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

**A Course In Commutative Algebra
**

Robert B. Ash

Preface

This is a text for a basic course in commutative algebra, written in accordance with the following objectives. The course should be accessible to those who have studied algebra at the beginning graduate level. For general algebraic background, see my online text “Abstract Algebra: The Basic Graduate Year”, which can be downloaded from my web site www.math.uiuc.edu/∼ r-ash This text will be referred to as TBGY. The idea is to help the student reach an advanced level as quickly and eﬃciently as possible. In Chapter 1, the theory of primary decomposition is developed so as to apply to modules as well as ideals. In Chapter 2, integral extensions are treated in detail, including the lying over, going up and going down theorems. The proof of the going down theorem does not require advanced ﬁeld theory. Valuation rings are studied in Chapter 3, and the characterization theorem for discrete valuation rings is proved. Chapter 4 discusses completion, and covers the Artin-Rees lemma and the Krull intersection theorem. Chapter 5 begins with a brief digression into the calculus of ﬁnite diﬀerences, which clariﬁes some of the manipulations involving Hilbert and Hilbert-Samuel polynomials. The main result is the dimension theorem for ﬁnitely generated modules over Noetherian local rings. A corollary is Krull’s principal ideal theorem. Some connections with algebraic geometry are established via the study of aﬃne algebras. Chapter 6 introduces the fundamental notions of depth, systems of parameters, and M -sequences. Chapter 7 develops enough homological algebra to prove, under approprate hypotheses, that all maximal M -sequences have the same length. The brief Chapter 8 develops enough theory to prove that a regular local ring is an integral domain as well as a Cohen-Macaulay ring. After completing the course, the student should be equipped to meet the Koszul complex, the AuslanderBuchsbaum theorems, and further properties of Cohen-Macaulay rings in a more advanced course.

Bibliography

Atiyah, M.F. and Macdonald, I.G., Introduction to Commtative Algebra, Addison-Wesley 1969 Balcerzyk, S. and Jozeﬁak, T., Commutative Noetherian and Krull Rings, Wiley 1989 Balcerzyk, S. and Jozeﬁak, T., Commutative Rings: Dimension, Multiplicity and Homological Methods, Wiley 1989 Eisenbud, D., Commutative Algebra with a view toward algebraic geometry, SpringerVerlag 1995 Gopalakrishnan, N.S., Commutatilve Algebra, Oxonian Press (New Delhi) 1984 Kaplansky, I., Commutative Rings, Allyn and Bacon 1970

2 Kunz, E., Introduction to Commutative Algebra and Algebraic Geometry, Birkh¨user a 1985 Matsumura, H., Commutatlive Ring Theory, Cambridge 1986 Raghavan, S., Singh, B., and Sridharan, S., Homological Methods in Commutative Algebra, Oxford 1975 Serre, J-P., Local Albegra, Springer-Verlag 2000 Sharp, R.Y., Steps in Commutative Algebra, Cambridge 2000 c copyright 2003, by Robert B. Ash. Paper or electronic copies for noncommercial use may be made freely without explicit permission of the author. All other rights are reserved.

Table of Contents

Chapter 0 Ring Theory Background

0.1 Prime Avoidance 0.2 Jacobson Radicals, Local Rings, and Other Miscellaneous Results 0.3 Nakayama’s Lemma

Chapter 1

1.1 1.2 1.3 1.4 1.5 1.6

Primary Decomposition and Associated Primes

Primary Submodules and Ideals Primary Decomposition Associated Primes Associated Primes and Localization The Support of a Module Artinian Rings

Chapter 2

Integral Extensions

2.1 Integral Elements 2.2 Integrality and Localization 2.3 Going Down

Chapter 3

Valuation Rings

3.1 Extension Theorems 3.2 Properties of Valuation Rings 3.3 Discrete Valuation Rings

Chapter 4

Completion

4.1 Graded Rings and Modules 4.2 Completion of a Module 4.3 The Krull Intersection Theorem 1

2 7.2 Chapter 5 5.5 5.1 Systems of Parameters 6.6 Dimension Theory The Calculus of Finite Diﬀerences Hilbert and Hilbert-Samuel Polynomials The Dimension Theorem Consequences of the Dimension Theorem Strengthening of Noether’s Normalization Lemma Properties of Aﬃne k-Algebras Chapter 6 Depth 6.4 Homological Methods Homological Dimension: Projective and Global Injective Dimension Tor and Dimension Application Chapter 8 Regular Local Rings 8.2 Regular Sequences Chapter 7 7.1 5.4 5.2 5.1 7.3 7.3 5.1 Basic Deﬁnitions and Examples Exercises Solutions .

in other words. but after the preliminary reduction (see the beginning of the proof). Note that P1 / / and P2 need not be prime for this argument to work. Suppose the result is false. Now for all i = 1. but P3 . . with I ⊆ P1 and I ⊆ P2 . but as ∈ P1 ∪ · · · ∪ Ps−1 . Thus a1 + a2 ∈ I ⊆ P1 ∪ P2 . ♣ / It may appear that we only used the primeness of Ps . with P1 and P2 not necessarily prime. so ai ∈ Pi . s − 1 we have ai ∈ Ps . We may assume that I is not contained in the union of any collection of s − 1 of the Pi ’s. P2 . and observe that a1 a2 · · · as−1 ∈ P1 ∩ · · · ∩ Ps−1 . / which does not belong to P1 ∪ · · · ∪ Ps−1 . / / so a cannot be in Ps . contradicting the hypothesis.1 Prime Avoidance Let P1 . . . contradicting a1 . be ideals in a ring R. . . else as would belong to this set. I is contained in the union of all the P ’s. Ps prime (if s ≥ 3). then we can avoid all the Pj simultaneously. a2 ∈ P1 . . . Then a1 ∈ P1 . (If so. a2 ∈ I. 1 . that is. Let I be any ideal of R. Similarly. we can simply replace s by s − 1.1. Ps . Thus a ∈ I and a ∈ P1 ∪ · · · ∪ Ps . / / / so a1 + a2 ∈ P2 . 0. hence a1 · · · as−1 ∈ Ps because Ps is prime.Chapter 0 Ring Theory Background We collect here some useful results that might not be covered in a basic graduate algebra course. i=1 Proof. then for some i we have I ⊆ Pi . if I ⊆ ∪s Pi . . Let a = (a1 · · · as−1 ) + as . 0. we can ﬁnd a single element in I that is in none of the Pj . for each j we can ﬁnd an element in I but not in Pj . . We will state and prove the contrapositive. By hypothesis. Now assume s > 2. a1 ∈ P2 . . . s ≥ 2. a2 ∈ P2 . so a1 + a2 ∈ P1 . But as ∈ Ps . it may very well happen that one of the other Pi ’s now occupies the slot that previously housed Ps .) Thus / for each i we can ﬁnd an element ai ∈ I with ai ∈ P1 ∪ · · · ∪ Pi−1 ∪ Pi+1 ∪ · · · ∪ Ps . . The idea is that if we can avoid the Pj individually.1 Prime Avoidance Lemma With I and the Pi as above. First assume s = 2.

♣ . is contained in some maximal ideal M. there is an element x ∈ S \ M. and therefore 1 ∈ M. a ∈ J(R). But then 1 ∈ M. then it must belong to M. Conversely. We have I ⊂ R iﬀ I. If N is another maximal ideal. Proof. then M + Ra = R. The “only if” part holds because any ideal is contained in its radical.3 Lemma Let S be any subset of R.2. and let a ∈ M. If 1 + ax is not a unit. and let I be the ideal generated by S. The contrapositive says that I = R iﬀ ∀M ∃x ∈ S such that x ∈ M. so 1 + ax is a unit. Proof. Thus every term in the sum belongs either to I or to J. proving the claim. Then 1 = (a + b)m+n = i+j=m+n m+n i j ab . ♣ 0. which is the ideal of nonunits. Then R is a local ring (a ring with a unique maximal ideal.2. and Other Miscellaneous Results Lemma Let J(R) be the Jacobson radical of the ring R. By (0. then it generates a proper ideal.1 Jacobson Radicals. But then a ∈ M. 1 ∈ I + J. hence ax ∈ M.2 CHAPTER 0. If 1 + a is not a unit. RING THEORY BACKGROUND 0. the intersection of all maximal ideals of R. Thus M ⊆ N . a contradiction. If x = −y.2. a contradiction. that is.1). and since both ideals are maximal. Thus for some b ∈ M and y ∈ R we have b + ay = 1.4 Lemma Let I and J be ideals of the ring R. ♣ 0. if a fails to belong to a maximal ideal M. If a ∈ M. Proof.2 0. necessarily M) if and only if every element of 1 + M is a unit. Conversely. then either i ≥ m or j ≥ n. Consequently. I ⊂ R iﬀ ∃M such that ∀x ∈ S we have x ∈ M. Then I = R iﬀ for every maximal ideal M. then 1 + ax = b ∈ M. assume that every element of 1+M is a unit. Proof. so 1 + ax cannot be a unit (else 1 ∈ M).2. Suppose R is a local ring.2. i Now if i + j = m + n. We claim that M ⊆ J(R).2 Lemma Let M be a maximal ideal of the ring R. Then I + J = R iﬀ √ I+ √ J = R. then ax ∈ M for every x ∈ R. In other words. Local Rings. Then a ∈ J(R) iﬀ 1 + ax is a unit for every x ∈ R. ♣ / 0. hence to I + J. they must be equal. then M = J(R) ⊆ M ∩ N . equivalently S. Assume a ∈ J(R). hence 1 + ax belongs to some maximal ideal M. Therefore R is a local ring. hence M = J(R). Thus assume that 1 = a + b with am ∈ I and bn ∈ J.

But by deﬁnition of I. we have (In − A)x = 0. we have equations of the form n n xi = j=1 aij xj . referred to as Nakayama’s lemma. Sometimes. Akizuki and and as a result. NAKAYAMA’S LEMMA 3 0. and M = N + IM . We can now prove that f is injective.3. we give an example of the determinant trick . we claim that IM = M .) Let I = (X).2 Theorem If M is a ﬁnitely generated R-module and f : M → M is a surjective homomorphism. Then 0 = [1 + Xh(X)]x = [1 + h(X)X]x = x + h(X)f (x) = x + 0 = x. and IM = M . where A = (aij ) and x is a column vector whose coeﬃcients are the xi . ♣ Here is a generalization of a familiar property of linear transformations on ﬁnitedimensional vector spaces. m = f (x) for some x ∈ M . then by the hypothesis that f is surjective. 0. a popular abbreviation for the lemma is 0. If In is the n by n identity matrix. ♣ In (0. Then there exists a ∈ I such that (1 + a)M = 0. “surjective” by “injective”.2). see (2. xn generate M .3. where ∆ is the determinant of (In − A).3. . . . hence ∆M = 0. x ∈ M . I an ideal of R contained in the Jacobson radical J(R). then f is an isomorphism. so m ∈ IM . with aij ∈ I. etc.1). (Thus X 2 x = f (f (x)). Proof. Thus (1+g)M = [1 + Xh(X)]M = 0. then M = 0. But if we look at the determinant of In − A. The equations may be written as j=1 (δij − aij )xj = 0. We can make M into an R[X]-module via Xx = f (x). .3. Since IM = M . . there exists g = g(X) ∈ I such that (1 + g)M = 0. NAK. then M = N . we obtain ∆x = 0. (b) If N is a submodule of the ﬁnitely generated R-module M . 0. and I an ideal of R such that IM = M . we see that it is of the form 1 + a for some element a ∈ I. For if m ∈ M . let f (x) = nx on is injective but not surjective. I an ideal of R contained in the Jacobson radical J(R).3.0.2) for another illustration. But X ∈ I. Proof. Let x1 .1 Theorem Let M be a ﬁnitely generated R-module. Premultiplying by the adjoint of (In − A). Suppose that x ∈ M and f (x) = 0. g must be of the form Xh(X) with h(X) ∈ R[X].3 Nakayama’s Lemma First.3.1. Thus ∆xi = 0 for all i. For example. If n ≥ 2. we cannot replace the integers. By (0. then f The next result is usually Krull are given some credit.3 NAK (a) If M is a ﬁnitely generated R-module. and therefore Xx = f (x) = m.

(ii) If {x1 + JM. By NAK. we deﬁne an equivalence relation on R × S by (a.4 Proposition Let R be a local ring with maximal ideal J. and assume that S is multiplicative. contradicting the assumption that V is n-dimensional. d) iﬀ for some s ∈ S we have s(ad − bc) = 0. . also called the ring of fractions of R by S. (iii) A generating set S for M with more than n elements determines a spanning set for V . S will / be the complement of a prime ideal. and the product of a/b and c/d is deﬁned as ac/bd. in other words.1 Construction of the Localized Ring If S is a multiplicative subset of the ring R.2. 1 + a is a unit by (0. The additive inverse of a/b is −(a/b) = (−a)/b. then S −1 R is a ﬁeld. which must contain a basis with exactly n elements. Section 2. xn + JM } is a basis for V over k. see TBGY. then {x1 . If a proper subset of the xi were to generate M . By (ii). If R is an integral domain and S = R \ {0}.8. Multiplying the equation (1 + a)M = 0 by the inverse of 1 + a. . The multiplicative identity is 1/1. so does ab. (a) By (0.4 Localization Let S be a subset of the ring R. . 1 ∈ S. b) ∼ (c.3. Since M is ﬁnitely generated over R. We would like to divide elements of R by elements of S to form the localized ring S −1 R.1). . b).1). so is S −1 R. If R is an integral domain. Proof. we get M = 0. so the xi generate M . We make the set of fractions into a ring in a natural way.4. Since the xi + JM generate V = M/JM . which coincides with s/s for every s ∈ S. and let V = M/JM . . Since I ⊆ J(R). (b) By hypothesis. (i) Since J annihilates M/JM . The additive identity is 0/1. which coincides with 0/s for every s ∈ S. In the case of interest to us. and if a and b belong to S. we have M = N + JM . the fraction ﬁeld of R. To summarize: S −1 R is a ring.3. RING THEORY BACKGROUND Proof. 0. then the corresponding subset of the xi + JM would generate V . V is a k-module. There is no diﬃculty when R is an integral domain. The sum of a/b and c/d is deﬁned as (ad + bc)/bd. and the result follows from (a). V is a ﬁnite-dimensional vector space over k. For full details. . that is. M/N = I(M/N ).4 CHAPTER 0. . 0. We will sketch the general construction for arbitrary rings R. (1 + a)M = 0 for some a ∈ I. a vector space over k. ♣ 0. (iii) Any two minimal generating sets for M have the same cardinality. 0 ∈ S. . If a ∈ R and b ∈ S. because in this case all division takes place in the fraction ﬁeld of R. Let M be a ﬁnitely generated R-module. S cannot be minimal. M = N . . Then (i) V is a ﬁnite-dimensional vector space over the residue ﬁeld k = R/J. xn } is a minimal set of generators for M . we deﬁne the fraction a/b as the equivalence class of (a. n (ii) Let N = i=1 Rxi . ♣ Here is an application of NAK.

If I is an ideal of R. In particular. with a ∈ R. so a/1 = b/s for some b ∈ I.4. and let a ∈ h−1 (S −1 I). then h is a monomorphism. s ∈ S. An integral domain can be embedded in its fraction ﬁeld. the left side is contained in the right side. then S −1 I is an ideal of S −1 R. Then h(a) = a/1 ∈ S −1 I. s. I is an ideal by the basic properties of preimages of sets. then h(a) = a/1 ∈ S −1 I. so R can be embedded in S −1 R. If J is another ideal of R. s ∈ S. where S consists of all non-divisors of 0 in R.1)]. 0. 0. Then h(a) = a/1 = (a/s)(s/1) ∈ J. s ∈ S}. If J is an ideal of S −1 R and I = h−1 (J). then I ⊆ h−1 (S −1 I). we have a ∈ I.4. then 1/1 = a/s for some a ∈ I. LOCALIZATION 5 There is a natural ring homomorphism h : R → S −1 R given by h(a) = a/1. There exists u ∈ S such that u(at − bs) = 0. Conversely. Since I is prime. (iv) S −1 I is a proper ideal iﬀ S ∩ I = ∅. so at = st ∈ S ∩ I. Proof. Let a/s ∈ S −1 I. Proof. then (i) S −1 (I + J) = S −1 I + S −1 J. let a/s ∈ J. if −1 S I = S −1 R. ♣ Ideals in S −1 R must be of a special form. with a ∈ I and s ∈ S. 0. then 1/1 = s/s ∈ S −1 I.4. Finally. t ∈ S. If S has no zero-divisors. Conversely.3 Lemma Let h be the natural homomorphism from R to S −1 R [see (0.4. Our goal is to study the relation between prime ideals of R and prime ideals of S −1 R. s t st st st To prove (iii).4. There exists t ∈ S such that t(s − a) = 0. Thus ast = bt ∈ I. if s ∈ S ∩ I. Thus assume that I is prime and disjoint from S. There exists t ∈ S such that t(as − b) = 0.4 Lemma If I is any ideal of R. where a ∈ I. (iii) S −1 (I ∩ J) = (S −1 I) ∩ (S −1 J). (ii) S −1 (IJ) = (S −1 I)(S −1 J). Then a/s = uat/ust = ubs/ust ∈ S −1 (I ∩ J). so a ∈ I and a/s ∈ S −1 I. so S −1 I = S −1 R. If a ∈ I. (ii) and (iii). deﬁne S −1 X = {x/s : x ∈ X. let a/s = b/t. The reverse inclusions in (i) and (ii) follow from a b at + bs a b ab + = . a ring R can be embedded in its full ring of fractions S −1 R. s ∈ S. Then a/1 = h(a) ∈ J. ♣ . Proof.0. and that in (i). then I is an ideal of R and S −1 I = J.2 Lemma If X is any subset of R. so a/s = (a/1)(1/s) ∈ J. The deﬁnitions of addition and multiplication in S −1 R imply that S −1 R is an ideal. ♣ Prime ideals yield sharper results. There will be equality if I is prime and disjoint from S. b ∈ J. with st ∈ I because / S ∩ I = ∅. = .

that is.2) part (iv). so H ⊆ I.4. (iii) implies (i): If I is the ideal of nonunits. S −1 (h−1 (Q)) = Q. (ii) implies (iii): If a and b are nonunits. RING THEORY BACKGROUND 0.5 Lemma If I is a prime ideal of R disjoint from S. hence is contained in the unique maximal ideal I. In this case. s. Then ab/st = c/u for some c ∈ I. a ring with a unique maximal ideal. By part (iv) of (0. then S −1 I is a prime ideal of S −1 R. u ∈ S. There exists v ∈ S such that v(abu − cst) = 0. Since I is / prime.4.8 Proposition For a ring R. ♣ 0. If H is any proper ideal. Let (a/s)(b/t) = ab/st ∈ S −1 I.6 Theorem There is a one-to-one correspondence between prime ideals P of R that are disjoint from S and prime ideals Q of S −1 R. and uv ∈ I because S ∩ I = ∅. Proof. hence a ∈ I or b ∈ I. then H cannot contain a unit. then S = R \ P is a multiplicative set. Q = S −1 (h−1 (Q)) = S −1 R. the following conditions are equivalent. contradicting the hypothesis. First. and h−1 (Q) is a prime ideal by the basic properties of preimages of sets. with a. Thus the maps P → S −1 P and Q → h−1 (Q) are inverses of each other.6 CHAPTER 0.5). 0. b ∈ R. (ii) There is a proper ideal I of R that contains all nonunits of R. If h−1 (Q) meets S. h−1 (S −1 P ) = P . Therefore I is the unique maximal ideal. then I is maximal.4. we give some conditions equivalent to the deﬁnition of a local ring. Thus abuv = cstv ∈ I. We are going to show that RP is a local ring. and call it the localization of R at P . Proof. so I = R.4). then I contains a unit. Therefore either a/s or b/t belongs to S −1 I. S −1 I is a proper ideal. By (0. then (a) is a proper ideal. and the result follows.4. given by P → S −1 P and Q → h−1 (Q). t ∈ S.2).4. (i) implies (ii): If a is a nonunit. S −1 P is a prime ideal. If not. because any larger ideal J would have to contain a unit.3).4. (i) R is a local ring. 0. ab ∈ I.4. a contradiction. (iii) The set of nonunits of R is an ideal. so are a + b and ra. ♣ The sequence of lemmas can be assembled to give a precise conclusion. and by (0. By (0.7 Deﬁnitions and Comments If P is a prime ideal of R. we write RP for S −1 R. Proof.4. then by (0. so J = R.4. ♣ .

Therefore S −1 P is a proper ideal containing every maximal ideal. t) equivalent if for some u ∈ S. I ⊆ P . If x.4. . then S −1 (M + N ) = S −1 M + S −1 N and S −1 (M ∩ N ) = (S −1 M ) ∩ (S −1 N ). st st In this way.1) to form the localization of M by S. s) is denoted by x/s. s) and (y. so it must be the unique maximal ideal.11 Localization of Modules If M is an R-module and S a multiplicative subset of R.4. Q = S −1 I for some prime ideal I of R that is disjoint from S = R \ P . Let Q be a maximal ideal of RP . Q = S −1 I ⊆ S −1 P .6). Proof.4. if M and N are submodules of an R-module L. s t st If a/s ∈ S −1 R and x/t ∈ S −1 M . because the product of an element of I and an arbitrary element of R belongs to I. and addition is deﬁned by x y tx + sy + = . Exactly as in (0. y ∈ M and s. Then Q is prime. If S −1 P = RP = S −1 R. then by (0. which is impossible. There is no ambiguity. In other words.9 Theorem RP is a local ring. t ∈ S.2) part (iv).4. we deﬁne ax ax = . The equivalence class of (x.2). S −1 M becomes an S −1 R-module.4.4. P is not disjoint from S = R \ P . we can essentially repeat the construction of (0.10 Remark It is convenient to write the ideal S −1 I as IRP .4. 0. so by (0.4. ♣ 0.0. we have u(tx − sy) = 0. LOCALIZATION 7 0. we call (x. Consequently. and thereby divide elements of M by elements of S.

Nilpotence means that for some positive integer n. (The radical of P n is the intersection of all prime ideals containing P n . R/Q = 0 and every zero-divisor in R/Q is nilpotent. nilpotence gives an M = a(an−1 M ) ⊆ N . an belongs to the annihilator of M/N . so x ∈ P n . (Note that rM (N ) is proper because λ1 is injective. and in fact it is a prime ideal. then xn ∈ P n .Chapter 1 Primary Decomposition and Associated Primes 1. √ A useful observation is that if P is a prime ideal.) If P = rM (N ). Equivalently.1 Primary Submodules and Ideals Deﬁnitions and Comments If N is a submodule of the R-module M . and injectivity gives an−1 M ⊆ N . because ann(R/I) = I. that is. Injectivity means that for all x ∈ M . Note that λa cannot be both injective and nilpotent. then P n = P for all n ≥ 1. For if λa and λb fail to be injective. denoted by rM (N ). it is an ideal of R. so M = N . an M ⊆ N . contradicting the assumption that N is proper. Thus if N is a primary submodule of M . we say that N is P -primary. a belongs to the radical of the annihilator of M/N . Conversely.) Specializing to M = R and replacing a by y. (Note that a ∈ ann(R/I) iﬀ aR ⊆ I iﬀ a = a1 ∈ I. Inductively. and a ∈ R. Since rM (N ) is the radical of an ideal. M ⊆ N . We say that N is a primary submodule of M if N is proper and for every a.1 1. then either x ∈ Q or y n ∈ Q for some n ≥ 1. Thus √ √ P n ⊆ P . √ If I is any ideal of R. one of which is P . Equivalently. let λa : M/N → M/N be multiplication by a. then rR (I) = I. if x ∈ P . denoted by ann(M/N ). If so. we deﬁne a primary ideal in a ring R as a proper ideal Q such that if xy ∈ Q. we have ax ∈ N ⇒ x ∈ N . λa is either injective or nilpotent. then rM (N ) is the set of all a ∈ R such that λa is not injective.) 1 . so does λab = λa ◦ λb .1.

then I is M-primary. We may assume that k = 2. 2. i = 1. It remains to show that rM (N ) = P .1. Thus 1 = 1k = (m + rb)k = mk + sb for some s ∈ R. The following result justiﬁes this process. ♣ I is a maximal ideal M.2). hence mk ∈ I for some k ≥ 1.1 Primary Decomposition Deﬁnitions and Comments A primary decomposition of the submodule N of M is given by N = ∩r Ni . so for some m ∈ M and r ∈ R we have m + rb = 1.2 If √ Lemma √ Proof.3 Deﬁnition The proper submodule N of M is irreducible if N cannot be expressed as N1 ∩ N2 with N properly contained in the submodules Ni . . an induction argument takes care of larger values. If a ∈ P . if a ∈ rM (N ) then a belongs to rM (Ni ) for i = 1. it follows that M + Rb = R. ♣ 1.3 Corollary If M is a maximal ideal. Assume for the moment that rM (N ) = P . and the result follows from (1. 1. ♣ We now prepare to prove that every submodule of a Noetherian module has a primary decomposition. We can always extract a reduced primary decomposition from an unreduced one. Now √ m ∈ M = I. so a ∈ rM (N ). If a ∈ R. .1). PRIMARY DECOMPOSITION AND ASSOCIATED PRIMES 1. and therefore a ∈ P.1.2.2.2. then there are positive integers n1 and n2 such that an1 M ⊆ N1 and an2 M ⊆ N2 . x ∈ M. Suppose that ab ∈ I and b does not belong to I = M. Nk are P -primary. Mn = M. Let N = N1 ∩ N2 and rM (N1 ) = rM (N2 ) = P . .2 Lemma If N1 . 1. where the i=1 Ni are Pi -primary submodules. then Mn is M-primary for every n ≥ 1. i=1 Proof. by discarding those Ni that contain ∩j=i Nj and intersecting those Ni that are P -primary for the same P . √ Proof. Then by maximality of M.2 CHAPTER 1. As we observed in (1. ax ∈ N . Multiply by a to get a = amk + sab ∈ I. 1. . then ∩k Ni is P -primary. Therefore an1 +n2 M ⊆ N . then since N1 and N2 are P -primary. and a ∈ rM (N ).2 1. we have / x ∈ N1 ∩ N2 = N . . Conversely.1. The decomposition is reduced if the Pi are distinct and N cannot be expressed as the intersection of a proper subcollection of the Ni .1. 2.

1 Associated Primes Deﬁnitions and Comments Let M be an R-module. We say that P is an associated prime of M (or that P is associated to M ) if P is the annihilator of some nonzero x ∈ M . Let ϕ = λi . Thus x ∈ p−1 (ker ϕ) = N1 (because y = p(x) ∈ ker ϕ). If x ∈ N1 ∩ N2 . so we can write N = N1 ∩ N2 . ♣ 1. so λi can’t be 0 (because a i a M ⊆ N ). and the result / follows. Suppose a a x ∈ ker ϕ ∩ im ϕ.3 1.2. Consequently. If not. and set N1 = p−1 (ker ϕ).3. so ker ϕ = 0.2 Proposition The prime ideal P is associated to M if and only if there is an injective R-module homomorphism from R/P to M . N1 and N2 can be expressed as ﬁnite intersections of irreducible submodules. then p(x) = 0 ∈ ker ϕ ∩ im ϕ.) Here is a useful characterization of associated primes. say a a at ker λi . then N has a primary decomposition. so x = ϕ(y) = 0. x ∈ N . we will show that N is properly contained in both N1 and N2 . (The standard notation is Ass(M). then N is primary. By deﬁnition of S.5 Theorem If N is a proper submodule of the Noetherian module M .4) applies. If S is nonempty. so p(x) = 0. ASSOCIATED PRIMES 3 1. so x ∈ N1 ∩ N2 . Please do not use this regrettable terminology. Similarly. then for some a ∈ R. but x ∈ N (because p(x) = y = 0). Therefore if N is a submodule of M . contradicting N ∈ S. . N ⊂ N2 . N ⊂ N1 .3. The set of associated primes of M is denoted by AP(M ). Since p is surjective. and P a prime ideal of R. hence a reduced primary decomposition. Let S be the collection of all submodules of M that cannot be expressed in this form.2. then AP(N ) ⊆ AP(M ). Now λa is not injective. and λa is not nilpotent. 1. hence so can N . a contradiction. then p(x) belongs to both ker ϕ and im ϕ. Finally. N must be reducible. N ⊂ N2 (with 0 = y ∈ im ϕ). then ker ϕ = ker ϕ2 and we claim that ker ϕ ∩ im ϕ = 0. Thus S is empty. so that (1. if x ∈ N . ♣ 1. so N is reducible. Then 0 = ϕ(x) = ϕ2 (y). λa : M/N → M/N is neither injective nor nilpotent. im ϕ = 0. By maximality of N .3. Conversely.1. there exists x ∈ M such that p(x) = y. We will show that N can be expressed as a ﬁnite intersection of irreducible submodules of M . Let p : M → M/N be the canonical epimorphism. Proof. and let x = ϕ(y).4 Proposition If N is an irreducible submodule of the Noetherian module M .2. The chain ker λa ⊆ ker λ2 ⊆ ker λ3 ⊆ · · · terminates by the ascending chain condition. Choose a nonzero element y ∈ ker ϕ. We will prove that N = N1 ∩ N2 . Proof. then S has a maximal element N (because M is Noetherian). in other words. N2 = −1 p (im ϕ). so y ∈ ker ϕ2 = ker ϕ.

and are sometimes unique.2). Then ∪{P : P ∈ AP(M )} ⊆ z(M ). the set of all r ∈ R such that rx = 0 for some nonzero x ∈ M . Consequently. If we can show that I is prime.4 CHAPTER 1. But P ∈ AP(Rx) ⊆ AP(M ) by (1.4 shows that the annihilator of any nonzero element of R/P is P .3. ♣ Now we prove a companion result to (1. so by (1. then x = 1x = 0. see (1. if an injective R-homomorphism from R/P to M exists. so rx = 0.2). Conversely. as desired. If P is the annihilator of x = 0. Then Rx = 0. Rx has an associated prime P = ann(bx). There are no nonzero elements in the zero module. By (1. hence no associated primes. We will show that P = annR (x).3. which is nonzero in R/P . . The converse holds if R is a Noetherian ring. and therefore r ∈ annR (x). 1.3. Suppose that Q = ann(r + P ) with r ∈ P .3) [assuming R Noetherian]. the set of elements r ∈ R such that rx = 0. ♣ 1.3. AP(R/P ) = {P }. with equality if R is Noetherian. for if I = R.3 Proposition If M = 0. ♣ Associated primes exist under wide conditions.3.3. we must show that Q = P . Proof. If rx = 0. The next result gives us considerable information about the elements that belong to associated primes.3.2). x = 0.4 Proposition For any prime ideal P . x = 0. Proof. Let ab ∈ I with a ∈ I.3. 1. ♣ 1. so a ∈ P . Then s ∈ Q iﬀ sr ∈ P iﬀ s ∈ P (because P is / prime). we have I ∈ AP(M ). so r ∈ P . let x be the image of 1 + P . P is an associated prime of R/P because there certainly is an Rmonomorphism from R/P to itself. If r ∈ P . Proof. r + P = 0. PRIMARY DECOMPOSITION AND ASSOCIATED PRIMES Proof. then AP(M ) is empty. The inclusion follows from the deﬁnition of associated prime. the desired homomorphism is given by r +P → rx. so b ∈ ann(ax).1). The ideal I must be proper.5 Remark Proposition 1. then by injectivity. Therefore a ∈ ∪{P : P ∈ AP(M )}. a contradiction. and the maximality of I gives I = ann(ax). Since ax = 0 we have abx = 0. Assuming that M = 0 and R is Noetherian. with ax = 0. / But I = ann x ⊆ ann(ax). that is.6 Theorem Let z(M ) be the set of zero-divisors of M . Thus assume a ∈ z(M ). By injectivity. x ∈ M. b ∈ I. there is a maximal element I = annR x in the collection of all annihilators of nonzero elements of M .3. If Q ∈ AP(R/P ).3. Then abx = 0 but ax = 0. then r + P = 0.

The image of the monomorphism is contained in the direct sum of ﬁnitely many components. the zero module can be expressed as ∩r Ni . then there is a monomorphism from R/P to ⊕Mj . 1. x ∈ M .7) when the index set is ﬁnite.3. By (1.3.3. Case 2: L = 0. (If h(r + P ) belongs to N .3. By (1. This takes us back to the ﬁnite case. The result follows from (1.2). AP(M1 ⊕ M2 ⊕ M3 ) ⊆ AP(M1 ) ∪ AP(M/M1 ) = AP(M1 ) ∪ AP(M2 ⊕ M3 ) ⊆ AP(M1 ) ∪ AP(M2 ) ∪ AP(M3 ).5). . / we have Pi = rM (Ni ) (see (1. and H is isomorphic to R/P via h. For example. Since Pi is ﬁnitely generated. ♣ We now establish the connection between associated primes and primary decomposition. Proof. Proof. Set H = h(R/P ) and L = H ∩ N . . In particular. and let h : R/P → M be a monomorphism. Let P be an associated prime of M . there are only ﬁnitely many associated primes. Then the map from H to M/N given by h(r + P ) → h(r + P ) + N is a monomorphism.7 Proposition If N is a submodule of M .2. Proof. so that P = ann(x).9 Theorem Let M be a nonzero ﬁnitely generated module over the Noetherian ring R. .1.1)). Renumber the Ni so that x ∈ Ni for 1 ≤ i ≤ j and x ∈ Ni for j + 1 ≤ i ≤ r. it must belong to H ∩ N = 0. ann x = P .1. Let P ∈ AP(M ). then x must belong to both H and N . . If L has a nonzero element x. there is a monomorphism from R/P to M/N . Then AP(M ) = i=1 {P1 . where Ni is Pi -primary.5).3. ♣ 1. Pini M ⊆ Ni for some ni ≥ 1. Therefore j r ( i=1 Pini )x ⊆ i=1 Ni = (0) . every proper submodule of M has a reduced primary decomposition. if P is an associated prime of the direct sum. so by deﬁnition of H. a ﬁnite set.) Thus H is isomorphic to a submodule of M/N .3. so that by (1. Pr }. so P ∈ AP(N ). ASSOCIATED PRIMES 5 1.8 Corollary AP( j∈J Mj = j∈J AP(Mj ).3. Thus x ∈ N and the annihilator of x coincides with the annihilator of some nonzero element of R/P . Thus P ∈ AP(M/N ). Since Ni is Pi -primary. Case 1: L = 0. as R/P is generated as an R-module by the single element 1 + P . and show that under wide conditions. x = 0. the right side is contained in the left side. In general. then AP(M ) ⊆ AP(N ) ∪ AP(M/N ).

a reduced primary decomposition of (0) in M/N is given by (0) = ∩r Ni /N . Pr . Then N is P -primary iﬀ AP(M/N ) = {P }.1)] the associated primes of M/N are determined by N . we show that each Pi is an associated prime. Without loss of generality. so we can choose x ∈ N2 ∩· · ·∩Nr with x ∈ N1 . as claimed.12 Deﬁnitions and Comments Let N = ∩r Ni be a reduced primary decomposition. . we may take i = 1. so λa is not injective and therefore must be nilpotent. . Now N1 is P1 -primary. the proof n will be complete upon showing that P1 is the annihilator of y.6 CHAPTER 1. so P1 ⊆ ann y. Proof. The “only if” part follows from the displayed equation above. PRIMARY DECOMPOSITION AND ASSOCIATED PRIMES so ∩j Pini ⊆ ann(x) = P .9). . so / n−1 n as in the preceding paragraph. N1 does not contain the intersection of the other Ni ’s. then ay ∈ N1 but y ∈ N1 . If N = ∩r Ni is a i=1 reduced primary decomposition of the submodule N . i=1 We say that Ni is an isolated (or minimal ) component if Pi is minimal. i=1 AP(M/N ) = {P1 . then N coincides with a P -primary submodule N . Ni is an embedded component (see Exercise 5 for an example). for some n ≥ 1 we have P1 x ⊆ N1 but P1 x ⊆ N1 . . ♣ We can now say something about uniqueness in primary decompositions. with associated primes P1 . Embedded components arise in algebraic geometry in situations where one irreducible algebraic set is properly contained in another. Otherwise. . By the correspondence theorem. Conversely. We have P1 y ⊆ P1 x ⊆ N1 r n r r and x ∈ ∩i=2 Ni . i = 1. Thus P1 y ⊆ ∩i=1 Ni = (0).11 Corollary Let N be a submodule of M (ﬁnitely generated over the Noetherian ring R). 1. Consequently.3. .3.3. To verify this.3. ♣ 1. . . r. and hence N (= N ) is P -primary. . if a ∈ R and ay = 0. . there is a j rather than an r on the left i=1 side of the inclusion. and Ni /N is Pi -primary. But [see (1. . let a ∈ P . Since the decomposition is reduced. . so that every associated prime must be one of the Pi . By (1. so P1 x ⊆ ∩i=2 Ni .10 First Uniqueness Theorem Let M be a ﬁnitely generated module over the Noetherian ring R.3. if P is the only associated prime of M/N . Conversely. Pi ⊆ P for some i ≤ j. (Take n−1 0 / P1 x = Rx and recall that x ∈ N1 . that is Pi does not properly contain any Pj . Proof. . We claim that Pi = P . Then ax = 0 and x ∈ Ni . j = i.) Since P is prime. then (regarding M and R as ﬁxed) the Pi are uniquely determined by N . Thus a ∈ rM (N1 ) = P1 . and Ni is Pi -primary. ♣ 1.) If we choose y ∈ P1 x \ N1 (hence y = 0). / a ∈ rM (Ni ) = Pi . (By our renumbering. so λa : M/N1 → M/N1 is not / injective and is therefore nilpotent. On the other hand. 1 ≤ i ≤ r. Pr }.

so sx = 0 and the proof is complete. then PS = P RS ∈ APRS (MS ). namely P . and MS the localization of the R-module M by S. Thus P ⊆ ann(sx).4. the set of prime ideals of R not meeting S. ♣ . so x/1 is indeed nonzero.2 Lemma The map of (1.) Proof. (By the above discussion. so by (1. Then (b/1)(x/t) = bsx/st = 0. If P ∈ APR (M ). the set of associated primes of the RS -module MS . As in (1. hence sax = 0 for all a ∈ P . Now b . Then (ai /1)(x/t) = 0. The set of associated primes of the R-module M will be denoted by APR (M ). On the other hand.4 Associated Primes and Localization To get more information about uniqueness in primary decompositions. then N = M . N = NP . and set M = MP . 1. By (1. the map P → PS is the restriction of a bijection and therefore must be injective. . If s is the product of the si . By (1.4. and N a P -primary submodule of M . an . RS the localization of R by S.4. M /N = 0.1) is surjective. ASSOCIATED PRIMES AND LOCALIZATION 7 1.2).4.4. hence is a bijection of APR (M ) ∩ C and APRS (MS ). Proof. . . In this section. we need to look at associated primes in localized rings and modules. ♣ 1. / belongs to P . no element of S can be a zero-divisor. then there exists t ∈ S such that tax = 0. If P is the annihilator of the nonzero element x ∈ M .3). and the set of all prime ideals of RS .3 Lemma Let M be a ﬁnitely generated module over the Noetherian ring R.1.1) and (1.1 Lemma Let P be a prime ideal not meeting S. (By (1. and it follows that a/s = at/st ∈ PS .11). . Conversely. there is a bijection between APRP (M/N )P (which coincides with APRP (M /N )) and the intersection APR (M/N ) ∩ C. suppose b annihilates sx. We need a subscript to distinguish this set from APRS (MS ). 1 ≤ i ≤ n. We conclude that b ∈ P .4.4. if (a/s)(x/1) = 0.3. and consequently b/1 = b /s for some b ∈ P and s ∈ S. Let P be generated by a1 .) For if a ∈ P and a/s ∈ PS . not meeting S = R \ P ). For each i there exists si ∈ S such that si ai x = 0. This means that for some u ∈ S we have u(bs − b ) = 0. hence ub . then (a/s)(x/1) = ax/s = 0.6). then sai x = 0 for all i. Suppose that PS is the annihilator of the nonzero element x/t ∈ MS . and the result follows. where C is the set of prime ideals contained in P (in other words. so b/1 ∈ PS . Proof. then PS is the annihilator of the nonzero element x/1 ∈ MS . which is not contained in P by hypothesis. Recall that P → PS = P RS is a bijection of C.4.3.3. so P = ann(sx). s cannot be a zero-divisor. S will be a multiplicative subset of the Noetherian ring R. Let P be any prime ideal of R. ♣ 1. But us ∈ P (because S ∩ P = ∅).1). and therefore so does ubs . If P ⊆ P . Thus APR (M/N ) ∩ C is empty. there is only one associated prime of M/N over R.

and since P1 depends on the ﬁxed ring R. r. Now NP1 = (N1 )P1 ∩ ∩r (Ni )P1 = (N1 )P1 ∩ MP1 = (N1 )P1 . See also TBGY. We will need the following result in order to proceed. But / / then by deﬁnition of I we have a ∈ I. then (regarding M and R as ﬁxed) Ni is uniquely determined by N . .3).5.5 Second Uniqueness Theorem Let M be a ﬁnitely generated module over the Noetherian ring R. . x ∈ N .5. As in (1. i=1 i = 1.4. 1. Suppose that P1 is minimal. it follows that N1 depends only on N . so assume x ∈ f −1 (N ). Choose a nonzero element x ∈ M .4.4. if P ⊆ P . It suﬃces to show that if MM = 0 for all maximal ideals M. i > 1. hence there exists a ∈ M (so a ∈ I) such that ax = 0. R \ P contains no zero-divisors of M/N .6). By (1. we have taken advantage of the isomorphism between (M/N )P and M /N . Proof.3). By hypothesis. (If = x/1 = 0. in other words. . so I is a proper ideal and is therefore contained in a maximal ideal M. we have R \ P ⊆ R \ P . By injectivity of g. ♣ 1. The result comes from the exactness of the localization functor. where f is the natural map from M to MP1 . then f (x) ∈ N by deﬁnition of f . then N = f −1 (N ).3).4. x + N is 0 in M/N . PRIMARY DECOMPOSITION AND ASSOCIATED PRIMES At the beginning of the proof of (1.4 Lemma In (1.5. where the technique is spelled out. and the result follows. so that P1 ⊇ Pi . and let I be the annihilator of x. APR (M/N ) = {P }.4.4) with P = P = P1 . i=2 Thus N1 = f −1 [(N1 )P1 ] = f −1 (NP1 ) depends only on N and P1 .3) with P = Pi .8 CHAPTER 1. If this is unfamiliar.4. then M = 0. we have (Ni )P1 = MP1 for i > 1.) If x ∈ N . where f is the natural map from M to M . . we have x = 0. Suppose that N = ∩r Ni is a reduced primary decomposition of the submodule N . Section 8. look ahead to the proof of (1.4. ♣ . Then 1 ∈ I (because / 1x = x = 0). if and only if MM = 0 for every maximal ideal M. Proof.3. P = P1 . Then f (x) ∈ N . Problem 5. then sx = 0 for some s ∈ S = R \ P . we have N1 = f −1 [(N1 )P1 ]. Since P ⊆ P . and Ni is Pi -primary. because all such zero-divisors belong to P . ♣ 1. Thus the natural map g : x → x/1 of M/N to (M/N )P ∼ M /N is injective. If Pi is minimal.3). a contradiction. By (1. 1. x/1 is 0 in MM . By (1. Proof. so f (x) + N is 0 in M /N . and since s is not a zero-divisor.1 Proposition M is the zero module if and only if MP = 0 for every prime ideal P .5 The Support of a Module The support of a module M is closely related to the set of associated primes of M .

and by the ﬁrst isomorphism theorem.5 Theorem If M is a ﬁnitely generated R-module. If MP = 0. Proof. Rxi ∼ R/ ann xi . ♣ / 1. To justify i=1 the last equality. We apply the localization functor to the exact sequence 0 → I → R → R/I → 0 to get the exact sequence 0 → IP → RP → (R/I)P → 0. We have / / cas = cb ∈ P . On the other hand. it follows that Supp M ⊆ Supp M .5. = i=1 Supp Rxi = V (ann xi ). = Thus P ∈ Supp R/I iﬀ RP ⊃ IP iﬀ IP is contained in a maximal ideal. Then MP = 0. hence 0 → MP → MP → MP → 0 is exact. Let M = Rx1 + · · · + Rxn . (R/I)P ∼ RP /IP . hence is an isomorphism. But this is equivalent to I ⊆ P .2 Deﬁnitions and Comments The support of an R-module M (notation Supp M ) is the set of prime ideals P of R such that MP = 0. necessarily P RP .1). so that MP = (Rx1 )P + · · · + (Rxn )P . THE SUPPORT OF A MODULE 9 1.5. Consequently. 1. then P ⊇ ann xi ⊇ ann M . then P ⊇ ann xi for some i.5. We conclude that a ∈ P . If I is any ideal of R. Conversely. ♣ Supports and annihilators are connected by the following basic result. By (1. and therefore P ∈ Supp M . Thus Supp M ⊆ Supp M . then MP = 0 (because MP → MP is surjective). By (1. There exists c ∈ P such that c(as − b) = 0. this is equivalent to M = 0.5. the Zariski topology on Spec R has the sets V (I) as its closed sets. To see this. 1.5. .3). a prime ideal. note that if P ∈ V (ann xi ). then Supp M = V (ann M ).1. ♣ i=1 And now we connect associated primes and annihilators. Proof. Let 0 → M → M → M Then Proof. Then Supp M = ∪n Supp Rxi . Then a/1 = b/s for some b ∈ P. and cs ∈ P . Let P belong to Supp M \ Supp M . if P ⊇ ann M = ∩n ann xi . with a/1 ∈ IP ⊆ P RP . since MP is isomorphic to a submodule of MP . s ∈ P . In algebraic geometry.4 Proposition → 0 be exact. we deﬁne V (I) as the set of prime ideals containing I. Supp M = Supp M ∪ Supp M .3 Proposition Supp R/I = V (I).5. so MP = 0. Thus Supp M = ∅ iﬀ MP = 0 for all prime ideals P . suppose a ∈ I. so the map MP → MP is injective as well as surjective.5. Therefore Supp M = ∪n V (ann xi ) = V (ann M ). But MP = 0 by assumption.

5.1.5.6). Take M = R in (1. Conversely. √ Proof. let P ∈ Supp M . Thus assume M = 0. P ∈AP(M ) Proof. Then AP(M ) ⊆ Supp M . Pr }.7 Corollary If R is a Noetherian ring. This contradicts (1.9).5) and (1.6 Proposition If M is a ﬁnitely generated module over the Noetherian ring R. If M = 0.5. then the nilradical of R is the intersection of all associated primes of R. if i=1 i=1 a belongs to this intersection.1)]. then by (1.3. Since ann R = 0. . If P does not contain any associated prime of M . If for n = max ni . ♣ Here is the most important connection between supports and associated primes. . Proof.8 Proposition Let M be a ﬁnitely generated module over the Noetherian ring R. let Q be the intersection of all associated primes. To prove that (2) implies (3). ♣ And now.8).2. a ∈ ∩r rM (Ni ) = ∩r Pi . ann R is the nilradical. Ni is Pi -primary and i=1 AP(M ) =√ 1 .6). P does √ not contain ann M . Thus for each i. where for each i. and the result to be proved is R = R. Consequently. and let P be any prime ideal of R. . then an M = 0. λa : M/Ni → M/Ni is nilpotent [see (1.5. By (1. The following conditions are equivalent: (1) P ∈ Supp M . hence P cannot contain the smaller ideal ann M . then √ P = ann M .8). (2) P ⊇ P for some P ∈ AP(M ). By (1. then √ all i there exists ni ≥ 1 such that ani M ⊆ Ni .3. there is a reduced primary decomposition (0) = ∩r Ni . so a ∈ ann M .5. To prove that (1) implies (2).5). so that (0) is a proper submodule. (3) P ⊇ ann M . then P does not contain the intersection of all associated primes (because P is prime).3).5. PRIMARY DECOMPOSITION AND ASSOCIATED PRIMES 1.9 Theorem Let M be a ﬁnitely generated module over the Noetherian ring R. with P = P .6)] ann M ⊇ ann M . associated primes and annihilators. √ Then P ⊇ P ⊇ Q = [by (1. 1. P contains some P ∈ AP(M ) ⊆ . We have AP(M ) ⊆ Supp M by (2) implies (1) in (1.5. then for some n ≥ 1 we have an M = 0. AP(M ) = ∅. and the minimal elements of AP(M ) and Supp M are the same. a connection between supports. Proof. then by (1) implies (2) in (1. If P is minimal in Supp M . . Conditions (1) and (3) are equivalent by (1.5.5. ♣ 1. {P If a ∈ ann M . 1.5).5.10 CHAPTER 1.5.

By minimality. P = P . hence [again by (1. where the Pj are prime ideals of R. . in other words.6. and ﬁnally Supp Mj ⊆ Supp M because Mj ⊆ M . .3. and in fact we will prove later in the section that an Artinian ring must be Noetherian.5. Any ﬁnite ring.4) and (1. Pj ∈ Supp Mj /Mj−1 . so that P is not minimal in Supp M . (Since M is Noetherian. 1. . n. then R is said to be an Artinian ring.6. . and in fact.1 Artinian Rings Deﬁnitions and Comments Recall that an R-module is Artinian if it satisﬁes the descending chain condition on submodules. Q ⊇ P ∈ AP(M ). By minimality. .10). .5. . Otherwise. If P ⊃ Q ∈ Supp M .9). ♣ 1. If M = M1 . ARTINIAN RINGS 11 Supp M .6 1. Continue inductively = to produce the desired chain. n. .2 Lemma If I is an ideal in the Artinian ring R. the minimal elements of all three sets AP(M ). . Pj ∈ Supp R/Pj .10).5.3. . P ⊃ Q ∈ AP(M ) ⊆ Supp M .6. . so by (1. By the ﬁrst isomorphism theorem. M has an associated prime P1 = ann x1 . ♣ Here is another way to show that there are only ﬁnitely many associated primes. a contradiction. . .8).5.3). ♣ 1.5.5. Take M1 = Rx1 ∼ R/P1 (apply the ﬁrst isomorphism theorem).10 Theorem Let M be a nonzero ﬁnitely generated module over the Noetherian ring R. .3.5. .11 Proposition In (1.4) and (1. .3. {P1 . Pn }. . AP(M ) ⊆ = {P1 . Now map R onto M2 /M1 by / r → rx2 + M1 . R satisﬁes the dcc on ideals. the process terminates in a ﬁnite number of steps. By (1. Thus P ∈ AP(M ). . Pn } in the proof of (1. . M2 /M1 ∼ R/P2 .9)). .7). . . .1.) For each j = 1. is both Noetherian and Artinian. The theory of associated primes and supports will help us to analyze Artinian rings. each Pj belongs to Supp M . .5. By (1. If the ring R is Artinian as a module over itself.4). Finally. x2 ∈ M1 . . Then there is a chain of submodules 0 = M0 < M1 < · · · < Mn = M such that for each j = 1. For any such chain. we have AP(Mj ) ⊆ AP(Mj−1 ) ∪ {Pj } by (1. P = P . = then the quotient module M/M1 is nonzero. Another inductive argument shows that AP(M ) ⊆ {P1 . contradicting P ⊃ Q ⊇ P . let P be minimal among associated primes but not minimal in Supp M . Let M2 = M1 + Rx2 . Mj /Mj−1 ∼ R/Pj . for example Zn . P must be a minimal associated prime. 1.3. Thus (replacing AP(M ) by {P1 . Pn }. Pn } and Supp M are the same. then by (1) implies (2) in (1. Proof. Note that Z is a Noetherian ring that is not Artinian. Supp Mj /Mj−1 ⊆ Supp Mj . Proof. then R/I is an Artinian ring.3)] has an associated prime P2 = ann(x2 + M1 ). with x1 a nonzero element of M . By (1.

Let I = N (R). The following two results reinforce this conclusion. hence M = Mi (because M and Mi are maximal).4 Proposition If R is an Artinian ring. Let Σ be the collection of all ideals K of R such that KL = 0. then since R is a domain. and (aL)L = aL2 = aL = 0. and the R-submodules are identical to the R/I-submodules. so for some i we have I i = I i+1 = · · · = L. so assume L = 0. Since the . ♣ 1. ba = 1. If M is any maximal ideal. Therefore. so for some t we have (at ) = (at+1 ). If at = bat+1 . R/P is a ﬁeld. Let Σ be the collection of all ﬁnite intersections of maximal ideals. contradicting our choice of a.6. Proof. hence P is maximal. The chain I ⊇ I 2 ⊇ I 3 ⊇ · · · stabilizes. in other words.6. But then M must contain one of the Mi (because M is prime). The chain of ideals (a) ⊇ (a2 ) ⊇ (a3 ) ⊇ · · · stabilizes. We have aL ⊆ Ra = K0 . Since R/I is a quotient of an Artinian R-module. From (a) = aL we get a = ab for some b ∈ L ⊆ N (R).6. We will show that the principal ideal (a) = Ra coincides with aL.3 Lemma An Artinian integral domain is a ﬁeld. it is also an Artinian R-module. Proof. then the nilradical N (R) is nilpotent. Thus R/I is an Artinian R/I-module. Then Ra ⊆ K0 (because K0 is an ideal). 1.5 Proposition An Artinian ring has only ﬁnitely many maximal ideals. Proof. ♣ 1. so aL ∈ Σ. In fact it is an R/I module via (r + I)(x + I) = rx + I. Let a be a nonzero element of the Artinian domain R.3). By minimality of K0 we have Ra = K0 . and RaL = aL = 0. We must produce a multiplicative inverse of a. so bn = 0 for some n ≥ 1. Let K0 be a minimal element of Σ.12 CHAPTER 1.6. so by minimality of I we have M ∩ I = I.6. so that R/I is an integral domain. Therefore a = ab = (ab)b = ab2 = · · · = abn = 0. If L = 0 we are ﬁnished. ♣ One gets the impression that the Artinian property puts strong constraints on a ring. By (1. PRIMARY DECOMPOSITION AND ASSOCIATED PRIMES Proof. the nilradical N (R) coincides with the Jacobson radical J(R). Artinian by (1.6. since L (as well as R) belongs to Σ. Proof.6.2). hence Ra ∈ Σ. Then Σ is nonempty and has a minimal element I = M1 ∩ · · · ∩ Mr (by the Artinian property). then M ⊇ M ∩ I ∈ Σ.6 Proposition If R is Artinian. an Artinian ring. Let P be a prime ideal of R. hence by (1. and choose a ∈ K0 such that aL = 0. then every prime ideal of R is maximal. ♣ 1. the Jacobson radical J(R) is nilpotent.4). By minimality of K0 we have aL = K0 = Ra. Then Σ is nonempty.

so the natural map from R to 1 R/Mk i k is an isomorphism. (3) Every prime ideal in the support of M is maximal. and we must show that it is local. a ﬁnite direct product of ﬁelds. ♣ We now prove a fundamental structure theorem for Artinian rings. with Mi /Mi−1 ∼ R/Pi .5. Powers of the Mi still satisfy the i r hypothesis of the Chinese remainder theorem. (1) ⇒ (2): As in (1.6.6.6.6. But Pi is prime.2). the intersection of the Mi coincides with their product.6. It is a standard result that an R-module M has ﬁnite length lR (M ) if and only if M is both Artinian and Noetherian. By maximality. the = hypothesis (1) implies that R/Pi has ﬁnite length for all i. But L is a power of the nilradical I.7). R/Pi is an Artinian ring. so the original chain will stabilize. Proof. is Artinian. In other words. .1. Since every associated prime is one of the Pi ’s [see (1.7 Theorem Every Artinian ring R is isomorphic to a ﬁnite direct product of Artinian local rings Ri . Mr .5). A sequence of exercises will establish the uniqueness of the Artinian local rings in the decomposition (1. M = Mi . We can relate this condition to associated primes and supports. 1. hence i i M ⊇ Mi (because M is prime).6. and the result follows. so R/Pi is an integral domain. Therefore each Pi is a maximal ideal. (2) Every associated prime ideal of M is maximal.3). 1. which is nilpotent by (1.6). Thus R/Pi is an Artinian R-module. By the Chinese remainder theorem.6.6. Proof.9 Proposition Let M be a ﬁnitely generated module over the Noetherian ring R. we must have L = 0. To see this. Since Mi /Mi−1 is a submodule of a quotient M/Mi−1 of M . ♣ i i 1. in particular. R/Mi is Artinian. . Thus for r r some k ≥ 1 we have ( 1 Mi )k = 1 Mk = 0. By (1. By (1. Thus the unique maximal ideal of R/Mk is Mi /Mk . R has only ﬁnitely many maximal ideals M1 . At some point. hence an Artinian R/Pi -module (note that Pi annihilates R/Pi ).8 Remarks A ﬁnite direct product of Artinian rings. project a descending chain of ideals onto one of the coordinate rings. A maximal ideal of R/Mk corresponds to a maximal ideal M of R with M ⊇ Mk . The intersection of the Mi is the Jacobson radical J(R). . . all projections will stabilize. ARTINIAN RINGS 13 assumption L = 0 has led to a contradiction. hence a ﬁeld by (1.10). . there is a chain of submodules 0 = M0 < · · · < Mn = M .5.10)]. the result follows.6. The following conditions are equivalent: (1) lR (M ) < ∞.

then AP(M ) = Supp M . (3) Every associated prime ideal of R is maximal. P ⊇ P for some P ∈ AP(M ).8). The “if” part follows because any module of ﬁnite length is Artinian and Noetherian. we get lR (M ) = n < ∞. By hypothesis. But length is additive. By additivity of length [as in (3) implies (1) in (1. Q is maximal.6. hence so is P . The “if” part follows from (1. Now consider the chain R = M0 ⊇ M1 ⊇ M1 M2 ⊇ · · · ⊇ M1 · · · Mk−1 ⊇ M1 · · · Mk = 0. Proof.13 Theorem The ring R is Artinian if and only if R is Noetherian and every prime ideal of R is maximal. and this vector space is ﬁnite-dimensional by the descending chain condition.11). By (1. and (2) implies (3) is immediate.5. hence R is Artinian. If lR (M ) < ∞.5. Thus Ti has ﬁnite length as an R/Mi -module.6.9). ♣ 1. hence R is Noetherian. if N is a submodule of M . it follows that Ti = M1 · · · Mi−1 /M1 · · · Mi is an Artinian R-module. (1) implies (2) by (1.5.4).6.8). so R/Pi is a ﬁeld. If R is Artinian. P contains some associated prime Q. so P = P ∈ AP(M ). then lR (R) < ∞ by (1.6.11 Theorem Let R be a Noetherian ring.6.6. the zero ideal is a ﬁnite product M1 · · · Mk of not necessarily distinct maximal ideals. we conclude that lR (R) < ∞.9). lR (Mi /Mi−1 ) = lR (R/Pi ) = 1 for all i. that is. PRIMARY DECOMPOSITION AND ASSOCIATED PRIMES (2) ⇒ (3): If P ∈ Supp M .11) and the hypothesis (3).6. then l(M ) = l(N ) + l(M/N ).4) or (1.12). ♣ 1. The following conditions are equivalent: (1) R is Artinian.12 Theorem The ring R is Artinian if and only if lR (R) < ∞.7). (2) Every prime ideal of R is maximal. Since any submodule or quotient module of an Artinian module is Artinian.6. P and P are both maximal. To prove that (3) implies (1).11). By (1. so let P ∈ Supp M . hence an Artinian R/Mi module. 1.6. By (1. Summing on i from 1 to n. lR (R) < ∞.9). every prime ideal of R is maximal. By (1. As in (1. (Note that Mi annihilates M1 · · · Mi−1 /M1 · · · Mi . then by (1. Proof. ♣ 1.) Thus Ti is a vector space over the ﬁeld R/Mi . (3) ⇒ (1): By (1. Proof. Proof.10 Corollary Let M be ﬁnitely generated over the Noetherian ring R.6. AP(M ) ⊆ Supp M . ♣ We can now characterize Artinian rings in several ways. Consequently. every Pi is maximal.9)].5.6. note that by (1. hence as an R-module.14 CHAPTER 1. Thus assume R Artinian. ♣ .6.6.

1. Proof.14 Corollary Let M be ﬁnitely generated over the Artinian ring R. By (1.6. Then lR (M ) < ∞.6. hence the module M is both Artinian and Noetherian. ♣ . ARTINIAN RINGS 15 1.13).6. R is Noetherian. Consequently. M has ﬁnite length.

we get ∆xi = 0 for all i. because it is a root of x2 − d. but f is not monic. and −cij elsewhere. and let α ∈ S.Chapter 2 Integral Extensions 2. if d is any integer. so 2/3 is not an algebraic integer. But then ∆ annihilates all of M . Notice that 2/3 is a root of the polynomial f (x) = 3x − 2.1. and a key step is the following result. then d is an algebraic integer. then α is integral over R iﬀ α is algebraic over R. We say that α is integral over R if α is a root of a monic polynomial with coeﬃcients in R.1 Integral Elements Deﬁnitions and Comments Let R be a subring of the ring S. . and recall that a module is faithful if its annihilator is 0. Then αxi ∈ IM . Expanding the determinant yields the desired monic polynomial. where ∆ is the determinant of A. . S and α be as above.1. we have Ax = 0. Thus n (δij α − cij )xj = 0. 2. If α is a complex number that is integral over Z. Multiplying on the left by the adjoint matrix. where A is a matrix with entries α − cii on the main diagonal. j=1 In matrix form. one proves that a rational number that is an algebraic integer must belong to Z. so we cannot conclude that 2/3 is an algebraic integer.1 2. sometimes called the determinant trick. . xn generate M over R. let x1 .2 Lemma Let R. Let M be a ﬁnitely generated R-module that is faithful as an R[α]-module. In a ﬁrst course in algebraic number theory. There are several conditions equivalent to integrality of α over R. then α is said to be an √ algebraic integer For example. so we are generalizing a familiar notion. Then α is a root of a monic polynomial with coeﬃcients in I. Let I be an ideal of R such that αM ⊆ IM . Proof. . so we may write αxi = n j=1 cij xj with cij ∈ I. ♣ 1 . so ∆ = 0. If R is a ﬁeld and S an extension ﬁeld of R. 1 ≤ i ≤ n.

Thus 1. . hence over A[b0 . then αn and all higher powers of α can be expressed as linear combinations of lower powers of α. .5 Lemma Let R be a subring of S. . . . By (2. . αn ∈ S. α2 . . ♣ 2. (3) R[α] is contained in a subring R of S that is a ﬁnitely generated R-module. every element of C is integral over B. and the following result will be helpful. . Then x is integral over A[b0 .4). so C = j=1 k=1 Ayj xk . bn−1 ]. ♣ We are going to prove a transitivity property for integral extensions. in other words.1.1. αn−1 generate R[α] over R.5).2) with I = R. . (2) R[α] is a ﬁnitely generated R-module. (4) There is a faithful R[α]-module M that is ﬁnitely generated as an R-module.1.2 CHAPTER 2.6 Transitivity of Integral Extensions Let A. . with α ∈ S. (4) implies (1): Apply (2. The following conditions are equivalent: (1) α is integral over R.1. Proof. with α1 . part (3). .1. . Proof.4). Going from n − 1 to n amounts to proving that if A. αn−1 ]. ♣ . . with C a ﬁnitely generated B-module and B a ﬁnitely generated A-module. .1. If α1 is integral over R.1. αn ] is a ﬁnitely generated R-module. By (2. If C is integral over B. then C is a ﬁnitely generated A-module. . . . x is integral over A. . If y ∈ R[α] and yM = 0. B and C be subrings of R. then y = y1 = 0. (2) implies (3): Take R = R[α]. . then R[α1 . . α. and B is integral over A. that is. . αM ⊆ M . 2. The n = 1 case follows from (2. part (2). . . α stabilizes M . then in particular. . Proof. . . . . (3) implies (4): Take M = R . Let x ∈ C. A[b0 . (1) implies (2): If α is a root of a monic polynomial over R of degree n. bn−1 . and αn is integral over R[α1 . α2 is integral over R[α1 ]. x] is a ﬁnitely generated A-module. This follows by a brief computation: r s r s C= j=1 Byj .3 Remark If αM ⊆ IM . . B = k=1 Axk . Each bi is integral over A. INTEGRAL EXTENSIONS 2.4 Theorem Let R be a subring of S. then C is integral over A. 2. . with xn + bn−1 xn−1 + · · · + b1 x + b0 = 0. . .1. B and C are rings. bi−1 ].

1. ♣ A domain that is an integral extension of a ﬁeld must be a ﬁeld. we have an + bc = 0.8 Proposition The integral closure Rc of R in S is integrally closed in S.1. Multiplying by bn . so that x is integral over Rc .6).1. 2. then just as in the proof of (2.1. b ∈ R and a and b are relatively prime. Thus b divides an . Rc ⊆ (Rc )c . Proof.2.1. As in the proof of (2.10 Proposition Let R be a subring of the integral domain S. Multiply the equation by an−1 to get a−1 = −(cn−1 + · · · + c1 an−2 + c0 an−1 ) ∈ R. and R is integrally closed in K. b is a unit. Assume that S is a ﬁeld. with S integral over R. ♣ We can identify a large class of integrally closed rings. then R is integrally closed. with c ∈ R. x is integral over R. By deﬁnition. there is an equation of the form (a/b)n + an−1 (a/b)n−1 + · · · + a1 (a/b) + a0 = 0 with ai ∈ R. there is an equation of the form (a−1 )n + cn−1 (a−1 )n−1 + · · · + c1 a−1 + c0 = 0 with ci ∈ R. 2.7 Deﬁnitions and Comments If R is a subring of S.6). y] is a ﬁnitely generated R-module. Then R is a ﬁeld if and only if S is a ﬁeld. x − y and xy belong to this module. the integral closure of R in S is the set Rc of elements of S that are integral over R. Thus x ∈ Rc .1.1. it follows from (2. 2.5) that R[x.1. If we simply say that R is integrally closed without reference to S. Let x belong to the fraction ﬁeld K of R. as the next result shows.9 Proposition If R is a UFD. Thus let x ∈ (Rc )c . But then x = ab−1 ∈ R. Note that R ⊆ Rc because each a ∈ R is a root of x − a. they are integral over R by (2. part (3). Since x + y. Proof. we assume that R is an integral domain with fraction ﬁeld K. If we take the integral closure of the integral closure. If the elements x and y of S are integral over R. The important conclusion is that Rc is a subring of S containing R. Write x = a/b where a. . and let a be a nonzero element of R. which cannot happen for relatively prime a and b unless b has no prime factors at all. We say that R is integrally closed in S if Rc = R. in other words.4). Since a−1 ∈ S.1. Proof. If x is integral over R. we get nothing new. INTEGRAL ELEMENTS 3 2.

Then b satisﬁes an equation of the form xn + an−1 xn−1 + · · · + a1 x + a0 = 0 with ai ∈ R. Similarly. R/I can be regarded as a subring of S/J. there is an element c ∈ R[b] ⊆ S such that bc = 1. let α/t ∈ ST . let b + Q ∈ S/Q. By (2. Let f be the R-linear transformation on this vector space given by multiplication by b. Thus if bz = 0 (with b = 0 by choice of b).10). Then P is a maximal ideal of R if and only if Q is a maximal ideal of S. then ST is integral over RT . The map a + P → a + Q is a well-deﬁned injection of R/P into S/Q. Suppose that Q is a prime ideal of S. with ci ∈ R. To see this.12 Proposition Let S be integral over the subring R.1.4 CHAPTER 2. S is a ﬁeld. INTEGRAL EXTENSIONS Now assume that R is a ﬁeld.1.4) part (2). R[b] is a ﬁnite-dimensional vector space over R. We can now invoke (2. As in (2. and if S is integral over R. Since R[b] is a subring of S. lying over the prime ideal P = Q ∩ R of R. in other words. We will analyze in great detail the relation between prime ideals of R and those of S.1. then I = J ∩ R is an ideal of R.10) to prove the following result.1. Moreover. R/P is a ﬁeld iﬀ S/Q is a ﬁeld. S/Q is integral over R/P . . But b + Q satisﬁes the same equation with ai replaced by ai + P for all i. with α ∈ S. Proof. 2. (We say that Q lies over P . If R is a subring of S and J is an ideal of S. ♣ 2. if S is integral over R and T is a multiplicative subset of R.1. and let Q be a prime ideal of S. Thus we can regard R/P as a subring of S/Q.11). Therefore if b ∈ S and b = 0. then S/J is integral over R/I. not necessarily prime.1. ♣ 2. t ∈ T .11 Preview Let S be integral over the subring R. Then there is an equation of the form αn + cn−1 αn−1 + · · · + c1 α + c0 = 0. and let b be a nonzero element of S. we have z = 0 and f is injective. z ∈ R[b]. f (z) = bz. To prove this. proving integrality of S/Q over R/P . because it is the preimage of Q under the inclusion map from R into S. By (2. and let P = Q ∩ R.) Then P is a prime ideal of R.1.1.13 Remarks Some results discussed in (2.11) work for arbitrary ideals. Consequently. it is an integral domain. Thus α cn−1 α n−1 c1 α c0 ( )n + ( + · · · + ( n−1 ) + n = 0 )( ) t t t t t t with cn−j /tj ∈ RT . But any linear transformation on a ﬁnite-dimensional vector space is injective iﬀ it is surjective. because P = Q ∩ R.

As at the beginning of the proof. But P RT is a maximal ideal of RT . Let T = R\P . Thus P1 ST and P2 ST lie over P RT . then x ∈ R ∩ Pi = P . Proof.2 Integrality and Localization Results that hold for maximal ideals can sometimes be extended to prime ideals by the technique of localization. Q ∩ RT must be the unique maximal ideal .13). contradicting the fact that P1 ST is a prime ideal. In the general case.2. INTEGRALITY AND LOCALIZATION 5 2. then Q ∩ R is maximal by (2.12). The prime ideals Pi . ST is integral over RT . it suﬃces to show that P1 ST = P2 ST . P1 ST and P2 ST are maximal by (2. then P1 = P2 . for otherwise RT ⊆ P1 ST and therefore 1 ∈ P1 ST . hence P1 ST = P2 ST . do not meet T .2. so are P1 and P2 .1. 2. In general. By (2. 2. that is. and the result follows.2 Lying Over Theorem If S is integral over R and P is a prime ideal of R. If P1 ⊆ P2 . Theorem 2.2 is also a good example of localization technique. Recall that ST is integral over RT by (2. and the vertical maps are canonical (f (r) = r/1 and g(s) = s/1). we localize with respect to P .1. If Q is any maximal ideal of S. P1 ∩ R = P2 ∩ R = P .1.1. R −−→ S −− g f R T − − → ST −− The horizontal maps are inclusions. The ﬁrst inclusion holds because P ⊆ P1 and RT ⊆ ST . 2. We have the following commutative diagram. If Q is any maximal ideal of ST .12). and let P1 and P2 be prime ideals of S that lie over the prime ideal P of R. so Q ∩ R must be P . The second inclusion is proper. If P is maximal. there is a prime ideal Q of S such that Q ∩ R = P.2.2. contradicting the deﬁnition of T . Proof. First assume that R is a local ring with unique maximal ideal P . then prime ideals of R can be lifted to prime ideals of S.13). as the next result demonstrates. so by the above claim. a multiplicative subset of R ⊆ S. We claim that P RT ⊆ (P1 ST ) ∩ RT ⊂ RT . because if x ∈ T ∩ Pi .1. ♣ If S/R is an integral extension.12). and similarly (P2 ST ) ∩ RT = P RT .2. By the basic correspondence between prime ideals in a ring and prime ideals in its localization. let T be the multiplicative set R \ P . (P1 ST ) ∩ RT = P RT . A good illustration follows. i = 1. then by (2.1 Proposition Let S be integral over the subring R. then as at the beginning of the proof.

. including the assumption that R is integrally closed. So it will be useful to get some practice with the idea of integral closure. f extends to a monomorphism g : L → C. x2 + P1 = x2 + Q1 . Let P be the kernel of f .1. . P is a prime ideal of R. But as above. . let x2 ∈ Q2 ∩ R. Since C is algebraically closed.11). m. which gives the desired extension of the Q-chain. Let L be the fraction ﬁeld of S/Q.2. because g extends f and f ◦ p|R = f .2. . there is a prime ideal Q of S such that Q ∩ R = P . hence x2 + P1 = y2 + Q1 for some y2 ∈ Q2 . the going down theorem. Consequently. To verify this.2. m = 1. . f −1 (P RT ) = P . ♣ 2. so x2 + P1 = x2 + Q1 ∈ (Q2 /Q1 ) ∩ R/P1 = P2 /P1 . where m < n. and let f be a ring homomorphism from R into an algebraically closed ﬁeld C.11). if x2 ∈ P2 then x2 + P1 ∈ Q2 /Q1 . We claim that Q2 ∩ R = P2 .1. . we have f −1 (P RT ) = Q∩R. Then f can be extended to a ring homomorphism g : S → C. If p : S → S/Q is the canonical epimorphism and g = g ◦ p. Proof.2. and therefore x2 ∈ Q2 .6 CHAPTER 2. then g is the desired extension of f . By induction. By (2. where Q2 is a prime ideal of S and Q1 ⊆ Q2 . There is an analogous result for ring extensions. By commutativity of the diagram.11). Thus assume P1 ⊆ P2 and Q1 ∩ R = P1 . Conversely. f induces an injective ring homomorphism f : R/P → C. we may apply the lying over theorem (2. Since P2 /P1 is a prime ideal of R/P1 .2. it suﬃces to consider the case n = 2. and a chain of prime ideals Q1 ⊆ · · · ⊆ Qm of S. Proof. Qn of S such that Qm ⊆ Qm+1 ⊆ · · · ⊆ Qn and Qi lies over Pi for every i = 1. (Note that if r ∈ R. ♣ It is a standard result of ﬁeld theory that an embedding of a ﬁeld F in an algebraically closed ﬁeld can be extended to an algebraic extension of F . x2 ∈ P2 . By (2. If Qi lies over Pi for i = 1.1)]. . so x2 − y2 ∈ P1 ⊆ P2 .3). By the basic localization correspondence [cf. ♣ In the next section.2. . S/Q is integral over R/P . we have an embedding of R/P1 into S/Q1 . INTEGRAL EXTENSIONS of RT . f −1 (Q ∩ RT ) = g −1 (Q ) ∩ R. namely P RT . .2).2) to produce a prime ideal Q2 /Q1 of S/Q1 such that (Q2 /Q1 ) ∩ R/P1 = P2 /P1 . which extends in the natural way to the fraction ﬁeld K of R/P . then f (r) ∈ Q ∩ RT iﬀ g(r) ∈ Q . Thus x2 + P1 = y2 + P1 for some y2 ∈ P2 . By (2. n. S/Q1 is integral over R/P1 . There will be extra hypotheses.4 Theorem Let S be integral over R. . . we will prove the companion result to (2. hence L is an algebraic extension of K. and suppose we have a chain of prime ideals P1 ⊆ · · · ⊆ Pn of R. By the factor theorem. 2. By (2.) If Q = g −1 (Q ). . so x2 − y2 ∈ Q1 .1. Since f maps into a ﬁeld. then there are prime ideals Qm+1 . .(2. and the result follows.3 Going Up Theorem Let S be integral over R.

To prove that (3) ⇒ (1).2. because rM ∈ M by construction. m = 1m = P aP rP m = 0. If α/t ∈ ST (α ∈ S. it follows from (2. Apply (2. Thus 1 can be written as a ﬁnite sum P aP rP where P is a maximal ideal of R and aP ∈ R. so there exists rP ∈ R \ P such that rP m = 0 in M .2. and denote by R the integral closure of R in S. t ∈ T .13) that (R)T is integral over RT . It is immediate that (1) ⇒ (2) ⇒ (3). .2. then (R)T is the integral closure of RT in ST . Thus I(m) must be R.6 Corollary If T is a multiplicative subset of the integrally closed domain R. Since R is integral over R. If P is a maximal ideal of R. but additional hypotheses will be needed and the analysis is more complicated.1. Then RT is the integral closure of RT in ST . and / in particular. but a slightly diﬀerent proof is given here. Therefore t0 α ∈ R.1). (2) MP = 0 for all prime ideals P of R.) 2.3 Going Down We will prove a companion result to the going up theorem (2.7 Proposition The following conditions are equivalent. The following result will be useful.2. so α/t = t0 α/t0 t ∈ (R)T .5.3. Consequently. (The same result was proved in (1. the fraction ﬁeld of R (and of RT ). There is an equation of the form α a1 α an ( )n + ( )( )n−1 + · · · + =0 t t1 t tn with ai ∈ R and ti . Proof.2. t ∈ T ) and α/t is integral over RT . ♣ Additional results on localization and integral closure will be developed in the exercises. 1 ∈ I(m). then RT is integrally closed. and multiply the equation by (tt0 )n to conclude that t0 α is integral over R. Proof. we must show that α/t ∈ (R)T . let m ∈ M .2. (3) MP = 0 for all maximal ideals P of R. so RT is integrally closed. Let I(m) be the ideal generated by the rP . (1) M = 0. Let t0 = i=1 ti . then m/1 is 0 in MP . ♣ 2. GOING DOWN 7 2. But ST = K. If T is a multiplicative subset of R. for an arbitrary R-module M .3). Proof.5 Lemma Let R be a subring of S. Then I(m) cannot be contained in any maximal ideal M.5) with R = R and S = K. ♣ n 2.

The coeﬃcients of f and g are in K and are symmetric polynomials in the roots. Step 2 : f ∈ I. we may invoke (2.3. sn .3 Proposition Let S be integral over the subring R. and is also a ﬁnitely generated R-module by (2. because an element that annihilates S1 annihilates 1 and is therefore 0. so that sn = i=1 ri si for some ri ∈ I and si ∈ S. we may clear denominators to produce a nonzero element r ∈ R such that rf ∈ IR[x] = I. so f ∈ I. Therefore f (s) = 0. r is not a zero-divisor of S. Then S1 = R[s1 . ♣ 2. hence h is a multiple of f .) Thus IK[x] is generated by a monic polynomial f . Since R is integrally closed. Proof. n ≥ 1. By hypothesis. Assume that no nonzero element of R is a zero-divisor of S. so there is a monic polynomial h ∈ R[x] such that h(s) = 0. . thus hs is just evaluation at s.3.3. . deﬁne a homomorphism hs : R[x] → S by hs (f ) = f (s). Since f ∈ IK[x].1 √ Let S be integral over the subring R. Then IS is the set of all s ∈ S satisfying an equation of integral dependence sm + rm−1 sm−1 + · · · + r1 s + r0 = 0 with the ri ∈ I. we can take a common denominator and . we have f (x) = i (x−ai ) and g(x) = j (x−bj ). Then the kernel I of hs is a principal ideal generated by a monic polynomial. s is integral over R. say h = f g. Conversely. and by hypothesis. then sm ∈ IS.5). with I an ideal of R. then IK[x] is an ideal of the PID K[x]. with g monic in K[x]. If f g ∈ R[x]. and assume that R is integrally closed. where R is an integrally closed domain. and IK[x] = 0 because s is integral over R.2 Lemma Let R be an integral domain with fraction ﬁeld K. . But R is integrally closed.1. . Step 3 : f generates I. Let f and g be monic polynomials in K[x]. If s satisﬁes such an equation. Proof. (If this is unclear. hence s. Since f generates IK[x]. By (2.3. ♣ 2. S1 is a faithful R[sn ]-module.2). hence are integral over R as well.) If s ∈ S. see the argument in Step 1 below. Now k k Lemma sn S1 = i=1 ri si S1 ⊆ i=1 ri S1 ⊆ IS1 . Then h ∈ I ⊆ IK[x]. In a splitting ﬁeld containing K. let sn ∈ k IS. If K is the fraction ﬁeld of R. then both f and g are in R[x]. (This is automatic if S itself is an integral domain. so s ∈ IS.2) to conclude that f and g belong to R[x]. Let q ∈ I ⊆ IK[x]. sk ] is a subring of S. Step 1 : f ∈ R[x].8 CHAPTER 2. INTEGRAL EXTENSIONS 2. Since the ai and bj are roots of the monic polynomial f g ∈ R[x]. By deﬁnition of I we have rf (s) = 0. and the result follows. Moreover. they are integral over R.1. √ Proof. satisﬁes an equation of integral dependence with coeﬃcients in I.

Deﬁne v(x) = rm xm + rm−1 rm−1 xm−1 + · · · + r1 rx + r0 .) Consequently. Since R/P1 is an integral domain and g. We apply (2. Thus Q1 ∩ R ⊆ P1 and Q1 = Q1 ∩ S ⊆ Q2 . so P1 ⊆ (P1 S) ∩ R ⊆ Q1 ∩ R. there is a monic polynomial g ∈ R[x] that generates the kernel of the evaluation map ht : R[x] → S. . By basic localization theory. Since f is monic. In checking that T is a multiplicative set. to produce a monic polynomial f (x) = xm + rm−1 xm−1 + · · · + r1 x + r0 with coeﬃcients in P1 such that f (rt) = 0. we must have g = xj for some j with 0 ≤ j ≤ m. n. . we have q1 f = 0. Let T be the subset of S consisting of all products rt. . If Qi lies over Pi for i = m. s ∈ Q1 iﬀ s/1 ∈ M. t belongs to the radical of P1 S. Thus f divides q in R[x]. M corresponds to a prime ideal Q1 of S that is disjoint from T . . With the aid of the note at the end of the last paragraph. We do this by taking the contraction of both sides of the inclusion P1 ST ⊆ M. . By induction. Proof. . we have (R \ P1 ) ∩ Q1 = (S \ Q2 ) ∩ Q1 = ∅. We refer to Q1 as the contraction of M to S. . r1 divides every coeﬃcient of q1 . By (2. which is a / contradiction (because 0 ∈ Q2 ). hence g.3. then the hypothesis that r is not a zero-divisor of S gives t = 0. Therefore v = ug for some u ∈ R[x].3. ♣ 2. Explicitly. we have P1 S ⊆ Q1 . Since ri ∈ P1 for all i = 0.4 Going Down Theorem Let the integral domain S be integral over the integrally closed domain R. it suﬃces to consider n = m = 2. m − 1. . But g ∈ ker ht . so q1 /r1 ∈ R[x].3). it is the preimage of M under the canonical map s → s/1. Since the contraction of P1 ST to S is P1 S. By (2. Therefore P1 ST is contained in a maximal ideal M. we show that T ∩ P1 S is empty. Note that R \ P1 ⊆ T (take t = 1).2. so v is not the zero polynomial. . Passing to residue classes in the polynomial ring (R/P1 )[x]. so t ∈ Q2 . Finally. ♣ . .3. . so g(t) = 0. . and if we pass to residue classes in the polynomial ring (R/Rr1 )[x]. . Now P1 ST is a proper ideal of ST . so for some positive integer l. . we have v = rm xm . which means that q1 itself must be 0. then by deﬁnition of T . / g(x) = xj + aj−1 xj−1 + · · · + a1 x + a0 with ai ∈ P1 . we have tl ∈ P1 S ⊆ P2 S ⊆ Q2 S = Q2 . T ∩ P1 S contains an element rt with r ∈ R \ P1 and t ∈ S \ Q2 . with 1 < m ≤ n. else 1 would belong to T ∩ P1 S. (Note that r ∈ P1 . Qm−1 such that Q1 ⊆ · · · ⊆ Qm and Qi lies over Pi for every i = 1. If rt = 0 for some r ∈ P1 (hence r = 0) / and t ∈ Q2 .1). First we prove the theorem under the assumption that T ∩ P1 S = ∅. We must show that P1 ⊆ Q1 ∩ R. . then there are prime ideals Q1 . i = 0. GOING DOWN 9 write q = q1 f /r1 with 0 = r1 ∈ R and q1 ∈ R[x].1). . .3. Thus r1 q = q1 f . If not. . r ∈ R \ P1 . is monic. Suppose we have a chain of prime ideals P1 ⊆ · · · ⊆ Pn of R and a chain of prime ideals Qm ⊆ · · · ⊆ Qn of S. we must make sure that it does not contain 0. as desired. n. and S \ Q2 ⊆ T (take r = 1).3. the leading coeﬃcient of q1 must be 0. . Then v(x) ∈ R[x] and v(t) = 0. t ∈ S \ Q2 . we have v = u g. Consequently. j − 1. . with I = P1 and s replaced by rt. This contradicts our choice of t (recall that t ∈ S \ Q2 ).

4. h) by (RP . Thus we may replace (R. suppose f ∈ I. If j is nonconstant. Let I = {f ∈ R[x] : f (α) = 0}. Thus we may assume that j is constant. as desired. let P be the kernel of h.2. If the constant is zero. Without loss of generality. If α is a nonzero element of K. Then h(f ) ∈ J. If f ∈ R[x] with f (x) = ai xi . and it follows that 1 ∈ J. Then P is a prime ideal. Say J = (j(x)). 3. and we can extend h to g : RP → C via g(a/b) = h(a)/h(b). or h can be extended to a ring homomorphism h : R[α−1 ] → C. g(RP ) ∼ RP /P RP . there exists f ∈ I such that h(f ) = 1.Chapter 3 Valuation Rings The results of this chapter come into play when analyzing the behavior of a rational function deﬁned in the neighborhood of a point on an algebraic curve. hence h(f ) is a multiple of j. then either h can be extended to a ring homomorphism h : R[α] → C. h(b) = 0. So we can assume that j = 0. We can then extend h to h : R[α] → C via h(α) = β. and h : R → C a ring homomorphism from R into an algebraically closed ﬁeld C. a ﬁeld (because P RP is a maximal = ideal). Here is another variation.1. we take h(f ) = h(ai )xi ∈ F [x]. 3. we generalized a result about ﬁeld extensions to rings. necessarily principal.1 Theorem Let R be a subring of the ﬁeld K. with β arbitrary. we may assume that R is a local ring and F = h(R) is a subﬁeld of C. then we may extend h exactly as above. and therefore h(f )(β) = 0. To verify that h is well-deﬁned. Proof. Our ﬁrst step is to extend h to a homomorphism of polynomial rings. it must have a root β in the algebraically closed ﬁeld C. so that f (α) = 0. The kernel of g is P RP . To see this. Consequently. Then J = h(I) is an ideal of F [x].1 Extension Theorems In Theorem 2. 1 . so by the ﬁrst isomorphism theorem. g).

Assuming that h has no extension to R[α−1 ].) ♣ 0 0 It is natural to try to extend h to a larger domain. We then carry out the same analysis with α replaced by α−1 . 4. Thus b0 ∈ M (else 1 ∈ M). It is therefore legal to multiply (2) by b−1 αs / 0 to get αs + b−1 b1 αs−1 + · · · + b−1 bs = 0 0 0 (3) Finally.1. Let K = k(x). . and this is where valuation rings enter the picture. and assume (without loss of generality) that r ≥ s. If so. 1. 3. either α or α−1 belongs to R. But h(a0 ) = 1 and h(ar b−1 bs ) = 0. Let K = k(x). with p a ﬁxed prime. We now return to the extension problem. i = 0 0. the unique maximal ideal of the local ring R. i > 0 (2) Take s minimal. where r ≥ 0. Take R to be the set of all rational functions pr m/n. This is essentially the same as the previous example. where g belongs to the ring R = k[[x]] of formal power series over k. i = 0 0. 2. but there are more interesting examples. so b0 is a unit. (The result of multiplying (3) by ar αr−s cannot be a copy of (1). Thus a nonzero element of K looks ∞ like f = i=r ai xi with ai ∈ k. Moreover. and therefore g. 3. we have s bi α−i = 0 with bi ∈ R and bi = h(bi ) = i=0 1. where r ≥ 0 and p divides neither m nor n.2 Deﬁnition A subring R of a ﬁeld K is a valuation ring of K if for every nonzero α ∈ K. VALUATION RINGS ai αi = 0 with ai ∈ R and ai = h(ai ) = i=0 1. hence f . We may write f = ar xr g. i > 0 (1) Choose r as small as possible. p is a ﬁxed polynomial that is irreducible over k and m and n are arbitrary polynomials in k[x] not divisible by p.3 Examples The ﬁeld K is a valuation ring of K. Take R to be the set of all rationals of the form pr m/n. Let K be the ﬁeld of formal Laurent series over k. r ∈ Z. it follows that b0 − 1 ∈ ker h ⊆ M. where k is any ﬁeld. we multiply (3) by ar αr−s and subtract the result from (1) to contradict the minimality of r. the constant term of g is 1.2 This gives a relation of the form r CHAPTER 3. Let K = Q.1. and ar = 0. and let R be the set of all rational functions f /g ∈ k(x) such that deg f ≤ deg g. Thus R is a valuation ring of K. r = s (hence αr−s = 1)and a0 = ar b−1 bs . can be inverted (by long division). Since h(b0 ) = 1 = h(1). 3.

V is integrally closed. . 4. either V [α] = V or V [α−1 ] = V . and pick a ∈ I \ J (hence a = 0). Let S be the set of all (Ri . V is a subring of K containing R. Suppose that I is not contained in J. A standard application of Zorn’s lemma produces a maximal extension (V. Similarly. hj ) if and only if Ri is a subring of Rj and hj restricted to Ri coincides with hi . This follows because a nonzero element α of K can be written as α/1 or as 1/α−1 . and if we multiply the above equation of integral dependence by α−(n−1) .3. hi ) ≤ (Rj . The fraction ﬁeld of V is K. We have b/a ∈ V (else a/b ∈ V . and h : R → C a ring homomorphism from R into an algebraically closed ﬁeld C. This follows from the deﬁnition of a valuation ring. then a + b ∈ M. PROPERTIES OF VALUATION RINGS 3 3. h is an extension of h. a contradiction). then either I ⊆ J or J ⊆ I. h). Thus the ideals of V are totally ordered by inclusion. Then h has maximal extension (V.4 Theorem Let R be a subring of the ﬁeld K. and the statement of each property will be followed immediately by its proof. Any subring of K containing V is a valuation ring of K. If a and b are nonzero nonunits. If r ∈ V and a ∈ M. where Ri is a subring of K containing R and hi is an extension of h to Ri . if b/a ∈ V .1. h has an extension to either V [α] or V [α−1 ]. hi ). ♣ 3. If b ∈ J. then α−1 ∈ V . then ra ∈ M. If not. If b = 0 we are ﬁnished. By maximality. h). In other words. with α integral over V . In addition.2. V is a local ring. 1. 3. We must show that α ∈ V . 5. Proof. If α is a nonzero element of K.2 Properties of Valuation Rings We have a long list of properties to verify. then a + b = b(1 + a/b) ∈ M (because if b(1 + a/b) were a unit. then b would be a unit as well). then by (3. we get α = −cn−1 − cn−2 α−1 − · · · − c1 αn−2 − c0 αn−1 ∈ V. so a = (a/b)b ∈ J.1. Throughout.1). Let α be a nonzero element of K. Partially order S by (Ri . else a would be a unit. for any maximal extension. V is a valuation ring of K. Then there is an equation of the form αn + cn−1 αn−1 + · · · + c1 α + c0 = 0 with the ci in V . Thus M is an ideal. If I and J are ideals of V . We will show that the set M of nonunits of V is an ideal. then either a/b or b/a belongs to V . so assume b = 0. we must show that b ∈ I. Therefore b = (b/a)a ∈ I. 2. V is a valuation ring of the ﬁeld K. The end of proof symbol will only appear at the very end. and there is no extension to a strictly larger subring. Therefore α ∈ V or α−1 ∈ V . If a/b ∈ V .

The integral closure R of R in K is the intersection of all valuation rings V of K such that V ⊇ R. either (a) ⊆ (b). a contradiction. it is also the fraction ﬁeld of VP . 8. if a belongs to (pm ) for every m ≥ 1. Consequently a ∈ V . Thus (a) = (pm ). a contradiction. hence a UFD. for if a ∈ V . multiply by a suﬃciently high power of a to get a monic equation satisﬁed by a. in which case a/b ∈ V . If (a) is an arbitrary ideal. so a ∈ R. If P is a prime ideal of the valuation ring V . note that each Vα is integrally closed by Property 4. the intersection of some (not necessarily all) valuation rings of K. Moreover. V /P is an integral domain. then V is a PID. that is. By hypothesis. then a = a1 = aa−1 b = b ∈ R . m=1 Since V is Noetherian. First note that if K is the fraction ﬁeld of V .4 CHAPTER 3. For the “if” part. Continuing inductively and using the fact that V is a PID. then / 1 = h(1) = h(aa−1 ) = h(a)h(a−1 ) = 0. for some prime p ∈ V . The result follows. then VP and V /P are valuation rings. and let h be the composition of the canonical map R → R /M = k and the inclusion k → C. Now by Property 5. . Conversely. In particular. the ideals of V are totally ordered by inclusion. 7. assume a ∈ R.) 9. For any such p. The result follows from Property 6. If a ∈ R. If V is a Noetherian valuation ring. But V is integrally closed by Property 4. But then p divides a. we have a = pm u for some positive integer m and unit u. we must have a = 0. But then I ⊆ (an ) ⊆ I. ∩∞ (pm ) = 0. Let R be an integral domain with fraction ﬁeld K. so assume a is a nonunit. h has a maximal extension to h : V → C for some valuation ring V of K containing R ⊇ R. m ≥ 0. Again using unique factorization. and h(a−1 ) = 0 by deﬁnition of h. every ideal is of the form (pm ). and p is prime because M is a prime ideal. an ideal I of V is ﬁnitely generated. Also. (If ba−1 = 1 with b ∈ R . Finally. then a is integral over R. VALUATION RINGS 6. If the ideals of V are partially ordered by inclusion. Let R be a subring of the ﬁeld K. (If a is integral over R. hence a belongs to each Vα . so a = pb. If b is a nonunit. . Then R is integrally closed if and only if R = ∩α Vα . and we get a = p2 c. then (a) = V if a is a unit. . we may renumber the ai so that (a1 ) ⊆ (a2 ) · · · ⊆ (an ). Conversely. let V be an integral domain with fraction ﬁeld K. say by a1 . or (b) ⊆ (a). a contradiction. By (3. Then a fails to / belong to the ring R = R[a−1 ]. hence over any valuation ring V ⊇ R. then p divides b. so a ∈ V . so the same is true of VP and V /P . so is p. in which case b/a ∈ V . an . then a is integral over each Vα .) It follows that a−1 belongs to a maximal ideal M of R .4). then α = a/b with a and b nonzero elements of V . the maximal ideal M of V is (p) for some p. hence has a fraction ﬁeld. (Note that if a is a unit. By Property 5. The “only if” part follows from Property 9. Now h(a−1 ) = h(a−1 ) since a−1 ∈ M ⊆ R .) Thus a−1 cannot be a unit in R .1. then V is a valuation ring of K. If α is a nonzero element of K. 10. .) ♣ . Let C be an algebraic closure of the ﬁeld k = R /M . then pm divides a for all m ≥ 1. hence so is R. (If a is a polynomial in a−1 . so I = (an ). a ∈ M.

with equality if and only if x = 0. so a−1 ∈ V . we take v(pr m/n) = r. 3.3 Proposition If v is a discrete valuation on the ﬁeld K.1. (a) v(x) = ∞ if and only if x = 0. The familiar absolute values on the reals and the complex numbers are archimedean. An arbitrary valuation ring corresponds to a generalized absolute value mapping into an ordered group rather than the real numbers. (c) v(x + y) ≥ min(v(x). as discrete valuations will be entirely adequate for us. proving that V is a valuation ring. 3.4 Deﬁnitions and Comments Discrete valuations do not determine all valuation rings. |x| ≥ 0.3. where c is a constant with 0 < c < 1.2 Examples We can place a discrete valuation on all of the ﬁelds of Subsection 3. If a ∈ V . |x + y| ≤ |x| + |y|. A discrete valuation induces a nonarchimedean absolute value via |x| = cv(x) . M is the ideal of nonunits and is therefore the maximal ideal of the valuation ring V .3. ♣ 3. However.3. In Example 3. 3. then V = {a ∈ K : v(a) ≥ 0} is a valuation ring with maximal ideal M = {a ∈ K : v(a) ≥ 1}. (b) v(xy) = v(x) + v(y). such that for every x. Since a is a unit of V iﬀ both a and a−1 belong to V iﬀ v(a) = 0. A discrete valuation on K is a surjective map v : K → Z ∪ {∞}. such that for every x.3.3. The deﬁning properties (a). 1.3. The absolute value is nonarchimedean if the third condition is replaced by a stronger version: 3 . / then v(a) < 0.3. A valuation ring V arising from a discrete valuation v as in . (b) and (c) of 3. Proof. ∞ v( i=r ai xi ) = r (if ar = 0). |y|).1 Discrete Valuation Rings Deﬁnitions and Comments An absolute value on a ﬁeld K is a mapping x → |x| from K to the real numbers.3. |x + y| ≤ max(|x|. We will not consider the general situation. DISCRETE VALUATION RINGS 5 3.3 3. In Examples 1 and 2. 2. our interest will be in nonarchimedean absolute values. v(y)). archimedean means not nonarchimedean. Here is where most of them come from. |xy| = |x| |y|. v(f /g) = deg g − deg f . In Example 4. As expected. y ∈ K.1 show that V is a ring. y ∈ K. so v(a−1 ) = v(1) − v(a) = 0 − v(a) > 0.

and therefore Mn ⊆ I. and consequently a ∈ (t). . Proof. contradicting the hypothesis that I is nonzero. then every nonzero element a ∈ K can be expressed uniquely as a = utn where u is a unit of V and n ∈ Z. by convention. As in the proof of (3.6). 3.6 CHAPTER 3.3. An element t ∈ V with v(t) = 1 is called a uniformizer or prime element. The last statement follows by Property 1 of Section 3. note that if a = utn .6). By (3. in particular. Also. with v(b) = k ≥ n by minimality of n. }. (t) ⊆ M. Choose a ∈ I such that n = v(a) is as small as possible.7 Proposition Every nonzero ideal I of the DVR V is of the form Mn . t. so that v(at−n ) = 0 and therefore at−n is a unit u. Since M is the unique maximal ideal. 3. that is. so at−1 ∈ V . Since k ≥ n we have b ∈ (tn ) = Mn .3. abbreviated DVR.5 Proposition Let t be a uniformizer in the discrete valuation ring V . Then t generates the maximal ideal M of V . Thus 1 = v(t) = v(c) + v(t ) ≥ 0 + 1 = 1. if t is any generator of M.8 Proposition Let I be a nonzero ideal of the discrete valuation ring R. K = S −1 V where S = {1. is 0. . A uniformizer tells us a lot about the DVR V and the ﬁeld K. Let n = v(a). If Mr = Ms with r < s.3. we have t = ct for some c ∈ V . so v(at−1 ) = v(a) − v(t) ≥ 1 − 1 = 0. Conversely. bt−k is a unit u . Proof. The uniqueness of n is a consequence of Nakayama’s lemma. Conversely. M0 = V . so b = u tk .3. ♣ We may interpret v(I) as the length of a composition series. proving that I ⊆ Mn . is determined by a. Since t ∈ M.6 Proposition If t is a uniformizer. M is principal. and hence u. which forces v(t ) = 1. then v(a) ≥ 1. VALUATION RINGS (3. Then v(I) = lR (R/I). 3.3. . so tn = u−1 a ∈ I. We write v(I) = n. then t is a uniformizer. Proof. ♣ A similar result holds for ideals.3) is said to be a discrete valuation ring.5). K = Vt . where M is the maximal ideal of V and n is a unique nonnegative integer. Now suppose M = (t ). By (3. t2 . a = utn . the composition length of the R-module R/I. so that n. ♣ 3.3. then v(a) = v(u) + nv(t) = 0 + n = n. hence M. M = (t). If a ∈ M. Thus Mr . To prove uniqueness. then Mr = Mr+1 = MMr .3. .3. let b ∈ I.2 and the observation that the elements of V are those with valuation n ≥ 0.

. .7)]. For some positive integer n.7).3. . say by a1 . . We claim that ri ≥ ni for some i. we have √ ( I)m ⊆ I.3. . with ani ∈ I. By basic properties of composition length. it is an R/M-module. . with l = lR . that is. hence R/I ⊃ M/I ⊃ M2 /I ⊃ · · · ⊃ Mn /I = 0. i = 0. 3. 3. √ Proof. I is ﬁnitely generated. √ 2. we have. . over the ﬁeld R/M. and t t m= i=1 ri < 1 + i=1 (ni − 1) = m.10 Proposition Let M be a maximal ideal of the Noetherian ring R. Q is M-primary. ♣ We are going to prove a characterization theorem for DVR’s. M/I M /I 7 Continuing in this fashion.3. The vector space is one-dimensional because the Mi . Then for some positive integer m. we have Mn ⊆ Q ⊆ M. and let Q be any ideal of R. n. Consequently. we get n−1 l(R/I) = i=0 l(Mi /Mi+1 ). then ri ≤ ni − 1 for all i. If not. . are distinct [see the proof of (3. In particular (take I = 0). l(R/I) = l( R/I M/I ) + l(M/I) = l(R/M) + l( 2 ) + l(M2 /I). a vector space. Q = M. ♣ t 1 3.3. √ Since R is Noetherian.3. 1.9 Proposition Let I be an ideal of the Noetherian ring R. hence ( I)m ⊆ I. The following conditions are equivalent: 1. √ a contradiction. By (3. Since M is generated by a uniformizer t. at . it follows that ti + Mi+1 generates Mi /Mi+1 . Since Mi /Mi+1 is annihilated by M. . we have R ⊃ M ⊃ M2 ⊃ · · · ⊃ Mn = I. l(R/I) = n. and some preliminary results will be needed. But then each product ar1 · · · art is in I. the nilradical of R is nilpotent. i t Then ( I)m is generated by all products ar1 · · · art with i=1 ri = m. DISCRETE VALUATION RINGS Proof.3. Our choice of m t 1 is t m=1+ i=1 (ni − 1).

Mn ⊆ Q ⊆ Q = M. 6. it would be contained in M. Now β −1 M ⊆ R. and a PID is a UFD in which every nonzero prime ideal is maximal. for if β −1 M ⊆ M.7) and (3.4) that β −1 is integral over R. 3.1.3. We have (1) implies (2) by deﬁnition of M-primary.3. Now let I be any nonzero ideal of R. but by the above claim we have I ⊆ ∩n=0 Mn . hence bn = bn+1 t. But t ∈ M. M is principal. for otherwise we are ﬁnished. By hypothesis. and we have a = bn tn = 0. By (3. (4) ⇒ (3): Let t be a nonzero element of M. 2. The dimension of M/M2 as a vector space over R/M is 1. then β −1 stabilizes a ﬁnitely generated R-module. ♣ Now we can characterize discrete valuation rings.1).8 CHAPTER 3.3. so (1 − ct)bn = 0. Thus there exists n ≥ 0 such that I ⊆ Mn and I ⊆ Mn+1 . Proof. Every nonzero ideal is a power of M. then a ∈ Rt = (t). (Note that a ∈ Mn−1 and Mn ⊆ (t). √ √ M = Mn ⊆ Q ⊆ M = M. proving the claim. β −1 is not integral over R.1).1.) Thus β −1 M is an ideal of R.9) with I = Q to get. (3) ⇒ (2): By hypothesis. and we claim that ∩∞ Mn = 0. To prove that (3) implies (2). R is a principal ideal domain. apply (3. for some n ≥ 1 we have Mn ⊆ (t) ⊆ M. see (1. because β −1 M = (a/t)M ⊆ (1/t)Mn ⊆ R. observe that by (1. so the radical of (t). R is integrally closed and every nonzero prime ideal is maximal. contradicting β −1 M ⊆ M. β −1 M = R. Then bn tn = bn+1 tn+1 .5).2). To prove that (2) implies (3). for some positive integer n. (2) ⇒ (4): This holds because a PID is integrally closed. The implication (2) ⇒ (1) follows from (1.10). contradicting the choice of a. Thus (bn ) ⊆ (bn+1 ) for all n. But then β −1 M ⊆ M.1. and consequently ct = 1. hence M is the principal ideal (β). n=0 Suppose that a belongs to Mn for all n. M is the only nonzero prime ideal. 3. Choose . Therefore β −1 ∈ R. VALUATION RINGS Proof.11 Theorem Let R be a Noetherian local domain with fraction ﬁeld K and unique maximal ideal M = 0. 4. a contradiction. and if it were proper. with a = bn tn for some bn ∈ R. and we may assume that (t) ⊂ M. (1) ⇒ (2): This follows from (3.1.3.3. and in fact (bn ) = (bn+1 ) for suﬃciently large n because R is Noetherian.) The following conditions are equivalent: 1. coincides with M. Thus bn must be 0. R is a discrete valuation ring. 5. Then I ⊆ M. Therefore bn = bn+1 t = ctbn for some c ∈ R. Thus for some n ≥ 2 we have / Mn ⊆ (t) but Mn−1 ⊆ (t). Choose a ∈ Mn−1 with a ∈ (t). M is a principal ideal (t). Consequently. so t is not a unit. / Since R is integrally closed. which is the intersection of all prime ideals containing t. and we conclude from the implication (4) ⇒ (1) in (2. and let β = t/a ∈ K. If β −1 = a/t ∈ R. (Thus R is not a ﬁeld.

M = (t). For the “only if” part. For ﬁxed a.3. n=0 n=0 we will have a ∈ (tn ) but a ∈ (tn+1 ) for some n. It follows that if β is a nonzero element of the fraction ﬁeld K.3. so we may choose t ∈ M \ M2 .11). a member of M. We cannot have n = 0 because (t) ⊆ M ⊂ R.3. where u is a unit of R and m is an integer. then v is a discrete valuation on K with valuation ring R. By hypothesis. Let a be any nonzero element of R. Then (a) ⊆ M. ♣. Now M(M/(t)) = (M2 + (t))/(t) = M/(t). This follows from the proof of (5) implies (3) in (3. If t ∈ M \ M2 . M = M2 . proving that I is principal. and by the proof of (3) ⇒ (2).8).2. The only possibility is n = 1. R is a local ring by Property 3 of Section 3.12 Corollary The ring R is a discrete valuation ring if and only if R is a local PID that is not a ﬁeld. that is. / / But then u is a unit.3.3. then every nonzero element a ∈ R is a unit. If R is a ﬁeld. Thus R(t+M2 )/M2 = M/M2 . In particular. To summarize. Proof.3. If we deﬁne v(β) = m.3. M is a principal ideal (t). (1) ⇒ (6): This follows from the proof of (3. t + M2 = M.3. is a nonunit). and since ∩∞ Mn = 0. contradicting our convention.11).3.11). (6) ⇒ (3): By hypothesis. note that a discrete valuation ring R is a PID by (1) implies (2) of (3. since Mn = (t)n = (tn ). 3. By the correspondence theorem. so tn = u−1 a ∈ I. But then t + M2 is a generator of the vector space M/M2 over the ﬁeld R/M. then β = utm uniquely. ♣ . Moreover. DISCRETE VALUATION RINGS 9 a ∈ I \Mn+1 .7). we have a = utn with u ∈ M (because a ∈ Mn+1 ). since R is a PID. ♣ 3. it is Noetherian. in other / / words. u is a unit. both u and n are unique (because t. hence M = (t). ∩∞ Mn = 0. M/(t) = 0. Thus the valuation v is trivial. then t is a uniformizer. Choose t ∈ M \ M2 . I ⊆ Mn = (tn ) ⊆ I.7). (t) = Mn for some n ≥ 0. and we cannot have n ≥ 2 by choice of t. (5) ⇒ (3): As in the proof of (3. The “if” part follows from (2) implies (1) in (3. so by NAK. the Noetherian hypothesis is not used here. (2) ⇒ (1): By hypothesis. hence v(a) = 0.13 Corollary Let R be a DVR with maximal ideal M. Let us agree to exclude the trivial valuation v(a) = 0 for every a = 0. (1) ⇒ (5): This follows from (3. Proof. possibly negative. Thus a = utn with u ∈ M.3. M = M2 .

the Rn are additive subgroups such that R = R0 ⊇ R 1 ⊇ · · · ⊇ R n ⊇ · · · with Rm Rn ⊆ Rm+n . we will be interested in the I-adic ﬁltrations of R and of M . in other words. Sometimes. where Rm Mn ⊆ Mm+n . The prototype is a polynomial ring in several variables. Note that the identity element of a graded ring R must belong to R0 . then 1a = a forces the degree of a to exceed n. In fact any metric space can be embedded as a dense subset of a complete metric space. 4. In this case. given respectively by Rn = I n and Mn = I n M . this procedure can be generalized to modules. The construction is a familiar one involving equivalence classes of Cauchy sequences.1 Graded Rings and Modules Deﬁnitions and Comments A graded ring is a ring R that is expressible as ⊕n≥0 Rn where the Rn are additive subgroups such that Rm Rn ⊆ Rm+n .Chapter 4 Completion The set R of real numbers is a complete metric space in which the set Q of rationals is dense. with Rd consisting of all homogeneous polynomials of degree d (along with the zero polynomial). If I is an ideal of the ring R and M is an R-module.1 4. a contradiction. each Mn is a submodule and we require that Rm Mn ⊆ Mm+n .) 1 . Rn is referred to as the nth graded piece and elements of Rn are said to be homogeneous of degree n. A ﬁltered module M = M0 ⊇ M 1 ⊇ · · · ⊇ · · · over the ﬁltered ring R may be deﬁned similarly. A graded module over a graded ring R is a module M expressible as ⊕n≥0 Mn . so that M0 = M . (Take I 0 = R. Now suppose that {Rn } is a ﬁltration of the ring R.1. We will see that under appropriate conditions. We call R a ﬁltered ring. For if 1 has a component a of maximum degree n > 0.

By induction hypothesis. then ab ∈ Rm+n+1 . It suﬃces to show that Rn ⊆ R for all n ≥ 0 (and therefore R = R ). the associated graded ring of R is deﬁned as gr(R) = n≥0 grn (R) where grn (R) = Rn /Rn+1 . Conversely. . we deﬁne scalar multiplication by (a + Rm+1 )(x + Mn+1 ) = ax + Mm+n+1 and it follows that (Rm /Rm+1 )(Mn /Mn+1 ) ⊆ Mm+n /Mm+n+1 . . If a ∈ Rn . . . ar of degree n1 . It is natural to ask for conditions under which a graded ring will be Noetherian. . then so is R0 . Xn ]. . If M is a ﬁltered module over a ﬁltered ring R. Then R is Noetherian if and only if R0 is Noetherian and R is a ﬁnitely generated R0 -algebra. if R is Noetherian. because R0 ∼ R/I where I is the ideal ⊕d≥1 Rd . If the condition on R0 holds. so assume as an induction hypothesis that Rd ⊆ R for d ≤ n − 1. . . . and since ai ∈ R we have a ∈ R . and the behavior of the subring R0 is critical. . If a ∈ Rm and x ∈ Mn . I is ﬁnitely = generated. If a ∈ Rm+1 and b ∈ Rn . . . 4. We must be careful in deﬁning multiplication in gr(R). .1. we deﬁne the associated graded module of M as gr(M ) = n≥0 grn (M ) where grn (M ) = Mn /Mn+1 . . so multiplication is well-deﬁned. say by homogeneous elements a1 . Thus gr(M ) is a graded module over the graded ring gr(R). Let R = R0 [a1 . then a + Rm+1 ∈ Rm /Rm+1 and b + Rn+1 ∈ Rn /Rn+1 . . . Proof. . hence R is Noetherian by the Hilbert Basis Theorem. then a can be expressed as c1 a1 + · · · + cr ar . . then R is a quotient of a polynomial ring R0 [X1 . We take (a + Rm+1 )(b + Rn+1 ) = ab + Rm+n+1 so that the product of an element of grm (R) and an element of grn (R) will belong to grm+n (R). If a ∈ Rm and b ∈ Rn . where n > 0.2 Associated Graded Rings and Modules If {Rn } is a ﬁltration of R. ar ] be the R0 -subalgebra of R generated by the ai . nr respectively. ♣ We now prepare for the basic Artin-Rees lemma. By hypothesis. . .2 CHAPTER 4. r) must be a homogeneous element of degree n − ni < n = deg a.1. . where ci (i = 1. ci ∈ R . We have R0 ⊆ R by deﬁnition of R .3 Proposition Let R = ⊕d≥0 Rd be a graded ring. COMPLETION 4.

Note that the I-adic ﬁltration is I-stable. M ∗ is ﬁnitely generated over R∗ . Deﬁne a graded ring R∗ and a graded R∗ -module M ∗ by R∗ = n≥0 I n. R∗ is a Noetherian ring. 2.3). 4. ∗ ∗ By deﬁnition. Proof. Since R is Noetherian. given by Nn = N ∩ Mn and (M/N )n = (Mn + N )/N respectively. GRADED RINGS AND MODULES 3 4. An I-ﬁltration with IMn = Mn+1 for all suﬃciently large n is said to be I-stable. and N ∗ = ⊕n≥0 Nn .6 Induced Filtrations If {Mn } is a ﬁltration of the R-module M . M is the union of the Mn over all n ≥ 0. let R∗ = ⊕n≥0 I n .4. then we have ﬁltrations induced on N and M/N . it follows that Mn is a ﬁnitely generated R∗ -module. 4.1. 1. ♣ 4. we conclude that {Nn } is I-stable. We say that {Mn } is an I-ﬁltration if IMn ⊆ Mn+1 for all n. Proof.1. I an ideal of R. the ﬁltration {Mn } is I-stable.5). As in (4.1. The following conditions are equivalent. and assume that M has an I-stable ﬁltration {Mn }. M is ﬁnitely generated over the Noetherian ring R and {Mn } is I-stable. Therefore M ∗ is ﬁnitely generated ∗ over R∗ if and only if M ∗ = Mm for some m. {Mn } is I-stable.1. and N is a submodule of M . and deﬁne i=0 ∗ Mn = M0 ⊕ · · · ⊕ Mn ⊕ IMn ⊕ I 2 Mn ⊕ · · · ∗ Since Nn is ﬁnitely generated over R.5 Proposition Let M be a ﬁnitely generated module over a Noetherian ring R.1.) By (4.5). Mm+k = I k Mm for all k ≥ 1. Let N be a submodule of M . I is ﬁnitely generated. Let Nn = ⊕n Mi .1. Again using (4.7 Artin-Rees Lemma Let M be a ﬁnitely generated module over the Noetherian ring R. where I is an ideal of R.1. M ∗ = ⊕n≥0 Mn .5). ♣ . M∗ = n≥0 Mn .4 Deﬁnitions and Comments Let M be a ﬁltered R-module with ﬁltration {Mn }. so R∗ is a ﬁnitely generated R-algebra. Then the ﬁltration {Nn = N ∩ Mn } induced by M on N is also I-stable. Equivalently. in other words. and suppose that {Mn } is an I-ﬁltration of M . so by (4. Therefore the submodule N ∗ is also ﬁnitely generated over R∗ .1. Now by hypothesis. (Elements of R∗ can be expressed as polynomials in a ﬁnite set of generators of I. Then M ∗ is ﬁnitely generated.1.

If i ≤ j. and let I be the ideal (p) where p is a ﬁxed prime. 4. The inverse limit of the Mn is the ring Zp of p-adic integers. . . we have a homomorphism fij from Mj to Mi . and the translations x + I m M form a basis for the neighborhoods of an arbitrary point x ∈ M . Now suppose that we have homomorphisms gi from an R-module M to Mi . . We can make M into a topological abelian group in which the module operations are continuous. . with R-module homomorphisms θn : Mn → Mn−1 . and the coherence of the gi forces the sequence (gi (x)) to be coherent.)The collection of modules and maps is called an inverse system.2.1 Completion of a Module Inverse Limits Suppose we have countably many R-modules M0 . g(x) = (gi (x)). 4. . (xi )(yi ) = (xi yi ). The above equation says that the I-adic topology on N coincides with the topology induced on N by the I-adic topology on M . . Then the gi can be lifted to a homomorphism g from M to M . Take R = Z. indexed by a directed set.1. The sets I m M are a base for the neighborhoods of 0. COMPLETION 4. in other words. . The inverse limit is denoted by lim Mn . Explicitly. as coherent sequences can be multiplied componentwise. then fij ◦ fjk = fik . Call the gi coherent if θi+1 ◦ gi+1 = gi for all i. but the ideas carry over to an arbitrary family of modules. .2 4. ←− Note that M becomes an R-module with componentwise addition and scalar multiplication of coherent sequences. The ﬁltration on N induced by the I-adic ﬁltration on M is given by Nm = (I m M )∩N . By Artin-Rees. The collection M of all coherent sequences is called the inverse limit of the inverse system. .4 CHAPTER 4. that is. with N a submodule of M . (xi ) + (yi ) = (xi + yi ) and r(xi ) = (rxi ). We assume that the maps can be composed consistently. The resulting topology is called the I-adic topology on M . if i ≤ j ≤ k. in other words. n ≥ 1. 1. i = 0.2. (We are restricting to the countable case to simplify the notation. A sequence (xi ) in the direct product Mi is said to be coherent if it respects the maps θn in the sense that for every i we have θi+1 (xi+1 ) = xi . Take Mn = Z/I n and θn+1 (a + I n+1 ) = a + I n . There is a basic topological interpretation of this result. M1 . for large enough m we have I k ((I m M ) ∩ N ) = (I m+k M ) ∩ N for all k ≥ 0. An inverse limit of an inverse system of rings can be constructed in a similar fashion.2 Examples 1.8 Applications Let M be a ﬁnitely generated module over the Noetherian ring R.

Notice that the condition xn − xm ∈ Mr is equivalent to xn + Mr = xm + Mr . or the process of completing an arbitrary metric space. Then x fails to belong to N iﬀ some neighborhood of x is disjoint from N . ♣ 4. m ≥ N . .4 Corollary The topology is Hausdorﬀ if and only if ∩∞ Mn = {0}. n = 1. If we go far out in a Cauchy sequence.2. This holds because separating distinct points x and y by disjoint open sets is equivalent to separating x − y from 0. 4. More precisely.1.3 Proposition If N is a submodule of M . By (4. x ∈ N +Mn / for some n. . n=0 Proof. . then for some z ∈ Mn we have y = x − z. Recalling the construction of the reals from the rationals. Let x be an element of M . then x = y + z. Thus y = x − z ∈ (x + Mn ) ∩ N . given a positive integer r there exists a positive integer N such that xn − yn ∈ Mr for all n ≥ N .3). the diﬀerence between terms becomes small. . . Let Mn = R/I n and θn (f + I n ) = f + I n−1 . (x+Mn )∩N = ∅ for some n. if y ∈ (x + Mn ) ∩ N . note that if x ∈ N + Mn . the T1 condition. . . let us try to come up with something similar in this case.2. . An element of Mn is represented by a polynomial f of degree at most n − 1. xn ] be a polynomial ring in n variables. COMPLETION OF A MODULE 5 2. . We have the following result. n=0 Proof. This suggests that the essential feature of the Cauchy condition is that the sequence is coherent with respect to the maps θn : M/Mn → M/Mn−1 . The ﬁltration determines a topology on M as in (4. Now let M be a ﬁltered R-module with ﬁltration {Mn }. and becomes exact under certain conditions. Let R = A[x1 . and the result follows.5 Deﬁnition of the Completion Let {Mn } be a ﬁltration of the R-module M . and I the maximal ideal (x1 . xn ]] of formal power series. in other words.4.2. . .8). ←− The functor that assigns the inverse limit to an inverse system of modules is left exact. . then the closure of N is given by N = ∩∞ (N + Mn ). so we are asserting that the Hausdorﬀ property is n=0 equivalent to points being closed. we deﬁne the completion of M as ˆ M = lim(M/Mn ). y ∈ N. with the Mn forming a base for the neighborhoods of 0. ∩∞ Mn = {0}. (We take the degree of f to be the maximum degree of a monomial in f . that is. Thus the inverse limit can be identiﬁed with the ring A[[x1 . 2. Equivalently. We identify the Cauchy sequences {xn } and {yn } if they get close to each other for large n. Conversely. xn ). . To justify the last step. . Motivated by this observation. Thus we can deﬁne a Cauchy sequence {xn } in M by the requirement that for every positive integer r there is a positive integer N such that xn − xm ∈ Mr for n. . .2. ♣ 4. z ∈ Mn . so x = y + z ∈ N + Mn .2.) The image of f in I n−1 is represented by the same polynomial with the terms of degree n − 1 deleted. .

proving the ﬁrst assertion. Let M = Mn and deﬁne an R. Proof. Then the sequence ˆ 0 → (M ) ˆ→ M → (M ) ˆ→ 0 . Let {Mn } be a ﬁltration of M . Now the maps (fn ) and (gn ) induce f = fn : M = Mn → M and g = gn : M → M = Mn . then dM is surjective.2.6 Theorem Let {Mn . θn }. 0 G Mn+1 G Mn fn+1 G Mn+1 G Mn gn+1 G Mn+1 G Mn θn+1 G0 θn+1 fn θn+1 gn 0 Then the sequence G0 0 → lim Mn → lim Mn → lim Mn ←− ←− ←− is exact. ♣ 4.6 CHAPTER 4. so that {Mn } induces ﬁltrations {M ∩ f −1 (Mn )} and {g(Mn )} on M and M respectively.7 Corollary Suppose that the sequence 0 GM f GM g GM G0 is exact. and assume that the diagram below is commutative with exact rows. The result is an exact sequence 0 → ker dM → ker dM → ker dM → coker dM . If θn is surjective for all n. θn } be inverse systems of modules. and {Mn . 0 GM dM f GM GM dM g GM GM dM G0 G0 0 GM f g We now apply the snake lemma. θn }.homomorphism dM : M → M by dM (xn ) = (xn − θn+1 (xn+1 )). and consequently the cokernel of dM is 0.2. COMPLETION 4. then 0 → lim Mn → lim Mn → lim Mn → 0 ←− ←− ←− is exact. which is discussed in detail in TBGY (Section S2 of the supplement). We have the following commutative diagram with exact rows. The second assertion follows. {Mn . If θn is surjective for all n. The kernel of dM is the inverse limit of the Mn .

the ﬁltrations {I n N } and {N ∩ I n M } of the submodule N are equivalent (Problem 5). 0 G M /(M ∩ f −1 (Mn+1 )) θn+1 G M/Mn+1 θn+1 G M /g(Mn+1 ) θn+1 G0 G M /(M ∩ f −1 (Mn )) G M/Mn G0 G M /g(Mn ) 0 Since θn is surjective for all n.9 Corollary ˆ Let {Mn } be a ﬁltration of the R-module M . so Cauchy sequences (and coherent sequences) can be identiﬁed with single points. and ˆ ˆ = M /Mn ∼ (M/Mn ) (M ). under the hypothesis of (4.2. Let Mn be the completion of Mn with ˆ ˆ respect to the induced ﬁltration on Mn [see (4.10 Remarks Two ﬁltrations {Mn } and {Mn } of a given R-module are said to be equivalent if they induce the same topology. Then Mn is a submodule of M and ˆ /Mn ∼ M/Mn for all n. and we have M /Mn ∼ M/Mn for every n.2.2. but the point is worth emphasizing. Since equivalent ﬁltrations give rise to the same set of Cauchy sequences. For example. to obtain the exact sequence ˆ ˆ 0 → Mn → M → (M/Mn ) ˆ→ 0.7) (with f the inclusion map and g the canonical epimorphism). We apply (4. 7 Proof.6) allows us to pass to the inverse limit. ˆ = M Proof. Therefore M is isomorphic to its comˆ ˆ = pletion.8)].4.2. COMPLETION OF A MODULE is exact.2. 4.1. ˆ= ˆ Now the mth term of the induced ﬁltration on M is Mm = (Mn + Mm )/Mn = Mn /Mn = 0 for m ≥ n. Exactness of the given sequence implies that the diagram below is commutative with exact rows. ♣ 4.8 Remark A ﬁltration {Mn } of an R-module M induces in a natural way a ﬁltration {N ∩ Mn } on a given submodule N .2. Thus M has the discrete topology. We have already noted this in (4. ˆ ˆ Thus we may identify Mn with a submodule of M . and a ﬁltration {(N + Mn )/N } on the quotient module M/N . ♣ 4.2.8). .7) with M = Mn and M = M/Mn . it follows that completions of a given module with respect to equivalent ﬁltrations are isomorphic.2. (4.

2).1) and (4. COMPLETION 4. then ∩∞ I n M = 0. 1 + a annihilates everything in the kernel N .3.2) gives precise information about this kernel.1 The Krull Intersection Theorem Deﬁnitions and Comments Recall from (4. contradicting the hypothesis that I is proper. n=0 ˆ Proof. ♣ 4. Suppose that a ∈ I. hence 1. so the above equation says that IN = N . and since R is a domain.1. which is ∩n=0 I M if the ﬁltration is I-adic.3.2. Set k = 1 to get I((I n M ) ∩ N ) = (I n+1 M ) ∩ N. If 0 = x ∈ N then (1 + a)x = 0. 4.8 CHAPTER 4. so if x ∈ N (hence (1 + a)x = 0). The completion of M with respect to the I-adic ﬁltration is called the I-adic completion. But N ⊆ I n+1 M ⊆ I n M . By (0. The kernel of this map is ∩n=0 Mn .1.8) that the I-adic topology on the R-module M is the topology induced on M by the I-adic ﬁltration Mn = I n M .3. then ∩∞ I n = 0.3. Then N is the set of elements x ∈ M such that x is annihilated by some element of 1 + I.2. Proof. I an ideal of R. I k ((I n M ) ∩ N ) = (I n+k M ) ∩ N. Conversely.1). The Krull intersection theorem (4.4 Corollary Let M be a ﬁnitely generated module over the Noetherian ring R.1).3. 1 + a = 0. Let N be the kernel of the natural map M → M .1. ♣ . hence x ∈ I n M for all n ≥ 0. we must show that for some a ∈ I. for all k ≥ 0. belongs to I. we have x = 0. Then x = −ax = −a(−ax) = a2 x = a2 (−ax) = −a3 x = a4 x = · · · . for some n we have. x ∈ M .4). Thus by (4. the I-adic n=0 topology on M is Hausdorﬀ.3.3 4.8). Proof. By (4. The intersection of the I n is the kernel N of the natural map from R to R. we can ﬁnd a single element of 1 + I that works for the entire kernel. If the ideal I of R is contained in the Jacobson radical J(R). and ˆ ˆ M the I-adic completion of M . there exists a ∈ I such that (1 + a)N = 0. 1 + a is a unit of R. Let a ∈ I ⊆ J(R) be such that (1 + a) annihilates the kernel N = ∩∞ I n M n=0 ˆ of the natural map from M to M . By (4. But then −1. ˆ There is a natural map from a ﬁltered module M to its completion M given by ∞ ∞ n x → {x + Mn }. and (1 + a)x = 0.2 Theorem Let M be a ﬁnitely generated module over the Noetherian ring R. In fact.3.3 Corollary If I is a proper ideal of the Noetherian integral domain R. By (0. 1 + a annihilates N for some a ∈ I. ♣ 4.

hn = H. and similarly the degree of wi is at most r. Then deg g1 = r and deg h1 = d − r. (1) . in particular the n=0 M-adic topology on R. Since g n has degree r. then ∩∞ Mn M = 0.4. . Since f = g 1 h1 . . The inductive step: Assume that gn and hn have been constructed. h1 ∈ A[X] such that g 1 = G and h1 = H. Let f (X)−gn (X)hn (X) = d i n i=0 ci X with the ci ∈ P . . A is a discrete valuation ring of a local ﬁeld. Thus the coeﬃcients of f − gn hn belong to P n . h = H and f = gh. Since G = g n and H = hn are relatively prime.1 Hensel’s Lemma Assume that f is a monic polynomial of degree d in A[X]. Assume that A is complete with respect to the P -adic topology. then the coset a + P ∈ k will be denoted by a.5 Corollary Let R be a Noetherian local ring with maximal ideal M. If f is a polynomial in A[X]. Thus the M-adic topology on M . every Cauchy sequence converges. .4. the degree of v i is at most d − r. Thus d d f (X) = i=0 ai X ∈ A[X]. and let k = A/P be the residue ﬁeld. If M is a ﬁnitely generated R-module. . . then reduction of the coeﬃcients of f mod P yields a polynomial f in k[X]. f (X) = i i=0 ai X i ∈ k[X]. Let r be the degree of G. deg hn = d − r.3. X i − vi (X)gn (X) − wi (X)hn (X) ∈ P [X]. in other words. n = 1. d there are polynomials v i and wi in k[X] such that X i = v i (X)g n (X) + wi (X)hn (X). Choose monic g1 . . Here is the precise statement. such that deg gn = r.3. But the statement and proof of the algebraic number theory result can be copied. If a ∈ A. so that deg H = d − r. Hensel’s lemma is about lifting a factorization of f from k[X] to A[X]. . g n = G. we have f − g1 h1 ∈ P [X]. for each i = 0. is Hausdorﬀ. hn ∈ A[X]. We will inductively construct gn . Proof.4 Hensel’s Lemma Let A be a local ring with maximal ideal P . where this result is most often applied. In algebraic number theory. The basis step: Let n = 1. and f (X) − gn (X)hn (X) ∈ P n [X]. HENSEL’S LEMMA 9 4. ♣ 4. 4. and that the corresponding polynomial F = f factors as the product of relatively prime monic polynomials G and H in k[X]. as follows.4). Proof. Since M = J(R).4. Then there are monic polynomials g and h in A[X] such that g = G. Moreover. this follows from (4. 2.

it follows that g n+1 = g n = G and hn+1 = hn = H. COMPLETION d gn+1 (X) = gn (X) + i=0 ci wi (X). hn+1 (X) = hn (X) + i=0 ci vi (X). completing the induction. the sequence of coeﬃcients of X i in gn (X) is Cauchy and therefore converges. Since the ci belong to P n ⊆ P . The ﬁnal term belongs to P 2n [X] ⊆ P n+1 [X]. so for any ﬁxed i.10 We deﬁne d CHAPTER 4. G and H are monic. By construction. Since g n = G and hn = H for all n. and similarly hn (X) → h(X). By the induction hypothesis. with the aid of Equation (1) above. with g(X). it must be exactly r. To simplify the notation we write gn (X) → g(X). and similarly the degree of hn+1 is d − r. the highest degree terms of g and h are of the form (1 + a)X r and (1 + a)−1 X d−r respectively. with a ∈ P . f − gn hn ∈ P n [X]. Since the degree of gn+1 is at most r. h(X) ∈ A[X]. The proof is complete. respectively.) By replacing g and h by (1 + a)−1 g and (1 + a)h. and we may let n → ∞ to get f = gh. the second grouped term belongs to P n P [X] = P n+1 [X].j ci cj wi vj . and. By deﬁnition of gn+1 . we have gn+1 − gn ∈ P n [X]. To check the remaining condition. f − gn+1 hn+1 = f − (gn + i ci wi )(hn + i ci vi ) = (f − gn hn − i ci X i ) + i ci (X i − gn vi − hn wi ) − i. (Note that 1 + a must reduce to 1 mod P . Since f. we can make g and h monic without disturbing the other conditions. we must have g = G and h = H. ♣ . the ﬁrst grouped term on the right is zero. Finishing the proof.

(We call ∆G the diﬀerence of G. . Add the equations G(a + 1) − G(a) = g(a). this charming subject is rarely taught these days.1 Lemma Let g and G be real-valued functions on the nonnegative integers.1. G(b + 1) − G(b) = g(b).Chapter 5 Dimension Theory The geometric notion of the dimension of an aﬃne algebraic variety V is closely related to algebraic properties of the coordinate ring of the variety. Then ∆k (r) = rk (r−1) . ♣ 1 . Later. It turns out to be needed in studying Hilbert and Hilbert-Samuel polynomials in the next section. Just compute: ∆k (r) = (k + 1)(r) − k (r) = (k + 1)k(k − 1) · · · (k − r + 2) − k(k − 1) · · · (k − r + 1) = k(k − 1) · · · (k − r + 2)[k + 1 − (k − r + 1)] = rk (r−1) .) Then b g(k) = G(k)|b+1 = G(b + 1) − G(a). . that is. ♣ 5. deﬁne k (r) = k(k − 1)(k − 2) · · · (k − r + 1). G(a + 2) − G(a + 1) = g(a + 1). Proof. . except in actuarial programs. the ring of polynomial functions on V . 5. We will see that under appropriate hypotheses. a k=a Proof. several concepts of dimension are equivalent. that is.2 Lemma If r is a positive integer. . and assume that ∆G = g.1. 5. we will connect the algebraic and geometric ideas. This relationship suggests that we look for various ways of deﬁning the dimension of an arbitrary commutative ring. G(k + 1) − G(k) = g(k) for all k ≥ 0.1 The Calculus of Finite Diﬀerences Regrettably.

Therefore n k 3 = [k (4) /4]|n+1 + 3n(n + 1)(2n + 1)/6 − 2n(n + 1)/2.3 Examples n k=1 k (2) ∆k (2) = 2k (1) . Then f is a polynomial-like function of degree r if and only if ∆f is a polynomial-like function of degree r − 1. so = [k (2) /2]|n+1 = (n + 1)n/2. k (3) = k(k − 1)(k − 2) = k 3 − 3k 2 + 2k. DIMENSION THEORY 5. The degree of f is taken to be the degree of g. In a similar fashion we can ﬁnd n k=1 k s for any positive integer s. 1 k 2 = k(k − 1) + k = k n + k (1) .1) and (5.1. so k 2 = [k (3) /3]|n+1 + (n + 1)n/2 = (n + 1)n(n − 1)/3 + (n + 1)n/2 1 k=1 = n(n + 1)(2n + 1)/6.1.5 Lemma Let f : N → Q. f (n) = g(n) for all suﬃciently large n (abbreviated n >> 0).1.2 Hilbert and Hilbert-Samuel Polynomials There will be two polynomial-like functions of interest.1. This follows from (5. 1 k=1 The ﬁrst term on the right is (n + 1)n(n − 1)(n − 2)/4. In other words. and is usually accomplished via the mean value theorem. (We can allow r = 0 if we take the degree of the zero polynomial to be -1.) ♣ 5.4 Deﬁnitions and Comments A polynomial-like function is a function f from the natural numbers (nonnegative integers) N to the rational numbers Q. (The analogous result from diﬀerential calculus that a function with zero derivative is constant is harder to prove. 5. so k 3 = k (3) + 3k 2 − 2k. and we begin preparing for their arrival.2 CHAPTER 5.2). so the sum of the ﬁrst n cubes is [n(n + 1)/4][n2 − 3n + 2 + 2(2n + 1) − 4] which simpliﬁes to [n(n + 1)/2]2 . such that f eventually agrees with a polynomial g ∈ Q[X]. 5.1.) Proof. along with the observation that a function whose diﬀerence is zero is constant. .

♣ 5. all with ﬁnite length. 5. Since M is Noetherian.3 Additivity of Length Suppose we have an exact sequence of R-modules 0 → A1 → A2 → · · · → An → 0. ci yi and ci zi must belong to ⊕m>n Mm . Nn is ﬁnitely generated over R.2. then each Mn is a ﬁnitely generated R0 -module. ♣ We will need the following basic property of composition length. m ≥ n.2.14). By (4. Then we have additivity of length.6. say by x1 . As above. that is. the length lRo (Mn ) of the R0 -module Mn is ﬁnite for all n ≥ 0. This is probably familiar for a short exact sequence 0 → N → M → M/N → 0. If y ∈ Mn .2. we can write ai = bi + ci where bi ∈ R0 and ci ∈ ⊕j>0 Rj .2. Proof. If M = ⊕n≥0 Mn is a ﬁnitely generated graded R-module.2 Corollary In (5. . . we can write xi = yi + zi with yi ∈ Mn and zi ∈ ⊕m>n Mm .1. xt . Since Nn = Mn ⊕ ⊕m>n Mm .) The general result is accomplished by decomposing a long exact sequence into short exact sequences. Section 7.) To see how the process works. where the additivity property can be expressed as l(M ) = l(N ) + l(M/N ). yt generate t Mn over R0 . R is a Noetherian ring. (See TBGY. 0 → A → B → coker f → 0. coker g ∼ im h (= ker i).1) and (1. Thus t t y= i=1 (bi + ci )(yi + zi ) = i=1 bi y i because the elements bi zi . HILBERT AND HILBERT-SAMUEL POLYNOMIALS 3 5. It suﬃces to prove that y1 .1 Proposition Let R = ⊕n≥0 Rn be a graded ring.3) and (1.2.2. . so our second short exact sequence is = 0 → im g → C → coker g → 0. consider an exact sequence GA f GB 0 Our ﬁrst short exact sequence is g GC h GD i GE G 0. (“Long” means longer than short.5. Proof.5. . .6. l(A1 ) − l(A2 ) + · · · + (−1)n−1 l(An ) = 0. Let Nn be the direct sum of the Mm . . Assume that R0 is Artinian and R is ﬁnitely generated as an algebra over R0 . Problem 5. Now coker f = B/ im f = B/ ker g ∼ im g (= ker h).1). Apply (5. . .13). But just as we decomposed xi above. then y is of the form i=1 ai xi with ai ∈ R. so the third short exact sequence is = 0 → im h → D → coker h → 0. . hence M is a Noetherian Rmodule.

we have the exact sequence G Kn G Mn λr G Mn+1 G Cn G 0. Applying additivity for short exact sequences. hence so is ∆h(M. then M is an R/I-module. ♣ 5. n). n ≥ 0.) By induction hypothesis. n) is polynomial-like of degree at most r − 1.2).6. n) − h(M. this becomes l(A) − l(B) + l(C) − l(D) + l(E) = 0 as desired.5 Deﬁnitions and Comments Let R be any Noetherian local ring with maximal ideal M. 0 Let K be the direct sum of the Kn and C the direct sum of the Cn . if Kn is the kernel. n) = 0 for n >> 0. ar ∈ R1 . as a function of n with M ﬁxed. h(K. By (5. where T is the graded subring of R generated over R0 by a1 .4 CHAPTER 5. h(M. Section 4. Problem 6. n). so by (5. . We argue by induction on r. h(K. If M is a ﬁnitely generated graded R-module. see TBGY. Now ar annihilates K and C.2.2. h(K. Then K is a submodule of M and C a quotient of M . . By hypothesis. is polynomial-like of degree at most r − 1. and Cn the cokernel. n) = lR0 (Mn ). n) − h(K. Proof.] . so λr (Mn ) ⊆ Mn+1 . If d is the maximum of the degrees of the generators. and let λr be the endomorphism of M given by multiplication by ar . n) are polynomial-like of degree at most r − 2. . DIMENSION THEORY But coker h ∼ im i = E.2. deﬁne h(M. Now assume r > 0. . n) + h(M.11) applies. n) and h(C. . Assume that R0 is Artinian and R is ﬁnitely generated as an algebra over R0 . [See (3.5). After cancellation.1) and (5. .3). Using slightly loose language. .10). so K and C are ﬁnitely generated T -modules.2. Choose a ﬁnite set of homogeneous generators for M over R. we call h the Hilbert polynomial of M . so we may replace the third short exact sequence by = 0 → im h → D → E → 0.4 Theorem Let R = ⊕n≥0 Rn be a graded ring.3. R/I is an Artinian √ ring. n) = 0 hence ∆h(M. (If an ideal I annihilates an R-module M . Thus K and C are ﬁnitely generated Noetherian graded R-modules. we have l(A) − l(B) + l(coker f ) − l(im g) + l(C) − l(coker g) + l(im h) − l(D) + l(E) = 0. n + 1) − h(C. then R = R0 .1.2. Then h. n) = h(C. An ideal I of R is said to be an ideal of deﬁnition if Mn ⊆ I ⊆ M for some n ≥ 1. ar−1 . . and note that I = M if and only if every prime ideal containing I is maximal. and therefore h(M.2. then Mn = 0 for n > d. n) and h(C. By (5. n ∈ N. n) are deﬁned and ﬁnite. ar belonging to R1 . of λr : Mn → Mn+1 . Equivalently. so (1. If r = 0. 5. with all generators a1 .

we have O(dn ) ≤ O(dn ) ≤ O(dn ). With the Iadic ﬁltrations on R and M . To see this.5). If I is generated over R by a1 . n) ≥ sM (M. . (Note that by deﬁnition of a graded ring. By (5. ar in I/I 2 generate grI (R) over R/I.4). n) + r(n) . with d3 = d1 . let t be a positive integer such that Mt ⊆ I ⊆ M. It follows that we may replace lR/I by lR in the above formula. . which allows us to produce elements in Rt for arbitrarily large t.6 The Hilbert-Samuel Polynomial Let I be an ideal of deﬁnition of the Noetherian local ring R. The Hilbert-Samuel polynomial satisﬁes a property analogous to (4. n) = sI (M. sI (M. n) + sI (M . Consequently ∆sI (M. from right to left.1. d2 and d3 . n) = lR (M/I n M ).2. If M is a ﬁnitely generated R-module. n) is the same for all possible choices. Then sI (M . n) = sI (M.7). Therefore all three 1 2 3 degrees coincide. HILBERT AND HILBERT-SAMUEL POLYNOMIALS 5 5. n) depends on the particular ideal of deﬁnition I. . . . Then for every n ≥ 1 we have Mtn ⊆ I n ⊆ Mn . so sM (M. . We deﬁne the Hilbert-Samuel polynomial by sI (M. then M/IM is a ﬁnitely generated module over the Artinian ring R/I.2.1. . so by (5. ar . hence has ﬁnite length over R/I. 5. . and suppose we have an exact sequence 0 → M → M → M → 0 of ﬁnitely generated R-modules. n). the associated graded ring and the associated graded module [see (4. the additivity of length. then the images a1 . n) is polynomial like of degree at most r. n) is ﬁnite.) By (5. An induction argument using additivity of length shows that sI (M.7 Theorem Let I be an ideal of deﬁnition of the Noetherian local ring R. grI (M ) = ⊕n≥0 (I n M/I n+1 M ). then N becomes an R/I-module via (a + I)x = ax. Ri Rj ⊆ Ri+j . Thus M/IM is Artinian (as well as Noetherian). n) = lR/I (I n M/I n+1 M ) < ∞.2. Again recall that if N is an R-module and the ideal I annihilates N . h(grI (M ). n). but the degree d(M ) of sI (M. n) = h(grI (M ).2)] are given by grI (R) = ⊕n≥0 (I n /I n+1 ).2.5.1. n + 1) − sI (M.4). tn) ≥ sI (M. If the degrees of these polynomial are. The Hilbert-Samuel polynomial sI (M. d1 . Now the sequence 0 → I n M/I n+1 M → M/I n+1 M → M/I n M → 0 is exact by the third isomorphism theorem.2. n) is polynomial-like of degree at most r − 1. h(grI (M ).

3 5. n) − lR (M /Mn ) = r(n) is polynomial-like of degree less than deg lR (M /Mn ) ≤ deg sI (M. which implies that lR (M /I n+m M ) ≥ lR (M /Mn+m ) ≥ lR (M /I n M ). If there is no upper bound on the length of such a chain. Proof. ♣ 5. The following sequence is exact: 0 → M /(M ∩ I n M ) → M/I n M → M /I n M → 0. the ﬁltration {Mn } is I-stable. n + m) and sI (M . where k is a ﬁeld. .7). . sI (M .8 Corollary Let M be a submodule of M . The left and right hand terms of this inequality are sI (M .6 CHAPTER 5. n) − lR (M /Mn ). will be taken as its Krull dimension. n). so IMn = Mn+1 for suﬃciently large n. n).2. sI (M. We have . Thus for every n ≥ 0 we have Mn+m = M ∩ I n+m M ⊇ I n+m M . noting that we can ignore r(n) because it is of lower degree than sI (M.3. Moreover.2. and it follows that sI (M . An example of an inﬁnitedimensional ring is the non-Noetherian ring k[X1 .1. as well as nonnegative for n >> 0. ]. DIMENSION THEORY where r(n) is polynomial-like of degree less than d(M ). and consequently I n+m M ⊆ Mn+m = I n Mm ⊆ I n M . n) and lR (M /Mn ) have the same degree and the same leading coeﬃcient. Then d(M ) ≤ d(M ). Then by additivity of length. Proof. The result now follows upon adding the equations sI (M.7). ♣ 5. n) respectively. n) − sI (M . Apply (5. .1 The Dimension Theorem Deﬁnitions and Comments The dimension of a ring R. n) = lR (M /Mn ) and r(n) = sI (M . we take n = ∞. By the Artin-Rees lemma (4. n) − sI (M . and the leading coeﬃcient of r(n) is nonnegative. denoted by dim R. Set Mn = M ∩ I n M . where M is a ﬁnitely generated module over the Noetherian local ring R. the maximum length n of a chain P0 ⊂ P1 ⊂ · · · ⊂ Pn of prime ideals of R. X2 . say. n ≥ m. n) = lR (M /Mn ) hence lR (M /Mn ) is polynomial-like.

such that for some elements a1 . Then by (1. . we deﬁne dim M = dim(R/ ann M ) if M = 0. and by (1. . because the ring of algebraic integers of a number ﬁeld is a Dedekind domain. By the above equivalent conditions. At the other extreme. and more generally an Artinian ring. 5. Formally. we deﬁne the height of I as the inﬁmum of the heights of prime ideals P ⊇ I. .5. M has at least one associated prime. If M = 0 we take δ(M ) = −1. and we take the dimension of the zero module to be −1. Every associated prime ideal of M is maximal.4).4. the minimal elements of AP(M ) and Supp M are the same. .3). has dimension 0 by (1. By (1. the height of P is the dimension of the localized ring RP . provided that the prime ideals in the chain contribute to M in the sense that they belong to the support of M . integrally closed integral domain in which every nonzero prime ideal is maximal. will be measured by length of chains of prime ideals. .5).5).6. called the Chevalley dimension δ(M ). By (1. . denoted by dim M .3.2 The Dimension of a Module Intuitively.9). There are several other ideas that arise from the study of chains of prime ideals.5. and the coheight of I as the supremum of the coheights of prime ideals P ⊇ I. (If I and J are ideals of R with I ⊆ J. We make the additional assumption that R is a local ring with maximal ideal M. Algebraic number theory provides many examples. a ﬁeld. X2 . THE DIMENSION THEOREM 7 the inﬁnite chain of prime ideals (X1 ) ⊂ (X1 . the following conditions are equivalent.5. 2. the dimension of an R-module M . ar belonging to M we have lR (M/(a1 . Since M is ﬁnitely generated. 4.6). . .) If I is an arbitrary ideal of R. 3. Every prime ideal in Supp M is maximal. A Dedekind domain is a Noetherian. We now assume that M is nonzero and ﬁnitely generated over the Noetherian ring R. lR (M/MM ) is ﬁnite. and S = (R/I)/(J/I). Thus dim M = sup{coht P : P ∈ Supp M } = sup{coht P : P ∈ AP(M )}. so S is an integral domain iﬀ R/J is an integral domain.5. The coheight of the prime ideal P (notation coht P ) is the maximum length n of a chain of prime ideals P = P0 ⊂ P1 ⊂ · · · ⊂ Pn .6. The length of M as an R-module is ﬁnite. P ⊇ ann M iﬀ P ∈ Supp M . we can assert that there is a smallest positive integer r. X2 ) ⊂ (X1 . By (1.3. By (0. and J/I is a = prime ideal of R/I iﬀ J is a prime ideal of R. A Dedekind domain that is not a ﬁeld has dimension 1. ar )M < ∞.3. Supp(M/MM ) = V (ann(M/MM )) which coincides with {M} by Problem 2. It follows from the correspondence theorem and the third isomorphism theorem that the coheight of P is the dimension of the quotient ring R/P . X3 ) ⊂ · · · . dim M = 0. .9). We deﬁne the height of a prime ideal P (notation ht P ) as the maximum length n of a chain of prime ideals P0 ⊂ P1 ⊂ · · · ⊂ Pn = P . 1. then S ∼ R/J.

[Since Ra + P ⊆ Q. so by (5. Since the chain Q ⊂ P2 ⊂ · · · ⊂ Qr is of length t−1.2) there is an injective homomorphism from R/P to M . We divide the proof into three parts. subsection S7. By NAK. . If t = 0.2. M has only ﬁnitely many associated primes. and assume that the result holds up to t − 1. and let a1 . DIMENSION THEORY 5. d(M ) ≤ δ(M ). Thus d(R/(Ra+P )) < d(R/P ). M = 0 so dim M = −1. then M = 0 and d(M ) = −1. n).) By Problem 9 of Chapter 1.5) is V (P )∩V (I) = V (Q) = Supp R/Q. we have (R/(Ra+P ))Q = 0. and consider prime ideals Q such that Ra + P ⊆ Q ⊆ P1 . 2.2) and (5. set Q = I + P .8) we have d(R/P ) ≤ d(M ). To prove this. Proof. where the map from R/P to itself is multiplication by a.3. The dimension dim M of the module M . ar be elements of M such that M/(a1 . the induction hypothesis implies that t−1 ≤ d(R/Q). . Let I be the ideal (a1 . n) = lR (M/Mn M ) = 0 for n >> 0. hence dim M is ﬁnite.) By (5. and we specify that the degree of the zero polynomial is -1.3. . We can pick a Q belonging to AP(R/(Ra + P )). . It follows from (5. If δ(M ) = −1. Now the sequence 0 → R/P → R/P → R/(Ra + P ) → 0 is exact. ar ) and let P be the annihilator of M . The Chevalley dimension δ(M ). it will follow that dim M = dim R/P ≤ d(R/P ) ≤ d(M ).3. hence d(R/P ) = −1 and we are ﬁnished. 2.8 CHAPTER 5. which by (1.3. (TBGY. see Problem 3.5. ﬁrst note that M/IM ∼ = M ⊗R R/I. (For convenience we take I = M. the maximal ideal of R. Therefore t ≤ d(R/P ). Then choose Q to be a minimal element in the support of R/(Ra+P ). ar )M has ﬁnite length.5. The degree d(M ) of the Hilbert-Samuel polynomial sI (M.2. . .2) there is an injective homomorphism from R/Q to R/(Ra+P ). Assume M = 0 and r = δ(M ) ≥ 0. By (1.8) we have d(R/Q) ≤ d(R/(Ra+P )). It suﬃces to show that for any chain of prime ideals P = P0 ⊂ · · · ⊂ Pt in R. sM (R/P.5. (The image of the map is Ra + P . 3.1 of the supplement.).2. The following quantities are equal: 1. the length t of the chain is at most d(R/P ).] By (1. then sM (M. 1.3. hence t − 1 ≤ d(R/(Ra + P )). . dim M ≤ d(M ). as desired.10). Thus assume t ≥ 1. . Choose a ∈ P1 \ P . and apply (1. n) = sM (R/P. (Note that the annihilator of R/I is I and the annihilator of R/Q is Q. By (1.3. If d(M ) = −1. n) + sM (R/(Ra + P ). . then R/P = 0 (because P is prime).) But the support . Thus assume d(M ) ≥ 0.9) or (1. . where I is any ideal of deﬁnition of R. We claim that the support of R/Q is {M}. If we can show that dim R/P ≤ d(R/P ). so by (5.3 Dimension Theorem Let M be a ﬁnitely generated module over the Noetherian local ring R. n) + r(n) where r(n) is polynomial-like of degree less than d(R/P ).1) that for some associated prime P we have dim M = coht P = dim R/P .9). .7). Supp M/IM = Supp M ∩Supp R/I. . and consequently t − 1 < d(R/P ).

9). ♣ 5.10). If dim M = −1 then M = 0 and δ(M ) = −1. where ai = ai + P . Pt be the associated primes of M whose coheight is as large as it can be.) Let P1 .4. In more detail. ar )M also has ﬁnite length. Thus δ(M ) ≤ r + 1. (If M/IM = 0. . The last assertion follows from the deﬁnition of Chevalley dimension in (5. n) is at most r. .9). hence sQ (M. so assume M = 0. the degree of the Hilbert-Samuel polynomial sQ (M. . . . / hence division by a is allowed. then M = 0 by NAK.6. By (5. of the number of generators of I. n) = sQ (M. Q is an ideal of deﬁnition of R. By (5. δ(N ) ≤ dim N . . Pi ⊂ M for every i. ar . because dim M = coht Pi and we are removing all the Pi . . then by (5. .6). so by (1.3. Q is M-primary. so δ(M ) = 0.2.2. in particular. (We have already noted in part 1 that dim M is ﬁnite.3. To see this. and consider M as an R-module. and let N = M/aM . coht Pi = dim M for all i = 1. Choose an element a in M such a belongs to none of the Pi . But by n the correspondence theorem. . . . . . and let a1 . contradicting our assumption. Then Supp N ⊆ Supp M \ {P1 . then MP = 0. Since the dimension of M is greater than 0.3. Let R = R/P . δ(M ) ≤ dim M . lR (M/Q M ) = lR (M/Qn M ).3).2.11).5)] I is an ideal of deﬁnition. . .3.2). ar )N has ﬁnite length. . . . Pt }. .5) and (3. If dim M = 0. proving the claim. Then R is a Noetherian local ring and Q is an ideal of deﬁnition of R generated by a1 . . . 3. Thus dim N < dim M . n). Proof.3. . CONSEQUENCES OF THE DIMENSION THEOREM 9 of M/IM is {M} by (1. a1 . . Now assume dim M > 0.4. that is. . ar )N = (apply the ﬁrst isomorphism theorem) .2).4 Consequences of the Dimension Theorem In this section we will see many applications of the dimension theorem (5.) Again by (1. .3. so M/(a.1). note that if NP = 0. dim R < ∞. also.5. R/I has ﬁnite length iﬀ (by the Noetherian hypothesis) R/I is Artinian iﬀ [by (5. Moreover. 5. M has ﬁnite length. the dimension of R is the minimum. . But M/(a. If M is a ﬁnitely generated R-module. .3). . . Let r = δ(N ). ar be elements of M such that N/(a1 . M ⊆ ∪1≤i≤t Pi . . AP(R/Q) = {M}. . then dim M < ∞. .1 Proposition Let R be a Noetherian local ring with maximal ideal M. over all ideals I of deﬁnition of R.1. δ(M ) ≤ r + 1 = δ(N ) + 1 ≤ dim N + 1 ≤ dim M. so by the prime avoidance lemma (0. . Q = Q/P . a1 . . . By the induction hypothesis. Finiteness of dimension follows from (5.6. . ♣ . NPi = 0 for all i because a ∈ Pi . In summary. It follows that d(M ) ≤ r. ar )M ∼ N/(a1 . t.

5).) By (5. of prime ideals of R will stabilize if and only if the chain AP0 ⊃ AP1 ⊃ AP2 ⊃ · · · of prime ideals of A stabilizes. (In other words.4.4. Then the chain P0 ⊃ P1 ⊃ P2 ⊃ .4 Theorem If R is a Noetherian ring. Proof. and R its M-adic completion.9). . The elements ai must belong to P . then IRP is an ideal of deﬁnition of RP√that is generated by n elements. and the result follows from (5. In particular. which is a contradiction because J must be proper.4. ar be elements of M such that a1 . n) = sM (R. If (b) holds. the height of M is at most r. n). . see (5. .2. the result follows. .5).10) as in the ﬁrst part of the proof. an /s with s ∈ R \ P .3. then the prime ideals of R satisfy the descending chain condition. ♣ 5. ar form a k-basis of M/M2 . . .3.3. else the ai /s will generate RP . an . (See (3. .4. Conversely. Take I to be the ideal of R generated by a1 . . . ˆ Proof. if P0 is a prime ideal of R. we conclude that I satisﬁes (b). we have dim R ≤ r by (5.4. the = ˆ ˆ ˆ ˆ two Hilbert-Samuel polynomials must have the same degree. .6 Krull’s Principal Ideal Theorem Let a be a nonzero element of the Noetherian ring R. . We may assume without loss of generality that R is a local ring.1) and (5. the result is immediate because dim R < ∞.2.2. If a is neither a unit nor a zerodivisor. ˆ Then dim R = dim R. ht P = dim RP ≤ n. By (5.2. Since M itself is an ideal of deﬁnition (see (5. DIMENSION THEORY 5.6). .3).4.10 CHAPTER 5.) Proof. by (5.4.3. . . (b) There is an ideal I of R that is generated by n elements.) ♣ 5. such that P is a minimal prime ideal over I. sM (R.10). . Let a1 .5)).5 Generalization of Krull’s Principal Ideal Theorem Let P be a prime ideal of the Noetherian ring R. .1). P is minimal subject to P ⊇ I. R/Mn ∼ R/Mn . let A be the localized ring RP0 . so by (5.2 Proposition Let R be a Noetherian local ring with maximal ideal M and residue ﬁeld k = R/M.1) there is an ideal of deﬁnition J of RP generated by n elements a1 /s. Invoking (3. Then dim R ≤ dimk (M/M2 ). . Proof. and note that if P is minimal over I iﬀ I = P . Explicitly. ♣ 5.3 Proposition ˆ Let R be Noetherian local ring with maximal ideal M. (Alternatively. ♣ 5. ar generate M.4). then every minimal prime ideal P over (a) has height 1. The following conditions are equivalent: (a) ht P ≤ n. By (4. a1 . . . . Since ht M = dim R. But if R is local as well as Noetherian. .3. .1). where ai = ai + M. Then by (0. if (a) holds then dim RP ≤ n. . Therefore d(R) = d(R).4.4.

. To prove equality. take M = R and N = R/(a) to conclude that dim R/(a) < dim R. because otherwise P has height 1 and we are ﬁnished. Then dim R = n.4.4. .6). . a1 . In the proof of part 3 of the dimension theorem (5.4. then M is the only prime ideal. the dimension is at least n because of the chain (0) ⊂ (X1 ) ⊂ (X1 .8 Corollary Let a be a non zero-divisor belonging to the prime ideal P of the Noetherian ring R.4. Then ht P/(a) = ht P − 1. In (5. we appeal to part 2 of the proof of (5. .9). a contradiction.10). . . We claim that RP = 0.3. ♣ 5. dim R ≤ 1 + s = 1 + dim R/(a). such that the images ai in R/(a) generate an M/(a)-primary ideal of R/(a). The unique maximal ideal is (X1 .3. On the other hand. X2 ) ⊂ · · · ⊂ (X1 .7). Then a. If a is a unit. but nothing is gained by dropping it. so by (1. so by (5.5. for if dim R = 0.4. P consists entirely of zero-divisors.3). 5. By (1. This allows us to ﬁnd elements a1 . and as in the proof of (5. hence dim R/(a) ≤ dim R − 1.4. . so the dimension of R is at most n.1) and (3. where Q is a minimal prime ideal over (a).7 Theorem Let R be a Noetherian local ring with maximal ideal M. Xn ). . a contradiction. hence P ∈ Supp R.4.9 Theorem Let R = k[[X1 . . CONSEQUENCES OF THE DIMENSION THEOREM 11 Proof.3. Proof. For if a/1 = 0.3. and let a ∈ M be a non zerodivisor. . . . It follows from (5. . . Then dim R/(a) = dim R − 1. with s = dim R/(a). We have dim R > 0. which contradicts the hypothesis that a is not a zero-divisor. ♣ . Proof. P is an associated prime of R. . then for some s ∈ R \ P we have sa = 0. M consists entirely of zero-divisors. .6). Xn ) of prime ideals. as generate an M-primary ideal of R. replace R by RP and R/(a) by (RP )Q .3). Thus assume ht P = dim RP = 0. then a cannot belong to the prime ideal P and the theorem is vacuously true.5) that ht P ≤ 1. .5. Proof. as ∈ M. ♣ 5. ♣ The hypothesis that a is not a unit is never used. Xn ]] be a formal power series ring in n variables over the ﬁeld k. . We may assume that P is minimal in the support of R. . .4.

. x2 + Y1r2 . xn ]. take ri = si . and therefore A is integral over B. . . . . . . 1. .12 CHAPTER 5. . Assume r = 1. . i > h(0). α = (a1 . . . . Ym ]. To study them we will need a stronger version of Noether’s normalization lemma. . . . The proof is by induction on r. We claim that there are positive integers ri (i = 2. . cα Y1a1 (x2 + Y1r2 )a2 · · · (xm + Y1rm )am − x1 = 0. (This is the standard normalization lemma. . . Ym ].5. . m. . A is integral over B = k[x1 . hence are integral over B) all Yi are integral over B. Ym ). xn ∈ A . xm )Y1j = 0 . . then (since the xi belong to B. and we have cβ Y1 f (β) + j<f (β) pj (x1 . . . r. xm + Y1rm ) − x1 = 0.2 Theorem Let A be a ﬁnitely generated k-algebra. . i = 2. . . . . . . we have x1 = g(Y1 . There exists a nonnegative integer n and elements x1 . am ). By our assumption that A is a polynomial ring. . For example. . . . xn ∈ A algebraically independent over k such that the following conditions are satisﬁed. and I1 ⊂ · · · ⊂ Ir a chain of nonzero proper ideals of A. For we may write A = A /I0 where A = k[Y1 . say α = β. let f (α) = a1 + a2 r2 + · · · + am rm . following Serre’s Local Algebra. . and we ﬁnd elements x1 . . Proof. . and pick the ri so that all the f (α) are distinct. . Now Y1 satisﬁes the equation x1 = g(Y1 .5. xh(i) . .5 5. . . . DIMENSION THEORY 5. If we write the polynomial g as a sum of monomials the above equation becomes cα Y α . . . . . 5. there is a positive integer h(i) such that Ii ∩ B is generated (as an ideal of B) by x1 . . Then there will be a unique α that maximizes f . . We ﬁrst consider the case in which I1 is a principal ideal (x1 ) = x1 A with x1 ∈ k.) 2. . If Ii is the preimage of Ii under the canonical map A → A /I0 . It suﬃces to let A be a polynomial ring k[Y1 . . cα = 0. . where s is greater than the maximum of the aj .1 Strengthening of Noether’s Normalization Lemma Deﬁnition An aﬃne k-algebra is an integral domain that is also a ﬁnite-dimensional algebra over a ﬁeld k. relative to the ideals I0 ⊂ I1 ⊂ · · · ⊂ Ir . . If we can show that Y1 is integral over B. In this section we will give the statement and proof. To produce the desired ri . . . . where xi = Yi − Y1ri . so g(Y1 . page 42. xm ]. . Aﬃne algebras are of great interest in algebraic geometry because they are the coordinate rings of aﬃne algebraic varieties. . . For each i = 1. . Ym ) / for some nonconstant polynomial g with coeﬃcients in k. satisfy the required conditions. m) such that A is integral over B = k[x1 . then the images of xi−h(0) in A.

. and let P be a prime ideal of R with I ⊆ P . . PROPERTIES OF AFFINE K-ALGEBRAS 13 so Y1 is integral over B. . . Then t = x1 u with u ∈ A. . Proof.5. . hence. . Finally. . xm ) = B = k[x1 . . hence Q = 0. . . Thus x1 A ∩ B = x1 B. Since B is isomorphic to a polynomial ring. .6. . xm ]. then Q is a minimal prime ideal over J. . Let S be the polynomial ring R[X]. Since A has transcendence degree m over k and an integral extension must be algebraic. and we have already taken care of m = 1 (because A is then a PID). u ∈ A ∩ k(x1 . an = = integral domain. xm satisfying the conditions of the theorem for k[t2 . . . tm satisfy the conditions of the theorem for the chain of ideals I1 ⊂ · · · ⊂ Ir−1 . (A . we have u ∈ B. and let s = h(r − 1). . tm are algebraically independent over k. . .2 Corollary Let I be an ideal of the Noetherian ring R. . If m = 0 there is nothing to prove. . there are elements x2 . Take xi = ti for 1 ≤ i ≤ s. . t2 . Let x1 be a nonzero element of I1 . there are elements xs+1 . where I1 is principal). . xm ]. and take J = IS and Q = P S. A = k[Y1 . tm ]. tm ∈ A such that x1 . xm are algebraically independent over k. . there are elements t2 . and the proof of the ﬁrst case is complete. t2 . we now consider the general case by induction on m. . the result follows. . A is integral over the polynomial ring C = k[x1 . . . . we may assume that R is a ﬁeld. . let t1 . . where Q and Q are prime ideals of S both lying above the same prime ideal P of R. Note that we have also shown that A ∩ k(x1 . Since A is integral over B. Suppose that the prime ideal Q1 of S is properly contained in Q. note that R[X]/P R[X] ∼ R[X]/P [X] ∼ (R/P )[X].6. .6 Properties of Aﬃne k-algebras We will look at height. . . . If we can show that I1 ∩ B = (x1 ) = x1 B. By the argument of the previous paragraph. so let t belong to I1 ∩ B. tm ]. The right-to-left inclusion follows from our assumptions about x1 . we take the inductive step from r − 1 to r. tm ] and the ideal I1 ∩ k[t2 . By modding out I. . . the second assertion will also hold. Thus the ﬁrst assertion of the theorem holds (in this ﬁrst case. dividing by x1 . it is a unique factorization domain and therefore integrally closed.1 Proposition Let S = R[X] where R is an arbitrary ring. ♣ 5. . tm ] satisfying the conditions for the ideal Ir ∩ k[ts+1 . . . 5. By what we have just proved. . . . we may assume that I = 0. and as we noted above. Then x1 . . xm satisfy the desired conditions. . coheight and dimension of aﬃne algebras. . . If Q ⊂ Q . . and x1 A ∩ C = x1 C. By localizing with respect to the multiplicative set R \ {0}. tm ]. By the induction hypothesis.6. Since P S is also 0. . . Ym ] is integral over B = k[x1 . Since R/P can be regarded as a subring of S/Q. . . then Q = P S. . . . To verify that Q is prime. it follows that x1 . . . xm ). Still assuming r = 1. . . If P is a minimal prime ideal over I. . and note that x1 ∈ k because I1 is / proper. . . xm ∈ k[ts+1 . we may assume without loss of generality that P = 0. Proof. ♣ 5. . But then every nonzero prime ideal of S is maximal. Then Q1 ∩ R ⊆ Q ∩ R = P S ∩ R = P . . . . . .

then dim R[X1 .14 CHAPTER 5. Proof. This follows from (5. and by (5.4) by induction. R Noetherian. The general result now follows by induction. its (geometric) dimension is the transcendence degree over k of the function ﬁeld (the fraction ﬁeld of the coordinate ring). . We now show that the geometric dimension coincides with the algebraic (Krull) dimension. ht Qn = ht Pn . Xn ].6. 5. DIMENSION THEORY polynomial belonging to R coincides with its constant term.6. Xn ] = n. . In particular. . Q = P S is a minimal prime ideal over J = IS. Xn ) is at most n. Proof. . Let n be the height of P . . . By the generalized Krull principal ideal theorem (5.3 Proposition As above.4. ♣ We may now prove a major result on the dimension of a polynomial ring. . . First consider i = n. . where R is a Noetherian ring. . ht Pj = ht Qj ≥ j. using (5. Since the height of Pj is at least j. But by choice of j. Let j be the largest index i such that Pi = Pi+1 .8). . let S = R[X]. . Then ht P = ht Q. Conversely. ♣ 5. By (5. .1). in view of the chain (X1 ) ⊂ (X1 . ♣ 5. ♣ If X is an aﬃne algebraic variety over the ﬁeld k. . We abbreviate transcendence degree by tr deg. The height of (X1 . We may assume that the Pi are not all distinct (otherwise dim R ≥ dim S ≥ dim S − 1). / Now consider a chain Q0 ⊂ Q1 ⊂ · · · ⊂ Qn of prime ideals of S. then Q0 ⊂ Q1 ⊂ · · · ⊂ Qn = Q. But the Q sequence can be extended via Qn ⊂ Qn+1 = Qn + (X). and let Pi = Qi ∩ R for every i = 0. Then ht(X1 . . . . if K is a ﬁeld then dim K[X1 .6 Corollary Let R = K[X1 .6. .3). 1.6. Xn ). (Note that X cannot belong to Qn because 1 ∈ Pn . hence dim S ≤ 1 + dim R. where K is a ﬁeld. so again by (5. Q = P S. . By (5. By (5.5 Corollary If R is a Noetherian ring. But J is also generated over S by n elements.6.4.) By minimality. the dimension of R.6. . if P0 ⊂ P1 ⊂ · · · ⊂ Pn = P ⊂ R and Qi = Pi [X]. Proof.5).2). then by (5.6. . a contradiction. Then dim S = 1 + dim R. Proof. . Xi ) = i. . If Qn = Pn S.6. Pj = Pj+1 ⊂ Pj+2 ⊂ · · · ⊂ Pn so ht Pj + n − j − 1 ≤ dim R.6. . . ♣ 5. . Let P0 ⊂ P1 ⊂ · · · ⊂ Pn be a chain of prime ideals of R. n. .6.5). Q1 = P S = Q. we have n − 1 ≤ dim R. 1 ≤ i ≤ n. .4.) It follows that dim S ≥ 1 + dim R. so ht Q ≥ ht P . . Q1 ∩ R = P . ht Q ≤ ht P . X2 ) ⊂ · · · ⊂ (X1 .1). and in fact the height is n. Xn ] = n + dim R.3). P a prime ideal of R. Qj = Pj S.4 Theorem Let S = R[X]. there is an ideal I of R generated by n elements such that P is a minimal prime ideal over I.

5.6. PROPERTIES OF AFFINE K-ALGEBRAS

15

5.6.7

Theorem

If R is an aﬃne k-algebra, then dim R = tr degk Frac R. Proof. By Noether’s normalization lemma, there are elements x1 , . . . , xn ∈ R, algebraically independent over k, such that R is integral over k[x1 , . . . , xn ]. Since an integral extension cannot increase dimension (see Problem 4), dim R = dim k[x1 , . . . , xn ] = n by (5.6.5). Let F = Frac R and L = Frac k[x1 , . . . , xn ]. Then F is an algebraic extension of L, and since an algebraic extension cannot increase transcendence degree, we therefore have tr degk F = tr degk L = n = dim R. ♣ It follows from the deﬁnitions that if P is a prime ideal of R, then ht P + coht P ≤ dim R. If R is an aﬃne k-algebra, there is equality.

5.6.8

Theorem

If P is a prime ideal of the aﬃne k-algebra R, then ht P + coht P = dim R. Proof. By Noether’s normalization lemma, R is integral over a polynomial algebra. We can assume that R = k[X1 , . . . , Xn ] with ht P = h. (See Problems 4,5 and 6. An integral extension preserves dimension and coheight, and does not increase height. So if height plus coheight equals dimension in the larger ring, the same must be true in the smaller ring.) By the strong form (5.5.2) of Noether’s normalization lemma, along with (5.6.6), there are elements y1 , . . . , yn algebraically independent over k such that R is integral over k[y1 , . . . , yn ] and Q = P ∩ k[y1 , . . . , yn ] = (y1 , . . . , yh ) Since k[y1 , . . . , yn ]/Q ∼ = k[yh+1 , . . . , yn ], it follows from (5.3.1) and (5.6.5) that coht Q = dim k[yh+1 , . . . , yn ] = n − h. But coht Q = coht P (Problem 5), so ht P + coht P = h + (n − h) = n = dim R. ♣

Chapter 6

Depth

6.1 Systems of Parameters

We prepare for the study of regular local rings, which play an important role in algebraic geometry.

6.1.1

Deﬁnition

Let R be a Noetherian local ring with maximal ideal M, and let M be a ﬁnitely generated R-module of dimension n. A system of parameters for M is a set {a1 , . . . , an } of elements of M such that M/(a1 , . . . , an )M has ﬁnite length. The ﬁniteness of the Chevalley dimension (see (5.3.2) and (5.3.3) guarantees the existence of such a system.

6.1.2

Example

Let R be a Noetherian local ring of dimension d. Then any set {a1 , . . . , ad } that generates an ideal of deﬁnition is a system of parameters for R, by (5.4.1). In particular, if R = k[[X1 , . . . , Xn ]] is a formal power series ring over a ﬁeld, then X1 , . . . , Xn form a system of parameters, since they generate the maximal ideal.

6.1.3

Proposition

Let M be ﬁnitely generated and of dimension n over the Noetherian local ring R, and let a1 , . . . , at be arbitrary elements of the maximal ideal M. Then dim(M/(a1 , . . . , at )M ) ≥ n − t, with equality if and only if the ai can be extended to a system of parameters for M. Proof. Let a be any element of M, and let N = M/aM . Choose b1 , . . . , br ∈ M such that N/(b1 , . . . , br )N has ﬁnite length, with r as small as possible. Then M/(a, b1 , . . . , br )M also has ﬁnite length, because (M/aM )/(b1 , . . . , br )(M/aM ) ∼ M/(a, b1 , . . . , br )M. = 1

2

CHAPTER 6. DEPTH

It follows that the Chevalley dimension of M is at most r + 1, in other words, δ(M/aM ) ≥ δ(M ) − 1. The proof will be by induction on t, and we have just taken care of t = 1 as well as the key step in the induction, namely dim(M/(a1 , . . . , at )M ) = dim(N/a1 N ) where N = M/(a2 , . . . , at )M . By the t = 1 case and the induction hypothesis, dim(N/a1 N ) ≥ dim N − 1 ≥ dim M − (t − 1) − 1 = dim M − t as asserted. If dim(M/(a1 , . . . , at )M ) = n − t with n = dim M , then we can choose a system of parameters at+1 , . . . , an for N = M/(a1 , . . . , at )M . Then N/(at+1 , . . . , an )N ∼ M/(a1 , . . . , at , at+1 , . . . , an )M = has ﬁnite length. Thus a1 , . . . , an form a system of parameters for M . Conversely, if a1 , . . . , at can be extended to a system of parameters a1 , . . . , an for M , deﬁne N = M/(a1 , . . . , at )M . Then N/(at+1 , . . . , an )N ∼ M/(a1 , . . . , an )M has ﬁnite length, hence = dim N ≤ n − t. But dim N ≥ n − t by the main assertion, and the proof is complete. ♣

6.2

Regular Sequences

We introduce sequences that are guaranteed to be extendable to a system of parameters.

6.2.1

Deﬁnition

Let M be an R-module. The sequence a1 , . . . , at of nonzero elements of R is an M sequence, also called a regular sequence for M or an M -regular sequence, if (a1 , . . . , at )M = M and for each i = 1, . . . , t, ai is not a zero-divisor of M/(a1 , . . . , ai−1 )M .

6.2.2

Comments and Examples

We interpret the case i = 1 as saying that a1 is not a zero-divisor of M , that is, if x ∈ M and a1 x = 0, then x = 0. Since (a1 , . . . , at )M = M , M = 0 and the ai are nonunits. It follows from the deﬁnition that the elements a1 , . . . , at form an M -sequence if and only if for all i = 1, . . . , t, a1 , . . . , ai is an M -sequence and ai+1 , . . . , at is an M/(a1 , . . . , ai )M -sequence. 1. If R = k[X1 , . . . , Xn ] with k a ﬁeld, then X1 , . . . , Xn is an R-sequence. 2. (A tricky point) A permutation of a regular sequence need not be regular. For example, let R = k[X, Y, Z], where k is a ﬁeld. Then X, Y (1 − X), Z(1 − X) is an R-sequence, but Y (1 − X), Z(1 − X), X is not, because the image of Z(1 − X)Y is zero in R/(Y (1 − X)).

5 Deﬁnition Let M be a nonzero ﬁnitely generated module over the Noetherian local ring R. . .2)) and therefore belong to the maximal ideal of R. and let a1 . .3).1. . where x belongs .2. .2.4 Corollary If R is a Noetherian local ring.3). the result follows from (6. dim M/(a1 . at−1 )M . Proof. we have. Since at is not a zero-divisor of N = M/(a1 . .3). the induction hypothesis says that a1 . . as in the proof of (6. The depth of M over R. = hence dim M/(a1 . where n = dim M . .2. a1 is part of a system of parameters for M . First let us show that a1 is not a zero-divisor of M/a2 M . .6. (Remember that the ai are nonunits (see (6.) By (6. Proof. It suﬃces to consider the transposition that interchanges a1 and a2 . . . . . .3 Theorem Let M be a ﬁnitely generated module over the Noetherian local ring R. . at )M. at−1 )M = n − (t − 1).1. . . We argue by induction on t.3). .2.1. By (6. . 6. . . Proof. But. Suppose a1 x = 0.7 Proposition Let M be a ﬁnitely generated module over the Noetherian ring R. at extend to a system of parameters for M . . is the maximum length of an M -sequence.2. .1. at } can be extended to a system of parameters for M .2. Since dim M is the number of elements in a system of parameters. .4.7). If a1 . Proof. at−1 is part of a system of parameters for M . as in the t = 1 case. . we have dim M/a1 M = dim M − 1 by (5.2. Then depth M ≤ dim M . at is an M -sequence. . . Since a1 is not a zero-divisor of M .3). written depthR M or simply depth M . a1 . Then any permutation of the ai is also an M -sequence. By (6. Take M = R in (6. . an be an M -sequence with all ai belonging to the Jacobson radical J(R).6 Theorem Let M be a nonzero ﬁnitely generated module over the Noetherian local ring R.3). We will see in the next chapter that any two maximal M -sequences have the same length. . If t > 1. then every R-sequence can be extended to a system of parameters for R. . dim N/at N = dim N − 1. at )M = dim N/at N = dim N − 1 = n − (t − 1) − 1 = n − t.2. . . .2. ♣ 6. N/at N ∼ M/(a1 . . then {a1 . . REGULAR SEQUENCES 3 6. . ♣ 6. ♣ 6.

4

CHAPTER 6. DEPTH

to M/a2 M . Then a1 x belongs to a2 M , so we may write a1 x = a2 y with y ∈ M . By hypothesis, a2 is not a zero-divisor of M/a1 M , so y belongs to a1 M . Therefore y = a1 z for some z ∈ M . Then a1 x = a2 y = a2 a1 z. By hypothesis, a1 is not a zero-divisor of M , so x = a2 z, and consequently x = 0. To complete the proof, we must show that a2 is not a zero-divisor of M . If N is the submodule of M annihilated by a2 , we will show that N = a1 N . Since a1 ∈ J(R), we can invoke NAK (0.3.3) to conclude that N = 0, as desired. It suﬃces to show that N ⊆ a1 N , so let x ∈ N . By deﬁnition of N we have a2 x = 0. Since a2 is not a zero-divisor of M/a1 M , x must belong to a1 M , say x = a1 y with y ∈ M . Thus a2 x = a2 a1 y = 0. But a1 is not a zero-divisor of M , hence a2 y = 0 and therefore y ∈ N . But x = a1 y, so x ∈ a1 N , and we are ﬁnished. ♣

6.2.8

Corollary

Let M be a ﬁnitely generated module over the Noetherian local ring R. Then any permutation of an M -sequence is also an M -sequence. Proof. By (6.2.2), the members of the sequence are nonunits, hence they belong to the maximal ideal, which coincides with the Jacobson radical. ♣

6.2.9

Deﬁnitions and Comments

Let M be a nonzero ﬁnitely generated module over a Noetherian local ring R. If the depth of M coincides with its dimension, we call M a Cohen-Macaulay module. We say that R is a Cohen-Macaulay ring if it is a Cohen-Macaulay module over itself. To study these rings and modules, we need some results from homological algebra. The required tools will be developed in Chapter 7.

Chapter 7

Homological Methods

We now begin to apply homological algebra to commutative ring theory. We assume as background some exposure to derived functors and basic properties of Ext and Tor. In addition, we will use standard properties of projective and injective modules. Everything we need is covered in TBGY, Chapter 10 and the supplement.

7.1

Homological Dimension: Projective and Global

Our goal is to construct a theory of dimension of a module M based on possible lengths of projective and injective resolutions of M .

7.1.1

Deﬁnitions and Comments

A projective resolution 0 → Xn → · · · → X0 → M → 0 of the R-module M is said to be of length n. The largest such n is called the projective dimension of M , denoted by pdR M . (If M has no ﬁnite projective resolution, we set pdR M = ∞.)

7.1.2

Lemma

The projective dimension of M is 0 if and only if M is projective. Proof. If M is projective, then 0 → X0 = M → M → 0 is a projective resolution, where the map from M to M is the identity. Conversely, if 0 → X0 → M → 0 is a projective resolution, then M ∼ X0 , hence M is projective. ♣ =

7.1.3

Lemma

If R is a PID, then for every R-module M , pdR M ≤ 1. If M is an abelian group whose torsion subgroup is nontrivial, then pdR M = 1. Proof. There is an exact sequence 0 → X1 → X0 → M → 0 with X0 free and X1 , a submodule of a free module over a PID, also free. Thus pdR M ≤ 1. If pdR M = 0, then by (7.1.2), M is projective, hence free because R is a PID. Since a free module has zero torsion, the second assertion follows. ♣ 1

2

CHAPTER 7. HOMOLOGICAL METHODS

7.1.4

Deﬁnition

The global dimension of a ring R, denoted by gldim R, is the least upper bound of pdR M as M ranges over all R-modules.

7.1.5

Remarks

If R is a ﬁeld, then every R-module is free, so gldim R = 0. By (7.1.3), a PID has global dimension at most 1. Since an abelian group with nonzero torsion has projective dimension 1, gldim Z = 1. We will need the following result from homological algebra; for a proof, see TBGY, subsection S5.7.

7.1.6

Proposition

If M is an R-module, the following conditions are equivalent. (i) M is projective; (ii) Extn (M, N ) = 0 for all n ≥ 1 and all R-modules N ; R (iii) Ext1 (M, N ) = 0 for all R-modules N . R We can now characterize projective dimension in terms of the Ext functor.

7.1.7

Theorem

If M is an R-module and n is a positive integer, the following conditions are equivalent. 1. pdR M ≤ n. 2. Exti (M, N ) = 0 for all i > n and every R-module N . R 3. Extn+1 (M, N ) = 0 for every R-module N . R 4. If 0 → Kn−1 → Xn−1 → · · · → X0 → M → 0 is an exact sequence with all Xi projective, then Kn−1 is projective. Proof. To show that (1) implies (2), observe that by hypothesis, there is a projective resolution 0 → Xn → · · · → X0 → M → 0. Use this resolution to compute Ext, and conclude that (2) holds. Since (3) is a special case of (2), we have (2) implies (3). If (4) holds, construct a projective resolution of M in the usual way, but pause at Xn−1 and terminate the sequence with 0 → Kn−1 → Xn−1 . By hypothesis, Kn−1 is projective, and this gives (4) implies (1). The main eﬀort goes into proving that (3) implies (4). We break the exact sequence given in (4) into short exact sequences. The procedure is a bit diﬀerent from the decomposition of (5.2.3). Here we are proceeding from right to left, and our ﬁrst short exact sequence is GM G0 where K0 is the kernel of . The induced long exact sequence is 0

i0

G K0

G X0

· · · → Extn (X0 , N ) → Extn (K0 , N ) → Extn+1 (M, N ) → Extn+1 (X0 , N ) → · · · R R R R

7).1).1. N ). the following conditions are equivalent. we will need the analog of (7. M ) → Extn+1 (M. R (iii) Ext1 (M. Now the second short exact sequence is G K1 i1 G X1 d1 G K0 G 0.1. R R R By (7. E) → Extn (M. N ) = 0 for all M and N . 0 We can replace X0 by K0 because im d1 = ker = K0 . M ). N ) → Extn+1 (X1 . INJECTIVE DIMENSION 3 Now if every third term in an exact sequence is 0. the outer terms are 0 for n ≥ 1. This technique is referred to as dimension shifting. Ext1 (Kn−1 . The associated long exact sequence is · · · → Extn (X1 . projective dimension has a dual notion.2.1. The given short exact sequence induces the following long exact sequence: · · · → Extn (M.2.8 Corollary gldim R ≤ n if and only if Extn+1 (M. N ) = 0 for all R-modules M .1. E) → · · · R R R R By (7. N ) → · · · R R R R ∼ and dimension shifting gives Extn (K0 . N ) ∼ Extn (K0 .2 Injective Dimension As you might expect. hence by the hypothesis of (3). we R R ∼ get Extn+1 (M.1. the map in the middle is an isomorphism. then the maps in the middle are both injective and surjective. M ) → Extn+1 (M. Kn−1 is projective.2 Proposition Let 0 → M → E → M → 0 be an exact sequence.6) applies. N ) → Extn (K1 . and the technique of dimension shifting is again useful.2. hence as in the proof of (7.1. 7. ♣ 7. N ) = 0 for all R-modules M and N . hence isomorphisms.6) for injective modules. Iterating this procedure. N ) = Extn−1 (K1 . To develop it. so as = R R we slide from right to left through the exact sequence. A proof is given in TBGY. ♣ . (ii) Extn (M. (i) N is injective. M ) ∼ Extn (M.6).8. Then for all n ≥ 1 and all R-modules M . N ). with E injective. By the deﬁnition (7. N ) → Extn+1 (K0 .1 Proposition If N is an R-module.2. ♣ R 7. Thus Extn+1 (M. This is precisely what we have here. subsection S5. Thus as we slide = R R through the exact sequence from left to right.1. the index of Ext drops by 1.1. N ). 7. R We are going to dualize (7.7). gldim R ≤ n iﬀ pdR M for all M iﬀ (by (1) implies (3) of (7.7. R Proof. N ) = Ext1 (Kn−1 . N ) = 0. N ) = 0 for all n ≥ 1 and all R-modules M . because X0 is projective and (7. we have Extn+1 (M.4) of global dimension. Proof.7)) Extn+1 (M. the upper index decreases by 1.

4 CHAPTER 7. this happens R iﬀ idR N ≤ n for all N . Cn−1 ) = R R for any M . Extn+1 (M. By (7.2) to each short exact sequence.1). R 3. To prove that (3) implies (4). and it ends at 1. If 0 → N → X0 → · · · → Xn−1 → Cn−1 → 0 is an exact sequence with all Xi injective. N ) = 0 for every R-module M .4). Use this sequence to compute Ext. gldim R ≤ n iﬀ pdR M ≤ n for all M .3).2. Cn−1 is injective. 0 → N → X0 → K0 → 0. Proof. .2.7). If (4) holds. More precisely.4) of global dimension. (If N has no ﬁnite injective resolution.1. If (1) is satisﬁed. ♣ . Exti (M. N ) = 0 for all i > n and every R-module M . N ) ∼ Ext1 (M. By the deﬁnition (7. ♣ 7. proving that (4) implies (1). Extn+1 (M.2. we have an exact sequence 0 → N → X0 → · · · → Xn → 0. If the index of Ext starts at n + 1. we set idR M = ∞. 1. The left side is 0 by hypothesis. The process is similar to that of (5. If we have (2). N ) = 0 for all M and N .) Just as in (7. . construct an injective resolution of N .2. idR N ≤ n.2. HOMOLOGICAL METHODS 7.3 Deﬁnitions and Comments An injective resolution 0 → N → X0 → · · · → · · · Xn → 0 of the R-module N is said to be of length n. 2. 7.1. but with emphasis on kernels rather than cokernels. denoted by idR M . we decompose the exact sequence of (4) into short exact sequences. R 4. 0 → Kn−2 → Xn−1 → Cn−1 → 0 We now apply the dimension shifting result (7. it drops by 1 as we go through each of the n sequences. The largest such n is called the injective dimension of M .2. idR N = 0 if and only if N is injective.5 Corollary The global dimension of R is the least upper bound of idR N over all R-modules N . then we have the special case (3). the following conditions are equivalent. 0 → K0 → X1 → K1 → 0. Cn−1 is injective.4 Proposition If N is an R-module and n is a positive integer. Proof. so the right side is also 0.2. By hypothesis. . by (7. but pause at step n − 1 and terminate the sequence by Xn−1 → Cn−1 → 0.1. with the Xi injective.2). . and conclude that (2) holds. The decomposition is given below. Equivalently. Extn+1 (M. then Cn−1 is injective. By (7.

N ) = 0 for all R-modules N . Therefore f is an isomorphism of F and M . Let M be a ﬁnitely generated R-module.1).3 Theorem Let R be a Noetherian local ring with maximal ideal M and residue ﬁeld k.1 Proposition If M is an R-module. . . . xn } be a minimal set of generators for M . Thus K ⊗R k = ker f = 0. i . en } and deﬁne an R-module homomorphism f : F → M via f (ei ) = xi . By NAK. N ) = 0 for all n ≥ 1 and all R-modules N . See Problems 3-6 for all the details. Now f is surjective by construction.1) K ⊗R k = K ⊗R (R/M) ∼ K/MK = so K = MK.6. . TOR AND DIMENSION 5 7. 1 In addition.3. and is injective by minimality of the generating set [see (0. Take a free module F with basis {e1 . K = 0. and this suggests that it would be proﬁtable to bring in the Tor functor as well. the following conditions are equivalent. . let {x1 . (i) pdR M ≤ n. . .3. n. ♣ 7. Since TorR (M. To prove the “if” part. 7.3. n (iii) TorR (M. .4) and the base change device below].2 Proposition Let R be Noetherian local ring with maximal ideal M and residue ﬁeld k. we have the short exact sequence 0 → K → F → M → 0. subsection S5. the following conditions are equivalent. If M is a ﬁnitely generated R-module. k) → K ⊗R k → F ⊗R k → M ⊗R k → 0 1 where the map f : F ⊗R k → M ⊗R k is induced by f . (i) M is ﬂat. Then M is free (⇐⇒ projective ⇐⇒ ﬂat) if and only if TorR (M. The “only if” part follows from (7. . hence M is free.3. (ii) TorR (M. i = 1. which is proved in TBGY. . if and only if M is ﬂat. 7.3 Tor and Dimension We have observed the interaction between homological dimension and the Ext functor.3. . 1 Proof. (ii) TorR (M. subsection S7.7. If K is the kernel of f . k) = 0. if R is a Noetherian local ring and M is ﬁnitely generated over R. We will need the following result. k) = 0. we can truncate the associated long exact sequence: 1 0 = TorR (M. then M is free if and only if M is projective. N ) = 0 for all i > n and every R-module N .3. But (TBGY. .

it suﬃces to show that Kn−1 is projective.3). all maximal M sequences have the same length. . Assuming (3). . Now we apply dimension shifting as in the proof of (7. . Then Extn (N. . and (2) follows from (7.3. n+1 (3) TorR (k. If an annihilates N . then pdR M ≤ n for all M .3. HOMOLOGICAL METHODS (iii) TorR (M. k) = 0 by hypothesis. . an be an M -sequence.3) gives TorR (k. For example. N ) = TorR (N. i = 1. n+1 Proof.4. (1) gldim R ≤ n.5).3). k) = 0. and let a1 .2). n annihilates N .6 CHAPTER 7.4 Corollary Let R be a Noetherian local ring with maximal ideal M and residue ﬁeld k. N ) = 0 for all ﬁnitely generated R-modules M and N .3. k) → TorR (K1 . Again by (7. = R . If (i) holds. the short exact sequence 0 → K1 → X1 → K0 → 0 [see(7.1. which implies that the map in the middle is an isomorphism. .4. the outer terms are 0.1. and if we use this resolution to compute Tor. it remains to prove that (3) implies (1). ♣ 7. If x is any element of N .1 Lemma Let M and N be R-modules. k) → TorR (K0 . k) = 0. . k) → TorR (X1 . . (2) TorR (M. . k) → · · · n n n−1 n−1 and as before. Let 0 → Kn−1 → Xn−1 → · · · → X0 → M → 0 be an exact sequence with all Xi projective. By (7.1) so that each ai .4. .7). k) ∼ TorR (M.7). For any positive integer n. we will prove that under a mild hypothesis. . an−1 )M is the zero map. . . ♣ 7. . Since (3) is a special case of (2).3. Since an is not a zero-divisor of M .2. hence gldim R ≤ n. the projective n+1 n+1 dimension of any R-module N is at most n. we have TorR (Kn−1 . If (1) holds. 7.7)] induces the long exact sequence · · · → TorR (X1 . There is no diﬃculty with (ii) =⇒ (iii) =⇒ (iv). N ) = 0 for every R-module N . Proof. n+1 Proof. . then M has a projective resolution of length n. (7. then an h(x) = h(an x) = h(0) = 0. an )M ). k) = 0 for all R-modules N . the following conditions are equivalent. M ) ∼ homR (N. . n+1 (iv) TorR (M. the result follows.1.4 Application As promised in (6. then the only R-homomorphism h from N to M = M/(a1 . so it remains to prove (iv) =⇒ (i). By = 1 n+1 (7.2 Proposition Strengthen the hypothesis of (7. ♣ 7. Kn−1 is projective. M/(a1 . Iterating.3. we get (ii).

.1).4. coincides with the grade of M on M . for example if R is Noetherian [see (1. R Proof. If R is a Noetherian local ring with maximal ideal M. . an )M . M ) ∼ Extn (N. . by (1. This is precisely the statement of the theorem. By induction hypothesis. Thus the depth of M . r + I to rx. . M ) R R R R where the label a1 indicates multiplication by a1 . Then homR (R/I. . M /(a2 . (If there are only ﬁnitely many associated primes. and I an ideal of R. I is contained in some associated prime of M0 . hence an isomorphism.3. M ) = 0. an is an M -sequence in I. then by (6. it will map 1+I to a nonzero element x ∈ M0 .4. Equivalently. Again using the induction hypothesis. We must check that the map is well deﬁned. M ) = 0.2. M ) a1 G Extn (N. but this follows because I annihilates x. . . In fact this map is zero.3. In (7. then by (0. . . another equivalent condition is that I is contained in the union of the associated primes of M0 . hence belong to M. M ) ∼ homR (R/I.2). But a + I = 0 + I since a ∈ I. we have = R R Extn−1 (N. induces the following long exact sequence: G Extn−1 (N. If a ∈ I. M/(a1 . . Extn (R/I. . By (7.4. ♣ 7.2). an )M = homR (N. . an−1 )M = 0 by (7.4 Theorem Let M be a ﬁnitely generated module over the Noetherian ring R. . M ) is isomorR phic to homR (N. so ax must be 0. .) Thus δ is injective. APPLICATION 7 Proof. the elements ai of an M -sequence are nonunits. with the map from M to M given by multiplication by a1 .4).4. and I an ideal of R such that IM = M . this says that if a1 . In view of (1.4. ♣ = R We prove a technical lemma to prepare for the main theorem. and in general. .) Proof.5 Remarks Under the hypothesis of (7. namely the smallest nonnegative integer n such that Extn (R/I. (The result is vacuously true for n = 1. ♣ R 7. Conversely.7. Extn−1 (N. . it can be extended to some an+1 ∈ I as long as Extn (R/I.3 Lemma Let M0 be an R-module. we have Extn−1 (N. Then Extn (R/I. an } be a set of generators of I. 7. The short exact sequence 0 → M → M → M/a1 M → 0.9)].4. . because a1 annihilates N .5). an )M ).2. then a + I is mapped to ax. . M ). M ) = 0 if and only if I is not R contained in the union of all associated primes of M0 . M/a1 M ) δ G Extn (N. hence δ is surjective. then we can construct a nonzero homomorphism by mapping 1 + I to x. .4.1.6).3).1). Consequently. . if x is a nonzero element of M0 annihilated by I. If there is a nonzero homomorphism from R/I to M0 . M0 ) = R where M0 = M/(a1 .3. M0 ) = 0 if and only if there is a nonzero element of M0 annihilated by I. as deﬁned in (6.4. we call the maximum length of an M -sequence in I the grade of I on M . M ) Extn−1 (N. . Then any two maximal M -sequences in I have the same length. . M ) ∼ homR (N. if M = M/a1 M . take N = R/I and let {a1 . M/(a1 . .1). . .

ar )M has ﬁnite length. if a1 . . . then the module M/(a1 .4).1) and (6. HOMOLOGICAL METHODS Again let M be ﬁnitely generated over the Noetherian local ring R. M ) = 0.3) and the maximality of M. .4. By (7.1. .8 CHAPTER 7. this happens iﬀ M is an associated prime of M . . . Note also that by (6. the depth of M is 0 if and only if homR (R/M. ar is an M -sequence of maximal length. .2.4. .3). By (7. .

hence dim R = 1 and R is regular. 3.3. Note that (3. condition (3).1. If R has dimension 0. k) be a Noetherian local ring.9). then by the dimension theorem [see (5. then by (3.2 Examples 1.) It appears that {X − 1. V ) is a regular local ring. . but this is not the case (see Problem 1). . Y − 1} is a minimal generating set for M. .) 8. we say that R is regular and that S is a regular system of parameters.4.11) assumes that R is an integral domain. Xd ]]. so (X 3 − Y 2 ) is a prime ideal. in other words. V ) of rational functions on V that are deﬁned at P .4.Chapter 8 Regular Local Rings In algebraic geometry. hence R is regular and {X1 . (The notation means that the maximal ideal is M and the residue ﬁeld is k = R/M. then R is regular iﬀ {0} is a maximal ideal. 2.1 Basic Deﬁnitions and Examples Deﬁnitions and Comments Let (R. 4.3. . In fact M is principal. (See Example 2 above. every generating set of M has at least d elements. If R has dimension 1. and note that R is a domain because X 3 − Y 2 is irreducible. but this is not a problem because we will prove shortly that every regular local ring is a domain. . Xd } is a regular system of parameters.11). and let R = K[X. M. . By (5. where K is a ﬁeld.1) to verify that S is indeed a system of parameters. localized at the maximal ideal M = {X − 1. Then P will be a nonsingular point of V if and only if O(P.) 1 .) If d is the dimension of R. .1. If M does in fact have a generating set S of d elements. the local ring of an aﬃne algebraic variety V at a point P is the set O(P. iﬀ R is a ﬁeld.1 8. dim R = d.1)]. (The overbars indicate calculations mod (X 3 − Y 2 ). Y ]/(X 3 − Y 2 ).1. (Check the deﬁnition (6. . 8. Let K be a ﬁeld whose characteristic is not 2 or 3. Let R = K[[X1 . Y − 1}. R is regular iﬀ R is a discrete valuation ring.

d. . .1. (2) a1 . ad be elements of M whose images in M form a regular system of parameters for R. . where 1 ≤ t ≤ d. Let a be the image of a in Mm /Mm+1 and let b be the image of b in Mn /Mn+1 . . . . . .1. i = t + 1. . and let at+1 .2 CHAPTER 8. .5 Proposition Let (R. ad + M2 } is a basis for M/M2 . ∩∞ Mn = 0. Now it takes two elements to generate M. . The following conditions are equivalent.4 Theorem A regular local ring is an integral domain. .2). ♣ 8. .1. and it follows that ab cannot be 0. i = 1.4) with M = R and I = M. hence a b = 0 (because the associated graded ring is a domain).3. . with the Madic ﬁltration [see (4. at+1 . Now assume (3). . . . at ). . . at .3 Proposition Let (R. . Then R is regular if and only if the dimension of R coincides with dimk M/M2 . . ad generate M. the image of ab in Mm+n+1 . at ∈ M. . . at ) is d − t.) Proof.1.1. By (6. . . . By the Krull intersection theorem. . .4) with J = M = M. at are linearly independent over k. Xd ]. . Conversely. Proof. Then a and b are nonzero. But the opposite inequality always holds (even if R is not regular). ad is a regular system of parameters for R. then the ai + M2 span M/M2 . Here is a convenient way to express regularity. . The proof of (8. . If R is regular and a1 . Y ) and drop the restriction on the characteristic of K. . The isomorphism identiﬁes ai with Xi . is isomorphic to the polynomial ring k[X1 . To prove that (1) implies (3). . . . .2)]. (Apply (0. . . . Let R be as in Example 4. Speciﬁcally.3) shows that the associated graded ring of R. M. . .3).) Thus R is regular. k) be a regular local ring of dimension d. the ai extend to a regular system of parameters iﬀ the ai extend to a k-basis of M/M2 . . . the dimension of R = R/(a1 . . . . .) Now let n=0 a and b be nonzero elements of R. . If x ∈ M. and let a1 . . k) be a Noetherian local ring. .3) shows that (1) and (2) are equivalent. but dim R = 1 (Problem 2). .3. condition (6).1. (3) R/(a1 . generate M = M/(a1 . then the ai generate M. . . But the d − t elements ai .4. at can be extended to a regular system of parameters for R. we have . . then modulo I = (a1 . . Let d be the dimension of R. . (1) a1 .1.3. But a b = ab. assume that a1 . at ) is a regular local ring of dimension d − t.11). . so dimk M/M2 ≤ d. hence R is regular. (See (3. M. 8. . . . d. and choose m and n such that a ∈ Mm \ Mm+1 and b ∈ Mn \ Mn+1 . where ai = ai mod M2 . The proof of (8. Thus R is not regular. REGULAR LOCAL RINGS 5. and is therefore a domain. 8. ♣ We now examine when a sequence can be extended to a regular system of parameters. the dimension of M/M2 as a vector space over k. by (5. at ). (Apply (4. for a prior appearance of this vector space. except that we localize at M = (X. if {a1 + M2 . Proof.

. . at+1 is not a zero-divisor of R. at extend to a regular system of parameters for R. R = R/(a1 . then by (8. By (8. ad . . . with (necessarily) d = dim R. It follows that a1 . at . By repeated applicaion of (5. The length of any such R-sequence is the dimension of R.1. . so R is regular. k) be a Noetherian local ring.1. x − t+1 ci ai ∈ I. . (To start the induction. at ) to be the zero ideal.4). . . ♣ . it follows that depthR = dim R. Proof. d ≤ depth R. ♣ 8. which has dimension 0. By deﬁnition of depth [see(6. . M. so these elements form a regular system of parameters for R. .6). depth R ≤ dim R. ♣ 8. ad .4. . . . ad . .1. The maximal ideal M of R can be generated by at+1 . . ad .1. . . If 1 ≤ t ≤ d.6 Theorem Let (R. at+1 is not a zero-divisor of R/(a1 . at+1 . . . .1.) Now assume that M is generated by the R-sequence a1 . . . . ad is an R-sequence. set t = 0 and take (a1 . the maximal ideal M of the regular local ring R can be generated by an R-sequence a1 . Proof. It follows that dim R = d.6). . . In other words. in other words. a1 . . . at ) is regular and has dimension d − t. BASIC DEFINITIONS AND EXAMPLES d d 3 x − t+1 ci ai = 0 for some ci ∈ R.1. . But R/M is the residue ﬁeld k. .2.7).5). . Then R is regular if and only if M can be generated by an R-sequence.5)]. . we have dim R/M = dim R − d. Thus R is regular (which we already know by hypothesis) and a1 . By (8. . . . . By induction. . . with a regular system of parameters a1 . at ). . . But by (6. . Since dim R = d. .7 Corollary A regular local ring is Cohen-Macaulay. .2.8. . . Assume that R is regular. ad generate M.

. . . . 3-5 discrete valuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-4 MS localization of M by S . . . . . . 5-4 sI (M. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1 k (r) analog of xr in the calculus of ﬁnite diﬀerences . . . . . . 3-3 absolute value . . . . . . . . . . . . . . . . . . . . . . .2-3 localized ring . . 1-8 N (R) nilradical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1 ﬁltration of a module . . . . . . . . . . . . . . . . . . . 3-5 {Rn } ﬁltration of a ring . . . . 1-8 V (I) set of prime ideals containing I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-4 Tor Tor functor . . . . . . . . . . . . .5-7 δ(M ) Chevalley dimension of the module M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .0-2 λa multiplication by a . . . . . . . . . . . . . . . . . 7-5 I-depth maximum length of an M -sequence in I . . . . . . . . . . . n) Hilbert polynomial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1 n >> 0 for suﬃciently large n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2 lim← Mn inverse limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3 z(M ) zero-divisors of M . . . . . . . . . . . 1-1 rM (N ) AP(M ) associated primes of M . . . . . . . . . . . . . 4-5 ∆G diﬀerence of G . . . . . . . . . . . . . 5-7 coht coheight. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6 Supp M support of M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-3 h(M. . . . 7-2 Ext Ext functor . . . . . . . . . . . . . . . . . . . . . . . 7-7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-14 pdR M projective dimension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-2 l length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2 gr(M ) associated graded module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-5 d(M ) degree of the Hilbert-Samuel polynomial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 radical of annihilator of M/N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1 {Mn } gr(R) associated graded ring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2 idR N injective dimension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-5 dim dimension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . n) Hilbert-Samuel polynomial . . . . . . . . . . . . . . . . . . . . . . . . 2-8 valuation ring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1 gldim R global dimension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-7 tr deg transcendence degree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-5 radical of an ideal I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-6 ht height . . 1-12 lR (M ) length of the R-module M . . . . . . . . . . . . . . . . .1 List of Symbols J(R) Jacobson radical. . . . . . . . . . . . . 1-13 Rc RT √ I V |x| v integral closure of R in a larger ring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4 ˆ M completion of a module .

8-1 . . . . . . . . . . . .2 (R. k) local ring with maximal ideal M and residue ﬁeld k . . . . . . M. . .

homological. 4-7 extension theorems. 4-5 contraction. 3-5 discrete valuation ring (DVR). see (5.1. 7-4.4) Dedekind domain. 4-1. 1-3 calculus of ﬁnite diﬀerences. 1-6 equivalent ﬁltrations. 4-1 graded piece. 2-7. 3-5 additivity of length. 7-5 formal power series. 3-5 archimedean. 7-2 going down. 5-9ﬀ. dimension shifting. 3-5 nonarchimedean.1 Index absolute value. 5-1 Chevalley dimension. . 1-6 ﬂat module. 4-1 ﬁltration. 3-3 faithful module. 7-1ﬀ. (7.3). 4-1 ﬁrst uniqueness theorem. 4-3 Artinian ring. 8-3 coherent sequences and homomorphisms. 2-1 dimension. 4-4 completion of a module. 5-6ﬀ. 6-3 determinant trick. 4-2 associated primes. 2-9 decomposing a long exact sequence into short exact sequences. 5-7 degree. graded ring. 2-6 grade. 4-1. 3-1. 5-12 Artin-Rees lemma. 2-1 ﬁltered ring and module. 7-7 graded module. 5-7 Cohen-Macaulay rings and modules. 5-3 aﬃne k-algebra. 4-1 depth.2. 5-8 consequences of. 2-9 going up. 1-11 associated graded rings and modules. discrete valuation. 7-3. 6-4. 7-6 dimension theorem. 3-5. 3-6 embedded component. 0-3. 5-7 coheight. 5-1. 5-11 global dimension.

2-3 inverse limit.2 Hausdorﬀ topology. 4-4 irreducible submodule. 5-10 Laurent series. 5-7 Hensel’s lemma. 4-3 ideal of deﬁnition. 0-4 lying over. 2-5 M -sequence. 0-2. 1-2 uniqueness of. 2-3 integral elements. 7-1 regular local ring. 3-2 length of a module. 8-1 . 4-5 height. 1-12 Krull dimension. 5-2 primary decomposition. 4-8 Krull’s principal ideal theorem. 5-6 Krull intersection theorem. 4-1 homological dimension. 7-1ﬀ. 4-3 I-stable ﬁltration. 5-11 nilradical. 5-4 Hilbert-Samuel polynomial. 1-1 prime avoidance lemma. 4-4 I-ﬁltration. 4-4 inverse system. 0-2. 1-8 primary ideal. 4-1 I-adic topology. 1-12 polynomial-like function. 1-6 Nakayama’s lemma (NAK). 0-3 Noether’s normalization lemma. 2-4. 5-4 injective dimension. 2-1 integrally closed. 1-3 reduced. 7-4 integral closure. 5-5 homogeneous elements. I-adic completion. 4-10 Hilbert polynomial. 6-2 minimal component. 3-6 P -primary submodule. 1-1 prime element. 1-13 local ring. extensions. 1-1 primary submodule. 1-6 Jacobson radical. 5-10 generalization. 1-2 isolated component. 4-8 I-adic ﬁltration. 1-6. 1-2 existence of. 0-1 projective dimension.

5-14 transitivity of integral extensions. 6-1 transcendence degree. 1-8 zero-divisors. 1-4 . 6-2 regular system of parameters. 2-3 uniformizer. 3-6 valuation ring. 2-2 support of a module. 1-8 system of parameters.3 regular sequence. 3-2 Zariski topology. 1-8 stabilizing a module. 0-4 second uniqueness theorem. 8-1 residue ﬁeld. 2-2 UFD (unique factorization domain).

where the Ri are Artinian local rings. If 2 P = (x. r. 4. 1 13. . 8. Continuing Problem 2. √ 11. but it is not primary. z). Let S be a multiplicative subset of R. Let R = k[x. z]/I where I = (xy − z 2 ). Let M be ﬁnitely generated over the Noetherian ring R. Show that in (1. y = y + I. and MS = S −1 M . Ii is Mi -primary. show that Q is not a power of a prime ideal. In Problems 10-13.7) for Artinian rings. Let x = x + I. Then −1 show that Mi = πi (Pi ) is a maximal ideal of R. and let P1 = (x). 10. Continuing Problem 6.1. y 2 ) is P -primary.6. I = 2 (x2 . . i = 1. Let R = k[x. 5.7). so by (1. What are the primary ideals of Z? 2. and let πi be the projection of R on Ri . y] where k is a ﬁeld.6. P2 = (x. . which is nilpotent. Q = (x2 . we consider uniqueness in the structure theorem (1. and identify P. . xy). z = z + I. 7. 1 r . Show that ∩r Ii is a reduced primary decomposition of the zero ideal.2). Let M and N be ﬁnitely generated modules over a local ring R. 12. Show that Q = (x. Use base change formulas in the tensor product to show that (M ⊗R N )S ∼ MS ⊗RS NS as RS -modules. and let P be a prime ideal in the support of M . y). the Ri are unique up to isomorphism. Show that M ⊗R N = 0 if and only if either M or N is 0. y. = 9. show that the result fails to hold if R is not local. y). Let R = 1 Ri . 3. Show that I i = Mi . Show that lRP (MP ) < ∞ if and only if P is a minimal element of AP(M ). Let Ii = ker πi . show that Supp(M ⊗R N ) = Supp M ∩ Supp N . y] where k is a ﬁeld.Exercises Chapter 1 1. 6. Show that each Ri has a unique prime ideal Pi . If M and N are ﬁnitely generated R-modules. Let R = k[x. Show that I = P1 ∩ P2 and I = P1 ∩ Q are both primary decompositions of I. 14. show that P is a power of a prime ideal and its radical is prime.

4. (a) The given sequence is exact. Continuing Problem 5. Let R = Z and S = Z[i]. Continuing Problem 6. (b) RP is integrally closed for every prime ideal P . and Y + I as y. and write (R.2. P prime. We may regard all localized rings RP as subsets of K.2 Chapter 2 1. / 3. coincides with the intersection of all RP . Y ]/I. ideal P . Show that f is injective [resp. 9.1). P maximal. Show that the ﬁelds RP /P RP and Frac R/P are isomorphic. Each is referred to as the residue ﬁeld at P . with maximal ideals MR and MS respectively. and in fact both intersections coincide with R. 5. Thus R is not integrally closed. Chapter 3 Let R and S be local subrings of the ﬁeld K. surjective] if and only if fP is injective [resp. Show that the following conditions are equivalent. MS ). (b) The sequence M P − − → MP − − → MP −− −− is exact for every prime ideal P . Let P be a prime ideal of R. (By (2. 8. Show that α = x/y is integral over R. Suppose we have a diagram of R-modules M −−→ M −−→ M −− −− with im f ⊆ ker g. where k is a ﬁeld and I is the prime ideal (X 2 − Y 3 ). We say that S dominates R. but α ∈ R. Let R = k[X. 7. (c) The localized sequence of (b) is exact for every maximal ideal P . if R is a subring of S and R ∩ MS = MR . show that MQ ⊆ RQ . Give an example of two prime ideals of S lying above the same prime ideal of R. show that S −1 M ⊆ P ∈max R fP gP f g S −1 RP 6. .) 2. If S is any multiplicative subset of R. the Gaussian integers. equivalently for every maximal. Write the coset X + I simply as x. where max R is the set of maximal ideals of R. if Q is any maximal ideal of R. surjective] for every prime. (c) RQ is integrally closed for every maximal ideal Q. there cannot be an inclusion relation between the prime ideals of S. Let f : M → N be an R-module homomorphism. Let R be an integral domain with fraction ﬁeld K. show that the intersection of all RP . MR ) ≤ (S. If R is an integral domain. Let M be the intersection of all RP for maximal ideals P . show that the following conditions are equivalent: (a) R is integrally closed.

We write grn (f ) instead of f n . MV ) is maximal with respect to the partial ordering induced by domination. Show that in (4. then f is n=0 injective. In other words. . The R-homomorphism f : M → N is said to be a homomorphism of ﬁltered modules if f (Mn ) ⊆ Nn for all n ≥ 0. . Let M and N be ﬁltered modules over the ﬁltered ring R. MR1 ) dominates (V. . h1 ) is an extension of (V. Then ker h1 is the unique maximal ideal MR1 . Let k be the residue ﬁeld V /MV . f induces a homomorphism f n : Mn /Mn+1 → Nn /Nn+1 via f n (x + Mn+1 ) = f (x) + Nn+1 . . Xn ). For each n. 6. 3. . so that maps go from Mn to Mn+1 . By (3. Put the I-adic ﬁltration on R. The grn (f ) extend to a homomorphism of graded gr(R)-modules. The direct limit of such a system is the disjoint union Mn . MV ). 5. If we reverse the arrows in the deﬁnition of an inverse system [see (4. where k is a ﬁeld. . 3. Show that (R1 . with sequences x and y identiﬁed if they agree suﬃciently far out in the ordering. Let R be the formal power series ring k[[X1 . we assume that gr(f ) is injective. Thus if Mn fn G Mn gn G Mn is exact for all n. Xn ].6) we proved that the inverse limit functor is left exact.4).1. 2. In (4. Conversely.2.1. . it suﬃces to show that (V. If V is a valuation ring of K. For the remainder of this problem and in Problems 3 and 4. Complete the proof by showing that (V. We deﬁne a homomorphism h : V → C. show that if in addition we have ∩∞ Mn = 0. we may assume R1 local and h1 (R1 ) a subﬁeld of C.2.2. 4. . Chapter 4 1. Xn ]]. Show that every local subring of a ﬁeld K is dominated by at least one valuation ring of K. .1). and let C be an algebraic closure of k. Show that the direct limit functor is always exact. Continuing Problem 2.1)]. h). We write gr(f ) = n≥0 grn (f ). we will show in Problems 2 and 3 that if (V.10). Show that the associated graded ring of R is the polynomial ring k[X1 . θn (xn ) = θn (yn ) for all suﬃciently large n. the two ﬁltrations {I n N } and {N ∩ I n M } are equivalent. 4.3 1. . show that f −1 (Nn ) ⊆ Mn for all n ≥ 0. if (R1 . h) is a maximal extension. we get a direct system. and M = lim Mn −→ . Show that Mn ∩ f −1 (Nn+1 ) ⊆ Mn+1 for all n ≥ 0. and exact under an additional assumption. . call it gr(f ) : gr(M ) → gr(N ). . 2. show that (V. h) is a maximal extension. by following the canonical map from V to k by the inclusion map of k into C. Continuing Problem 3. where I is the unique maximal ideal (X1 . MV ) is maximal. then V is a valuation ring. As in (3.

ˆ ˆ 10. Show that V (ann(M/MM )) = {M}. 7. ht J ≤ ht I. 6. Show that I n /I n+1 ∼ (I)n /(I)n+1 . Show that R is a ﬂat R-module. the ring S is integral over the subring R. and let M [resp. show that the topology induced on M by {Mn } must be Hausdorﬀ. Let M be nonzero and ﬁnitely generated over the local ring R with maximal ideal M. 8. ˆ ˆ show that R is a local ring with maximal ideal M. Note that M is an R-module via {an } {xn } = {an xn }. Show that I is contained in the Jacobson radical J(R). Deﬁne an R-module ˆ ˆ homomorphism hM : R ⊗R M → M by (r. In diﬀerential calculus. Show that the completion of M is always complete. then the sequence M gn . Establish the following.) f GM g GM is exact. If I is an arbitrary ideal and P a prime ideal of R. In Problems 12-14. show that hM is surjective. that is. coheight and dimension. show that (R/I)P = 0 iﬀ P ⊇ I. then by deﬁnition of height. What is the analogous statement in the calculus of ﬁnite diﬀerences? 2. If M is complete with respect to the ﬁltration {Mn }. Show that (I)n ∼ (I n ). and I = J ∩ R. In Problems 4-7. coht I = coht J. show that hM is an isomorphism. J is an ideal of S. Let M be an R-module. 4. ˆ ˆ R]. the exponential function ex is its own derivative. we have ht P + coht P ≤ dim R. = ˆ ˆ ˆ 15. ˆ In Problems 12-16. = ˆ ˆ = 13. m) → r m. Let R be a local ring with maximal ideal M. (The maps f and g are induced by the fn and ˆ ˆ 7. then ht J = ht I. ˆ= ˆ 12. Show that I ∼ R ⊗R I ∼ RI. If P is a prime ideal of R. = ˆ = ˆ ˆ ˆ = ˆ 9. 3.4 is the direct limit of the Mn (similarly for M and M ). If R and S are integral domains with R integrally closed. ˆ ˆ 14. If R is the M-adic completion of R. Thus if R is complete (R ∼ R). M ∼ M . . In Problem 7. Let R be the I-adic completion of the Noetherian ring R. R is assumed Noetherian. dim R = dim S. then M is complete (M ∼ M ). If M is ﬁnitely generated over R. In Problems 8 and 9 we show that the inequality can be strict. R] be the I-adic completion of M [resp. 5. R is the I-adic completion of the ring R. if in addition R is Noetherian. ˆ 16. Chapter 5 1. 11.

and K = ker g. G ··· Let R be a Noetherian local ring with maximal ideal M and residue ﬁeld k = R/M. where I = (XY. at are part of a system of parameters for R. and let R = S/I where I = (XY. and suppose that the elements a1 . . . with exact rows. members of a regular sequence (Section 6. 2 → 0. . Y.5 even if R is Noetherian. XZ). so that ht P + coht P < dim R. Deﬁne f : Z4 → Z4 by 0 → 0. X + Y } is a system of parameters. Let R be a Noetherian local ring with maximal ideal M. Show that G Z4 f G Z4 f G Z4 ··· resolution of M of inﬁnite length. and deﬁne uM : M ⊗R M → M via uM (a ⊗ x) = ax. . Z]] be a formal power series ring over the ﬁeld k. X = X + I ∈ R. but Z is a zerodivisor. f (x) = 2x mod 4. Let Z4 = Z/4Z.e. Let M = 2Z4 ∼ Z2 (also a Z4 -module). let F be a free R-module with basis e1 . . If M is generated by x1 . . for example. i. a ∈ M. en . Show that {Z. . . hence projective. 2. g : F → M . at ) is prime and has height t. 1 ≤ i ≤ n. G M⊗F G0 G M⊗M M⊗K GK GF GM 0 Applying the snake lemma. Deﬁne X = X + I. then M is free. We are going to show in Problems 3. The following diagram is commutative. .e. Let P be the prime ideal (Y . Let M be a ﬁnitely generated R-module. Z)]] be a formal power series ring over the ﬁeld k. Z) of R. . Y = Y + I. 9. Chapter 7 1. 2. then Cn = 0 for all n. and deﬁne g : Z4 → M = by 0 → 0. G Cn+1 ∂n+1 f G Z4 g GM G 0 is a free. Given an exact sequence ··· G An fn G Bn gn G Cn ∂n G An−1 fn−1 G Bn−1 gn−1 G Cn−1 Show that if the maps fn are all isomorphisms. Show that the dimension of R is 2. Show that P has height 0 and coheight 1. 3 → 1. xn . XZ).4 and 5 that if uM is injective. we have an exact sequence ker uM δ uK uF uM G coker uK f∗ G coker uF g∗ G coker uM . i. Y. 1 → 2. Chapter 6 1.2) cannot be zero-divisors.. . . If the ideal P = (a1 . where f : K → F. 2 → 0. show that ht P + coht P = dim R. a free Z4 -module. . 8. . 3 → 2. x ∈ M . Z = Z + I. On the other hand. let S = k[[X. g(x) = x mod 2. . 1 → 1. Deﬁne a homomorphism g : F → M via ei → xi . Use an overbar to denote cosets mod I. .. . and let R = S/I. Let S = k[[X. We have an exact sequence 0 → K → F → M → 0.

and let M be an R-module. . . so that its kernel B is nonzero. We can deﬁne a homomorphism of graded k-algebras ϕ : k[X1 . n) has degree r − 1.1. . It follows that the Hilbert polynomial has degree less than r − 1.2. Xr ]. Show that coker uM ∼ k ⊗R M . . . Show that g is injective.1. Now assume that ϕ is not an isomorphism. Show that h(n) = n+r−1 − lk (Bn ). . k) be a Noetherian local ring.) In Problems 3-5. Then An is isomorphic as a k-vector space to I = (X1 . Show that the global dimension of a ring R is the least upper bound of pdR (R/I). 7. Xr ] → grM (R) via ϕ(Xi ) = ai + M2 . Problem 2 for terminology. Let f : R → S be a ring homomorphism. completing the proof.2). (a) TorR (M. where the ai generate M. . Xr ]. . Compute the Hilbert polynomial of grM (R) and show that it has degree r − 1. Example 4. 3. . 1 (b) TorR (M. show that X − 1 and Y − 1 are associates. if and only if M is ﬂat. 2. the Hilbert-Samuel polynomial sM (R. = 6.2)]. (See Chapter 4. Show that coker uK = 0. it is an isomorphism. 8. Equivalently. . S) = 0 and M ⊗R S is a ﬂat S-module. . M can be replaced by R/I. M. and similarly coker uF ∼ k ⊗R F . = = 4.1). Assume that ϕ is an isomorphism. n) has degree r. hence M ∼ F . and M is free. N ) = 0 for all S-modules N .1. and let An be the set of homogeneous polynomials of degree n in k[X1 . In (8. Xr ). 4. Show that M is free if and only if M is projective. Let (R. Let M be a ﬁnitely generated module over the Noetherian local ring R. . . . 5. I an arbitrary ideal of R.6 3. Then B becomes a graded ring ⊕n≥0 Bn with a grading inherited from the polynomial ring A = k[X1 . Prove that the following conditions are equivalent. We have an exact sequence 0 → Bn → An → Mn /Mn+1 → 0. . . we are going to show that ϕ is an isomorphism if and only if the Hilbert polynomial h(n) = h(grM (R).2). r−1 5. and let grM (R) be the associated graded ring with respect to the M-adic ﬁltration [see(4. . Justify the assertions made in Example 5 of (8. Since g is surjective by deﬁnition. Show that the polynomial-like functions on the right side of the above equation for h(n) have the same degree and the same leading coeﬃcient. where I ranges over all ideals of R. 1 Chapter 8 1. Show that in (7. 9.

k) be a Noetherian local ring of dimension d. Let (R. . . . . Xd ]. . Show that R is regular if and only if the associated graded ring grM (R) is isomorphic as a graded k-algebra to k[X1 . M.7 6.

/ / 1 . 2 4.] If Mk = 0. c ∈ k. Note also that the results are consistent with the ﬁrst uniqueness theorem. y). But x ∈ Q n and x ∈ P0 for n ≥ 2. and k = R/M the residue ﬁeld. Then Mk ⊗k Nk = = (k⊗R M )⊗k (k⊗R N ) = [(k⊗R M )⊗k k]⊗R N = (k⊗R M )⊗R N = k⊗R (M ⊗R N ) = 0. By Problem 6. 2. If Q = P0 with P0 prime. Since y ∈ P0 but y ∈ Q. which is 0 if n and m are relatively prime. this happens iﬀ MP = 0 or NP = 0.1. By Problem 8. P0 = (x. an integral domain. a domain). Let Mk = k ⊗R M = (R/M) ⊗R M ∼ M/MM . Thus Q is primary. y) is a polynomial multiple of xy. x ∈ 2 P . then g(x. P is prime since R/P ∼ k[y]. We have Z/nZ ⊗Z Z/mZ ∼ Z/(n. (M ⊗R N )S ∼ RS ⊗R (M ⊗R N ) ∼ (RS ⊗R M )⊗R N ∼ MS ⊗R N ∼ (MS ⊗RS RS )⊗R N ∼ = = = = = MS ⊗RS (RS ⊗R N ) ∼ MS ⊗RS NS . Since Mk and Nk are ﬁnite-dimensional vector spaces over a ﬁeld. / 2 5. y)y 2 (or f (x. y)y). Thus P2 and Q are P2 -primary.2). R/Q ∼ k[y]/(y 2 ). then Q = P0 . P2 is = √ 2 2 maximal and P2 = Q = P2 . Nk = 0 implies N = 0. = 8. so by NAK. so = they are nilpotent. P ∈ Supp M or P ∈ Supp N .Solutions to Problems Chapter 1 1. m)Z. hence P1 is P1 -primary. so f (x. Since r(Q) = P = (x. note that if f (x. √ n 3. Thus P is a prime power and = / its radical is the prime ideal P . (M ⊗R N )P ∼ MP ⊗RP NP as RP -modules. Now P1 is prime (because R/P1 ∼ k[y]. For the reverse inclusions. and this equals (k s )r = k rs . p prime. y)x = g(x. Assume M ⊗R N = 0. so by Problem 2.1. one of them must be 0. = 9. Let M be the maximal ideal of R.] 2 6. The primary ideals are (0) and (pn ). y). Q is P -primary. y) must involve y. and zero-divisors in this ring are of the form cy + (y 2 ).1) and (1. y ∈ P. y) must involve x and f (x. because x y = z 2 ∈ P . But it is not primary. Similarly. M = 0. We have I ⊆ P1 ∩ P2 and I ⊆ P1 ∩ Q by deﬁnition of the ideals involved. y)x = g(x. we have Q = P0 / n / and we reach a contradiction. Thus P ∈ Supp(M ⊗R N ) / = iﬀ MP ⊗RP NP = 0. that is. so Q = P0 for n ≥ 2. then M = MM . [k r ⊗k k s = (k ⊗k k s )r because tensor product commutes with direct sum. 7. [See (1.

the second assertion follows from (1. Now RP is a local ring with maximal ideal P RP .5. P contains some P ∈ AP(M ). the decomposition is primary. the zero ideal is the intersection of the Ii . and consequently X − Y f (X. and Ii + Ij = R for i = j. Y )Y 3 can produce X. consider the exact sequence 0 − − → im f − − → ker g − − → ker g/ im f − − → 0 −− −− −− −− Applying the localization functor. hence (x/y)2 = y. By hypothesis. 12. apply Problem 3 to the sequence 0 − − → M − − → N. In the injective case. If α ∈ R. then α = x/y = f (x. this is X + I = Y f (X.6. The result follows. the Ii are unique (for a given R). so by (1. 2. y). Since Ii consists of those elements that are 0 in the ith coordinate. y) for some polynomial f in two variables with coeﬃcients in k.8). Then x ∈ Mi iﬀ πi (x) ∈ Pi iﬀ πi (xn ) = 0 iﬀ xn ∈ Ii . and (b) implies (c) is immediate. which is 0 for every prime ideal P . Written out longhand. 4.5. Thus Mi + Mj = R. hence divisible by (2 + i)(2 − i) = 5. by (c). 11. in other words.4. By (1.6). Y ) ∈ I = (X 2 . By the bijection of (1. Since the preimage of a prime ideal under a ring homomorphism is prime. proving (a). Y 3 ). Thus Q1 ∩ Z = (5). the Mi are distinct and hence minimal.2). Thus x = yf (x. (ker g/ im f )P = (ker g)P /(im f )P = ker gP / im fP . Thus α2 − y = 0. so α is integral over R. Since Ri ∼ R/Ii . By (1. Y )X 2 + h(X. we get the exact sequence 0 − − → (im f )P − − → (ker g)P − − → (ker g/ im f )P − − → 0 −− −P − −P − −− for every prime ideal P . But by basic properties of localization. and under the assumption that lRP (MP ) is ﬁnite. P must coincide with P .6. Q2 = (2 − i). lRP (MP ) < ∞ iﬀ there is no Q ∈ AP(M ) such that Q ⊂ P . −− −− f i π i π . and Ii ⊇ ∩j=i Ij .4). An integer divisible by 2 + i must also be divisible by the complex conjugate 2 − i.4.6. ker g = im f .2 10. 14. P ∈ Supp M . so the Mi are distinct and the decomposition is reduced. the Ri are unique = up to isomorphism. By Problem 12. Chapter 2 1. ker g/ im f = 0. To prove that (c) implies (a). we have (a) implies (b). This is impossible because there is no way that a linear combination g(X. We have x2 = y 3 . the length lRP (MP ) will be ﬁnite iﬀ every element of APRP (MP ) is maximal.9). By the second uniqueness theorem (1. Since the localization functor is exact. √ Now Ii ⊆ Ii = Mi . Let Q1 = (2 + i). 3. 13. By Problem 11. and similarly Q2 ∩ Z = (5). Say Pin = 0.1).6. The ﬁrst assertion follows from (1. and the result follows. Y ) + I.5).4) and (1.

MR ). f Chapter 3 1. and let α be a nonzero element of R. each of which is integrally closed (in the same fraction ﬁeld K). Thus R is integrally closed. Therefore V = R1 . then α = αα−1 b = b ∈ V . MV = MR1 ∩ V . as α and its inverse both belong to R. so if we identify terms of degree at least d + 1 with 0. the result follows. ker h = MV . we have R ⊆ M . 2. iQ is surjective. we have the following chain of inclusions. MV ) is maximal with respect to domination. hence α−1 ∈ V ⊆ R. Chapter 4 1.) Thus α−1 ∈ MV = MR ∩ V . and (b) immediately implies (c). This follows because S −1 (∩Ai ) ⊆ ∩i S −1 (Ai ) for arbitrary rings (or modules) Ai . so assume α ∈ V . that is. MR ) is not dominated in this way. Since Q is an arbitrary maximal ideal. Taking S = R \ Q and applying Problem 5. K is the fraction ﬁeld of Ai . note that if for every i. 9. so α−1 is not a unit of R. so (V. 2. By deﬁnition of h. Since R is contained in every RP . 4. where the Ai are domains that are integrally closed in K. we get an isomorphism between I d /I d+1 and the homogeneous polynomials of degree d. −− −− 5. i is surjective by Problem 4. 7. But R ⊆ ∩P prime RP ⊆ M . f ∈ R. then it is a maximal element in the domination ordering. MV ) = (R1 . 2. If α ∈ V we are ﬁnished. a. 8. hence RQ ⊆ MQ for every maximal ideal Q.2. so R = M . then f (x) + Nn+1 = 0. This tells you exactly how to / construct the desired isomorphism. RQ = MQ . Since R1 ⊇ V . and the proof is complete. ker h = (ker h1 ) ∩ V . and the result follows. α−1 is not a unit of V . Then either α or α−1 belongs to V . 6.6). MR1 ). in particular. The elements of the ﬁrst ﬁeld are a/f + P RP and the elements of the second ﬁeld are (a + P )/(f + P ). If x ∈ Mn and f (x) ∈ Nn+1 . By Problem 6. The implication (a) implies (b) follows from (2. 3. f ∈ P . then ∩i Ai is integrally closed.2. This is a contradiction. If (R. (V. where in both cases. Let i : R → M and iQ : RQ → MQ be inclusion maps. By hypothesis. Since h1 extends h.3 and in the surjective case. apply Problem 3 to the sequence M − − → N − − → 0. / Just as in the proof of Property 9 of Section 3. (If b ∈ V and bα−1 = 1. It follows from Problem 7 that R is the intersection of the RQ . where P ranges over all maximal ideals of R: MQ = (∩P RP )Q ⊆ ∩P (RP )Q ⊆ (RQ )Q = RQ . Assume that (V. To prove that (c) implies (a). Take the direct sum over all d ≥ 0 to get the desired result. . hence R itself is a valuation ring. so x ∈ Mn+1 . MV ) ≤ (R. We have f ∈ I d iﬀ all terms of f have degree at least d.

(See TBGY. ∼ ˆ ˆ 7.2. 5. we have g ◦ f = 0. Since R ⊗R R = R and tensor product commutes with direct sum. see TBGY. so by Problem 8. so by Problem 8. hN is surjective. ˆ∼ ˆ 14.4). we have f −1 (0) ⊆ f −1 (∩Nn ) = ∩f −1 (Nn ) ⊆ ∩Mn = 0. Thus the functor ˆ ˆ R ⊗R — is exact.7. Using the additional hypothesis and Problem 3. By hypothesis. the result is true for n + 1. proving that R is ﬂat. By Problem 2.9). with exact rows. If x belongs to Mn for every n. Thus for some xn ∈ Mn we have yn = fn (xn ). By (4. and the image of is ﬂat over R by Problem 10. with M ﬁnitely generated. Then all vertical maps are isomorphisms. No generality is lost. with exact rows. Since hF is an isomorphism. By (4. then x converges to 0. If g(y) = 0.8) we have (I m+k M ) ∩ N = I k ((I m ) ∩ N ) ⊆ I k N ⊆ (I k M ) ∩ N. which is contained in Mn by the induction hypothesis.1). 4. Since gn ◦ fn = 0 for all n.2. 6. then y is represented by a sequence {yn } with yn ∈ Mn and gn (yn ) = 0 for suﬃciently large n. R = ˆ this map is RI. the ﬁrst row remains exact. In general. The elements xn determine x ∈ M such that y = f (x). (10.2.) 9. (I n ) = RI n = (RI)n = (I)n . Take inverse limits in (4. ∼ ˆ ˆ 13. proving exactness.2. Since g is surjective and hF is an isomorphism. The result holds for n = 0 because M0 = M and N0 = N . Thus the following diagram is commutative. 8. let x ∈ f −1 (Nn+1 ). part (ii). hM is an isomorphism when M is free of ﬁnite rank.7) for the last row.8. Since R ˆ ˆ ˆ ⊗R I → R ⊗R R ∼ R is injective. so if we augment the ﬁrst row by attaching 0 → on the left.4 3. hM is injective by the four lemma. ˆ 11. the topology is Hausdorﬀ. By Problem 12. hence x and ˆ 0 are identiﬁed in M .1. it follows ˆ that hM is surjective. Since M is isomorphic to its completion. . If it is true for n. G R⊗ F G R⊗ M G0 ˆ ˆ R ˆ R R ⊗R N 0 GN ˆ hN ˆ f GF ˆ hF g ˆ GM ˆ hM G0 See (4. we may regard M as the set of constant sequences in M . Consider the diagram for Problem 7. Since Nn+1 ⊆ Nn . it follows that x belongs to f −1 (Nn ). hI : R ⊗R I → I is an isomorphism. we have an exact sequence g GN f GF G0 GM 0 with F free of ﬁnite rank. ˆ ˆ ˆ 12. N is ﬁnitely generated. I is ﬁnitely generated. 4. 10. The following diagram is commutative.

By (4. By (4.9). 1 annihilates M/MM .2. hence P = M by maximality of M.2. If P is a prime ideal containing ann(M/MM ).1). M = 0. . 5. hence I is a prime ideal of R. 4. R is complete with respect to the I-adic topology. hence M is a maximal ideal. Since an + an+1 + · · · + am ∈ (I)n for all n. 7.2. 3. so MM = M .9). and distinctness is preserved by (2.3. The contraction of a chain of prime ideals of S contained in J is a chain of prime ideals of R contained in R.1). S/J is integral over R/I. Thus ht J ≤ ht I. Let S = R \ P . This will be true unless ann(M/MM ) = R. the series 1 + a + a2 + · · · + an ˆ converges to some b ∈ R. ∼ ˆ ˆ ˆ ˆ ˆ 16.2. 6. R/M = R/M.1) and Problem 4 that coht I = dim R/I = dim S/J = coht J. any chain of distinct prime ideals of R can be lifted to a chain of distinct prime ideals of S. distinct by (2. Since ˆ ˆ ˆ ax belongs to I for every x ∈ R. it follows from (5. Then P = Q ∩ R ⊇ I. 2. ˆ ˆ 15.4)]. there is nothing to prove. A chain of distinct prime ideals of S contracts to a chain of prime ideals of R. M is contained in every maximal ideal. then I = J ∩ R is a prime ideal of R. so by Lying Over [see (2.3)]. Conversely. Thus ht J ≥ ht I. The function 2n is its own diﬀerence. contradicting the hypothesis. A descending chain of distinct prime ideals of R starting from I can be lifted to a descending chain of distinct prime ideals of S starting from J.1). a ∈ J(R). In this case.) As in the previous problem. Chapter 5 1. we have ht J ≤ ht Q ≤ ht P = ht I. and it follows that M is the ˆ unique maximal ideal of R. (If the height of I is inﬁnite. Thus a ∈ I ⇒ 1 − a is a unit in R.2. If J is a prime ideal of S. Thus with the aid of the above proof for J prime. and let P be a prime ideal of R such that P ⊇ I and ht P = ht I. so dim S ≥ dim R. For any ideal J. Now let J be any ideal of S. we must show that M ⊇ ann(M/MM ). By NAK.5 0 G I n /I n+1 G (I)n /(I)n+1 ˆ ˆ G R/I n+1 G R/(I)n+1 ˆ ˆ G R/I n G R/(I)n ˆ ˆ G0 0 G0 The second and third vertical maps are isomorphisms by (4. By ˆ ˆ Problem 15.2)] there is a prime ideal Q containing J that lies over P . Since S/J is integral over the subring R/I. so the ﬁrst vertical map is an isomorphism by the short ﬁve lemma. Suppose ˆ ˆ that a ∈ I.2.9) and Problem 9.2. by Going Down [see (2. Then (R/I)P = 0 iﬀ S −1 (R/I) = 0 iﬀ S −1 R = S −1 I iﬀ 1 ∈ S −1 I iﬀ 1 = a/s for some a ∈ I and s ∈ S iﬀ I ∩ S = 0 iﬀ I is not a subset of P . and we can ˆ ˆ let n approach inﬁnity to get (1 − a)b = 1. First assume J is a prime ideal of S. By (0. Thus dim R ≥ dim S.3. By Going Up [see (2. 1 + ax is also a unit. Now (1 − a)(1 + a + a2 + · · · + an ) = 1 − an+1 . then P ⊇ M. let Q be a prime ideal of S with Q ⊇ J. so R/M is a ﬁeld.2.

K = MK. 8. Chapter 6 1.) By (6. Z is a zero-divisor. Z) is the unique maximal ideal of S. By Problem 6.9). (The inverse is 1 ⊗ xi → 1 ⊗ ei .2). Y. Since Z X = 0.4). 4. Z .4. 2. 2.1).1.6.4. 6. If M is ﬂat. But ker uM is 0 by hypothesis. R The last term is 0 by hypothesis. δ is surjective. X + Y } is a system of parameters. Chapter 7 1. But this = map is just uM .3). so δ = 0. Y ) ⊂ (X. and the result follows from Problems 3-5. If M = (X. and Y (X + Y ) = Y . which forces Cn = 0. which induces. The height of P is 0 because the ideals (Y ) and (Z) are not prime.3. 7.1). dim R/P = coht P . Thus 0 = im f = ker g. Let J be the ideal (Z. Taking the inﬁmum over Q. then the map M ⊗R M → R ⊗R M ∼ M via a ⊗ x → ax is injective. X ∈ (Y ) and Z ∈ (Y ). 2}. then the functor N → N ⊗R M is exact. Y . This says. Since free implies projective implies ﬂat always. 9. but X Z = 0 ∈ (Y ). Y Z) ⊆ J ⊆ M. it suﬃces to show that ﬂat implies free. S7. N ). and ht P ≥ ht I by deﬁnition of height. by Baer’s criterion (TBGY 10. N ) → HomR (I. im f . ht J = ht I. N ) → HomR (R. hence the map i∗ : HomR (R. which is an isomorphism. so by NAK we have K = 0. N ) → HomR (I. The chain of prime ideals (X) ⊂ (X.1. By (6. ht Q ≥ ht P . N ) → Ext1 (R/I. 2 2 2 2 then M = (X .7) and (5. we have ht J ≥ ht I. Z) gives dim R ≥ 2. (Note that 2 2 X Y = X Z = 0. ∼ 3. the exact sequence HomR (R/I. We have the short exact sequence 0 → I → R → R/I → 0. Since XY (or equally well XZ). Thus gn is the zero map. K is ﬁnitely generated. so ker ∂n = im gn = 0. Note that ker f. Z) and is not a zero-divisor. If M is the maximal ideal of R. P / / = has coheight 1 by (5. . that N is injective. Since M is a Noetherian R-module. Since R/P ∼ k[[X]] has dimension 1. Therefore ∂n is an injective zero map. Since f ∗ is the zero map. 5. By Problem 4. so g is injective. {Z. We have g ∗ : 1 ⊗ ei → 1 ⊗ xi . we have dim R ≤ dim S/(XY ) = dim S − 1 = 2 by (5.3. and the result follows. By (5. Y. Y .6 By what we have just proved. For example. We have im ∂n = ker fn−1 = 0 and ker gn = im fn = Bn .) Thus im f ∗ = ker g ∗ = 0. This follows from the base change formula R/I ⊗R M = M/IM with I = M (see TBGY. dim R/P = dim R − t = dim R − ht P . for any R-module N . belongs to the maximal ideal (X.1). X(X + Y ) = X . X + Y ). so J is an ideal of deﬁnition. N ) is surjective. and ker g are all equal to {0. This forces coker uK = 0.

N ). Ext1 (R/I. namely TorR (M. 1 1 As before. the coheight of (X 3 − Y 2 ) is 1. By exactness. and similarly for the other two tensor products. we have TorR (M. Given an exact sequence as in (7. so localization may reduce the dimension. F ) → TorR (M. The maximal ideal is not principal because X and Y cannot both be multiples of a single polynomial. Cn−1 is injective. and is therefore injective. F ) = 0 1 for every free S-module F . The left side is at least equal to the right side. so assuming that the right side is at most n. Thus X − 1 and Y − 1 are associates. By = R R (7.4) part 4. we use (5. idR N ≤ n.1) and Problem 7. To ease the notation we will omit all the overbars and adopt the convention that all calculations are mod (X 3 − Y 2 ).1. By (5. The corresponding (truncated) long exact sequence is 0 = TorR (M. which is n+r−1 r−1 = (n + r − 1)(n + r − 2) · · · (n + 2)(n + 1) (r − 1)! This is a polynomial of degree r − 1 in the variable n. Apply Tor to the exact sequence 0 → N → N → N → 0 to get the exact sequence 0 = TorR (M. Now X 2 + X + 1 and Y + 1 are units in R because they do not vanish when X = Y = 1. By R (7. we must count the number of such monomials.7). Cn−1 ). If N is an arbitrary S-module. Thus localization at (X. it suﬃces to show that idR N ≤ n for all N . Y ) has no eﬀect on dimension. Y ). Cn−1 ) = 0.7). Since K(Y ) has transcendence degree 1 over K and K(X. The map whose domain is (M ⊗R S) ⊗S K is induced by the inclusion of K into F . (In general. M ⊗R S is ﬂat. (we are adjoining a root of X 3 − Y 2 ).3. Now assuming (b). assuming that the characteristic of K is not 2 or 3. only the second assertion of (b) requires proof. N ) → M ⊗R K → M ⊗R F → M ⊗R N → 0.2. it follows that the dimension of K[X. Since monomials of degree n in r variables form a basis for the polynomials of degree n. Thus the kernel of the map. N ) ∼ Ext1 (R/I.1). because Tor commutes with direct sums. 2.2. prime ideals of a localized ring AP correspond to prime ideals of A that are contained in P . By deﬁnition. is zero.) 3. . 9.2.6. Y )/(X 3 − Y 2 ) is algebraic over K(Y ). we replace M ⊗R K by (M ⊗R S)⊗S K. To show that dim R = 1. 1 Chapter 8 1.7 8. so dim R = 1. We have (X 2 + X + 1)(X − 1) = X 3 − 1 = Y 2 − 1 = (Y − 1)(Y + 1).4). Y ]/(X 3 − Y 2 ) is 1. the Hilbert polynomial is the composition length lk (I n /I n+1 ). (X. and the corresponding sequence of prime ideals is (X 3 − Y 2 ). because M ⊗R S is a ﬂat S-module by hypothesis. dimension shifting yields Extn+1 (R/I. and similarly for the other two tensor products. N ) → M ⊗R N → M ⊗R N → M ⊗R N → 0. 1 We may replace M ⊗R N by (M ⊗R S) ⊗S N . we have a short exact sequence 0 → K → F → N → 0 with F free. so by (7. If (a) holds.

we have bAn ⊆ Bn+d for n ≥ 0. we may deﬁne the graded k-algebra homomorphism ϕ of Problems 3-5 with r = d. . By (8.) By deﬁnition of a graded ring. ϕ is an isomorphism.2. 5. an isomorphism of graded k-algebras induces an isomorphism of ﬁrst components. b is referred to as a homogeneous element of degree d. = But the k-vector space on the left has a basis consisting of all monomials of degree 1.3). (Frequently.1. Xd ])1 ∼ M/M2 . the result follows. R is regular. .8 4. If R is regular. Then lk (Bn+d ) ≥ lk (bAn ) = lk (An ) ≥ lk (Bn ). . in other words. This follows from Problem 3 and additivity of length (5.3). 6. Since the Hilbert polynomial has degree d. . we have dimk M/M2 = d. Conversely. (k[X1 . Fix a nonzero element b ∈ Bd . Since lk (An ) = n+r−1 r−1 . . Since there are exactly d of these.

Sign up to vote on this title

UsefulNot useful- Chapter 12 Arithmetic of Power Series
- Alexander Schrijver - Theory of Linear and Integer Programming.pdf
- Pre-Calculus Semester 2 Cheat Sheet
- Numerical Analysis, 2nd Edition, Timothy Sauer
- Christophe Hohlweg, Carsten Lange and Hugh Thomas- Permutahedra and Generalized Associahedra
- 321
- Topologie Cards
- Lectures on Communicative Algebra
- Undergraduate Topology
- algoritmic notes
- Hausdorff Topological Spaces
- Smarandache V−Connected spaces
- l0k-Chapter8Solutions.pdf
- l0k-Chapter6Solutions.pdf
- Munkres Solution S13
- Topological Spaces
- Left Multiplicative Generalized Derivations on Right Ideal in Semi Prime Rings
- topology chap 1
- Linus Kramer, Saharon Shelah, Katrin Tent and Simon Thomas- Asymptotic Cones of Finitely Presented Groups
- Smirnov
- Saharon Shelah- Many Partition Relations Below Density
- Solved Excersices of Introducion of Conmutavie Algebra Michale Atiyah
- 8b1b7fa930ce30fc30c8
- Renling Jin and Saharon Shelah- Essential Kurepa Trees Versus Essential Jech-Kunen Trees
- Math 215 Lecture Notes
- Note Am Ica 0.1.2
- Exercise 2.1
- Menachem Kojman and Saharon Shelah- A ZFC Dowker space in aleph-omega+1
- Paul C. Eklof and Saharon Shelah- The Structure of Ext(A,Z) and GCH
- Modules
- A Course in Commutative Algebra - R. Ash (2003) WW