ 0. This part is not used in the theory of solvability by radicals and may be skipped in a first reading.VI. We have: 2 3 n σ◦u = σ(x0)·σ + σ(x1)·σ + … + σ(xn−2)·σ 2 3 n = (x1 − x)·σ + (x2 − x)·σ + … + (xn−1 − x)·σ 2 3 n 2 n−1 n = x1·σ + x2·σ + … + xn−1·σ + x·σ − x·(σ +σ + … +σ +σ ) = u − x·TrL/K. Thus. The cyclic extensions of degree p of a field of characteristic p > 0 are described in what follows. Let x0. x = σ(y) − y. 3. …. Let z ∈ L* such Lemma. σ. ! σ(a ) = ω a = a . The argument is essentially the same as in the multiplicative version (3. x1 := x + σ(x0) = x + σ(x).n and a K contains a primitive nth root of unity. such that p . Proof. then TrL/K(x) = TrL/K(σy) − TrL/K(y) = 0. Thus. If x = σ(y) − y. n−1 xn−1 := x + σ(xn−2) = x + σ(x) +…+ σ (x) ( = TrL/K(x) = 0) 2 n−1 Let u = x0σ + x1σ +…+ xn−2σ ∈ EndK(L). [K(a) : K] = n and K(a) = L. The propositions 3. n n . n−1 Suppose TrL/K(x) = x +σ(x) + … + σ (x) = 0. by Dedekind's n−1 are L-linearly independent. …. …. If t := z/TrL/K(z). since TrL/K(y) = TrL/K(σy). a is fixed by the subgroup <σ > = G. n−1 We have TrL/K = id + σ +…+ σ ≠ 0 since. Then: TrL/K(x) = 0 ⇔ there exists y ∈ L* such that x = σ(y) − y. let σ be a generator of G(L/K) and let x ∈ L*. then TrL/K(t) = 1 and thus σ(u(t)) = u(t) − x.

so f is irreduci! ble. whence a − a ∈ L = K. 3. G(K(a)/K) has order p. …. Because p p p p σ(a − a) = (σa) − σa = (a + 1) − a − 1 = a − a. by induction. f = X − X − b and let a ∈ Ω \ K be a root of p p f. so σa = a + 1. since n σ a = a + n·1. We have TrL/K(1) = [L : K]·1 = p·1 = 0. The proof also shows that.4). with a − a =: b ∈ K. Let σ be a generator of G := G(L/K). p Conversely. where p char p > 0. Then f is irreducible in K[X] and. a has p distinct conjugates in L. we have K(a) = L. for any b ∈ K. Then there exists p a ∈ L. namely a. Applying the additive Hilbert Satz 90. σ a = a + n·1. . K) = p = deg f. Note that it is enough to suppose that K ⊆ L is Galois. K). Then there exists a K-isomorphism σ : K(a) → K(a + 1). so Irr(a. Because |G(K(a)/K)| = p ≥ [K(a) : K]. for any root a ∈ Ω of f. K ⊆ K(a) is Galois and p = [K(a) : K] (use V. p p <σ> a − a is fixed by σ. p Conversely. Proof. σa = a + 1. having prime order. let b ∈ K. for any n ∈ N. its Galois group must then be cyclic.302 VI. 21 Otto Schreier (1901-1929). Since deg Irr(a. Then 2 n σ a = a + 2 and. p−1 σ a = a + (p − 1).4. let b ∈ K such that f = X − X − b does not split over K. Applications of Galois Theory 3.9 Remark.8 Theorem. X − X − b is either irreducible in K[X] or splits over K. so it is cyclic. the extension K ⊆ K(a) is cyclic of degree p. (Artin-Schreier21) Let K be a field of characteristic p > 0 and let K ⊆ L be a cyclic extension of degree p. German mathematician. there exists a ∈ L such that 1 = σa − a. with σ(a) = a + 1. for any n ∈ N. such that L = K(a). Then a + 1 is also a root : (a + 1) − (a + 1) − b = a − a − b = 0. we deduce that f = Irr(a. Therefore σ ∈ G(K(a)/K) and ord σ = p. Thus. K) has degree > p. Since K(a) ⊆ L and [L : K] = p.

If G is a group such that {ordx | x ∈ G} is a finite set. such that G(L/K) is an Abelian group whose exponent divides n. if K contains a primitive nth root of unity. the theory is considerably more complicated. This type of theorems are known as Kummer Theory. a theory that describes all Abelian extensions of K using only the internal structure of K is the Class Field Theory. lead to structure theorems of finite Abelian extensions of the field K. An extension is called a Kummer extension if it is n-Kummer for some n ∈ N*. (see e. the exponent of the extension is the exponent of its Galois group. 3. If K ⊆ L is a finite Galois extension. An n-Kummer extension is a finite Galois extension K ⊆ L. one of the major accomplishments of 20th century Algebra. combined with the fact that any Abelian finite group is a product of cyclic groups. 22 Ernst Eduard Kummer (1810-1893). The theory we developed so far allows us to obtain such theorems in two situations (described by the hypotheses in prop.8): .3 Cyclic extensions and Kummer extensions 303 The theorems that characterize cyclic extensions. NEUKIRCH [1986]). In some cases (when K is a “local field” or when K is a finite extension of Q).German mathematician.10 Definition. 3. Let n ∈ N* and let K be a field containing a primitive nth root of unity.for Abelian extensions of exponent n. denoted exp G.g. by Lagrange's theorem. then its exponent is defined and exp G divides |G|.VI. then the LCM of {ordx | x ∈ G} is called the exponent of G.6. respectively 3. We describe now Kummer theory.22 If K does not contain a primitive nth root of unity. . . If G is finite.for Abelian extensions of exponent p. if char K = p.

since exp G = n = |C1|·…·|Cm|. If σ ∈ G(L/K). Then L is the composite of a finite set of cyclic intermediate fields K ⊆ Li. which implies that K ⊆ L is Galois. …. …. ∀i. biω. bm}. so it is an n-Kummer extension. Proof. X − ai has n distinct roots in L: bi. 3. Let G := G(L/K). biω . Then: k k j k j+k (σ ◦τ)(bi) = σ(biω ) = σ(bi)σ(ω ) = biω ω = biω = (τ ◦σ)(bi). Proof. The finite Abelian group G is a direct product of cyclic subgroups: G = C1×…×Cm. a) Let L = K(b1. n am ).11 Lemma.304 VI. with j ∈ N depending on σ and i. But j σ(bi) = biω .12 Proposition. so L = E. Let τ ∈ G(L/K). The extension Li/K is Galois and G(Li/K) ≅ G/Hi ≅ Ci. Let L/K be an Abelian finite extension of exponent n. So. Applications of Galois Theory If K contains a primitive nth root of unity and b ∈ K*. For such an extension. with bi ∈ L and bin = ai .…. 1 ≤ i ≤ m. 3. ! n For any extension K ⊆ L and any subset S of K. bm). σ is determined by its values on the generators {b1. where S := n L* ∩ K*is a subgroup in (K*. and each degree [Li :K] divides n. . it follows that |Ci| divides n. 1 ≤ i ≤ m. 3. 1 ≤ i ≤ m. Then a finite extension K ⊆ L is n n-Kummer if and only if there exist a1. am ∈ K* \ K* such that L = K( n a1 . let n S := {x ∈ L| x ∈ S}. ·). L is a splitting field over K of the separable polynomial ∏1≤ i ≤ m(X n − ai).6 says that K( n b )/K is a cyclic extension and its degree divides n. n n −1 Since ω ∈ K. Let E be the composite L1…Lm. We have G(L/E) = H1∩…∩Hm = {id}. so Li/K is cyclic and [Li :K] divides n. …. L = K( n S ). …. Take the subgroup Hi := ∏ j≠i Cj and let Li be the fixed subfield of Hi. Let n ∈ N* and let K be a field containing a primitive nth root of unity ω. with k τ(bi) = biω .

∀x ∈ L* with x ∈ K*. The homomorphism ϕ induces an isomorphism n n n tn(L*/K*) ≅ Imϕ = (L* ∩ K*)/K* . K* is a subgroup). But L = K(b1.…. So. where each Li is a cyclic extension of K. L ⊆ K( n S ). Then the canonical group homomorphism n n n n ϕ : tn(L*/K*) → K*/K* . with bi ∈ L* and bin = ai ∈ K*. which is a subgroup in K*/K* . for some c ∈ K. since ω n di is a primitive di-th root of unity in K. let 1 be its identity element and let n ∈ N*. Then: n tn(L*/K*) = {xK* | x ∈ L*. ·) and K( n S ) ⊆ L.13 Definition. Let (H.3 and 3. 3.ϕ(xK*) = x K* . then. L = L1…Lm. …. Li = K(bi). j for some j.14 Proposition. So. In order to investigate the structure of the Galois group of an n-Kummer extension. We use this notion for an extension K ⊆ L. for the factor group L*/K* (L* is a group with respect to multiplication. By 3. bm). ! 1 ≤ i ≤ m. The last statement is justified as follows: S is a subgroup in (K*.6.6. x ∈ K*}. absurd). . bm) and clearly bin =: ai ∈ K*. Then L = K(b1. with bi ∈ L* and bidi ∈ K*. with di|n. An easy check shows that tn(H) is a subgroup in H (it is the kernel n of the homomorphism x & x ) and exp tn(H) divides n.11. bi = cω ∈ K and Li = K(bi) = K. so n σ = id and the exponent of G(L/K) divides n. Also. G(Li/K) has order di. G(L/K) is Abelian. Let K ⊆ L be an n-Kummer extension. so ai ∉ K* (if ai = c . 3. σ (bi) = biω = bi. so ai ∈ S. we need some background.VI. We have n n di > 1 (otherwise K = Li). This result generalizes 3. n Let tn(H) := {x ∈ H | x = 1}. Suppose now that K ⊆ L is n-Kummer. ·) be an Abelian group with the operation denoted multiplicatively. 3. included in L.3 Cyclic extensions and Kummer extensions n jn 305 so σ ◦τ = τ ◦σ. For the extension K ⊆ Li we can apply prop. is injective. Thus.

ϕ is well defined: if xK*∈ tn(L*/K*). so −1 n n n n n n (xy ) ∈ K* . x ∈ K* and x K* = 1K*. If exp G divides |C|.23 23 Hom(G. 3. Applications of Galois Theory n Proof. a) Hom(G. so there exists a ∈ K* such that x = a . Then x = aζ. It is immediate that ϕ preserves multiplication. The class xK* is independent on the choice of the root x of n X − y. (H. If K ⊆ L is n-Kummer. The definition of ψ is correct: ∀y ∈ L* ∩ K*. ∃x := n n n y ∈ L* such that x = y. x K* = y K* . since any other root is of the form xα.15 Lemma. C*) ≅ Hom(G.306 VI. n n n n n n n Imϕ = {x K* | x ∈ L*. Denote by 1 the neutral elements. Hence x ∈ K* . so xαK* = xK*. ! The main theorem says that the n-Kummer extensions of K are in n one-to-one correspondence with the finite subgroups of K*/K* . ·) be a cyclic group. i. with α ∈ Un(K). Let (G. C) ≅ G. then xy ∈ K*. if x. then n n G(L/K) ≅ tn(L/K) ≅ (L* ∩ K*)/K* . (C* is the multiplicative group of nonzero complex numbers). −1 y ∈ L* such that xK* = yK* ∈ tn(L*/K*). . C*) is called the dual of the group G. C*) ≅ G. stated in the following lemma. The inverse of ϕ is n n n ψ : (L* ∩ K*)/K* → tn(L*/K*). We need some group theoretical facts. So. then x ∈ K*. Let x K* = 1K* for x ∈ L* with n n n n n x ∈ K. n n n Let us show Kerϕ = {1K*}. x ∈ K*} = {x K* | x ∈ L* ∩ K*} n n = (L* ∩ K*)/K* . ·). ψ(yK* ) = n y K*. ·) be finite Abelian groups and let (C. where ζ is an nth root of unity (ζ ∈ K by hypothesis). then Hom(G.e.

Un) = G'. ∀x ∈ H} = {1}. the invariant factors theorem says that G = G1×…×Gm. C*). we have: p(ab. ∀α ∈ G'. G' = Hom(G. So. Suppose first that G is cyclic of order n and let g ∈ G be a generator. Thus. Let exp G = n. x)p(b. Un). Imα ⊆ Un. Since Cn ≅ Un. the group of complex nth roots of unity. n n a) Let G' := Hom(G. where αs(g) = ω . ∀x ∈ H is a group homomorphism. is a group homomorphism. Then s α is determined by α(g). we have Hom(G. (*) Then. px : G → C. then C includes a unique subgroup with n elements Cn (Cn is cyclic) and. If G is an arbitrary Abelian finite group. C) ≅ Hom(G1. C) p(a. A) = {ϕ : G → A |ϕ is a homomorphism} is an Abelian group with respect to the law (αβ)(x) = α(x)β(x). x)p(a. ∀a ∈ G. x) = 1. Proof. Cn). Let us prove that G' ≅ G. Cn) ≅ Hom(G. y). ∀s ∈ Z. C) = Hom(G. Hom(G. as above. cyclic of order n. Then exp G divides |C| and η : H → Hom(G. The function η : H → Hom(G. with Gi cyclic groups. x) = p(a. Define s ϕ : Z → G'. A). ϕ(s) = αs. Z/nZ ≅ G'. Then α(x) = α(x ) = 1. A quick check shows that ϕ is a surjective group homomorphism and Kerϕ = nZ. β ∈ Hom(G. which means that G' is cyclic of order n. for some s ∈ N. C) ≅ G. x). (**) . px(a) = p(a. ∀x ∈ G.3 Cyclic extensions and Kummer extensions 307 b) Let p : G × H → C be a bilinear function: for any a. C). for any x ∈ G. y ∈ H. and α(g) = ω . for any Abelian group (A. so: H ≅ Hom(G. η(x) = px. x) = 1. Hom(G. isomorphic to G. Let α ∈ G' and let ω generate Un. ∀x. c) Suppose that the bilinear function p is nondegenerate: {x ∈ H | p(a. If n divides |C|. x). xy) = p(a. ∀α.VI. b ∈ G. Note first that. ∀x ∈ G. C)×…×Hom(Gm. But Hom(G1×…×Gm. ∀a ∈ G} = {1} and {a ∈ G | p(a.·). So. C) is an isomorphism.

then there is a canonical isomorphism: tn(L*/K*) ≅ Hom(G(L/K). But exp G | m. But G and H are finite. So. m m m p(x . For any x ∈ G. Un) aK* & χa. Let x. G ≅ Hom(G. x = 1. y ∈ H and let a ∈ G. We have η(xy)(a) = pxy(a) = p(a.20) and each Hom(Gi. Thus Hom(G. for any aK* ∈ tn(L*/K*). where χb(σ) = σ (n b ) n b . Therefore. xy) = p(a. there is a canonical isomorphism n n (L* ∩ K*)/K* ≅ Hom(G(L/K). c) Ker η = {x ∈ H | η(x) = 1} = {x ∈ H | p(a. ∀x ∈ G. Un). C) ≅ G1×…×Gm = G. n n n ∀ bK* ∈ (L* ∩ K*)/K* . C) ≅ Gi. where χa(σ) = σ(a)/a. as shown. so G is isomorphic to a subgroup of H and |G| ≤ |H|. y) = 1. b) There exists an order preserving bijection between the set of all n-Kummer extensions of K included in Ω and the finite subgroups of n K*/K* : . So η is injective. n bK* & χb. Applications of Galois Theory (see II. in particular |H| ≤ |G|. The situation is symmetric in G and H. so H is isomorphic to a subgroup of G. Therefore. (multiplicative Kummer Theory) Let K be a field containing a primitive nth root of unity ω. Let m = |C|. ∀a ∈ G} = {1} (p is nondegenerate). Also. x) = 1. so |G| = |H| and G ≅ H. y) = η(x)(a)·η(y)(a). b) Condition (*) means that px is a homomorphism. a) If K ⊆ L is an n-Kummer extension.3.308 VI. ∀σ ∈ G(L/K). Because p is nondegenerate. ∀σ ∈ G(L/K).16 Theorem. ! 3. C) by a). y) = p(x. G(L/K) is isomorphic to the finite subgroup n n *n (L ∩ K*)/K* of K*/K* . x)p(a. so η(xy) = η(x)η(y). C) ≅ G. ∀y ∈ H. H ≅ Imη ≤ Hom(G.

we have n n n x = a ∈ K.3 Cyclic extensions and Kummer extensions n 309 n the extension K ⊆ L corresponds to the subgroup (L* ∩ K*)/K* .VI. we have: στ ( x ) στ ( x ) τ ( x ) ⎛ σ ( x ) ⎞ τ ( x ) σ ( x ) τ ( x ) p(στ. ψ(yK* ) = n y K*. Also. yK*). It is easily seen that p(σ. xK*) = 1. τ ∈ G(L/K) and ∀xK*. then x is fixed by any σ ∈ G(L/K). a) Let Un = Un(K) be the group of the nth roots of unity in K (we know that Un = Un(Ω) is a cyclic group of order n). xK*)p(τ. so σ(x) is a root of X − a. p(σ. xK*) .14.15. xK*) = = ⋅ =τ⎜ = ⋅ = ⎟⋅ x τ (x) x x x ⎝ x ⎠ x We used first that στ = τσ (G(L/K) is Abelian).xK*·yK*) = p(σ.15. Using ψ : (L* ∩ K*)/K* → tn(L*/K*). yK*∈ tn(L*/K*). If xK* ∈ tn(L*/K*) with p(σ. n ∀xK* ∈ tn(L*/K*). so σ = id. ∀x ∈ L with x ∈ K. Define: p : G(L/K)×tn(L*/K*) → Un. for any σ ∈ G(L/K) and any x ∈ L* with xK*∈ tn(L*/K*). with K* ≤ S ≤ K*. via the isomorphism in n n Lemma 3. 3. so it is fixed by τ. xK*) is independent on the choice of the representative of the class xK*. But n L = K({x ∈ L*| x ∈ K}) (see 3. so x ∈ K. p is bilinear. xK*)p(σ. Indeed.c). xK*) = 1. Proof. corresponds to the extension K ⊆ K( n S ). then that σ(x)/x ∈ K.xyK*) = σ(xy)/xy = (σ(x)/x)·(σ(y)/y) = p(σ. we obtain the other isomorphism. the isomorn phism in 3. thus. Thus. p is nondegenerate: if σ ∈ G(L/K) and p(σ. Finally.12). The degree of the extension is equal to the order of its corresponding subgroup. ∀σ ∈ G(L/K). xK*) := σ(x)/x. p(σ.a) implies n n p(σ. The definition is correct: ∀x ∈ L* with xK*∈ tn(L*/K*). the subgroup S/K* . Un). ∀σ. We have tn(L*/K*) ≅ Hom(G(L/K). σ(x)/x ∈ Un. then σ(x) = x. xK* = 1K*. so (σ(x)/x) = 1.

and a1. n To S/K* corresponds K( n S ). For any extension K ⊆ L. ∀xK* ∈ ∆. then K ⊆ K{ n S } = K{ n a1 . For the other inclusion. ∆ ≅ Hom(G(L/K). They are also inverse to each other. equal to L (by 3. As in a). xK*) = 1. But ∆' ≅ Hom(G(L/K). ∀x ∈ S.12). xK* ) := σ( n x )/ n x . χb(σ) = σ (n b ) n b . Thus. Since ∆ ⊆ ∆' and the groups are finite. Its associated n-Kummer extension is K ⊆ K( n S ) =: L. where ∆ := S/K* is a finite group. the subgroup corresponding to the extension K ⊆ L is finite. Applications of Galois Theory n n tn(L*/K*) ≅ (L* ∩ K*)/K* ≅ G(L/K). xK* ) = 1. so ∆ ⊆ ∆'. ∀σ ∈ G(L/K). let K* ≤ S ≤ K*. they are equal. Let n n n (L* ∩ K*)/K* =: ∆'. as we now prove. ∀bK* ∈ S/K* .c). If xK* ∈ S/K* is fixed by any with p(σ. ! Using the same methods. n n b) Any finite subgroup of K*/K* can be written uniquely as S/K* . an analogous result can be obtained for Abelian extensions of exponent p of a field K having characteristic p > 0. S ⊆ L* ∩ K*. n n Conversely. define n p : G(L/K) × ∆ → Un. (L* ∩ K*)/K* ≅ G(L/K). a finite group. Consequently. if K* ≤ S ≤ K*. so x ∈ K and x ∈ K .310 VI. Un) via the isomorphism n n *n bK & χb. n Let K ⊆ L be n-Kummer. by a). …. xK* = 1K* . From 3. n Evidently. am ∈ K* are the n representatives of the classes in S/K* . ∀σ ∈ G(L/K). then n x n n *n n σ ∈ G(L/K). n n Conversely. so σ = id. n am } is an n-Kummer extension (see 3. Un).12). ∀σ ∈ G(L/K). then n n σ( n x ) = n x . these correspondences are inclusion preserving. one shows that p is correctly defined and bilinear. we use the following notations: . Clearly. n ∀xK* ∈ ∆.15. It is n n nondegenerate: if σ ∈ G(L/K) and p(σ. We have to prove that ∆ = ∆' (⇔ S = L* ∩ K*). n n with K* ≤ S ≤ K* and [S : K* ] finite. S/K* is finite.…. p(σ. n n If K ⊆ L is n-Kummer. Its associated subgroup is S = L* ∩ K*. But L = K( n S ).

12. Then K ⊆ L is Abelian of exponent p if −1 and only if there exist a1. …. am ∈ K + \ P(K +) such that L = K(P (a1). Let K be a field of characteristic p > 0 and let K ⊆ L be a finite extension. 3. The proofs are proposed as an exercise. ! a subgroup in K +/P(K +). P (am)).19 Theorem. + tP(L+/K +) := {x + K + ∈ L+/K + | P(x) ∈ K }.VI. …. ∀x ∈ L+ cu P(x) ∈ K +. the roots of X − X − a are b. +). Fp ). For such an extension. (Additive Kummer Theory) Let K be a field of characteristic p > 0 and let Fp be its prime subfield. 3. 3. p P : Ω+ → Ω+ is the group homomorphism given by P(x) = x − x. then there exists a canonical isomorphism + tP(L+/K +) ≅ Hom(G(L/K). a subgroup in K +. Fp+ is a cyclic group of order p.14.18 Proposition. ∀x ∈ Ω. a) If K ⊆ L is Abelian of exponent p. 3. is injective. The homomorphism ψ induces an isomorphism tP(L+/K +) ≅ Imψ = (P(L +) ∩ K +))/P(K +). The following results are the “additive” versions of 3. . ψ(x + K +) = P(x) + P(K +). −1 p P (a) denotes any root in Ω of X − X − a (3. The additive homomorphism P plays the role of the multiplicative n + homomorphism x & x in the multiplicative theory.17 Proposition. ! 3. L = K(P (S)). where S := P(L+) ∩ K +. if b is a p root. Then the canonical group homomorphism ψ : tP(L+/K +) → K +/P(K +). Let K be a field of characteristic p > 0 and let K ⊆ L be a finite Abelian extension of exponent p.3 Cyclic extensions and Kummer extensions 311 L+ is the additive group (L.16. b + 1. b + (p − 1)). For any a ∈ K .8 ensures that. −1 −1 ….

4. For any n ∈ N*. Therefore G(L/K) is isomorphic to the finite subgroup (P(L+) ∩ K +)/P(K +) of K +/P(K +).312 VI. 12 2 ) and all its intermediate fields. Exercises 1. ∀ b + P(K +) ∈ (P(L+) ∩ K +)/P(K +). corresponds to the extension K ⊆ K(P (S)). Let p1. the subgroup S/K +. then Q(ζm) ∩ Q(ζn) = Q. Determine the degree of Q ⊆ Q{{ d | 1 ≤ d ≤ 20}. .…. Generalization. 3. n) = 1. …. Let ω be a primitive 12th root of unity. ∀σ ∈ G(L/K). Then Q ⊆ Q{ p1 . Fp ). Find its Galois group. ∀ a + K + ∈ tP(L+/K +). n ∈ N. if m. (m. where χa(σ) = σ(a) − a. Determine the Galois group of Q ⊆ Q(ω. where χb(σ) = σ(P (b)) − P (b). pn be distinct primes. pn } has den gree 2 . Applications of Galois Theory a + K + & χa. b) The set of Abelian extensions of exponent p of K included in Ω is in a bijective inclusion preserving correspondence with the finite subgroups of K +/P(K +): the extension K ⊆ L corresponds to the subgroup (P(L+) ∩ K +)/P(K +). −1 −1 b + P(K +) & χb. let ζn be a primitive nth root of unity. Prove that. We also have a canonical isomorphism + (P(L+) ∩ K +)/P(K +) ≅ Hom(G(L/K). ∀σ ∈ G(L/K). with P(K +) ≤ S ≤ K +. 2. −1 The degree of the extension is equal to the order of the correspond! ing subgroup.

except for the four vulgar operations. Construct a 6-Kummer extension of degree 18 of C(X. In the 16th century. the formulas for the roots of any polynomial equation of degree 3 or 4 were discovered. using only the four arithmetic operations and radicals (of any order). Independently from Ruffini (whose intricate proof failed to convince many fellow mathematicians). and one cannot sustain that transcendental operations might be involved”. ∀t ∈ K. Let K = C(X.).4 Solvability by radicals The classical problem of “solving by radicals an algebraic equation” requires expressing the roots of a given polynomial as a function of the coefficients of the polynomial. Euler24.C. who proved the impossibility of solving the “general” equation of degree 5 by radicals. Prove that K ⊆ L is a 4-Kummer extension and determine its degree and all its intermediate fields. A typical example is the formula for the roots of the equation of second degree (known to the Babylonians.VI. the Norwegian mathematician Niels 24 Leonhard Euler (1707-1783). These ideas were proven to be false at the beginning of the 19th century by Paolo Ruffini. Y). Y ]). These successes led many mathematicians of that time to think that such formulas exist for algebraic equations of any degree. . Y) (the fraction field of the polynomial ring C[X. For instance. 4 Y 3 }. n Let L = K{ 4 XY . VI. n t denotes a root of T − t (in some extension of K). about 1900-1600 B. 6. where.4 Solvability by radicals 313 5. famous Swiss mathematician. believed that “the expressions for the roots contain no other operations than radicals. around 1749.

Evariste Galois. where α = a . β. if a. Q ⊆Q ( a ) ⊆ Q( a )( 3 b − 3 a = L. αβ . If n1 = … = nm = n. The ideas and results of Galois had a decisive contribution to the development of Algebra. creates the notion of group (of permutations) and formulates a general criterion of solvability by radicals of a polynomial equation. using what we call today the Galois group of the polynomial. Abel published a memoir in which he describes a class of polynomials solvable by radicals (namely the polynomials whose Galois group is commutative). The commutative groups are called today Abelian. We say that the polynomial f is solvable by radicals over K (or that the equation f = 0 is solvable by radicals over K) if there exists a radical extension L that includes the splitting field of f over K. These definitions articulate rigorously the notion of “formula using radicals and the four operations”. 0 ≤ i < m. For instance. β . A field extension K ⊆ L is called a radical extension if there exists a finite tower of intermediate fields K = K0 ⊆ K1⊆ …⊆ Km = L. Any element of L is expressed in a basis of L 2 2 over Q (for instance 1. In 1829. cu Ki+1 = Ki(ai) and aini ∈ Ki. unaware of Abel's results. K ⊆ L is called an n-radical extension. β = 3 b − 3 a ) as an “expression using radicals and the four operations”. there exists ai ∈ Ki+1 and ni ∈ N*. such that.1 Definition. Let f ∈ K[X] and let E be the splitting field of f over K. Applications of Galois Theory Henrik Abel gave in 1824 (at the age of 22) a clear and rigorous proof of this impossibility. ) is a 6-radical extension. in his honor. for any i. two months before his death. 4. The Galois group G(E/K) is called the Galois group of the polynomial f over K. αβ.314 VI. In 1830. α. b ∈ Q. .

b. with ain ∈ Ki and Ki+1 = Ki(ai).6). where n = LCM(n1. xn).c). . for each i. the formula for the roots. a polynomial is solvable by radicals over K if and only if its Galois group over K is solvable. …. There exists a tower K = K0 ⊆ K1⊆ …⊆ Km = L and ai ∈ Ki+1. Indeed. then K ⊆ M is (n-)radical. b. ) b.c) ⊆ Q ( a . 4. We want to prove the following result (Galois' criterion of solvability by radicals): If char K = 0. m c) If K ⊆ L is such that L = K(x1. is solvable by radicals over Q(a. xi ∈ K * for some m ∈ N . Note that. c ∉ Q. The converse is false: if K ⊆ M is n-radical and K ⊆ L ⊆ M. b) The (n-)radical extensions are transitive: If K ⊆ L and L ⊆ M are (n-)radical. but K ⊆ L is not necessarily radical (see example 5. a ≠ 0.VI. …. then the equation may not be solvable over Q. c ∈ C. if a.2 = −b ± b 2 − 4ac shows that the roots xi are in the 2a 2-radical extension Q(a. nm). then L ⊆ M is n-radical. then K ⊆ L is m-radical.4 Proposition. 4.3 Remarks.4 Solvability by radicals 2 315 4.2 Examples. The normal closure of an n-radical extension is an n-radical extension. x1.b. Let K ⊆ L be n-radical and let N be its normal closure. Indeed the extension R ⊆ C is radical (why?) and C is algebraically closed. where. then K ⊆ L is n-radical. so it includes the splitting field of f over R. where a. b) Any polynomial f ∈ R[X] is solvable by radicals over R. solvability by radicals means “solvability by radicals over Q of polynomials in Q[X]”. as in the definition above. Historically. a) The equation aX + bX + c = 0. c ) ( b 2 − 4ac . 0 ≤ i < m. b. Proof. a) If K ⊆ L is a radical extension.

Since K ⊆ M is normal.19). K). where c1. It is enough to prove that M ⊆ N is n-radical (and then apply the transitivity of n-radical extensions). K). xr are n the roots of g := Irr(a. So. and N = K(x1. By induction. Let b ∈ Km −1 such that am = b.316 VI. cin = σ i ( am ) = σi(b). ct) = N is ! n-radical. Extend this homomorphism to a K-automorphism σi : Ω → Ω such that σi(am) = ci (see IV. …. xr). The following proof of Galois' characterization of solvability by radicals uses basic facts on solvable groups (any subgroup and any factor group of a solvable group are solvable. which can be found in the Appendices. 1 ≤ i ≤ r. if an extension K ⊆ L is contained in a radical extension K ⊆ E. …. For a fixed i. and N is the splitting field over K of the family of polynomials {Irr(ai. where x1. So. ct). we may assume (taking the normal closure) that K ⊆ E is radical and normal. so K ⊆ N is n-radical. with a = b ∈ K. 1 ≤ i ≤ t . 4. Since g | X − b. Applications of Galois Theory We prove by induction on m that K ⊆ N is radical. K(am) and K(ci) are K-isomorphic via an isomorphism that takes am in ci. then n L = K(a). The theorem follows by setting L = the splitting field of f over K in the statement below: .18 and n IV. we have xin = b. K ⊆ M is n-radical. let M be the normal closure of K ⊆ Km −1. Thus. The polynomial f is solvable by radicals over K if and only if the Galois group of f over K is solvable. 1 ≤ i ≤ t.5 Theorem. any finite solvable group has a normal series with the factors cyclic groups). n …. σi(b) ∈ M. ….2. cin ∈ M. N = M(c1. ∀i. Proof. and therefore M ⊆ M(c1. K)| 1 ≤ i ≤ m}. If m > 1. (Galois) Let K be a field of characteristic 0 and let f ∈ K[X]. …. K)| 1 ≤ i ≤ m − 1}. Thus. Note that M is the splitting field over K of the family of polynomials {Irr(ai.2. If m = 1. ct are the roots of h := Irr(am.

the group G := G(M(ω)/K) has the following series 1 = Hm ⊆ Hm−1 ⊆ … ⊆ H1 ⊆ H0 ⊆ G. Thus.VI.4 we may suppose that M/K is normal. since char K = 0. K ⊆ K(ω) is an Abelian extension (by 3. with ain ∈ Ki and Ki+1 = Ki(ai). We use the characterization of the extensions of the form K ⊆ K( n a ) (prop.3). there exists a sequence K(ω) = K0 ⊆ K1⊆ …⊆ Km = M(ω) and ai ∈ Ki+1. there exists a radical extension K ⊆ M with L ⊆ M if and only if G(L/K) is a solvable group. G(M/L) . since char K = 0). Ki −1 ⊆ Ki is normal. so H0 . it is separable (thus Galois) and n-radical. so Hi .3 implies that Ki ⊆ Ki+1 is cyclic. G and G/H0 ≅ G(K(ω)/K) is Abelian. By 4. In the tower of extensions K ⊆ K(ω) ⊆ M(ω). we need that ω ∈ K (ω is a primitive nth root of unity in Ω. Let Hi = G(M(ω)/Ki). Thus. G(M/K) and G(L/K) ≅ G(M/K)/G(M/L). Hi −1 and Hi −1/Hi ≅ G(Ki/Ki −1) is cyclic.1) and K(ω) ⊆ M(ω) is n-radical. Let us show that G(L/K) is solvable. Let K ⊆ L ⊆ M with L/K normal and M/K n-radical (for some n ∈ N). K ⊆ M(ω) is n-radical.4 Solvability by radicals 317 If char K = 0 and K ⊆ L is a normal extension. By Galois Theory. In order to apply this proposition. ω exists. We claim that this is a solvable series. also. Since K ⊆ M and M ⊆ M(ω) are n-radical. If we show that G(M/K) is solvable. . 0 ≤ i < m. Let us show that G(M/K) has a solvable series (a normal series with the factors Abelian groups). extensions that come up in the definition of a radical extension. 3. G(L/K) will be solvable. ∀i. Because ω ∈ Ki. Consider the extension K ⊆ M(ω). as a factor of a solvable group. 1 ≤ i ≤ m. Let us prove this claim. Therefore G is solvable. 3. ∀i. K ⊆ K(ω) = K0 is normal.

suppose G(L/K) is a solvable group.…. Applications of Galois Theory Conversely. Passing to the fraction field K(X1. So. Since F = S(X1. The extension K ⊆ L is Galois.…. Xn] → K[X1. so it is cyclic.6 Example (The general equation of degree n).…. 1 ≤ i ≤ m. there exists ai ∈ Ki+1(ω). and Hi −1/Hi ≅ G(Ki/Ki −1) is cyclic. Let ω be a primitive nth root of unity in Ω.3. 4. The extension Ki −1(ω) ⊆ Ki(ω) is Galois and G(Ki(ω)/Ki −1(ω)) is isomorphic to a subgroup of G(Ki/Ki −1) (we used V.….8). Xn) = F. Xn ∈ F. K ⊆ L(ω) is radical (and clearly L ⊆ K(ω)). −1 ∀σ.…. 1 ≤ i ≤ m. the subfield of symmetric rational fractions in F (where s1. ∀i. …. It is immediate that (στ)* = σ *τ * and id* = id. we remark that any σ ∈ Sn induces a K-algebra homomorphism σ * : K[X1. Xn).…. with ain ∈ Ki(ω) and Ki+1(ω) = Ki(ω)(ai). Thus. so the extension S ⊆ F is separable and normal. sn are the fundamental symmetric polynomials in X1. Xn).…. so they behave like indeterminates. F is the splitting field of P over S.318 VI.4. The following polynomial in the indeterminate X n n−1 n−2 n + s2 X − … + (−1) sn P := (X − X1)…(X − Xn) = X − s1 X has coefficients in S and its roots are X1. Xn] such that σ *(Xi) = Xσ(i). τ ∈ Sn. Let K be a field. To prove this.4. By 3. Let Ki be the fixed field of Hi. ∀i. K(ω) ⊆ L(ω) is radical. The polynomial P is called the generic polynomial of degree n. we obtain an isomorphism . the symmetric group on n elements. Consider S := K(s1. Xn) of rational fractions in n indeterminates. so Ki −1 ⊆ Ki is Galois. with Hi . ∀i. the equation P(x) = 0 is called the general equation of degree n. The name comes from the fact that s1. sn). since ! K ⊆ K(ω) is radical. The Galois group G := G(F/S) is isomorphic to Sn. let n ∈ N* and let F be the field K(X1. The roots of P are distinct.…. There exists a series of subgroups 1 = Hm ⊆ Hm−1 ⊆ … ⊆ H1 ⊆ H0 ⊆ G.….3). σ * is an isomorphism and its inverse is (σ )*. Hi −1 and Hi −1/Hi cyclic. sn are algebraically independent over K (see IV. 0 ≤ i < m.

4 Solvability by radicals 319 (denoted also by σ *) σ * : F → F.…. given a field K and a finite group G. for any solvable finite group G. Since σ *(si) = si.VI. Xn). Shafarevitch (proven around 1954 in several papers totaling more than 100 pages) has as a consequence that. If a1. there exists a Galois extension of Q. we obtain formulas that solve by radicals any polynomial equation of degree n ≤ 4 with coefficients in K. Thus. solvable by radicals 5 (like X = 0 …).2. More generally. 1 ≤ i ≤ n. Also note that P is irreducible over S (the root X1 of P has n = deg P conjugates X1. Sn is not solvable. an ∈ K. …. Back to our example. does there exist a Galois extension K ⊆ L whose Galois group is G? This problem is extremely difficult in the general case.….…. So. G has at most n! elements. a theorem of I. .16). σ * ∈ G(F/S). sn). In other words. The results of this type are known as Constructive Galois Theory. an. For example. replacing the “indeterminates” s1.…. Sn has n! elements. so Sn ≅ G. But there exist particular equations of degree ≥ 5. 1 ≤ i ≤ n (K[s1. σ & σ * defines a homomorphism (clearly injective) from Sn to G. there exists formulas that express by radicals the roots of P if n ≤ 4. recalling that Sn is solvable for any n ≤ 4 implies the solvability of the extension S ⊆ F (if char K = 0).…. In conclusion. given a field K. sn with a1. σ * fixes the subfield S = K(s1.R. sn] → K such that ϕ(si) = ai. On the other hand. whose Galois group is G. …. For n ≥ 5. In other words. sn] is isomorphic to the K-algebra of polynomials in n indeterminates and use the universality property of this K-algebra). as the Galois group of the splitting field of f a polynomial of degree n (see Example IV. so the general equation of degree n is not solvable by radicals over K (this is the Abel-Ruffini Theorem). we constructed a (transcendental…) extension S of K such that there exists a Galois extension S ⊆ F whose Galois group is Sn. there exists a unique K-algebra homomorphism ϕ : K[s1.

D(g) := a where a is the leading coefficient of g. xn of g. j ∈ {1.320 VI. ∀i. We use the discriminant to obtain data on the Galois group G of the polynomial g. The discriminant of g is 2n − 2 ∏i < j (xj − xi)2. For any σ ∈ G = G(L/K). assuming g has only simple roots (it is separable). 5. xn be the roots of g (in some algebraic closure Ω of K). …. xn be its (distinct) roots in L. …. g has a double root if and 2 only if the discriminant D = b − 4ac is zero.5 Discriminants. let g ∈ K[X] be a polynomial of degree n and let x1. If x1. Applications of Galois Theory VI. x2 are the roots of 2 2 g. Let L be the splitting field of g over K and let x1. xn may change ∆ in −∆. Thus. It is useful to define also ∆ := ∏i< j (xj − xi) ∈ Ω. This suggests the following general definition: 5. g has a double root if and only if D(g) = 0. K)). …. n}. ∆ is defined up to a sign. ∆ ∉ K and relabeling the roots x1. resultants The notion of discriminant of a polynomial of degree 2 is 2 well-known. …. then D = a (x1 − x2) (easy exercise). Note that D(g) is a symmetric polynomial in the roots x1. In general.2 Definition. then the discriminant of the element x over K is DK(x) := D(Irr(x. The mapping ϕ : G → Sn is an injective group homo- . a) Let K be a field. To this end. Moreover. b) If K ⊆ L is a field extension and x ∈ L is algebraic over K. This implies that D(g) ∈ K and that D(g) is independent on the labeling of the roots of g. …. let ϕ(σ) ∈ Sn be the permutation given by ϕ(σ)(i) = j ⇔ σ(xi) = xj. we see the elements of the Galois group as permutations of the roots.1 Definition. If g = aX + bX + c ∈ R[X].

if a permutation (a subgroup of permutations) has a property that is invariant under any conjugation25. If τ ∈ Sn. We say that σ is even if ε(σ) = 1 ⇔ ϕ(σ) is an even permutation in Sn. Keep the previous notations and suppose char K ≠ 2. −1 Indeed. These definitions are independent on the labeling of the roots of g.3 Proposition. Define the signature of σ.c)) and identifies G with a subgroup of Sn.15. Let g ∈ K[X] be a polynomial without multiple roots and let σ ∈ G(L/K). x'n. We say that σ is odd if ε(σ) = − 1. This shows that a reordering of the roots leads to replacing ϕ(σ) −1 with a conjugate permutation τ ϕ(σ)τ. the conjugation by τ is the automorphism κτ : Sn → Sn defined by −1 κτ(η) = τ ητ. 2 b) All σ ∈ G(L/K) are even ⇔ ∆ ∈ K ⇔ D(g) ∈ K . Also. Thus. Thus. ∀η ∈ Sn. also. …. 5. σ ∈ G(L/K) is odd ⇔ σ(∆ ) = −∆. If 1 ≤ k ≤ n. We have ϕ'(σ)(i) = j ⇔ σ(x'i) = x'j −1 ⇔ ϕ(σ)(τ(i)) = τ( j). thus ϕ'(σ) = τ ◦ϕ(σ)◦τ. ε(τ ◦ϕ(σ)◦τ) = ε(ϕ(σ)) (see above). ε(σ) := ε(ϕ(σ)). If we reorder the roots (say x'1. we say that σ is a cycle of length k if ϕ(σ) is a cycle of length k.5 Discriminants. 25 . resultants 321 morphism (see IV.3. where x'i = xτ(i) for some τ ∈ Sn) then ϕ is replaced by ϕ' : G → Sn. In other words: σ ∈ G(L/K) is even ⇔ σ(∆ ) = ∆. where ε(ϕ(σ)) denotes the signature of the permutation ϕ(σ) ∈ Sn. we define: Let σ ∈ G(L/K).VI. then this property can be transferred to the corresponding element in G(L/K) (respectively to a subgroup in G(L/K)). the conjugate of a cycle of length k is also a cycle of length k. Then: a) σ(∆ ) = ε(σ)∆.

We show now that τ *(d) = ε(τ )d. j) := ⎨ 1 if τi < τj ⎪ ⎩ −1 if τi > τj ⎧ ⎪ Note that {(Xj − Xi) | 1 ≤ i < j ≤ n} = {(Xτi − Xτj)·ε(i. ∆ ≠ −∆ and K ⊆ L is a Galois extension. Let g = a0 + a1 X +…+ an −1 X + X ∈ K[X] and let x1. …. Applications of Galois Theory c) Gal(L/K(∆)) = {σ ∈ G(L/K) | σ is even}. which means that σ is even.…. Let d := ∏i< j (Xj − Xi) ∈ K[X1. a) In our hypotheses. so τ *(d) = ∏i< j (Xτj − Xτi) = ∏i< j ε(i. …. this implies ∆ ∈ K. xn). for any σ ∈ G(L/K). Xn] → K[X1. Let σ ∈ G(L/K). σ(∆) = ∆. For any m ∈ N. the Galois correspondences take the intermediate field K(∆) to the subgroup −1 G(L/K) ∩ ϕ (An) of even elements in G(L/K). 5.322 VI.e. let n −1 n tm := x1m + … + xnm ∈ K. xn) = ε(σ)∆. j)(Xj − Xi) = ε(τ )d. j) | 1 ≤ i < j ≤ n}. In other words. b) If G(L/K) ⊆ An (i. ∆ ∈ K. Then . 2 2 2 If D(g) ∈ K . Then: σ(∆) = {ϕ(σ)*(d)}(x1. then D(g) = a . Xn]. …. define: ε(i. for some a ∈ K. so 2 2 ∆ = a ⇒ ∆ = a or ∆ = −a. Since K ⊆ L is Galois.4 Proposition. Anyway.…. But D(g) = ∆ . We need a method to compute the discriminant of a polynomial without knowing its roots. any σ ∈ G(L/K) is even). xn be the roots of g in Ω. xn) = ε(ϕ(σ))d(x1.…. so 2 2 D(g) = ∆ ∈ K . We have ∆ = d(x1. Proof. thus. ∀σ ∈ G(L/K). …. For any 1 ≤ i < j ≤ n. Xn] denote the unique K-algebra homomorphism such that τ *(Xi) = Xτi. If τ ∈ Sn. ∆ ≠ 0. let τ * : K[X1. then σ(∆) = ∆. is c) G(L/K(∆)) = {σ ∈ G(L/K) | σ(∆) = ∆} = {σ ∈ G(L/K) | σ −1 ! even} = G(L/K) ∩ ϕ (An).

The discriminant can also be computed using the resultant. − 2 p − 3q 2 p 2 t . if m > n. a0 = q. t1 = 0. t3 = − 3q. Let us compute the discriminant of a polynomial of 3 degree 3. a2 = 0. t1 = − an − 1. where A is the Vandermonde matrix: 1 A= 1 x1 … x1n-1 x2 … x2n-1 t0 t1 ! ! 1 ! * ! xn … xnn-1 The matrix A·A is the matrix in the statement. X + pX + q. −tm = an − 1 tm − 1 + an − 2 tm − 2 + … + an − m + 1 t1 + an − m m. ∆ = ∏i< j (xj − xi) = det A. Proof. We have n = 3. Viète's relations coupled with Newton's identities (see Appendix) ! yield the formulas for the tm.5 Example. −tm = an − 1 tm − 1 + an − 2 tm − 2 + … + a0 tm − n. defined below. a1 = p. 5. Thus 3 0 − 2p 3 3 2 D(X + pX + q) = 0 − 2 p − 3q = −4p − 27q .5 Discriminants.VI. if 2 ≤ m ≤ n. resultants 323 t1 … tn − 1 t2 … tn D(g) = det ! * ! tn − 1 tn … t2n−2 The sums tm can be computed by recurrence using the following relations (Newton's identities): t0 = n. We have 2 t0 = 3. and t 2 2 det( A·A) = det(A) = ∆ = D(g). t4 = 2p . Newton's identities yield: t2 = −2p.

where a. Then. then (by Proposition 5. which is cyclic of degree 3) and g is solvable by radicals. in Ω[X]. On the other hand. g = X − 3X + 1 has D(g) = 81 ∈ Q and is irreducible (g has no roots in Q and has degree 3). Since the extension is normal. but such L. g is solvable by radicals over Q because its Galois group G(L/Q) is a subgroup of S3. the splitting field of g over Q. Let K be a field and let g. We want to give an example of polynomial g ∈ Q[X] solvable by radicals over Q. Assume that the roots of g are x1. We claim that any irreducible polynomial g ∈ Q[X]. a normal extension Q ⊆ L of degree 3 cannot be radical. n ∈ N. h = b(X − y1)…(X − yn). so bω and 2 bω ∈ L. h ∈ K[X] be nonzero. Indeed. of degree 3. such that the splitting field of g over Q has degree 3. Thus. h) := a b · n m (xi − 1≤ i≤ m. if cos(2π/9) 3 g = X + pX + q ∈ Q[X] is irreducible and has the discriminant 3 2 D(g) = −4p − 27q a square of a rational number. where ω is a primitive third root of unity.…. where b ∈ L is a root of h = X − a. absurd. Applications of Galois Theory 5. Assuming Q ⊆ L is radical.7 Definition. .4.8). It follows that 2 = [Q(ω) : Q] divides 3 = [L : Q]. b ∈ K* and m. the definition 3 implies L = Q(b). The resultant of two polynomials g and h ∈ K[X] appears in the problem of finding conditions under which g and h have a common nonconstant factor (equivalently. has this property.6 Example. is not a radical extension of Q.…. More generally.1≤ j≤ n ∏ yj). ω ∈ L.324 VI.3) the Galois group of g is A3 (the alternating group on 3 elements. all roots of h are in L. 5. yn ∈ Ω. The resultant of g and h is the product Res(g. so Q ⊆ Q(ω) ⊆ L. For 3 2 instance. for some a ∈ Q. g = a(X − x1)…(X − xm). they have a common root in the algebraic closure Ω of K). xm ∈ Ω and the roots of h are y1. An example of such polynomial is the minimal polynomial of (see exercise V. which is solvable.

The following formulas hold: deg(h) (R1) Res(g. h) = 0 ⇔ g and h have a common root in Ω ⇔ GCD(g. g') = 0. where r = 0 or deg r < m = deg h. deg(r) A polynomial g has multiple roots if and only if g and its formal derivative g' have common roots. n ∈ N. r) (r is the remainder of the division of h by g: h = gq + r. h ∈ K[X] be nonzero. h) ∈ K and provides an algorithm that computes the resultant using divisions with remainder. R2 is clear from the definition. . Finally. 1 ≤ i ≤ m. Proof. keeping in mind that the GCD of g and h in Ω[X] is the same as their GCD in K[X]. h) = a n − deg(r) 1≤i ≤ m mn (R2) (R3) Res(g. using R3 and R4. ! c) By induction on min(deg g. From R1. ∀g ∈ K[X]. 0) = 0. h) = a ∏ h( xi ) Res(h. g') differ only by ± the dominant coefficient of g.5 Discriminants. D(g) and Res(g. since g(xi) = 0. R1 follows by multiplying these relations. h) ∈ K. h) is a nonconstant polynomial in K[X]. Res(0. r ∈ K[X]. Let g. n n n Res(g. Thus D(g) = 0 ⇔ Res(g. h) = a ∏i h(xi) = a ∏i (g(xi)q(xi) + r(xi)) = a ∏i r(xi). Bur Res(g. where q.8 Proposition. b) Let g = a(X − x1)…(X − xm). h) Res(g. c) Res(g. g) = Res(g. 5. g) = (−1) Res(g. Let h = gq + r. 1 ≤ i ≤ m. b ∈ K* and m. ∀b ∈ K a constant polynomial. a) Clear. deg r < deg g or r = 0) m (R4) Res(g. with a.VI. In fact. h = b(X − y1)…(X − yn). deg h). Then: a) Res(g. resultants 325 By convention. b) = b . b) We have h(xi) = b(xi − y1)…(xi − yn). r) = a ∏i r(xi) and R3 follows. R1 implies R4. The next proposition shows that Res(g. for any i.

the product above contains (xi − xj) and (xj − xi). Prove that solving the equation p(X) = 0 is equivalent to solving an equation in Y n n−2 + … + b1 Y + b0. Exercises 1. 2. Then D(g) = a (−1) Proof. then there exists b ∈ L such that L = K(b) and b ∈ K.326 VI. so 2m−1 m(m−1)/2 (−1) ∏i<j (xi − xj) 2 = a(−1)m(m−1)/2D(g). g'). and y is a root of g. such that the characteristic exponent of K does not divide n. j). by setting Y = X + a/n. where xi ∈ Ω. deg f = n. with the help of resultants one can find a polynomial that has the root x + y (see Exercise 9).9 Proposition. Deduce that a normal extension of degree p of Q cannot be radical. 3. Then g' = a∑i∏j≠i (X − xj). Let K ⊆ L be an extension of prime degree p. If x is a root of f. Let g = a(X − x1)…(X − xm). 1 ≤ i < j ≤ m. if K contains a primitive n!-th root of unity. Let g ∈ K[X] have degree m ∈ N and let a ∈ K* −1 m(m−1)/2 Res(g. Let K be a field of characteristic 0 and let f ∈ K[X] be a polynomial solvable by radicals over K. g') = a The resultant of two polynomials can be also computed by means of a determinant formed with their coefficients (see the exercises). be its dominant coefficient. then the splitting field of f over K is a radical extension of K. ! Res(g. There are m(m − 1)/2 such couples. m−1 m−1 Res(g. Applications of Galois Theory 5. By R1. Let K be a field and let p = X + an − 1 X + … + a1 X + a0 ∈ K[X]. of the form Y + bn − 2 X n n−1 . Prove that. g') = a ∏i g'(xi) = a ∏i {a∏j≠i (xi − xj)}. If the extension is p radical. For each couple (i.

(Solving the cubic equation)26 Let K be a field whose characteristic is not 2 or 3 and let the equation 3 f(x) = x + px + q = 0. but was not published. The condition uv = − p/3 implies that the roots of f are: 3 3 3 A + 3 B. an equation of degree − q ± q 2 + 4 p 3 27 3 . sive de regulis algebraicis”. 26 . q. 5. Let 2 in u is obtained. Niccolò Fontana “Tartaglia” (1500-1557) rediscovered the formula in 1535 and communicated it to Girolamo Cardano (1501-1576). ω23 A + ω3 B . r ∈ K. Verify the details in the following steps. Then 3 u = − q/2 ± γ . B = − q/2 − γ and let ω be a primitive 3rd root of unity. where γ = − D/108. ω3 A + ω23 B.5 Discriminants. Then f(x) = 0 is rewritten: 2 ⎛ x 2 + p ⎞ = − qx − r + p ⎜ 2⎟ 4 ⎝ ⎠ c) Let u ∈ K. one obtains the equation 3 3 u + v + q + (3uv + p)(u + v) = 0. a student of Cardano. Then any solution x of the equation satisfies 2 The formula was obtained around 1515. whose solutions are u 3 = 2 3 2 D = − 4p − 27q be the discriminant of the equation f = 0. resultants 327 4. where p. (Solving the quartic equation)27 Let K be a field whose characteristic is not 2 or 3 and let the equation 4 2 f(x) = x + px + qx + r = 0. q ∈ K. b) Requiring u + v + q = 0 and 3uv + p = 0.1526). where p. c) Let A = − q/2 + γ . who published it in his book “Ars Magna. b) Let q ≠ 0. leading to a solution of this equation: a) Solve the equation if q = 0. by Scipione del Ferro (? . Prove the following assertions: a) Setting x = u + v.VI. 27 The formula was found by Lodovico Ferrari (1522-1565).

7. Applications of Galois Theory 2 2 ⎛ x 2 + p + u ⎞ = − qx − r + p + u 2 + 2ux 2 + pu ⎜ ⎟ 2 4 ⎝ ⎠ The right hand side is a perfect square of some polynomial in x if and only if p2 q ⎞ − qx − r + + u 2 + 2ux 2 + pu = ⎛ x 2u − ⎜ ⎟ 4 2 2u ⎠ ⎝ d) The last equality is equivalent to 8u 3 + 8 pu 2 + 2 p 2 − 8r u − q 2 = 0 . for any d ∈ Z a positive squarefree integer. Show that g is irreducible. if ∆ = − 4p − 27q is the discriminant of g.…. n ∈ N*. 8. then Q(α. 3 3 .328 VI. Let D the square matrix in Mm + n(R[X1. where ε ∈ {− 1.…. ∆ ) is the splitting field of g over Q. Consider the polynomials in X. Let g = X + X + 3 ∈ Q[X]. Yn]: j f := a ∏1 ≤ i ≤ m (X − Xi) = ∑0 ≤ j ≤ m ajX j g := b ∏1 ≤ i ≤ n (X − Yi) = ∑0 ≤ j ≤ n bjX . Y1. Deduce that Q ⊆ Q(α) is normal if and only if ∆ is the square of a rational number. e) Let u be a solution of the cubic resolvent.….…. (The resultant as a determinant) Let R be a domain. Yn]). 1}. d ) = Q(α). Then the 4 solutions of f = 0 are: εq −u p u ± − − . b ∈ R and let m. let a. Y1. with coefficients in R[X1. and let α ∈ R be a root 3 2 of g. Xm. Let α be the real root of g and let K be the splitting field of g over Q. Let g = X + pX + q ∈ Q[X] be irreducible. Show that K ∩ Q(α. Prove that. (called the cubic resolvent of Ferrari). x =ε 2 2 2 2 2u 2 ( ) 6. Xm.

More precisely. show that det D = Res{ f. …. and let h := Res{ f.VI. View f and gZ as polynomials in X (with coefficients in K[Z]). g} = det D. the roots of h are xi + yj. y) ∈ R | ∃t ∈ R ⎨ 2 ⎩ y = y (t ) = t − t + 1 . g} ∈ R[a0. b0. This means that C = {(x. Let Z be an indeterminate and let gZ := g(Z + X) ∈ K[Z][X]. …. xn} are the roots of f and {y1. ⎡ Y1m+ n −1 … … Y1 1⎤ ⎢ ⎥ … … ⎢ ⎥ ⎢ Ynm+ n −1 … … Yn 1⎥ M = ⎢ m +n −1 ⎥ … … X 1 1⎥ ⎢ X1 ⎢ ⎥ … … ⎢ m +n −1 ⎥ … … X m 1⎥ ⎢Xm ⎣ ⎦ Computing det(MD) in two ways. 9. K) and let g = Irr(y. 1 ≤ j ≤ m. let f = Irr(x. y be algebraic elements over K. 1 ≤ i ≤ n. g} b) Prove that Res{ f. ym} are the roots of g. Prove that x + y is a root of h. The point is to prove that Res{ f. gZ} ∈ K[Z]. …. …. K). 10. am. Let C be the curve in R given by the parametric representation 2 ⎧ x = x (t ) = t 2 + t 2 . bn]. a) Let M be the Vandermonde matrix of dimension m + n. where {x1. Let K be a field and let x. t ∈ R. resultants 329 bn ⎤ ⎡ am ⎥ ⎢a a bn −1 bn ⎥ ⎢ m −1 m ! bn −1 ⎥ ⎢ ! am −1 ! ⎥ ⎢ ! ! ! am b0 ! ! ⎥ ⎢ D= ⎢ a0 b0 bn ⎥ ! ! am −1 ! ⎥ ⎢ a0 ! ! ! bn −1 ⎥ ⎢ ⎢ ! ! ! ⎥ ⎥ ⎢ a0 b0 ⎦ ⎣ There are n columns of a's and m columns of b's (the empty places contain 0's).5 Discriminants.

g(x(t)). y = y(t)}. find an implicit ∀t ∈ R.) 2 2 . (Hint.. conversely. g(x. Y] such that: ∃t ∈ R such that x = x(t). y = y(t)). y(t)) = 0. Applications of Galois Theory such that x = x(t).e. y) ∈ R . y) ∈ C ⇔ the polynomials t + t − x and t − t + 1 − y have a common root. Using resultants.330 VI. (x. a polynomial g ∈ R[X. y) = 0 implies that 2 representation of the curve (i. and. ∀(x.

we write this also I ≤ J. They include themes less likely to be included in a first course on Algebra and themes that are necessary. The reader should feel free to consult the appendices as needed.Appendices The following appendices contain a part of the background required for reading the book. 331 . where I + J denotes the sum of the ideals I and J. y ∈ I and ∀r ∈ R. An ideal I is called a proper ideal if I ≠ R. Prime ideals and maximal ideals All rings considered are assumed commutative rings with identity. we denote this by I ≤ R. j ∈ J} (the ideal generated by I ∪ J). Throughout this section. If I and J are ideals in R with I ⊆ J. we have x + y ∈ I and rx ∈ R). I + J := {i + j | i ∈ I. sup(I. inf(I. 1.e. but auxiliary to the main ideas in the book. J) = I + J. R denotes a commutative ring with identity. J) = I ∩ J. If I is an ideal of the ring R (i. The set of all ideals of a ring is a lattice with respect to inclusion: for any two ideals I and J of R. when reading the main text. ∀x.

An ideal M of R is called a maximal ideal if M ≠ R and there exist no proper ideals of R that strictly include M. (0) is its only proper ideal. Then. b) I is a maximal ideal if and only if the factor ring R/I is a field. this means that n is irreducible. a ∈ I or b ∈ I. for any J ≤ R. In the ring Z. so ab ∈ I. assume that R/I is a domain and let a. d) If K is a field. in other words. c) The ring R is a domain (has no zero divisors) if and only if (0) is a prime ideal. Proof. An ideal P of the ring R is called a prime ideal if P ≠ R and. by means of the factor ring. for any x. Since I is prime. y ∈ P. m|n implies m ∼ n or m = 1. let nZ be a maximal ideal. Let I be a proper ideal of R. then p is a prime number. If αβ = 0 ∈ R/I. nZ ⊆ mZ implies nZ = mZ or mZ = Z. Therefore. for some n ∈ Z. if pZ is a prime ideal. Here is a useful characterization of maximal (respectively prime) ideals. then (a + I)(b + I) = 0 + I. R/I has no zero divisors. In other words. b ∈ R) be elements in R/I. Since nZ ≠ Z.332 Appendices 1. 1. This implies . xy ∈ P implies x ∈ P or y ∈ P. β = b + I (where a. M ≤ J implies M = J or J = R. b ∈ R with ab ∈ I. any ideal is of the form nZ. Conversely. ∀m ∈ Z.2 Examples. Conversely. (0) is also a maximal ideal and a prime ideal. The converse is left to the reader. a) Let I be a prime ideal. b) An ideal I is maximal in the ring R iff I is a maximal element of the set (ordered by inclusion) of all proper ideals of R. Let α = a + I. a) If p ∈ Z is a prime number. is a prime ideal in Z. hence it is a prime.3 Theorem. Indeed. This implies that nZ is maximal if and only if n is a prime number.1 Definition. 1. hence a + I = α = 0 + I or b + I = β = 0 + I. then the ideal generated by p in Z. a) I is a prime ideal if and only if the factor ring R/I is a domain. denoted pZ.

Hence. Then the ideal generated by I and a. includes strictly I. x + I ≠ 0 + I. Krull's Lemma (II. so a + I = 0 + I or b + I = 0 + I. a commutative ring with identity. ! 1. 1 + I = (ra + i) + I = ra + I = (r + I)(a + I). so x + I is invertible in R/I.2. I + Ra. Thus. 1 + I = (r + I)(x + I). Any maximal ideal is a prime ideal. the prime nonzero ideals are exactly the maximal ideals and they are the ideals generated by irreducible elements. +. It follows that 1 ∈ J ⇔ J = R. there exists x ∈ J. Algebras.4 Corollary. For some r ∈ R. If R is a principal ideal domain and not a field. b) Suppose I is a maximal ideal in R. since I is maximal. I + Ra = R. because the factor ring Z [ X ] ( X ) ≅ Z is a domain. whence there exists i ∈ I such that 1 = rx + i. x ∉ I. We want to show that any nonzero element of R/I is invertible. 1 ∈ R is written i + ra. Polynomial and monoid algebras We fix (R. for some i ∈ I and r ∈ R. Algebras. Let α = a + I ≠ 0 + I. so a + I is invertible. ! The converse is false: the ideal (X) of the ring Z[X] is prime and not maximal. In particular. ·).20) says that any proper ideal is included in some maximal ideal.1. a ∈ I or b ∈ I. but not a field. . If R/I is a field and J is an ideal strictly including I. Thus. 2. Polynomial and monoid algebras 333 that (a + I)(b + I) = 0 + I. hence a ∉ I.

334 Appendices The R-algebra A is called associative (respectively with identity. An R-algebra is a ring (A. s ∈ R. a) Let A be an associative R-algebra with identity element e. if A is an associative ring with identity.1 Definition. such that. ∀r ∈ R. ∀a ∈ A. ∀r ∈ R. Proof.2 Proposition. b) Exercise. α(1) = 1e = e (since A is an R-module). 2. Thus. a ∈ A. Then the function α : R → A defined by α(r) := re. We are interested in R-algebras that are associative and have an identity. ·) (not necessarily associative or having an identity). then. which is simultaneously an R-module. Cen(A). α is an identity r ∈ R. +. ! . Also. ∀r ∈ R. α(R) ⊆ Cen(A)). a) If r. ∀a. ∀r ∈ R is an identity preserving ring homomorphism and α(r)a = aα(r). then A is an R-algebra if we define the R-module multiplication by ra := α(r)a. b) Conversely. α(r)α(s) = (re)(se) = r(e(se)) = r(se) = (rs)e = α(rs). and α : R → A is an identity preserving ring homomorphism such that α(R) ⊆ Cen(A). is the set of all elements in A that commute with any element: Cen(A) := {a ∈ A | ab = ba. ∀a ∈ A (in other words. If α(r)a = (re)a = r(ea) = ra = r(ae) = a(re) = aα(r). by definition: α(r + s) = (r + s)e = re + se = α(r) + α(s) preserving ring homomorphism. For this type of algebras there is the following characterization (often taken as the definition): 2. ∀b ∈ A} Cen(A) is a subring of A (easy exercise). b ∈ A: r(ab) = (ra)b = a(rb). Recall that the center of the ring A. commutative) if the ring A has the corresponding property.

Let A and B be two R-algebras.4 Examples. A ring homomorphism ϕ : A → B that is simultaneously an R-module homomorphism is called an R-algebra homomorphism.5 Definition. see the corresponding property for rings or for modules). ∀a ∈ C ⇒ ra ∈ C. then a ring homomorphism ϕ : A → B is an R-algebra homomorphism if and only if ϕ ◦α = β. a) The ring Mn(R) of all square matrices of type n×n with entries in R is an associative R-algebra with identity (noncommutative if n ≥ 2). Naturally.3 Remark. we consider only associative algebras with identity. It is immediate that the intersection of a family of subalgebras of A is still a subalgebra of A (for the proof. and α. A subset C of the R-algebra A is called an R-subalgebra of A if C is a subring in A and also a submodule in R A: ∀r ∈ R. the subalgebra generated by S is denoted by R[S] and it is the set of all . Algebras. For commutative R-algebras. the subalgebra generated by S as the intersection of all subalgebras of A that include S. If A and B are associative and have identities. This allows to define. b) The polynomial ring R[X] is a commutative R-algebra. Polynomial and monoid algebras 335 2. In what follows. are the structural homomorphisms.2. The structural homomorphism takes r ∈ R to the matrix having r on the diagonal and 0 elsewhere. If K ⊆ L is a field extension. Which are the structural homomorphisms (equivalently. which is the module structure) for these examples? 2. respectively β. L is a K-algebra. a ring A can have several different R-algebra structures (corresponding to different structural homomorphisms). given a subset S of A. 2. The homomorphism α : R → A in the proposition above us called the structural homomorphism of the associative R-algebra with identity A.

this product extends by (G) linearity to any two elements of R : ⎛ ⎞ ⎛ ⎞ ⎜ ∑ a g g ⎟ ⋅ ⎜ ∑ bh h ⎟ = ∑ a g bh gh ⎠ ( g .h )∈ G×G ⎝ g∈ G ⎠ ⎝ h∈ G The rigorous construction is described in the sequel. If I is an ideal of the R-algebra A. The idea is the follow(G) ing: we define on R (the free R-module on the set G) a multiplication law that is associative and distributive.1. We fix a commutative ring with identity R and a monoid (G. In particular. We shall now describe the construction of the monoid algebra80 R[G].14). An ideal I of the ring A is also called an ideal of the R-algebra A. ·). then the factor ring A/I is canonically an R-algebra. then ϕ(A) is a subalgebra of B. g∈ G where ag ∈ R. The product between g and h ∈ G (seen as elements in the basis of (G) (G) R ) is gh (element in the basis of R ). any ele(G) ment in R is written uniquely as a finite sum ∑ ag g . More precisely.336 Appendices polynomial expressions in the elements of S. Recall that (G. ·) is a monoid (or a semigroup with identity) if the operation “·” is associative and has an identity element e. This algebra is called the factor algebra of A relative to the ideal I. prop. where π : A → A/I is the canonical surjection. 80 Called group algebra if G is a group. IV. with coefficients in R (cf. we obtain the construction of polynomial algebras (in a finite or infinite number of indeterminates). for any g ∈ G and supp(ag)G is finite. If ϕ : A → B is an R-algebra homomorphism. . its structural homomorphism being π ◦α. and that coincides with the multiplication in G for the elements of G.

Therefore. Endowed with these operations. since all terms in the sum in the definition are zero. So. We must also show that ϕ +ψ and ϕ ·ψ have finite support.v) ∈ G×G such that ϕ(u)ψ(v) ≠ 0 is included in supp(ϕ)×supp(ψ). supp(ϕψ) is included in {uv | u ∈ supp(ϕ) and v ∈ supp(ψ)}. Algebras. the set of all (u. Indeed. Polynomial and monoid algebras 337 The support of a function ϕ : G → R is the set supp(ϕ) := {g ∈ G | ϕ(g) ≠ 0}. R[G] is an R-module. "+" and " · " are correctly defined and are composition laws on R[G]. then (ϕ ·ψ)(g) = 0. R[X]. which is finite. ∀r ∈ R. (r ·ϕ)(g) := rϕ(g). ∀g ∈ G. A function in R[G] is called a function of finite support. ψ ∈ R[G]. The first equality clearly defines a function ϕ +ψ : G → R.1 Remark.v )∈G×G uv = g ∑ϕ (u )ψ (v ) . The reader is encouraged to verify that if - . ∀g ∈ G. (ϕ +ψ)(g) := ϕ(g) +ψ (g) (ϕ ·ψ)(g) := ( u . ∀ϕ ∈ R[G]. Define R[G] := {ϕ : G → R | supp(ϕ) is finite}. a finite set. We define an external operation "·" : R × R[G] → R[G]. For ϕ ·ψ: if g ∈ G \ {uv |u ∈ supp(ϕ) and v ∈ supp(ψ)}. The construction we described generalizes (and is inspired from) the usual construction of the polynomial ring in one indeterminate over R. namely the free R-module of basis G (if we disregard the multiplication in R[G]).2. It is not as obvious that ϕ ·ψ is correctly defined: we must show that the sum defining ϕ ·ψ has a finite number of nonzero terms. We have: supp(ϕ +ψ) ⊆ supp(ϕ) ∪ supp(ψ). We define on R[G] the following operations: ∀ϕ. which is finite. 2.

We check the associativity of multiplication. Im i is included in the center of R[G] (so. ·) is an associative ring with identity. 2. That entitles us to write r instead of ψr (identifying r ∈ R with its image ψr ∈ R[G]). The other ring axioms are left to the reader. is an injective ring homomorphism. We also have: 2. we obtain the same thing. ∀ g ∈ G. if h = e It is clear that ηg. so (ϕψ)η = ϕ(ψη). we define the following elements in R[G]: ⎧0. if h ≠ g ∀g ∈ G. ∀r ∈ R. ∀h ∈ G.t .v )∈G 2 uv = g ⎛ ⎞ ⎜ ⎟ ∑ 2 ⎜ ∑ϕ2 (s )ψ (t ) ⎟η (v ) = ∑ϕ3(s )ψ (t )η (v ) . ((ϕψ )η )(g ) = ∑ (ϕψ )(u )η (v ) = (u . ∀h ∈ G. ⎩1. ! It is necessary to show that this construction satisfies the demands stated at the beginning of this section and. To this end. (R[G]. Proof. define ηg : G → R by η g ( h ) = ⎨ .v )∈G ⎟ uv = g ⎝ st =u stv = g ⎠ Computing (ϕ(ψη))(g). η ∈ R[G] and let g ∈ G. ·) is (N. if h = g ⎧0.338 Appendices (G. define ψr : G → R by ψ r ( h ) = ⎨ . a) The function i : R → R[G]. ∀r ∈ R. The existence of neutral elements for addition and multiplication is proven below. ψr ∈ R[G].2 Proposition. R[G] is an R-algebra of structural homomorphism i). + . then R[G] is exactly the ring R[X]. if h ≠ e ∀r ∈ R. +). given by i(r) = ψr. . ψ.t )∈G ( s . the additive monoid of the natural numbers. Furthermore.v )∈G ⎜ ( s . ⎩ r. ( u .3 Proposition. Let ϕ.

is an injective monoid homomorphism. In conclusion. Proof. h ∈ G and any r ∈ R. j(g) = ηg. for any couuv =g ψ r (u )ψ s (v ) = 0. b) The injectivity is easy. a) It is evident that ψr + s = ψr + ψs. x ≠ gh. ψϕ(g) is identified with ag = ϕ(g). if h = g d) Any element ϕ ∈ R[G] is written as a finite sum: ϕ = ∑ψ ϕ ( g )η g = ∑ a g g . ⎩ r. ∀g. ∀r. e) The zero of the ring R[G] (the neutral element for addition) is ψ0 (written as a sum of the type ∑ a g g as the sum with one term 0e). so (ψr·ψs)(e) = ψr(e)·ψs(e) = rs. g ∈ supp (ϕ ) g∈ supp (ϕ ) support). ∀g ∈ G. then. The injectivity is clear. where (ag)g ∈ G and g∈ G (bg)g ∈ G have finite support. g∈G The identity of the ring R[G] (the neutral element for multiplication) is ηe = 1e. we have u ≠ e or v ≠ e. Polynomial and monoid algebras 339 b) The function j : G → (R[G]. v) ∈ G×G such that uv = g. Therefore. We write g instead of ηg (identifying g ∈ G with its image ηg ∈ R[G]). If g ≠ e. (ηgηh)(x) = ∑η g (u )ηh (v ) = 0. ψr·ψs = ψrs. then (ψr·ψs)(g) = 0. h ∈ G. . Also. ·). s ∈ R. since uv = x ≠ gh imuv =x plies u ≠ g or v ≠ h. Also. for any g ∈ G (thus (ag)g ∈ G has finite The writing of ϕ is unique: if In the second sum. if g ≠ e. if h ≠ g c) For any g.2. we obtain ψr·ψs(g) = ∑ψ r (u )ψ s (v ) . We prove that ηgηh =ηgh. ϕ(g) is denoted ag. (ηgηh)(gh) = 1. Computing ψr·ψs. (ψr·ηg)(h) = ⎨ . ⎧0. ηg is identified with g and g∈ G ∑ ag g = ∑ bg g . ∀g ∈ G. For any x ∈ G. Algebras. then ag = bg. ple (u.

if h ≠ g . as We used in the last equality that (ψ ϕ ( g )η g ) ( h ) = ⎨ ⎩ϕ ( g ) . if h ∉ supp (ϕ ) . ⎟ ⎟ ⎜ ⎟ ⎜ ⎜ ⎠ ⎠ g∈ G ⎝ uv = g ⎠ ⎝ g∈ G ⎝ g∈ G of R having finite support.340 Appendices c) Exercise. The addition obeys the rule ∑g∈G agg + ∑g∈G bgg = ∑g∈G (ag + bg)g. ⎧0. ⎟ ⎜ ⎠ ⎝ g∈ G e) We show that ηe is the identity of the ring R[G]. ⎞ ⎛ The uniqueness ensues from ⎜ ∑ a g g ⎟(h ) = ah . R[G] meets the conditions stated in the introduction. ! 2. if h ∈ supp (ϕ ) = ϕ ( h ) ⎩ ⎧0. We can identify a ∈ R with the sum with one term a·e. R[G] is isomorphic to the free R-module of basis G. ∀h ∈ G. by c). d) For any h ∈ G. where (ag)g∈G is a family of elements (left and right) distributive relative to addition and obeys the rule (1·g)·(1·h) = 1·(gh). . if h = g seen at c). The elements of R[G] can be seen as “formal” finite sums of the form ∑g∈G agg. ∑g∈G agg = 0 ⇔ ag = 0. we can identify g ∈ G with 1·g. the sums being always finite. Any element of R[G] is written uniquely ∑g∈G agg. a) By construction. ηgηe = ηge = ηg = ηeηg.4 Remarks. For any g ∈ G. In particular. we have ⎛ ⎞ ⎜ ∑ ag g ⎟ ( h ) = ∑ ψ ϕ ( g )η g g∈ supp(ϕ ) ⎝ g∈ suppϕ ⎠ ( ) ( h ) = ⎨ϕ ( h ) . also. The general case is proven using d) and the distributivity. the multiplication is Thus. ∀g ∈ G. Therefore: ⎞ ⎞ ⎛ ⎞ ⎛ ⎛ ⎜ ∑ a g g ⎟ ⋅ ⎜ ∑ bg g ⎟ = ∑ ⎜ ∑ (au bv ) ⎟ g .

Polynomial and monoid algebras 341 b) If G is a commutative monoid. as part b) of the proposition shows. n . jn) := (i1 + j1. We make the connection with the usual form of a polynomial. Setting Denoting η1 by X. Often encountered terminologies are “unknown” or “variable” instead of “indeterminate”. The classical polynomial algebras. jn) ∈ N . For (G. 0 elsewhere). I. +). where the addition is defined component-wise: (i1. ηi = (η1) . j ∈ N. …. ∀i ∈ N. II. an arbitrary element in R[N] is a function ϕ : N → R with finite ϕ(i) =: ai.2. ·) = (N. in + jn). …. Consider the commutative monoid (N . then R[G] is not commutative. +) (for a fixed n ∈ N*). the general form of an element f in R[N] is i f = ∑ aiηi . Algebras. …. ∀(i1. then R[G] is also a commutative ring. one obtains the usual form: f = ∑i∈N ai X = a0 + a1 X + … + an X i n where n = max {i ∈ N | ai ≠ 0} is the degree of f. in) + (j1. …. ….1. n}. An element of R[N ] is called a polynomial (in n indeterminates).…. ∀i ∈ N.…. i∈ N support (a sequence of elements of R with finite support). 81 n (j1. let n ei := (0. For each i ∈ {1. n The R-algebra R[N ] is called the polynomial algebra in n n indeterminates81 with coefficients in R.0) ∈ N (1 on the i-th place. If G is not commutative. one obtains the usual construction of the polynomial algebra in one indeterminate with coefficients in R. Indeed. Since ηiηj = ηi + j. …. in). for any i.

…. if s = t ( S) es (t ) = ⎨ . with coefficients in R. +). (i + j)(s) = i(s) + j(s). ∀s ∈ S. having finite sup- n port. ∀t ∈ S and let Xs be the element ηe s ∈ R[N ].…. j ∈ N . if s ≠ t ( S) . ⎧0. define es ∈ N . Continuing the analogy with N . +) with (N . For any s ∈ S. The polynomial g is thus a sum of products between a nonzero elei ment of R and a term. is a family of elements in R. we see the elements of N ( S) ( S) n ( S) as “multi-indices” and ( S) use notations like i. of the form ai1…in X 1i1 … X nn . n where N := { f : S → N | the support of f is finite}. ⎩1. To n ( S) this end. g is a linear combination of terms.…. …. The ( S) R-algebra R[N ] is called the polynomial algebra in S indeterminates with coefficients in R.342 Appendices n as a sum of ei's (in other words. where S is an arbitrary nonempty set. The addition in the monoid N is defined naturally as follows: for any i. A product of indeterminates (of the form X 11 … X nn ) is called a Any element in N is written uniquely . then ai1…in X 1i1 … X nn is called a monomial of g.… .up to a order of the terms n term. III. The previous construction can be generalized to the polynomial algebra in S indeterminates. The axioms of a commutative monoid are readily checked. n The element ηei ∈ R[N ] is denoted by Xi and is called an indetermii i nate. Any polynomial g ∈ R[N ] is then written uniquely as a finite sum: n g= where ( ai1…in ) i . Thus. R[N ] is denoted usually by R[X1. en} generate the monoid N ). Xn]. i (1 ) ∈ 'n ( i1 . If g is like above i and ai1…in ≠ 0. j. we replace the monoid (N . {e1.in )∈' n ∑ i ai1…in X 1i1 … X nn .

( S) Therefore.…. ηi = ∏ X sms . if i∈F ( S) ∪i∈F supp(i) =: {s1. ·). where (ms)s∈S s∈S 82 is a family of natural numbers indexed by S. where F is a finite subset of N . ∀s ∈ S. S]. The R-algebra R[N ] is usually denoted as R[(Xs)s∈S] or R[Xs]s∈S or R[X. ·) be a monoid and let i : R → R[G]. n 1 (the (am …m )(m . …. In other words. j : G → R[G] be the canonical mappings defined at 2.mn )∈N ∑ n am1…mn X sm1 … X smn . Algebras. then f= where 1 ( m1 . mi is defined as (mi)(s) := m·i(s). any polynomial in S indeterminates is a polynomial ( S) in a finite set of indeterminates in S. j) has the following universality property: for any R-algebra T having the structural homomorphism α : R → T and any homomorphism of monoids β : G → (T.2. there exists a unique homomorphism of R-algebras ϕ : R[G] → T such that ϕ ◦i = α and ϕ ◦j = β: j G⎯ ⎯→ R[G ] β ϕ T 82 For any m ∈ N and i ∈ R[N ]. Polynomial and monoid algebras 343 Any element i ∈ N ( S) is written uniquely as i = ∑ ms e s . i. sn} (a finite subset of S). (The universality property of the monoid algebra) Let (G. having finite support. (S) . 2.3. An arbitrary polynomial f ∈ R[N ] has a s∈supp ( i ) unique writing as f = ∑ aiηi . The triple (R[G].m )∈N n the 1 sum n is finite family of elements of R n has finite support).….5 Theorem.

Thus. j). ∀i ∈ {1. A direct check shows that ϕ given by the above is a ring homomorphism and satisfies the stated requirements. this theorem reads: 2. For the classical polynomial algebras. ∀g ∈ G. ∀s ∈ S. ⎟ ⎠ g∈ G g∈G ⎞ which shows that ϕ is uniquely determined by α and β. If ∑g∈G agg ∈ R[G]. For any n-uple a = (a1. ·) is a homomorphism of monoids) satisfies the same universality property as (R[G]. S]) Let S be a nonempty set. . For any mapping γ : S → A there exists a unique R-algebra homomorphism evγ : R[X. a) (The universality property of the polynomial algebra R[X]) For any a ∈ A there exists a unique R-algebra homomorphism eva : R[X] → A such that eva(X) = a.an) ∈ A there exists a unique R-algebra homomorphism eva : R[X1. S] → A such that evγ(Xs) = γ(s). γ : R → A is its structural homomorphism and δ : G → (A.7). ∀r ∈ R and ϕ(g) = β(g). Suppose ϕ is as stated. ! The universality property of the monoid algebra determines this algebra up to an unique isomorphism: if (A.…. i. using the remarkable fact that the monoid (N. Let A be an R-algebra. +) is generated by n elements). Xn] → A such that eva(Xi) = ai. δ) (where A is an R-algebra. b) (The universality property of the polynomial algebra R[X1. c) (The universality property of the polynomial algebra R[X. γ.344 Appendices Proof. +) is generated by one element (respectively n (N .….….….6 Theorem. then there exists a unique R-algebra isomorphism ϕ : R[G] → A such that ϕi = γ and ϕj = δ (cf. n}. ϕ(r) = α(r). then ϕ⎜ ⎜ ⎛ ⎝ g∈ G ∑ a g g ⎟ = ∑ϕ (a g )ϕ ( g ) = ∑α (ag )β ( g ) . n Xn]) Let n ∈ N*. 5.

in )∈N n a = (a1. (f·g)(a) = f(a)·g(a). and any (i1 . a) For any f = i =0 ∑ bi X i ∈ R[X]. j(es) = Xs. there exists a unique R-algebra homomorphism evγ : R[X. vγ (Xs) = β(es) = γ(s). …. c) The mapping γ induces a homomorphism of monoids i(s ) ( S) ( S) β : N → (A.2. then v◦j = β. in our case. ….in )∈Nn ∑ i i bi1…in a11 … ann . ! . ∀f. Nevertheless. The uniqueness part of the universality property implies v = evα. we sketch a proof for every case. an) ∈ A . where β is the Let us prove the uniqueness of evα. let eva( f ) = n i =0 ∑ bi a i n (usually called the value of f at a and denoted f (a)).….S] → A such that vγ ◦j = β. let n eva( f ) = f (a1. ∀s ∈ S.…. (f + g)(a) = f(a) + g(a). ·). Xn]. an ) = the value of f at a = (a1. Algebras. Polynomial and monoid algebras 345 Proof. Of course. The fact that eva is a homomorphism amounts to that. ∀s ∈ S. where j : N → R[X.…. If v : R[N ] → A is an homomorphism defined above. ∀i ∈ N s∈supp ( i ) Applying the universality property of the monoid algebra ( S) R[N ] = R[X. A standard check proves these familiar properties and the rest of the claims. an). (i1 . β(i) = ∏ γ ( s ) . i b) For any f = ∑ bi1…in X 1i1 … X nn ∈ R[X1. g ∈ R[X]. S]. S] is the ( S) canonical mapping.… . a) follows from b). ( S) Thus R-algebra homomorphism with v(Xs) = γ(s). which is a particular case of c).

Similarly. Of course.…. In what follows we discuss the important notion of degree of a polynomial. Indeed. 1 ≤ i ≤ n. 1 ≤ i ≤ n − 1.…. A useful property of the R-algebras R[X1.…. ϕβ = id.….…. Proof. Xn] has the same properties. R[X1. so ϕ is an ! isomorphism. Xn]. Xn] → R[X1.b): there exists a unique R-algebra homomorphism ϕ : R[X1.7 Theorem. Xn−1][Xn]. Xn]. .b) implies that βϕ = id. Xn] ≅ R[X1. βϕ : R[X1.…. Use 2. Xn]. with ϕ(Xi) = Xi. The universality property of the A-algebra A[Xn] shows that there exists a unique A-algebra homomorphism β : A[Xn] → R[X1.…. The previous theorem formalizes and gives a precise meaning to the procedure of assigning the values a1.…. Xn] → R[X1. Xn. We have βϕ = id. …. Conversely. The uniqueness part in 2.…. Xn] is an A-algebra with α its structural homomorphism. Xn−1] → R[X1.6.b) applied to R[X1. n If aX is a monomial in R[X] (with a ≠ 0).…. yields a unique R-algebra homomorphism α : R[X1. Xn] is an R-algebra homomorphism such that βϕ(Xi) = Xi. sometimes used to define them by recurrence on n. Xn−1][Xn]. Thus. such that β(Xn) = Xn.…. Xn] → R[X1.…. with α(Xi) = Xi. is the following: 2. Let A := R[X1. β is also an R-algebra homomorphism. 2. For an arbitrary polynomial g ∈ R[X]. Let n ≥ 1. and id : R[X1.…. …. n is called the degree of n aX .346 Appendices The homomorphism eva (respectively eva) above is called the evaluation homomorphism.…. Xn−1]. 1 ≤ i ≤ n. Then there exists a canonical isomorphism of R-algebras: R[X1. Xn−1].6.…. an to the indeterminates X1.6.

Xk) = deg (g.….…. the natural number n is called the degree of g. Sometimes deg 0 is not defined. deg (gh. if R is a domain. X k ) := ik (the expo- i i If aX 11 … X nn is a monomial in R[X1. Xk) + deg (h. Thus.…. For any g ∈ R[X1. and i nent of Xk in the monomial). the degree of g is the greatest degree of the monomials of g. then its degree in Xk is deg ( aX 1i1 … X nn . deg (g. unless otherwise specified. deg h). Xk). Polynomial and monoid algebras n 347 g = a0 + a1X + … + anX .2. The elements a0. If the leading coefficient an is 1. Xn] (where a ≠ 0). a0 is called the constant term. then the degree is additive: ∀g. Xn]. Xn] is the largest total degree of its monomials. where an ≠ 0. the total degree of an arbitrary polynomial g ∈ R[X1. then the polynomial g is called monic.…. also denoted deg X k ( aX 1i1 … X nn ) . Xn]. …. i i The total degree of the monomial aX 11 … X nn is i1 + … + in. i 1 ≤ k ≤ n. Usually. an is called the leading coefficient of g. If R is a domain. deg (h. A polynomial whose monomials have all the same degree is called a homogeneous polynomial. We define by convention deg 0 = −∞. Algebras. deg (g + h. deg (gh) = deg g + deg h. If R is a domain. h ∈ R[X1. Xk). . an ∈ R are called the coefficients of the polynomial g. the “degree” of a polynomial in several indeterminates is its total degree. h ∈ R[X1. among these. Xk) is the greatest degree in Xk of the monomials of g. Xk)). Also: deg (g + h) ≤ max(deg g. then the degree in Xk satisfies the same relations as above: ∀g. Xn]. denoted deg g. Xk) ≤ max(deg (g.…. The total degree satisfies relations similar to the above.

348

Appendices

3. Symmetric polynomials
Let R be a fixed commutative ring with identity. Let n ∈ N* and let σ ∈ Sn, the symmetric group of all permutations of n objects. Then R-algebras there exists a unique homomorphism of ϕσ : R[X1,…, Xn] → R[X1,…, Xn] such that ϕσ(Xi) = Xσ(i), ∀i = 1,…, n (see 2.6, the universality property of the polynomial R-algebra R[X1,…, Xn]). If g ∈ R[X1,…, Xn], then ϕσ(g) = g(Xσ(1),…, Xσ(n)). If R is a domain and K is its field of fractions, consider K(X1,…, Xn) (the field of fractions of the domain R[X1,…, Xn], called the field of rational fractions in the indeterminates X1,…, Xn with coefficients in K). Then ϕσ extends to a unique field homomorphism (denoted also ϕσ) ϕσ : K(X1,…, Xn) → K(X1,…, Xn). For any g, h ∈ R[X1,…, Xn], h ≠ 0, ϕσ(g/h) = ϕσ(g)/ϕσ(h).
3.1 Definition. Let g ∈ R[X1,…, Xn]. We call g a symmetric polynomial in R[X1,…, Xn] if, for any σ ∈ Sn, we have ϕσ(g) = g. If R is a domain and K is its field of fractions, a rational fraction g/h ∈ K(X1,…, Xn) is called symmetric if, for any σ ∈ Sn, ϕσ(g/h) = g/h. 3.2 Example. In R[X1, X2, X3], the following polynomials are symmetric: X1 + X2 + X3, X1 X2 X3, 2 2 2 2 2 2 X1 X 2 + X1 X 3 + X 2 X1 + X 2 X 3 + X 3 X1 + X 3 X 2 . The polynomial X1 + X2 is not symmetric in R[X1, X2, X3] (but it is symmetric in R[X1, X2]). 3.3 Remarks. a) Let S = {g ∈ R[X1,…, Xn] | ϕσ(g) = g, ∀σ ∈ Sn} be the set of symmetric polynomials. Then S is an R-subalgebra of R[X1,…, Xn]. For instance, if g, h ∈ S, then

3. Symmetric polynomials

349

The other conditions are checked similarly. It is easy to see that the symmetric rational fractions form a subfield in K(X1,…, Xn). i i b) If aX 11 … X nn is a monomial of the symmetric polynomial g,
i i then aX σ1 (1) … X σn( n ) is also a monomial of g, for any σ ∈ Sn.

ϕσ(g + h) = ϕσ(g) + ϕσ(h) = g + h, ∀σ ∈ Sn.

sk := ∑{∏i∈I Xi | I ⊆ {1, …, n}, |I| = k}. is called the fundamental (or elementary) symmetric polynomial of degree k in R[X1,…, Xn] .

3.4 Definition. Let n ∈ N* and let 0 ≤ k ≤ n. The polynomial

In other words, sk is the sum of all products of k distinct indeterminates chosen among {X1,…, Xn}; thus sk has n monomials. By k

()

convention, s0 = 1 and sk = 0 if k > n. The polynomial sk is homogeneous of degree k (indeed, all its monomials have degree k). Since sk obviously depends on the number of indeterminates, the notation sk(X1,…, Xn) is sometimes used to avoid confusions. For example, the fundamental symmetric polynomials in 4 indeterminates are: s0 = 1 s1 = X1 + X2 + X3 + X4 s2 = X1 X2 + X1 X3 + X1 X4 + X2 X3 + X2 X4 + X3 X4 s3 = X1 X2 X3 + X1 X2 X4 + X1 X3 X4 + X2 X3 X4 s4 = X1 X2 X3 X4 The fundamental symmetric polynomials appear in the relations between the coefficients of a polynomial and its roots (Viète's relations).
3.5 Theorem. a) Let n ∈ N* and let sk = sk(X1,…,Xn). R[X1,…, Xn][X] the following relation holds: n n−1 n−2 n (X − X1)…(X − Xn) = X − s1 X + s2 X − … + (−1) sn.

In

350

Appendices

b) If R is a subring of the domain S and n g = a0 + a1X + … + anX ∈ R[X] has the roots x1, …, xn ∈ S, then k ansk(x1, …, xn) = (–1) an − k. Proof. a) Induction on n (exercise). b) There exists a unique R-algebra homomorphism ϕ : R[X1,…, Xn][X] → S[X] such that ϕ(Xi) = xi and ϕ(X) = X. We have, by a): ϕ(an(X − X1)…(X − Xn)) = an(X − x1)…(X − xn) = n n−1 n−2 n an(X − s1 X + s2 X − … + (−1) sn). But an(X − x1)…(X − xn) = g (in the field of fractions K of S, these polynomials have the same roots and the same leading coefficient). ! The relations follow by identifying the coefficients.
3.6 Theorem. (The fundamental theorem of symmetric polynomials) Let g be a symmetric polynomial in R[X1,…, Xn]. Then g is a polynomial of the fundamental symmetric polynomials: there exists a unique polynomial h ∈ R[X1,…, Xn] such that g = h(s1, …, sn). In other words, denoting by S the R-subalgebra of symmetric polynomials in R[X1,…, Xn], the unique R-algebra homomorphism ψ : R[X1,…, Xn] → S with ψ(Xi) = si (for 1 ≤ i ≤ n) is an isomorphism. i i Proof. Let T := {X 11 … X nn (i1,…, in ) ∈ N n } be the set of all terms in R[X1,…, Xn]. Define a total order relation on T (the lexicographic ori k i der) by: X 11 … X nn ≤ X 1k1 … X n n ⇔ ∃r, 1 ≤ r ≤ n, such that it = kt, ∀t < r and ir ≤ kr. For instance, we have X1 > X2 >… > Xn and 2 1 < X 37 < X 2 X 32 < X 1 < X 1 X 2 . This order relation is total and it is compatible with term multiplication: ∀λ, µ, ν ∈ T, µ ≤ ν implies λµ ≤ λν (one can prove that it is the unique total order on T, compatible with multiplication, such that X1 > X2 >… > Xn). Moreover, T is well ordered by the lexicographic order (any nonempty subset of T has a smallest element), as the following lemma shows. Therefore, one can make induction proofs on this ordered set (as is this proof).

3. Symmetric polynomials

351

The lexicographic order induces a preorder relation83, denoted also " ≤ ", on the set {aλ | λ ∈ T, a ∈ R, a ≠ 0} of all monomials in R[X1,…, Xn], by aλ ≤ bµ ⇔ λ ≤ µ. The proof of this is straightforward. If p ∈ R[X1,…, Xn], there exists a unique greatest monomial of p (with respect to the lexicographic preorder), called the leading monomial of p and denoted by lm(p). The following property holds: If p, q ∈ R[X1,…, Xn], such that lm(p) = aλ, lm(q) = bµ, where λ, µ ∈ T, a, b ∈ R and ab ≠ 0, then lm(pq) = lm(p)lm(q) = abλµ. Indeed, any monomial of pq is a sum of monomials of the form forma rα·sβ, where rα is a monomial of p and sβ is a monomial of q. But α ≤ λ and β ≤ µ, so αβ ≤ λβ ≤ λµ. Thus, abλµ = lm(pq). We proceed to the proof of the theorem. Let g be a symmetric i i polynomial and let lm(g) = aX 11 … X nn . Then i1 ≥ i2 ≥ …≥ in (if not, i i i i then there exists a k such that ik < ik+1, and aX 11 … X kk +1 X kk+1 … X nn is also a monomial in g, strictly greater than lm(g), contradiction). We want to find a polynomial p of the form as1j1 … snjn , such that lm(p) = lm(g). Using the above property, j j lm ( as1j1 … snjn ) = aX 1j1 ( X 1 X 2 ) 2 … ( X 1 … X n ) n This monomial equals lm(g) if and only if j1 + … + jn = i1, j2 + … + jn = i2, …, jn = in. Thus jn = in, jk = ik − ik + 1, for 1 ≤ k < n. The polynomial g1 := g − as1j1 … snjn is symmetric and lm(g1) < lm(g). If lm(g1) = 0, then g1 = 0 and we are finished. If lm(g1) ≠ 0, replace g by g1 and apply the procedure above. The algorithm terminates in a finite number of steps because any strictly decreasing sequence of terms must be finite, as the next lemma shows. This concludes the existence part of the proof. Let us prove the uniqueness (or, equivalently, Kerψ = 0). Suppose there exists a nonzero polynomial p ∈ R[X1,…, Xn] such that
83

A relation that is reflexive and transitive, but not necessarily antisymmetric.

352

Appendices

ψ(p) = p(s1, …, sn) = 0. We claim that there exists a unique nonzero monomial λ of p such that lm(ψ(p)) = lm(λ(s1, …, sn)). Indeed, if i i α = X 11 … X nn and β = X 1j1 … X njn are distinct terms, then:
i i lm(α(s1, …, sn)) = X 11 +…+in … X nn ≠ X 1j1 +…+ jn … X njn = lm(β(s1, …, sn)).

Thus, there exists a unique nonzero monomial λ of p such that lm(λ(s1, …, sn)) = max {lm(α(s1, …, sn)) | α is a monomial of p}. Since p(s1, …, sn) = ∑{α(s1, …, sn) | α is a monomial of p}, we lm(p(s1, …, sn)) = lm(λ(s1, …, sn)) ≠ 0, contradicting that have p(s1, …, sn) = 0. !
3.7 Lemma. a) Let (A, ≤) and (B, ≤) be well ordered sets. Then A×B is well ordered by the (lexicographic) order defined as (a, b) ≤ (a', b') if and only if a < a' or (a = a' and b ≤ b'). b) In a well ordered set (A, ≤) there exist no infinite strictly decreasing sequences. c) For n ∈ N, the set Tn of the terms in R[X1,…, Xn] is well ordered by the lexicographic order (thus any strictly decreasing sequence of terms must be finite). Proof. a) Recall that the ordered set (A, ≤) is called well ordered if for any nonempty subset S of A, ∃α ∈ S such that α ≤ a, ∀a ∈ S (α is unique with this property and is called the smallest element of S. Thus, A is well ordered if any nonempty subset has a smallest element). Let ∅ ≠ S ⊆ A×B. Since S1 := {a ∈ A | ∃b ∈ B cu (a, b) ∈ S} ≠ ∅, and A is well ordered, there exists its smallest element α ∈ S1 (so, ∀(a, b) ∈ S, α ≤ a). Let S2 := {b ∈ B| (α, b) ∈ S}. There exists the smallest element β of S2. Then (α, β) is the smallest element of S: ∀(a, b) ∈ S, we have α < a (thus (α, β) < (a, b)) or α = a, in which case b ∈ S2, so β ≤ b. b) Let (an)n ≥ 1 be a decreasing sequence of elements in A. Then the set {an | n ≥ 1} has a smallest element ak. For any n ≥ k, we must have then ak ≤ an; since an ≤ ak (the sequence is decreasing), an = ak and the sequence is not strictly decreasing.

3. Symmetric polynomials
n

353

c) Induction on n. If n = 1, T1 = {X | n ∈ N} is isomorphic as an ordered set to (N, ≤), which is well ordered. If n > 1, then Tn, ordered lexicographically, is isomorphic to Tn−1 × T1 with the order defined as in a). By induction, Tn−1 is well ordered and, by a), Tn−1 × T1 is well ordered.

The theorem 3.6 extends easily to symmetric rational fractions.
3.8 Corollary. (The fundamental theorem of the symmetric rational

fractions) Let R be a domain and let K be its field of fractions. If p,
q ∈ R[X1,…, Xn], q ≠ 0, are such that p/q is a symmetric rational fractions, then there exist the polynomials f, g ∈ R[X1,…, Xn] such that p f ( s1 ,… , sn ) . In other words, the subfield of the rational symmet= q g ( s1 ,…, sn ) ric fractions of the field K(X1,…, Xn) is K(s1,…, sn).
Proof. If q is a symmetric polynomial, then p is symmetric, being the product q·(p/q) in the subfield of the symmetric rational fractions. From 3.6, it follows that p, q ∈ R[s1,…, sn]. If q is not symmetric, let s = ∏σ ∈Snϕσ(q). Then s is symmetric and p p ∏σ ≠id ϕσ (q ) = , q s

and we are in the conditions of the first case.

!

The symmetric polynomial tm := X 1m +…+ X nm ∈ R[X1,…, Xn] (m ∈ N) is expressible using the fundamental symmetric polynomials s1, …, sn. The following identities allow a recursive computation of tm as a polynomial of s1, …, sn.
3.9 Proposition. (Newton's identities) Let m ∈ N. In R[X1,…, Xn] the following relation holds: m−2 m−1 sm − 1 t1 + (−1) msm. tm = s1 tm − 1 − s2 tm − 2 + … + (−1)

354

Appendices

Proof. If m > n, the convention sk = 0 for k > n truncates the formula above (there are only n terms). Let r ≤ n and let (a1, …, ar) be an r-uple of natural numbers with a1 ≥ a2 ≥ … ≥ ar. Let s(a1, …, ar) be the unique symmetric polynomial a in R[X1,…, Xn] having the leading monomial X 1a1 X 2 2 … X rar . s(m, 0, …, 0) = X 1m +…+ X nm = tm, For instance, s(1, 1, 0, …, 0) = X1 X2 + X1 X3 + … = s2. To simplify notations, let 1i := (1, …, 1) (1 appears i times) and (a, 1i) := (a, 1, …, 1) (1 appears i times); also, we omit writing a sequence of 0's: s(m, 0, …, 0) = s(m), s(1, 1, 0, …, 0) = s(1, 1) = s2, s(1i, 0, …, 0) = s(1i) = si. The following relations can be verified easily: s1 tm − 1 = tm + s(m − 1, 1) s2 tm − 2 = s(m − 1, 1) + s(m − 2, 1, 1) s3 tm − 3 = s(m − 2, 1, 1) + s(m − 3, 1, 1, 1) … More generally, for any i ≤ min{m − 1, n}, si tm − i = s(m − i + 1,1i) + s(m − i,1i). If m ≤ n and i = m − 1, then sm − 1t1 = s(2,1m − 2) + msm. If m > n = i, then sn tm − n = s(m − n + 1,1n − 1). Newton's identities follow by using the relations above in the sum ∑1≤ i< m (−1)i−1si tm−i. !

4. Rings and modules of fractions
The method of construction of the field Q from the domain Z generalizes naturally to any commutative ring R (although in general

A subset S of R is called a multiplicatively closed set (or a multiplicative set) if: a) 1 ∈ S. s) | a ∈ R. (b. 4. and the product of two denominators is also a denominator. n For example. t) ∈ R × S. In what follows. 4. all nonzero elements in Z become invertible. b) 0 ∉ S. It is thus natural to define a concept corresponding to the notion of “set of denominators”. s ∈ S} we define the relation: ∀(a.t) ⇔ ta − sb = 0. (a. b) ∀s.3 Remark. such that the images by ϕ of all elements in S are invertible in T (T is thus an R-algebra and ϕ is its structural homomorphism). The reason for adopting the . there is no need that all nonzero elements in a ring R become invertible in some “extension” of R. s). 0 cannot be one. {2 | n ∈ N} are multiplicatively closed sets in Z. s). In the classic case of the construction of Q we have: R = Z. In Q. all rings are commutative with identity and R denotes such a ring. s) ∼ (b.1 Definition. Z \ 2Z. S = Z* and the following relation is used: ∀(a. These conditions are natural for a “set of denominators”: 1 must be a denominator. On the set R × S = {(a.4. 4.2 (prove this!). we construct a ring T and a ring homomorphism ϕ : R → T. For a given ring R and a multiplicatively closed set S ⊆ R. Z \ {0}. t ∈ S ⇒ st ∈ S. t) ∈ R × S. t) ⇔ ∃u ∈ S such that u(ta − sb) = 0. Rings and modules of fractions 355 the result will not be a field).2 Definition. s) ∼ (b. define: (a. (b. In many cases. In this case the relation obtained coincides with the one in definition 4.

so (a. s ∈ S}.t) and (b. Proof. Let (a. s The factor set R × S/∼ (the set of all equivalence classes) is denoted by S R: It easy to see that −1 −1 S R := { a/s | a ∈ R. t) ∼ (c. t) ∈ R × S | (b. t). 4. s) ∼ (b. s). s) ∈ R × S.u). s) a with respect to "∼" is denoted by or a/s and is called a fraction or a s quotient (of denominator s and numerator a). (c. (b.4 Proposition. s) ∼ (c. a ta = . having in mind the usual rules of addition and multiplication of fractions. The equivalence class of (a. define: a b ta + sb + := s t st a b ab ⋅ := s t st . Then ∃v. For any (a. u) ∈ R × S such that (a. s). s)}. We prove only the transitivity. vwt ∈ S. (b. 4.2 is to handle the more general case when the multiplicatively closed set S possibly contains zero divisors. w ∈ S such that v(ta − sb) = 0 and w(ub − tc) = 0. We obtain uwvta − uwvsb + svwub − svwtc = 0 ⇔ vwt(ua − sc) = 0. s ts −1 We define on S R two operations.t ∈ S. The relation " ∼ " is an equivalence relation on R × S.u). Thus: a = a/s = {(b. ∀s. t) ∼ (a.5 Definition.356 Appendices definition 4. ! Since S is multiplicatively closed. Multiply the first equality by uw and the second by sv and add. Let (a. ∀a ∈ R. t) ∈ R × S.

Indeed. Let (a. −1 −1 . s). t') ∈ R × S. The operations defined above on S R are cor−1 rectly defined and S R becomes a commutative ring with identity. the homomorphism ϕ is injective ⇔ S contains no zero divisors. called the canonical homomorphism (thus S R is an R-algebra). t) ∼ (b'.6 Proposition. Rings and modules of fractions −1 357 4. is a ring −1 homomorphism. S R is a field. Multiply the first of these equalities by tt'v and the second by ss'u and add them. 0/1 = a/s. s't'). st) ∼ (t'a' + s'b'. then S R is the zero ring (with only one element. for this reason the condition 0 ∉ S is imposed in the definition of a multiplicatively closed set. The rest of the proof (the multiplication is correctly defined. Moreover. 1 s −1 The mapping ϕ : R → S R. ϕ(a) = a/1. such that (a. ! Note that any s ∈ S is taken by ϕ into an invertible element in −1 S R: ϕ(s) = s/1 has the inverse 1/s ∈ S R. Proof. (b'. s') and (b. s'). and denoted Q(K). s) ∼ (a'. (a'. ∀s ∈ S. ∀a ∈ R. t). We obtain vu((ta + sb)s't' − (t'a' + s'b')st) = 0. ∀s ∈ S). We must show that (ta + sb. −1 If 0 ∈ S. Let u. In the important case when R is a domain and S = R \ {0}. (b. check−1 ing the axioms for the ring S R) is left to the reader. a/1 = 0/1 ⇔ ∃u ∈ S such that ua = 0. ∀s ∈ S. v ∈ S such that u(s'a − sa') = 0 and v(t'b − tb') = 0. −1 The elements 0 and 1 in S R are: 0 0 0 = = . We check that the addition is correctly defined. ∀a ∈ R. 1 s 1 s 1 = = . called the field of fractions (or field of quotients) of R. −1 The ring S R is called the ring of fractions (or the ring of quotients) of R with respect to the multiplicatively closed set S.4. t').

there exists a unique ring homomorphism g : S R → T such that γ = gϕ. the field of quotients of the polynomial ring K[X1. with g ≠ 0. b) For any field K. ….8 Theorem. S R) is universal relative to this property. Then S R is a −1 commutative ring with identity and ϕ : R → S R is a ring −1 homomorphism such that ϕ(s) is invertible in S R. This fact is stated rigorously as follows: 4. such that γ(s) is invertible in T. (The universality property of the ring of fractions) −1 Let S be a multiplicatively closed subset of R. −1 ∀s ∈ S. ∀s ∈ S.7 Example. −1 Proof. Similarly. Moreover. Its elements are “fractions” of the form . −1 the pair (ϕ. ! . For any domain R. −1 More generally. …. Xn) and is called the field of rational fractions in n indeterminates with coefficients in K. The reader can easily verify that g is correctly defined. that it is a ring homomorphism and it is the only one with γ = gϕ. g ∈ K[X]. ∀a ∈ R. Q(R) is the “smallest” field that “includes” R. ∀s ∈ S.358 Appendices 4. T) where T is a commutative ring with identity and γ : R → T is a ring homomorphism. the field of quotients of the polynomial ring K[X] is denoted by K(X) and is called the field of rational fractions f with coefficients in K. g where f. S R is the “smallest” ring that includes R (if S contains no zero divisors) such that all elements in S are invertible in −1 S R. Xn] is denoted by K(X1. a) The field of quotients of Z is Q. c) The field of quotients of Z[X] is (isomorphic to) Q(X) (prove this!). namely: For any pair (γ. Define g(a/s) = γ(a)(γ(s)) .

(b.4. (a. s) ∼ (b. t) ⇔ ∃u ∈ S such that u(ta − sb) = 0 (cf. t) ∈ M × S.2). −1 Then there exists a unique ring isomorphism h : S R → B such ! that hϕ = β. Let M be an R-module and let S be a multiplicatively closed subset of R. . there exists a unique R-algebra homomorphism g : S R → T. such that γ(s) is invertible in T for any −1 s ∈ S. Let S be a multiplicatively closed subset of R. S M becomes an ⎨s ⎬ ⎩ ⎭ Abelian group with the addition: ∀x. we define on M × S the equivalence relation: ∀(a. the above property reads: For any commutative R-algebra (γ. Rings and modules of fractions −1 359 The “complete” structure of S R is that of R-algebra. s). there exists a unique ring homomorphism g : B → T such that γ = gβ. Assume B is a commutative ring with identity and β : R → B is a homomorphism satisfying the property: For any commutative ring with identity T and any ring homomorphism γ : R → T such that γ(s) is invertible in T. 4. y ∈ M. where γ : R → T is the structural homomorphism. In this setting. Denoting by the equivalence class s −1 −1 x of (x. Given a multiplicatively closed subset S ⊆ R and an R-module M. Τ).10 Proposition. t ∈ S. s ∈ S ⎫ . ∀s ∈ S. −1 The following result is proven exactly like in the case of S R: 4. the canonical homomorphism ϕ being the structural homomorphism. The above construction can be applied to an R-module M. the universality property of the ring of fractions determines the ring of fractions up to a (unique) isomorphism: 4.9 Theorem. As expected. s) ∈ M × S and S M := ⎧ x ∈ M . Then the relation " ∼ " defined above is an x equivalence relation on M × S. with minor modifications. ∀s.

is called the canonical homomorphism. let −1 I := ϕ (J) (an ideal in R). If now J ≤ S R. Then −1 S I := {a/s | a ∈ I. If S is an arbitrary multiplicatively closed set. . s ∈ S} is an ideal in S R. ∀x ∈ M. dr = s implies d ∈ S and r ∈ S. any ideal in −1 −1 S R is of the form S I. ∀d. The connection between the ideals of R and the ideals of the ring of fractions is very close. ! −1 −1 −1 A similar connection exists between the submodules of M and the −1 submodules of S M (can you state and prove it?). 4. Let I be an ideal in the ring R. We have a/1 ∈ J ⇔ ∃s ∈ S such that −1 a/s ∈ J. ∀s. Then −1 S I = {a/s | a ∈ I. ∃s ∈ S such that dr = s}. It is immediate that S I ≤ S R if I ≤ R. The homo−1 morphism ϕM : M → S M. S M is an S R-module with the multiplication defined by: ∀a ∈ R. let S' := {d ∈ R | ∃r ∈ R. S is saturated ⇔ S = S'. ! ⋅ := s t st The S R-module S M is called the module of fractions (or quotients) of M relative to the multiplicatively closed subset S. −1 −1 −1 Proof. Evidently. for some ideal I in R. Moreover. a x ax . t ∈ S. A multiplicatively closed subset S ⊆ R is called saturated if all the divisors of the elements in S are also in S: ∀s ∈ S. Thus ϕ (J) = {a ∈ R | ∃s ∈ S such that a/s ∈ J}.12 Definition.360 Appendices x y tx + sy . r ∈ R.11 Proposition.∀x ∈ M. An immediate property is: 4. + := s t st −1 −1 Moreover. s ∈ S} = J. S' is called the saturate of the multiplicative set S. ϕM(x) = x/1.

Ker ϕ = {0/1}. s ∈ S. ∃r ∈ R such that ur ∈ S and ura = 0.4. b) We denote the equivalence relation on S' × R (defined as in 4. s) ∼ (b. the equivalence class of (a. −1 Therefore. so a//d = ar//dr = ϕ(ar/dr). −1 ϕ(a/s) = a//s. so S I = R. Proof. with p ∈ P. −1 −1 b) There exists a canonical isomorphism S R ≅ S' R. s) in R × S' is denoted by −1 a//s ∈ S' R (in order to distinguish from the fraction a/s which may −1 −1 −1 be in S R). Let P be a prime ideal in the ring R. a/s = 0/1. t). then ab ∉ P. Thus. then (a. t). We have −1 Ker ϕ = {a/s ∈ S R | a//s = 0//1}. Proof. But a//s = 0//1 ⇔ ∃u ∈ S' such that ua = 0.14 Proposition. d ∈ S'. ϕ is surjective: if a//d ∈ S' R. where I ∩ S = ∅ (⇔ I ⊆ P). so −1 −1 J ⊆ S P. s) ≈ (b. Rings and modules of fractions 361 The following property says that the any ring of fractions can be constructed using a saturated multiplicative set. a) Check the definition. Also. Then −1 S := R \ P is a multiplicative subset in R and the ring of fractions S R has a unique maximal ideal (it is a local ring). Thus. Indeed. ! An important example of a multiplicative set and its corresponding ring of fractions is the following: 4.e. The definition of a prime ideal can be stated as follows: for any a. ∀a/s ∈ S R. then S I = S R. Then: a) S' is a saturated multiplicative set of R.2) with ≈ . which means that S is multiplicatively −1 −1 closed. If I ≤ R.13 Proposition. b ∉ P. then there exists r ∈ R such that dr ∈ S. the proper ideals −1 −1 of S R are of the form J = S I. ϕ is a ring homomorphism. with a ∈ R. Clearly. It is clear then that r ∈ S'. 4. i. if s ∈ I ∩ S. −1 −1 then s/1 ∈ S I and it is invertible. Define the canonical mapping ϕ : S R → S' R. with I ∩ S ≠ ∅. . The definition is independent on the choice of the representatives a and s: if (a. Let S be multiplicative set in the ring R. But S P is a proper ideal: if 1/1 = p/s.

4. 5.15 Proposition. where ϕM : M → S M and ϕN : N → S N are the canonical homomorphisms. ∀x ∈ M. the con- .Mod is exact: if the sequence A ⎯u B ⎯v C ⎯→ ⎯→ is exact in R-Mod. contra−1 −1 dicting S = R \ P. S − is an additive functor: for any R-module homomorphisms u1. S P is the unique maximal ideal in S R. −1 Moreover. u2 : M → N. Thus. Categories. −1 −1 S u is easily seen to be an S R-module homomorphism and it is −1 the unique S R-module homomorphism with the property that −1 −1 −1 (S u)◦ϕM = ϕN◦u. ∀s ∈ S. −1 −1 −1 S (u1 + u2) = S u1 + S u2. ! If u : M → N is an R-module homomorphism. functors The category language is nowadays all-pervading throughout mathematics. The functor S − : R-Mod → S R.362 Appendices then ∃u ∈ S such that u(s − p) = 0 ⇒ us ∈ P ⇒ u ∈ P or s ∈ P. then we define the mapping −1 −1 −1 −1 S u : S M → S N. Introduced in 1942 by MacLane and Eilenberg. Thus. then the sequence −1 −1 ⎯u ⎯v S −1 A ⎯S ⎯→ B ⎯S⎯→ S −1C −1 ! is exact in S R.Mod. (S u)(x/s) := u(x)/s. −1 −1 The unproven statements above are proposed as exercises. for a fixed multiplicatively closed subset S. we defined a functor: −1 −1 S − : R-Mod → S R-Mod.

One advantage of this approach is given by generality: a result that holds in any category is valid in the category of groups. A detailed presentation of category theory is to be found for instance in HERRLICH.5. and also in the category of topological spaces etc. unifies (to a certain extent) and standardizes the language of mathematics. Besides many results and clarifications brought in almost all areas of mathematics. and ignoring the elements of these structures. we briefly describe the concept of class. usually a natural notion of homomorphism arises (for example. For instance. Categories. the category theory simplifies. field. functors 363 cept of category marks a new step of abstraction in mathematics. endowed with a certain structure). also. or ring homomorphism). Any element of a set is thus also a set. The philosophy in the category theory is to study the class of all structures of a certain type (for instance the class of all rings) using the homomorphisms between these structures. In this theory. topological space etc). from the abstract notion of integer one passes to the notion of set of integers Z (as a new object of study. all objects are sets. ring. For a certain type of structure. The notion of class is introduced in the axiomatic set theory (in order to avoid the paradoxes generated by considering “very large sets”). We present here some basic concepts on categories that are useful for a better understanding of several topics. This leads to the study of structures given by axioms (such as the structures of group. Before proceeding to the definition of a category. Generalizing key properties of Z yields the concept of ring. the notion of set and the relation "∈" (“belongs to”) are primary notions (are undefined). the familiar notion of group homomorphism. . STRECKER [1979].

B) is also written u : A → B or A ⎯u B . If P and Q are classes. but there exist classes that are not sets (for instance. namely y). Similarly. B) is given. ∧ and the quantifiers ∀ and ∃.for every couple (A. B). A is the domain (or source) and B is ⎯→ In the Gödel-Bernays-von Neumann theory. a predicate with one variable). by analogy with the set language. For details. The fact that u ∈ HomC(A. if P(x) is a predicate.e. Roughly speaking. defined by the predicates P(x) and Q(x). Of course. Any set A is a class (corresponding to the predicate "x ∈ A"). see 5. 84 . sets) x for which P(x) is true. defined by the predicate "x = x"). B ∈ Ob C. a (possibly empty) set HomC(A.1 Definition. 85 We do not define here this formal language. (in the Zermelo-Fraenkel theory84) a class is an expression of the formal language of the set theory85 that contains exactly one free variable (in other words. Instead of writing "P(x) is true" one writes "x ∈ P". We say that P ⊆ Q if the proposition ∀x(P(x) → Q(x)) is true. B) are called morphisms (or arrows) from A to B. in which x and z are bound variables and y is a free variable. An expression with no free variables is a proposition. The sets are exactly the classes that are elements of some class. For instance "(∀x(x ∈ y))∧ (∃z(y = z)∨¬(z = a))" is an expression. . ¬. the class of all sets. the notion of class is a primary notion. their union P ∪ Q is the class P(x) ∨ Q(x). it consists of expressions (strings of symbols) formed from the atomic expressions (of the type x = y or x ∈ y) by using the logical operators ∨.a class Ob C. then this expression is a predicate (it has one free variable. The elements of HomC(A. A category C consists of the following data: .364 Appendices Formally. one can define the analogue for classes of the usual set operations. where A. The elements of Ob C are called the objects of the category C. If a is assumed to be a constant. their intersection P ∩ Q is the class P(x) ∧ Q(x). the class P defined by P(x) is intuitively the “collection” of all objects (i.

B. The composition of morphisms in Set is the usual function composition. The class ∪{HomC(A.u) is denoted by v◦u (or simply vu) and is called the composition of the morphisms v and u. B)∩HomC(C. The identity morphism of A is the identity function of A. Categories. 5. B)| A. B. it is the class defined by the predicate H(u) = "∃A∃B(A ∈ Ob C ∧ B ∈ Ob C ∧ u ∈ Hom(A. C. w : C → D. D ∈ Ob C. 3) ∀A ∈ Ob C. HomSet(A. B ∈ Ob C} is denoted Hom C and is called the class of the morphisms of the category C. (A.86 . C). b) In the definition of a category. In any category C the following axioms must be satisfied: 1) Any morphism has a unique domain and a unique codomain: for any A. Its objects are sets.2 Remark. functors 365 the codomain (or sink) of u. C. If A and B are sets. v : B → C. we have u◦1A = u and 1A◦v = v. there exists a morphism 1A : A → A (called the identity morphism of A) such that. The notion of category can be defined using only the concept of morphism. 86 Formally. The image of the couple (v. B))".3 Examples. B) is the set of all functions ϕ : A → B. a) The category Set of all sets. we have w(uv) = (wu)v (denoted usually by wuv). ∀B ∈ Ob C and ∀u : A → B. D ∈ Ob C and ∀u : A → B. D) implies HomC(A. C)×HomC(A. 2) The composition of morphisms is associative: ∀A. B) ≠ (C. . 5. a) The identity morphism of an object A is unique: if j: A → A is an identity morphism of A. the morphisms are essential.5. D) = ∅. C) of objects of C there exists a function defined on HomC(B. B) with values in HomC(A. One can identify an object A ∈ Ob C with its identity morphism 1A. then j = j◦1A = 1A. v : B → A.for any triple (A. B.

set ⎧ {(a. the category of topologi- b) The category Gr of groups. with the usual composition. b) = ⎨ ⎩ ∅. For any a. else The reader is invited to define the composition of morphisms and check the axioms 1) − 3). . H) is the set of group homomorphisms from G to H. the composition of the morphisms a. ·) be a monoid (the operation is associative and has a neutral element).Ring: the rings with the ring homomorphisms. The reader can easily produce other examples of categories. if a ≤ b HomA(a. the category of fields. In the same way one defines the category Mod-R of right R-modules.Poset: the (partially) ordered sets. Define a category G as follows: Ob G is a set with one element (for instance Ob G = {G}). Define a category A. G) = G (the morphisms are the elements of G). Ob Gr is the class of groups and HomGr(G. H ∈ Ob Gr. The category R-Mod has as objects left R-modules.Ab: the Abelian groups. b)}. with the group homomorphisms . the composition is the usual function composition. the morphisms being the order preserving mappings. The identity morphism is the identity element. and HomG(G. As in Set. based on hers/his mathematical background (the category of semigroups. the category of finite sets. and the morphisms are R-module homomorphisms. b ∈ Ob A. b ∈ G is a·b (where · is the operation on G).Ringu: the rings with identity. d) One can define similarly the following categories: .366 Appendices e) Let (G. with Ob A := A. ≤) be a set equipped with a preorder relation (a transitive and reflexive relation). . the morphisms are ring homomorphism preserving the identity. d) Let (A. . c) Let R be a ring with identity. ∀G.

Often the writing A ∈ C replaces A ∈ Ob C. An object F is called a final object if ∀A ∈ C. a) An object I ∈ C is called an initial object if ∀A ∈ C. ∀A. |HomC(A. the set of morphisms between two arbitrary objects. Gr is a subcategory of Set. B). the composition of morphisms and check axioms 1)-3). B). w ∈ HomC(B.an isomorphism if there exists v : B → A such that uv = 1B and ~ vu = 1A (v is called then the inverse of u). B ∈ Ob C.5 Definition. . ∀v. An object that is simultaneously initial and final is called a zero object.).an epimorphism if ∀C ∈ C. 5. . 5. vu = wv implies v = w. In each situation it is necessary to state exactly the class of the objects of the category.a bimorphism if it is both a monomorphism and an epimorphism.5. the continuous mapping being the morphisms etc. moreover. The notation A → B is used to denote an isomorphism. A category C is called a subcategory of a category D if Ob C ⊆ Ob D and. . w ∈ HomC(B. |HomC(I. B ∈ Ob C. b) A morphism u : A → B is called: . Ab is a full subcategory of Gr.a monomorphism if ∀C ∈ C. ∀v. if no confusion arises. We define now some remarkable objects and morphisms in a category C that generalize familiar concepts. Categories. C). uv = uw implies v = w. I)| = 1. but not a full subcategory. functors 367 cal spaces. We call C a full subcategory of D if ∀A. HomC(A. the composition of two morphisms in C is their composition in D. HomC(A. B) = HomD(A. For instance. A). A)| = 1 (there exists a unique morphism I → A).4 Definition. . B) ⊆ HomD(A.

ϕ and ψ are isomorphisms. Likewise. The same remark holds for Ab and R-Mod. B ∈ C are called isomorphic (written A ≅ B) if there exists an isomorphism A → B. d) In the category Ann of rings with identity. c) The monomorphisms in Set (as in Gr. The morphism ψϕ : A → A is equal to 1A (because there exists a unique morphism A → A). But B is an initial object. and is not a surjective function or an isomorphism. ! It is important to point out that various “universality properties” that some objects satisfy are simply a restatement of the fact that those objects are initial (or final) objects in certain categories.368 Appendices Two objects A. 5. Let C be a category and let a A. namely the groups having one element (necessarily the neutral element of that group). there exists a unique morphism ϕ : A → B. Set has no zero objects. inverse to each other. b) In Set the empty set ∅ is the only initial object87. In this situation.7 Proposition. there exists a unique function ∅ → A. Proof. Ab. So. thus there exists a unique morphism ψ : B → A. B be initial ob~ jects in C. The relation of isomorphism on the class Ob C is an equivalence relation. any set with one element is a final object. 5. R-Mod) are the morphisms that are injective functions. the inclusion Z → Q is a monomorphism and an epimorphism. . Which are the epimorphisms? In these categories the isomorphisms coincide with the bimorphisms. a) In Gr there exist initial objects. ϕψ = 1B.6 Examples. Then there exists a unique isomorphism A → B. the propositions of the type “the universality property of … determines … up to a unique isomorphism” merely translate for a spe87 For any set A. Since A is an initial object. These are also final objects (thus they are zero objects) in Gr. namely the function ∅.

5. Categories, functors

369

cific category the assertion “between any two initial (final) objects in a category there exists a unique isomorphism”.
5.8 Examples. a) The direct sum. Let (Mi)i∈I be a family of objects in R-Mod. Let S be the category whose objects are couples of the form (S, (σi)i∈I), where S ∈ R-Mod and σi : Mi → S are morphisms in R-Mod, ∀i ∈ I. If (S, (σi)i∈I), (T, (τi)i∈I) ∈ S, define the morphisms in S between these objects as the R-module homomorphisms ϕ : S → T such that ϕσi = τi, ∀i ∈ I. Check the axioms for a category and the following assertion: (S, (σi)i∈I) is a direct sum of the family (Mi)i∈I, of canonical injections (σi)i∈I, is tantamount to saying that (S, (σi)i∈I) is an initial object in S. b) The direct product. Let (Mi)i∈I be a family of objects in R-Mod. Then: (P, (πi)i∈I) is a direct product of the family (Mi)i∈I ⇔ (P, (πi)i∈I) is a final object in a certain category (describe it!).

An important principle, often invoked, is the principle of duality.
5.9 Definition. If C is a category, the dual category C° is defined as follows: Ob C° := Ob C; if A ∈ Ob C, let A° be the object A seen in C°. For any A, B ∈ Ob C, let HomC°(B°, A°) := HomC(A, B). A morphism u : A → B in C is denoted u° : B° → A° in C°. The Composition in C° of the morphisms u° : B° → A° and v° : C° → B° is defined by u°v° := (vu)°, where vu is the composition of u : A → B and v : B → C in C. Evidently, there exists 1A° = (1A)°. Intuitively, the dual of the category C is obtained by “reversing the arrows” in C (and reversing the order of composing the arrows).

Let P be a statement formulated in terms of objects and morphisms. For each category C, one obtains a proposition, denoted P(C). Let P°

370

Appendices

be the dual statement (obtained from P by reversing the arrows and the order in composing the arrows)88. The principle of duality is the following: If P is valid in any category, then P° is valid in any category. Similarly, any notion (definition) in a category has a dual notion, obtained by reversing the arrows and the order in composing the arrows. A notion that coincides with its dual is called autodual.
5.10 Example. a) The dual of the notion of initial object is the notion of final object. b) The dual of the notion of monomorphism is the notion of epimorphism. c) The notion of isomorphism is autodual. d) We saw that: for any category C, any two initial objects in C (if any) are isomorphic. By dualization, one obtains (no new proof needed): for any category C, any two final objects in C (if any) are isomorphic.

The intuitive concept of “morphism of categories” is the notion of functor.
5.11 Definition. Let C and D be categories. A covariant functor F from C to D, denoted F : C → D, is a couple F = (F', F"), where F' : Ob C → ObD, F": Hom C → Hom D, such that: (F1) ∀A, B ∈ Ob C, F"(HomC(A, B)) ⊆ HomD(F'(A), F'(B)); in other words, if u : A → B, then F"(u) : F'(A) → F'(B). (F2) F preserves the composition of morphisms: ∀A, B, C ∈ Ob C and ∀u : A → B, v : B → C, then F"(v◦u) = F"(v)◦F"(u). (F3) F preserves the identity morphisms: ∀A ∈ C, F"(1A) = 1F"(A).

88

in other words, P°(C) is the same thing as P(C°) interpreted in C.

5. Categories, functors

371

A contravariant functor F : C → D satisfies F3 and the duals of F1 and F2: (F1*) F"(HomC(A, B)) ⊆ HomD(F'(B), F'(A)), ∀A, B ∈ Ob C. (F2*) ∀A, B, C ∈ Ob C and ∀u : A → B, v : B → C, then F"(v◦u) = F"(u)◦F"(v). A contravariant functor “reverses the arrows”. A contravariant functor from C to D is the same as a covariant functor from C to D° (or from C° to D). This is the reason why the results on covariant functors transfer to contravariant functors. In what follows, “functor” means “covariant functor”. The distinction between the components F' and F" of the functor F is usually dropped, denoting F(A) instead of F'(A) and F(u) instead of F"(u). Besides, F is perfectly determined by F", its action on morphisms, cf. 5.2). A functor F : C → D is called: - faithful if ∀A, B ∈ C, FA,B : HomC(A, B) → HomD(FA, FB) is an injective function. - full if ∀A, B ∈ C, FA,B : HomC(A, B) → HomD(FA, FB) is a surjective function. - fully faithful if it is full and faithful. We are interested in categories in which the morphisms are functions between sets:
5.12 Definition. The category C is called concrete if there exists a covariant faithful functor F : C → Set.

The category Gr is concrete: define the functor U : Gr → Set, by associating to any group its underlying set89 and sending any group
89

group and · : G×G → G is the group operation. Thus, a function from a group (G, ·)

A group G is, formally, a couple (G, ·), where G is the underlying set of the

372

Appendices

homomorphism u to u, seen as a function between the underlying sets. Then U is a faithful functor (which is not full). U is called a “forgetful functor”: it “forgets” the group structure. Similarly, Ab, Ring, R-mod are concrete categories (why?). In concrete categories there is a notion of free object on a set X (compare the following definition with the universality property of the free R-module on a set X ).
5.13 Definition. Let C be a concrete category and let F : C → Set be a covariant faithful functor. The object L in C is called free (relative to F) over the set X ⊆ F(L) if, for any A ∈ C and any function γ : X → F(A), there exists a unique morphism in C, g : L → A, such that F(g)|X = γ. If C is one of the categories Gr, Ab, R-mod, Ring, and F is the forgetful functor, the definition reads: L ∈ C is free over the set X ⊆ L if, for any object A ∈ C and any function γ : X → A, there exists a unique morphism in C, g : L → A, such that g|X = γ.

The following notion allows comparing two functors and is the analogue of the concept of homomorphism of algebraic structures.
5.14 Definition. Let F, G : C → D be covariant functors. A natural transformation (or a functor morphism) α : F → G is given if for any object A ∈ C, there exists a morphism in D, αA : F(A) → G(A), such that, for any u : A → B morphism in C, αB◦F(u) = G(u)◦αA, i.e. the following diagram (of morphisms in D) is commutative:

to a group (H, *) is not the same thing as a function between their underlying sets G and H. For this reason, Gr is not a subcategory of Set.

6. Solvable groups

373

F (u ) F ( A) ⎯⎯ ⎯ ⎯⎯→ F (B )

αA

αB

G (u ) G ( A) ⎯⎯ ⎯ ⎯⎯→ G (B ) If , for any A ∈ C, αA : F(A) → G(A) is an isomorphism in D, then α : F → G is called a natural isomorphism, in which case the functors F and G are called naturally isomorphic.

Many “canonical isomorphisms” in module theory (for instance) express the fact that there exists a natural isomorphism between some functors.

6. Solvable groups
The notion of solvable group is closely connected with the solvability by radicals of a polynomial. In what follows, (G, ·) is a group and its neutral element is denoted by 1. The trivial subgroup {1} is also denoted 1. The notation H ≤ G signifies “H is a subgroup of G”, and H . G means “H is a normal subgroup of G”.
6.1 Definition. The group G is called a solvable group if there exists a finite chain of subgroups of G: (S) 1 = G0 ≤ G1 ≤ … ≤ Gn = G, such that: i) Gi−1 . Gi, ∀i, 1 ≤ i ≤ n; ii) The factor groups Gi/Gi−1 are Abelian, ∀i, 1 ≤ i ≤ n.

374

Appendices

A chain of subgroups (S) that satisfies (i) is called a normal series of G.90 A chain (S) satisfying i) and ii) is called a solvable series of G. The number n is called the length of the series (S). The groups Gi/Gi−1 are called the factors of the normal series (S).
6.2 Examples. a) Any Abelian group is solvable. b) Let n ∈ N* and let Sn be the group of all permutations on n objects (also called the symmetric group on n objects). Let An be the alternating group on n objects, namely the subgroup of Sn formed by the even permutations. An is a subgroup of index 2 in Sn (it is the kernel of the sign homomorphism ε : Sn → {− 1, 1}), thus it is normal in Sn. The group S3 is solvable (and not Abelian!), a solvable series being 1 ≤ A3 ≤ S3. Indeed, A3 has 3 elements, so it is Abelian; A3 is normal in S3, and the factor S3/A3 is Abelian, since it has 2 elements. c) S4 is solvable. A solvable series is 1 ≤ V ≤ A4 ≤ S4, where V = {1, (12)(34), (13)(24), (14)(23)}91. Check the details! d) Any nonabelian simple group (i.e. having no proper normal subgroups) is not solvable.

Before studying solvable groups, let us recall some elementary results in group theory.
6.3 Theorem. (The fundamental isomorphism theorem) Let ϕ : G → G' be a group homomorphism. Then there exists a canonical isomorphism G/Kerϕ ≅ Imϕ

xKerϕ & ϕ(x)
90

!

Some authors use in this case the term “subnormal” and call a “normal series” a chain (S) of normal subgroups of G. 91 V is also called the Klein group (the “Viergruppe”).

6. Solvable groups

375

6.4 Theorem. (Second isomorphism theorem) Let G be a group, let H and N be subgroups of G, such that N is a normal subgroup in the group (H, N) generated by H ∪ N in G. Then (H, N) = HN = {hn | h ∈ H, n ∈ N} = NH, N ∩ H . H and there is a canonical isomorphism HN/N ≅ H/(N ∩ H). !

If H, N are subgroups of G satisfying the conditions in the previous theorem, HN or NH denotes the subgroup generated by H ∪ N.
6.5 Theorem. (Third isomorphism theorem) Let G be a group, let A and B be normal subgroups of G such that A ≤ B. Then B/A . G/A and there is a canonical isomorphism G A G ≅ . ! B A B 6.6 Proposition. (The modularity property) Let G be a group and let A, B, N be subgroups of G, such that N is normal and B ≤ A. Then ! A ∩(BN) = B ∩(AN). 6.7 Proposition. Let G be a group. a) If G is solvable, then each subgroup of G is solvable. b) If G is solvable, then each factor group of G is solvable. c) If H . G and H and G/H are solvable, then G is solvable. In other words, if 1 → H → G → F → 1, 92 is an exact sequence of groups and group homomorphisms, then G is solvable if and only if H and F are solvable.

The notion of exact sequence of groups and group homomorphisms is defined exactly as in the case of modules.

92

Gi and ϕ is surjective. Gi i We used the fundamental and the second isomorphism theorems. where H = G0 ≤ G1 ≤ … ≤ Gn = G. ! . If 1 = G0 ≤ G1 ≤ … ≤ Gn = G is a solvable series of G. where G' ≤ G. G and Gi−1 ≤ Gi. Fi−1 . which is Abelian. Since Gi−1 . ∀x ∈ Hi) has kernel Hi−1. Fi. Thus Hi−1 . The canonical homomorphism ϕ : Hi → Gi/Gi−1 (ϕ(x) = xGi−1. Then Ker (ϕ series of F. We claim that 1 = H0 ≤ H1 ≤ … ≤ Hn = H is solvable series of H. There exists a surjective homomorphism ϕ : G → F. with H ≤ G. a solvable series for G/H is of the form 1 = G0/H ≤ G1/H ≤ … ≤ Gn/H = G/H. so Fi ≅ Gi/Ker (ϕ G ) = Gi/(H ∩ Gi) ≅ (GiH)/H. c) Any subgroup of G/H is of the form G'/H. Therefore: Fi GH H GH G (G H ) Gi Gi ≅ i ≅ i = i i −1 ≅ = (Gi −1H ) ∩ Gi Gi −1 ( H ∩ Gi ) Fi −1 Gi −1H H Gi −1H Gi −1H We used successively: the third. If 1 = H0 ≤ H1 ≤ … ≤ Hm = H is a solvable series for H. 1 ≤ i ≤ n. let H≤G and let Hi := Gi ∩ H. Fix i. Hi and Hi/Hi−1 ≅ Imϕ ≤ Gi/Gi−1. Let ) = H ∩ Gi. let Fi := ϕ(Gi). b) Let F be a factor group of G. But: Gi Gi Gi −1 ≅ . Gi −1 ( H ∩ Gi ) Gi −1 ( H ∩ Gi ) Gi −1 The last group is Abelian (it is a factor group of Gi/Gi−1). Therefore. then 1 = H0 ≤ H1 ≤ … ≤ Hm = G0 ≤ G1 ≤ … ≤ Gn = G is a solvable series for G. a) Let 1 = G0 ≤ G1 ≤ … ≤ Gn = G be a solvable series of G. We claim that 1 = F0 ≤ F1 ≤ … ≤ Fn = F is a solvable H = Kerϕ. the second isomorphism theorems and the modularity property for H .376 Appendices Proof.

Since H and G/H have orders less than |G|. If the series has length 2. This is an easy consequence of the structure theorem of the Abelian finite groups. let G be finite solvable. generated by x ∈ G.8 Proposition. Conversely. Proof. for any d dividing n. which is essential in the theory of solvability by radicals. G such that H and G/H are solvable and the previous result applies. 6. ! 6. as above.9 Proposition. Assume now that G is not Abelian. 6. The definition of solvability implies the existence of a proper normal subgroup H of G such that H is solvable and G/H is Abelian. there exists H . Let G be a group. If |G| ≤ 3. A finite group is solvable if and only if it has a normal sequence whose factors are cyclic groups of prime orders. there exists a n/d subgroup of order d (generated by x ).c). Apply now an induction. Continue by induction on the length of the series. Proof. then G is solvable. As in the proof of prop. we are finished. Thus. We suppose the statement is true for any solvable group whose order is less than |G| and we prove for G. We give a proof that does not use this result. Assume first that (G. with cyclic factors of prime orders. . The factor group G/H and H have orders less than |G|.10 Lemma. gluing these series yields a solvable series of G.6. all is clear. If n is prime. If G is Abelian. ! The finite solvable groups have the following property.7. A finite Abelian group has a normal sequence whose factors are cyclic groups of prime orders. a group with the property above is solvable. Solvable groups 377 6. Of course. taking a prime divisor p of n. If G has a normal series whose factors are solvable groups. if not. ·) is cyclic of order n. there exists a subgroup H of order p. then use the following lemma. Proof. the induction hypothesis says that H and G/H have each a normal series with cyclic factors of prime orders.

12 Proposition.11 Proposition. thus solvable.e)). Let G be a finite group. n Proof. More precisely. ! The next proposition says that among the symmetric groups Sn. The proof is not included. Since C(G) is Abelian. (where S is a system of representatives of the conjugacy classes of G with at least 2 elements). Using an induction argument. G and the factor group G/C(G) is still a p-group. the solvable groups are exactly those in Example 6.15. C(G) ≠ 1. Thus. .378 Appendices If G is not cyclic. in the conjugacy classes formula (IV. a) If n ≤ 4. and solvable. the cyclic subgroup C generated by x is not equal to G. Then G is solvable. Indeed. then Sn is solvable. so |C(G)| is a multiple of p (and is nonzero. ! 6. Consequently. then Sn is not solvable. |G| = |C(G)| + ∑a∈S [G : C (a )] . let x ∈ G. of orders less than |G|. 1 ≠ C(G) . If n = 1. then the center of G. If n > 1. b) If n ≥ 5. Apply an induction for C and G/C. there exists a normal series of G whose factors are cyclic groups of order p. since 1∈ C(G)). the alternating subgroup An is simple (has no proper normal subgroups) and ! noncommutative. In particular.2. we obtain that G is solvable. x ≠ 1. so it is not solvable. |G| and [G : C(a)] are powers of p. we obtain that G/C(G) is solvable.3. Let |G| = p . then G is cyclic. whose order is less than |G|. being often encountered in introductory Algebra texts (see for instance HUNGERFORD [1974]) 6. whose order is a power of a prime p (such a group is called a p-group).

........... 223 alternating group......................................................223 Chinese remainder theorem.....334 characteristic exponent..... 374 annihilator......................................................371 dual............................................ 198 algebraic closure (relative) .................................172 characteristic polynomial ........ 177 algebraic closure (absolute) ...... 175 algebra .......................163 center of a ring ......................................161 of an endomorphism...........................................290 of a matrix ......... 334 algebra homomorphism ............ 104..........................................75 category........Index A adjoining elements to a field..............364 concrete ...................................................290 characteristic of a ring...... 364 Artin-Schreier theorem ......... 34 associated in divisibility .. 133 arithmetically equivalent matrices121 arrow.................................307 nondegenerate....................................... 178 algebrically independent (set)....173 characteristic matrix .....75 canonical injections..............................307 bimorphism .....86 canonical projections...........................367 C canonical homomorphism ........................ 12 autodual ..................... 113 basis of a module.................... 189 algebraic generators ......141 379 ................ 302 ascending chain condition................... 370 B basis change matrix.......81 canonical surjection.............. 226 algebraic integer ....... 335 algebraic (element) .369 Cayley-Hamilton theorem ......161 Chevalley theorem .............................. 27 algebraic number .............................................................106 bilinear function .....

.......................................................... 365 congruent modulo n.........22 divisor ........... 276 constructibile real number .......................... 277 content of a polynomial ...................................................................320 divisibility .............................. 26 conjugate elements ............... 240 constant term ....................... 278 constructibile point .................................. 364 coefficient of a polynomial................................................. 11...92 of modules...........................306 duality ................ 347 constructibile complex number ................. 172 degree of a polynomial ................................165 Eisenstein criterion ..................................................................... 363 codomain ...............................123 endomorphism .............. 96 comaximal ideals.....................................81 of modules.. 96 composite of fields .121 direct product of homomorphisms ................154 elementary transformations......... 213 cyclotomic polynomial ........47 element integral ....................................................................................................................... 262 degree of an algebraic element................................. 347 denominator..........................90 direct summand........ 14 coproduct................... 175 composition of morphisms .........................................147 elementary divisors (of a matrix) 154 elementary divisors (of an endomorphism) ...................83 direct sum of modules.... 41 coordinates of an element in a basis .............................320 of an element.........380 class .....369 E eigenvalue .......................................................... 86 cyclotomic extension...................165 eigenvector.... 141 complement .............................................................52 elementary divisors .. 291 determinant of an endomorphism162 diagonal matrix ...........68 .....10 division with remainder theorem ......... 346..................... 235 conjugate extension ... 106 coprime.... 181 of an extension............. 364 dual of a group . 214 D Dedekind's Lemma ......... 356 determinant........................................................ 90 complex of modules ................90 discriminant of a polynomial ................ 347 cokernel ..........................................11 domain .....................85 of submodules .....................

...........................................174 fixed field ..233 fixed subfield...................................................................................................................................................................172 formal derivative ........ 76.. 172 finitely generated ...........195 Fp 171 fraction ................................ 169 algebraically closed......371 forgetful...............................................371 functor morphism .. 314 simple......372 fully faithful.............................266 fundamental theorem of symmetric polynomials ......................................................... 303 n-radical .......................350 G Galois connections .................... 175 transcendental ....... 23 Euclidian domain.............371 covariant ..................................................211 Galois group of a field extension .................... 170 Abelian............ 48 field.....356 free resolution .... 153 endomorphism ring.........210 Frobenius theorem.................................... 188 extension of a module............................................................................ 170 field generated by a set....... 224 radical ..................................................... 61 epimorphism .. 267 Kummer ..................233 Galois Field ...............................................................................105 Frobenius endomorphism................................... 348 field extension.......................... 314 purely transcendental ...................................371 plin........ 299 algebraic...... 303 exponent of a group .....................................314 ....................... 221 extension... 336 factorization algorithm.......................................109 free subset of a module .................................. 346 exponent (of a group)............................ 97 external direct sum of modules............................................................. 299 finite..............................370 fidel ......... 187 cyclic...................................381 cyclic.....233 of a polynomial. 213 evaluation homomorphism ......233 Galois group of a polynomial.........163 functor contravariant......... 198 of rational fractions ..........................................372 fundamental isomorphism theorem77 Fundamental Theorem of Algebra205 fundamental theorem of Galois Theory ....................................... 367 Euclidian Algorithm ................. 22 Euler phi function ................ 87 F factor algebra ................ 175 Galois......

............................... 114 Gr 366 greatest common divisor.......... R) ......................................171 internal direct sum .18 isomorphism..71 K-homomorphism .258 integer division theorem .......................... K) ..367 Irr(x................................................152 inverse (of a morphism) ................................ 331 maximal ......11 intermediate field ........................ 318 generic polynomial of degree n ...........................157 Jordan cell........................................ 15 general equation of degree n........................... 373 group of units ................................................ 342 initially constructibile point........171 proper ..... 14 group solvable....................... 331 idempotent...... 300 additive version... 97 I ideal ......................................... 13 H Hilbert Satz 90............. 318 GL(n.............. 12 GCD .........303 additive............ 332 prime............................................................................................................... 71 independent family of submodules90 indeterminate .................................................................... 365 image ................131 invariant subspace...................367 J Jordan canonical form of a matrix164 Jordan canonical matrix ... 14 GCD-domain ............ 332 proper........................................... 244 extension.................... 245 inseparable degree..................................................................................................154 Invariant factors theorem ....................................44 integers modulo n............. 275 inseparable element ..............................308 L Lagrange interpolation polynomial54 ... 91 identity morphism...............................171 Kummer Theory...... 20..................................................................................................156 K kernel .....137 invariant factors (of a matrix) ...90 invariant factors ...................382 Gauss integers .........154 invariant factors (of an endomorphism) ..................26 integral domain ...................... 301 homomorphism of module extensions ..............................................311 multiplicative ..........179 irreducible ..........

.............................................. 104....................11 multiple root... 121 matrix companion of a polynomial ..........71 monic polynomial.....374 numerator ........... 154 Mod-R ....................................................................................................................45 multiplicative set ...........................14 N natural isomorphism.................347 monoid ....................35 norm ........................................................................ 140 factor .............................................. 14 lexicographic order .. 156 matrix of a homomorphism ....................................367 .....................342 monomorphism .............373 natural transformation ... 360 right ...355 multiplicity of a root......................360 multiplicatively closed set........... 350 linear combination ........364 multiple ... 112 minimal polynomial....... 187 M matrix canonically diagonal ........................................................................................ 65 free ...................... 63 linearly dependent........... 366 module ............. 106 indecomposable ............... 104...........................59 simple .................................................. 14 leading coefficient ............. 121 elementary.......................................................................................383 lattice ...................76........................................ 367 morphism ..................68 module isomorphism..............................353 Noetherian ring .... 75 finitely generated .........238 normal series ...........................372 Newton's identities ............................................ 59 decomposable............................. 105 Lüroth’s theorem ..45.....27............... 176 LCM ..................................................... 105 linearly independent.................................................................................................. 291 normal closure......336 monomial .............................. 59 of fractions .......................................................355 saturated ................71 module homomorphism ............. 140 left ..................... 347 leading monomial .80 module automorphism.............356 O object final .241 normal extension ......... 194 mutually prime .............. 351 least common multiple............................... 122 in Smith normal form........................ 179 minimal polynomial (of a matrix)154 minimal polynomial (of an endomorphism) ....

......248 sequence.........................................27 remainder ................................ 364 opposite of a ring.....................22 resultant...........212 primitive ........... 27 symmetric ........................................52 of quadratic integers.................. 175 primitive polynomial ................................28 R-Mod ................................ 13 p-submodule ....................... 261 p-dimension........................96 exact ........245 polynomial ............................................................................................................................... 246 p-group . 367 zero .. 18 prime field ............................... 347 monic ..................................................................... 255 Q quadratic integer ............ 261 perfect field ................. 378 PID ..97 sequence of modules ..................................207 quotient .......................................... 348 polynomial algebras .................. 130 P p-basis........... 172 primitive element............................96 . 41 principal ideal.....208 quaternions.177 root of unity ............................................ 133 purely inseparable...................244 extension . 31 polynomial elementary symmetric.................................. 341 polynomial function .. 349 fundamental symmetric ............. 54 prime .......................................194 simple............................................................. 349 homogeneous ....................... 31 projective linear group.................................................................................... 254 purely inseparable closure .... 20 principal ideal domain ..........................................27 quaternion group .......253 separable degree.......... 272 proper divisor ....................................................................................................324 ring integrally closed ........194 root of a polynomial.......................................... 372 initial................384 free............................................................ 367 object (in a category)........111 rational integers................. 22.................................................................................................. 60 order .......366 root multiple ...... 356 R rank of a free module .............. skew field of ..........................244 separable closure.213 S separable element.........................................................

.... 83 of the direct sum ............. 170 proper.............. 291 trace of a matrix ......... 364 smallest element .................133 torsion submodule ............................... 335 subalgebra.... 337 symmetric group .107 of the monoid algebra. 183 splitting field........... 45 sink .............................................................177 transcendental number ...133 total degree........229 transcendental (element) ................................86.....................27...... 364 split epimorphism ................36 u-indecomposable space............. 352 solvable by radicals..............226 transcendence degree ....................385 semiexact ....133 torsion-free module ............................................................................................................................................................................... 100 split monomorphism .......................................................... 201 squarefree .............................................13 universality property of the direct product ................................................................152 u-invariant subspace...................................... 152 similar matrices..... 63 subring generated by a set ...................162 transcendence basis ......... 365 similar endomorphisms.....167 U UFD.................... 65 generated by a set.............................. 67 minimal . 67 proper... 11 structural homomorphism (of an algebra) .................................. 100 split short exact sequence ................... 367 full.............................. 64 maximal ..........................................152 unique factorization domain. 367 subextension ............................... 314 solvable series.......20....................................................178 trigonable matrix .......................................................374 system of generators.............................................................................. 374 source......................342 torsion module........ 170 submodule....161 trace of an endomorphism ..... 335 subcategory. 335 generated by a set.......................................................36 unit .... 96 Set ..............85 of the free module.................... 171 subfield ............................................ 331 support..................... 152 simple root ............... 100 splits................................64 T term ................ 62 cyclic.........................343 ............................................................................................................................347 trace..................................................175 sum of ideals ........82....................

.........219 well ordered ......................................... 344 of the ring of fractions . 13 W Warning Theorem ..386 of the polynomial algebra ........ 358 U(R) ..................................223 Wedderburn's Theorem..................352 .....................

ION.P. Bucureşti. I. N. 6. Masson. Masson. Ed. F. BOURBAKI.. K. M. ION. [1997] Théorie de Galois. Algebra. Fasc. BOURBAKI. 17. [1984] Capitole de teoria algebrică a numerelor.. VI. 15. Macmillan. Category Theory.R. [1987] 19 Lecţii de teoria grupurilor. Paris. Bucureşti.. Academiei R. Berlin. Bucureşti. New York. Springer Verlag.. [1981] Algèbre. Ştiinţifică şi Enciclopedică. Paris. Ellipses. FULLER. ŞAFAREVICI. 11. 13. GH. 20. Ed. C. Livre II: Algèbre. HALL. N. Berlin. ESCOFIER. Algèbre multilinéaire. 387 . Bucureşti. [1985]. N. Paris.R. 10. Bucureşti. Teoria numerelor. RAIANU. T. New York. NĂSTĂSESCU. Chapitre 3. 2. Heldermann Verlag. [1985] Eléments de mathématique: Algèbre commutative. Didactică şi pedagogică. I.W. Ed. NIŢĂ.R. Ş. [1958] Eléments de mathématique. Ed.W. Algèbre linéaire. Ed. [1974].. ALBU. [1997] Théorie de Galois. N. 4. AYAD. T. J. BOURBAKI. Ed. Paris.. Chapitre 2. ANDERSON. Springer-Verlag.. Chapitres 1 à 4..P. [1984] Complemente de algebră. Bucureşti. [1961] Corpuri de funcţii algebrice şi varietăţi algebrice. second edition. [1997] Théorie de Galois.S. Bucureşti. RADU. Masson. I. Bucureşti.D.Bibliography 1. GALBURĂ. 19. Universităţii Bucureşti. Livre II: Algèbre. Didactică and pedagogică. BOREVICI. [1981a] Algebra. MANOLACHE. [1972] Algebră. New York. 3. N. T. Paris. 7. HERRLICH. BOURBAKI.D. Ed. T. Ellipses. [1974] Rings and categories of modules. ION. ALBU.R. Fasc.I. Paris. Ştiinţifică and enciclopedică.D. 8. Hermann.. C. H. STRECKER G. 122 exercices corrigés. 14.E. ALBU. Academiei R. M. Chapitres 4 à 7. N. HUNGERFORD. Springer-Verlag. [1986] Field Arithmetic. [1959] The Theory of Groups. FRIED. I. [1967] Eléments de mathématique. 12. M. 5. VII.. Paris. [1984] Lecţii de algebră comutativă. M. GH. Z. 16. GOZARD. 9. JARDEN. GALBURĂ.. 18. I. Ed. [1979]. Universităţii Bucureşti. Hermann.

Paris. Algebraic numbers. I. Springer-Verlag. Academiei R. Academiei R. 41. 23.. M. Universităţii Bucureşti. 30. [1968] Inele locale. Iaşi. SCORPAN.. [1996] Field and Galois Theory. RADU. [1974] Introducere în teoria mulţimilor. Ed Universităţii „Al.. I. TOFAN. GH. [1982] Tratat de algebră modernă. 42. Bucureşti. Bibliography ION. Blaisdell Publishing. [1981] Elemente de teoria mulţimilor and logică matematică. Springer-Verlag. Hermann. MACCARTHY. I. SAMUEL. [1981b] Probleme de algebră. POPESCU. Academiei R. Junimea.. D. REGHIŞ. SPIRCU. NITA. C. Ed. Reading Massachussetts. J. Springer-Verlag. [1994] Abstract Algebra with Applications. Freeman and Co. 22. NITA... Fields and Rings.. [1986] Class Field Theory. 39. Didactică and pedagogică. N.. NIŢĂ. Bucureşti. 24. M.. Langages géometrique et algébrique. Bucureşti. New York. 40. Sociedade de Matemática de São Paulo. I. KAPLANSKY.. P. [1986] Bazele Algebrei. 37. [1971] Categorii abeliene. [1973]. VRACIU. Addison-Wesley. 31.. 28. Module. NĂSTĂSESCU. Bucureşti. vol. San Francisco. Facla. . Timişoara. C. C. vol. Ed. LANG. W. Algebraic Extensions of Fields.388 21. M. SPINDLER. Berlin. C. J.H..S. A. C. Bucureşti. I. D.P. N. [1964]. 35. NEUKIRCH. Ed. NIŢĂ. Waltham. Bucureşti. [1996] Introducere în teoria axiomatică a mulţimilor.. Paris. [1963] Anneaux factoriels. Ed. PARENT. [1968] Théorie algébrique des nombres. 45. Ed. V. vol II. [1964]. Ed.. 44. Academiei R. MORANDI. 34. C. N. Ed. Bucureşti. Tehnică. Gauthier-Villars. N. Categorii.R. Dekker.R. 26. 27. LAFON. RADU. C. Basic Algebra I. P. 36. Paris. Ed. [1966]. 25. Academiei R.R. [1976] Inele.S. JACOBSON.S. T. Tehnică. Bucureşti.S. C. Cuza”. New York. [1974]. [2000] Introducere în algebra omologică. SAMUEL. Chicago. NĂSTĂSESCU. RADU. [1977] Algèbre commutative. Didactică şi pedagogică. Ed. S. Massachusets.R. [1979] Teoria calitativă a ecuaţiilor algebrice. [1988] Algebra categoriilor şi functorilor.P. K. GH. I.D. Ed. Ed. [1974] Probleme de structuri algebrice. 32. II. Bucureşti. 43.S.R. C. I. vol. [1978] Exercices en théorie des nombres. Lectures in Abstract Algebra III. POPESCU. 38.J.. The University of Chicago Press. NĂSTĂSESCU. 33. PURDEA. JACOBSON. NĂSTĂSESCU. N. P. Theory of Fields and Galois Theory. Iaşi. Hermann. GONTINEAC. 29. New York. RADU. P. Ed. Bucureşti.

[1996] Polynomial Algorithms in Computer Algebra. R. ZARISKI. Universităţii „Al. [1958] Commutative Algebra. [1987] Galois' Theory of Algebraic Equations. O.. VAN DER WAERDEN. I. [1993] Introducere în teoria grupurilor. Algebra I (Achte auflage der Modernen Algebra) Springer-Verlag. [1971].L. New York. A . 48. Dover Publications. 55.-P. Ed. TIGNOL.L. [1967]. B. SAMUEL. [1960] Commutative Algebra. Ed Universităţii „Al. Springer Wien New York. J. 53. Algebra II (Fünfte auflage der Modernen Algebra) Springer-Verlag. History of Algebra. 50. I. 49. 52. 47.J. ŞTEFĂNESCU. F. [1950] Algebraic Curves. Iaşi. I. P. Berlin. vol. M. VAN DERWAERDEN. Cuza”. Springer-Verlag. B. Princeton. Solvable389 groups 46. SAMUEL.L. I. [2000] Capitole speciale de structuri algebrice. vol. Iaşi.. O. WALKER. P. Berlin. Princeton. II. Cuza”. [1985]. 51. TOFAN. VAN DER WAERDEN. B. WINKLER.6.. Berlin. ZARISKI. Longman Scientifical and Technical. 54.

Sign up to vote on this title
UsefulNot useful