How statement and proof provenance work
The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.
- Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
- AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
- AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.
These labels describe origin, not correctness: citations and verification chips remain separate evidence.
Krull Dimension and Height Theorems
1 · Prerequisites
- Algebraic Extensions, Extension Degree, and Finite Fields
- Associated Primes and Primary Decomposition
- Binary Operations, Monoids, Groups and Subgroups
- Chain Conditions, Semisimple Modules and the Wedderburn–Artin Theorem
- Compactness in Metric Spaces
- Construction of the Natural Numbers
- Construction of the Real Numbers via Cauchy Sequences
- Construction of the Real Numbers via Dedekind Cuts
- Cosets, Index and Lagrange's Theorem
- Countability and Uncountability
- Determinants of Matrices over a Commutative Ring
- Divisibility, Euclidean Domains, Principal Ideal Domains and Unique Factorisation
- Finite Counting, Factorials and Binomial Coefficients
- Finite Fields and Cyclotomic Extensions
- Foundations of the Real Numbers for Analysis
- Free Modules, Exact Sequences, Projective and Injective Modules
- Group Actions, Orbits, Stabilisers and Cayley's Theorem
- Group Homomorphisms and the Isomorphism Theorems
- Ideals, Quotient Rings and the Isomorphism Theorems for Rings
- Integral Extensions and Going Up
- Linear Independence, Bases and Dimension
- Localisation of Modules and Support
- Modules, Submodules, Quotient Modules and the Isomorphism Theorems
- Noether Normalisation and Nullstellensatz
- Noetherian Rings and Hilbert Basis
- Normal Subgroups and Quotient Groups
- Order, Zorn's Lemma, and the Axiom of Choice
- Polynomial Rings, the Division Algorithm and Roots
- Prime Spectra and Radicals
- Relations, Functions, and Quotients
- Rings, Subrings, Integral Domains and Fields
- Roots, Rational Powers, and Classical Inequalities
- Sequences and Limits
- Simple Field Extensions and the Construction of the Complex Numbers
- Symmetric Groups, Cycle Decomposition and the Sign Homomorphism
- Tensor Products of Modules
- The Determinant of a Linear Operator, Cofactors and Cramer's Rule
- The Field of Fractions and Localisation
- The ZFC Axioms and the Basic Set Constructions
- Vector Spaces, Linear Subspaces, Span and Direct Sums
2 · Summary
This page turns the earlier spectrum-and-height definitions into working dimension theory. It proves Krull's principal ideal theorem and the height theorem, builds their local converse, and packages systems of parameters as the local radical form of dimension.
The second half records the one-variable polynomial dimension jump, the affine-domain dimension equals transcendence-degree theorem, and the affine dimension formula. The examples page isolates the boundary cases where zero divisors, localization, and noncatenary behavior matter.
3 · Logical flowchart
4 · Definitions, theorems and proofs
Minimal primes are exactly the primes of height zero
Statement
Let be a commutative ring and let . Then is minimal if and only if .
Facts & Assumptions
Given: A commutative ring and a prime ideal .
The height of is the supremum of the lengths of strict prime chains ending at (Height equals local dimension).
Proof
If is minimal, there is no strict prime ideal properly contained in . Therefore every strict chain ending at has length , and [L1] gives .
If , then [L1] says no strict chain of positive length ends at . In particular there is no prime ideal properly contained in , so is minimal.
The two implications prove that minimal primes are exactly the primes of height zero.
In a domain, every prime chain below a prime begins at (0)
Statement
Let be an integral domain and let . Then is a prime ideal of contained in . Consequently every strict prime chain below can be extended downward to a strict chain beginning at .
Facts & Assumptions
Given: An integral domain and a prime ideal .
An integral domain is a nonzero commutative ring in which implies or (Zero divisor, and integral domain: a commutative ring with and no zero divisors).
A prime ideal is a proper ideal such that implies or (Prime ideals and maximal ideals in a commutative ring).
Proof
By [L1], is nonzero, so . If , then , and [L1] gives or . Thus or , so [L2] shows that is prime.
Every ideal contains , hence . Therefore any chain of primes below can be extended by adjoining at the bottom if it is not already present.
So every prime chain below begins at after at most one downward extension.
Height in a quotient measures chains between two primes
Statement
Let be a commutative ring and let be prime ideals. Then the height of in is the supremum of the lengths of strict prime chains
in .
Facts & Assumptions
Given: A commutative ring and prime ideals .
Prime ideals of correspond exactly to the prime ideals of containing , with strict inclusions preserved (Prime ideals of a quotient ring are exactly the prime ideals containing the ideal).
Height is the supremum of the lengths of strict prime chains ending at the chosen prime (Height equals local dimension).
Proof
By [L1], strict prime chains in ending at are in bijection with strict prime chains in beginning at and ending at . Corresponding chains have the same length.
Applying [L2] to the prime of the quotient ring , the height of is exactly the supremum of the lengths of those quotient chains. Step 1.1 translates that supremum into the displayed chains in .
Therefore height in the quotient is the relative chain length from to .
Reduce the principal ideal theorem to a Noetherian local domain
Statement
Let be a Noetherian commutative ring, let , and let be a prime ideal minimal over . Then for every minimal prime of the localized quotient
is a Noetherian local domain. If , then the image of in is nonzero and the maximal ideal of is minimal over that principal ideal. Consequently the principal ideal theorem is reduced to bounding the maximal ideal of such a local domain by .
Facts & Assumptions
Given: A Noetherian commutative ring , an element , and a prime ideal minimal over .
Minimal primes are exactly the primes of height zero (Minimal primes are exactly the primes of height zero).
Quotients and localizations of Noetherian rings are Noetherian (Every quotient and every localisation of a Noetherian ring is Noetherian).
A quotient by a prime ideal is an integral domain ( is an integral domain if and only if is a prime ideal).
Localization at a prime ideal is local, with maximal ideal the extended prime ( is local with unique maximal ideal ).
Prime ideals of a quotient and of a localization correspond by inclusion-preserving bijections (Prime ideals of a quotient ring are exactly the prime ideals containing the ideal, Prime ideals of a localization are exactly the primes disjoint from the denominator set).
Proof
Let be a minimal prime of . By [L1], . The quotient is a Noetherian domain by [L2] and [L3], and then [L4] makes a Noetherian local domain.
If , then already has height by step 1.1 and there is nothing left to prove. Assume now that . Since is minimal over , the element cannot lie in ; otherwise the prime would also contain and minimality would force . Thus the image of in , and hence in , is nonzero.
Let be the maximal ideal of . By [L5], primes of correspond to primes of that lie between and . Because is minimal over , the prime is minimal over the image of in , and after localizing there is no smaller prime of containing . Hence is minimal over .
Let be any strict prime chain. If there is nothing to bound. If , then , so minimality of over gives . By [L2]–[L4], is a Noetherian local domain in which the image of is nonzero; [L5] also shows that its maximal ideal is minimal over that image. The original chain induces a strict chain of length ending at this maximal ideal. Therefore a height bound of in every reduced local-domain case forces . Since the original chain was arbitrary, .
The symbolic-power step inside the principal ideal theorem
Statement
Let be a Noetherian local domain, and let be nonzero. If is minimal over , then every prime ideal strictly contained in is zero. In particular .
Facts & Assumptions
Given: A Noetherian local domain and a nonzero element such that is minimal over .
In a domain, the zero ideal is prime (Zero divisor, and integral domain: a commutative ring with and no zero divisors, Prime ideals and maximal ideals in a commutative ring).
Quotients and localizations of Noetherian rings are Noetherian, and every ideal of a Noetherian ring is finitely generated (Every quotient and every localisation of a Noetherian ring is Noetherian, A commutative ring is Noetherian exactly when every ideal is finitely generated, exactly when its ideals satisfy the ascending chain condition, and exactly when every nonempty set of ideals has a maximal member).
The nilradical of a Noetherian ring is nilpotent (The nilradical of a Noetherian ring is nilpotent).
Prime ideals of a localization correspond to primes disjoint from its denominator set (Prime ideals of a localization are exactly the primes disjoint from the denominator set).
If for a finitely generated module, then for some (Determinant trick for Nakayama).
In a local ring, every element outside the unique maximal ideal is a unit; in particular is a unit for in the maximal ideal (Assuming the Axiom of Choice, a nonzero commutative ring is local exactly when its nonunits form an ideal, exactly when one of and is a unit for every ).
Proof
Since is both maximal and minimal among primes containing , it is the only such prime. Hence the quotient is Noetherian by [L2] and has the single prime . Its nilradical is therefore , so [L3] gives for some . Each layer is a finite-dimensional vector space over : it is finitely generated by [L2] and is annihilated by . A descending chain of ideals in induces descending chains in these finitely many finite-dimensional layers, so all layers, and therefore the original chain, stabilize. Thus is Artinian.
Let be prime. Then . For , define the symbolic power These form a descending chain. By step 1.1, the ideals in stabilize, so for some , If , write with . In the element is a unit, while ; hence . Therefore
The quotient module is finitely generated by [L2], and step 2.1 says . Apply [L5] with . It gives for some . Fact [L6] makes a unit, so and . Localizing this equality at gives
Apply [L5] in the local ring to the finite module and the ideal . Step 3.1 says that ideal times the module is the module, so [L6] gives . Because is a domain, this forces , and contraction through [L4] gives . Thus every prime strictly below is zero, and [L1] yields .
Krull's principal ideal theorem
Statement
Let be a Noetherian commutative ring, let , and let be a prime ideal minimal over . Then .
Facts & Assumptions
Given: A Noetherian commutative ring , an element , and a prime ideal minimal over .
The principal-ideal bound reduces to the case of a Noetherian local domain whose maximal ideal is minimal over one nonzero element (Reduce the principal ideal theorem to a Noetherian local domain).
In that reduced local-domain situation, every prime properly below the maximal ideal is zero, so the maximal ideal has height at most (The symbolic-power step inside the principal ideal theorem).
Proof
By [L1], choose a minimal prime and pass to the Noetherian local domain whose maximal ideal is minimal over the image of .
Fact [L2] applies to , so its maximal ideal has height at most .
The reduction packaged in [L1] identifies this with the desired bound upstairs in .
A Noetherian local domain has dimension zero exactly when it is a field
Statement
Let be a Noetherian local domain. Then if and only if is a field.
Facts & Assumptions
Given: A Noetherian local domain .
A local ring has a unique maximal ideal (A local ring is a nonzero commutative ring with a unique maximal ideal).
In a domain the zero ideal is prime (Zero divisor, and integral domain: a commutative ring with and no zero divisors, Prime ideals and maximal ideals in a commutative ring).
Krull dimension is the supremum of lengths of strict prime chains (Krull dimension of a nonzero ring).
Proof
Suppose . By [L2], the chain is a prime chain. If it were strict, [L3] would give , contradiction. Hence , so every nonzero element is outside the maximal ideal and therefore is a unit. Thus is a field.
Conversely, if is a field, its unique maximal ideal is . Therefore the only prime ideal is , and [L3] gives .
So a Noetherian local domain has dimension zero exactly when it is a field.
A minimal prime over a principal nonzerodivisor has height one
Statement
Let be a Noetherian commutative ring, let be a nonzerodivisor, and let be a prime ideal minimal over . Then .
Facts & Assumptions
Given: A Noetherian commutative ring , a nonzerodivisor , and a prime ideal minimal over .
Every prime minimal over a principal ideal has height at most (Krull's principal ideal theorem).
The principal-ideal reduction passes to a Noetherian local domain whose maximal ideal is minimal over the image of (Reduce the principal ideal theorem to a Noetherian local domain).
A Noetherian local domain has dimension zero exactly when it is a field (A Noetherian local domain has dimension zero exactly when it is a field).
Proof
By [L1], .
Apply [L2] to a minimal prime of . Because is a nonzerodivisor, , so . In the reduced local domain , the image of lies in the maximal ideal. If that maximal ideal had height , then [L3] would make a field, forcing to be a unit, contradiction. Hence the maximal ideal of has height , so has height at least .
Steps 1.1 and 2.1 give .
Choose the first generator's minimal prime inside the target prime
Statement
Let be a Noetherian commutative ring, let , let
and let be a prime ideal minimal over . Let be the minimal prime ideals over . If is not one of the and if
is a strict prime chain, then there exists a strict prime chain of the same length ending at whose first proper subprime is not contained in any . For such a chain one can choose
and then is minimal over .
Facts & Assumptions
Given: A Noetherian commutative ring , an integer , the ideal , a prime ideal minimal over , the minimal primes over , and a strict chain .
A Noetherian ring has finitely many minimal primes over any ideal (A Noetherian ring has finitely many minimal prime ideals, Minimal primes over a proper ideal exist).
Finite prime avoidance lets us choose an element outside a finite union of prime ideals once the ambient ideal is not contained in that union (An ideal contained in a finite union of prime ideals lies in one of them).
Every prime minimal over a principal ideal has height at most (Krull's principal ideal theorem).
Proof
By [L1], the family is finite. Repeatedly applying the two-step prime-avoidance argument to the triples for produces a strict chain of the same length ending at whose first proper subprime is not contained in any . Relabel so that the resulting chain is again with for every .
Since is not contained in the finite union , [L2] gives . In particular , so contains .
Let be prime minimal over . Because contains , it contains one of the minimal primes . By the choice of , one has , so . If , then is a strict chain in the quotient ring , so the prime has height at least . But is minimal over , hence is minimal over the principal ideal generated by the image of , contradicting [L3]. Therefore .
Hence is minimal over , with chosen from the first proper subprime of a chain of the same length.
Quotienting by the first minimal prime reduces the remaining height count
Statement
In the situation of Choose the first generator's minimal prime inside the target prime, choose so that is minimal over , and let
be a strict prime chain with . Then in the prime is minimal over the ideal generated by the images of , and the quotient chain
is strict of length .
Facts & Assumptions
Given: A Noetherian commutative ring , an ideal with , a prime minimal over it, an element as in Choose the first generator's minimal prime inside the target prime, and a strict chain with .
The chosen element makes minimal over (Choose the first generator's minimal prime inside the target prime).
Prime ideals of a quotient correspond to prime ideals upstairs containing the quotient ideal, with strict inclusions preserved (Prime ideals of a quotient ring are exactly the prime ideals containing the ideal).
Proof
By [L1], every prime ideal of containing and lying inside is equal to . Hence is minimal over the images of in .
Because for every , [L2] sends the given chain to The inclusions stay strict, because equality between consecutive quotient primes would contract back to equality between the corresponding primes of .
Thus quotienting by leaves a prime minimal over generators and reduces this chosen chain by exactly one step.
Krull's height theorem
Statement
Let be a Noetherian commutative ring, let
be an ideal generated by elements, and let be a prime ideal minimal over . Then .
Facts & Assumptions
Given: A Noetherian commutative ring , an -generated ideal with , and a prime ideal minimal over .
The case is Krull's principal ideal theorem (Krull's principal ideal theorem).
For , one may replace the first generator by an element from the penultimate prime of a chain so that is minimal over (Choose the first generator's minimal prime inside the target prime).
After quotienting by that chosen , the image is minimal over generators and a chain ending at loses one step (Quotienting by the first minimal prime reduces the remaining height count).
Proof
If , [L1] gives .
Assume and that the theorem is known for -generated ideals. Suppose for contradiction that . Then there exists a strict prime chain with . By [L2], after replacing by a suitable element , the prime is minimal over .
By [L3], the quotient prime is minimal over the -generated ideal generated by the images of , and the above chain descends to a strict chain of length ending at in . This contradicts the induction hypothesis for generators.
Therefore the assumption is impossible, and .
Height is bounded by the minimal number of local generators
Statement
Let be a Noetherian commutative ring and let . Then
where denotes the minimal number of generators.
Facts & Assumptions
Given: A Noetherian commutative ring and a prime ideal .
The localization is a Noetherian local ring with maximal ideal (Every quotient and every localisation of a Noetherian ring is Noetherian, is local with unique maximal ideal ).
A prime minimal over an ideal generated by elements has height at most (Krull's height theorem).
The height of is the dimension of (The height of a prime ideal).
Proof
By [L1], the maximal ideal of the local ring is . If it is minimally generated by elements, then it is certainly minimal over the ideal generated by those same elements. Applying [L2] inside gives .
By [L3], , so step 1.1 says .
This is the claimed local-generator bound on height.
Select generators witnessing the converse height theorem
Statement
Let be a Noetherian commutative ring and let be a proper ideal of height . Then there exist elements such that for every the ideal has height exactly .
Facts & Assumptions
Given: A Noetherian commutative ring , a proper ideal , and an integer .
A Noetherian ring has only finitely many minimal primes over a given ideal (A Noetherian ring has finitely many minimal prime ideals).
Finite prime avoidance lets us choose an element outside finitely many forbidden prime ideals (An ideal contained in a finite union of prime ideals lies in one of them).
The principal ideal theorem and the height theorem bound the height of a prime minimal over chosen generators by (Krull's principal ideal theorem, Krull's height theorem).
Proof
If , the empty list works.
Assume . The prime ideals of height are exactly the minimal primes of , hence are finite by [L1]. None of them contains , because . Therefore [L2] provides outside every height-zero prime. Any prime minimal over must then have height at least , while [L3] gives height at most . Thus has height exactly .
Suppose and have already been chosen so that has height exactly . The minimal primes over that ideal are finite by [L1]. Among them, collect those of height ; none can contain , because has height . By [L2], choose outside all of those primes. Then every prime minimal over has height at least , because otherwise it would sit inside one of the excluded height- minimal primes. On the other hand [L3] bounds its height by . Hence has height exactly .
Steps 1.1, 1.2, and 2.1 build the required list .
Converse to Krull's height theorem in localised form
Statement
Let be a Noetherian commutative ring and let have finite height . Then in the local ring there exist elements such that the maximal ideal is minimal over . Equivalently, is minimal over an -generated ideal after localizing at .
Facts & Assumptions
Given: A Noetherian commutative ring and a prime ideal with .
The localization is a Noetherian local ring with maximal ideal (Every quotient and every localisation of a Noetherian ring is Noetherian, is local with unique maximal ideal ).
By definition, (The height of a prime ideal).
In a Noetherian ring, a proper ideal of height contains elements whose successive generated ideals have heights (Select generators witnessing the converse height theorem).
Proof
By [L1] and [L2], the local ring has maximal ideal and dimension . Applying [L3] to the proper ideal produces elements such that has height . Writing each with and , and replacing by the associate , we may assume with .
Let be a prime ideal minimal over . Because has height , the prime has height . The maximal ideal also contains , so . If the inclusion were strict, a strict chain of length ending at would extend by one more step to a chain ending at , contradicting . Therefore , so the maximal ideal is minimal over .
Therefore becomes minimal over an ideal generated by elements after localizing at .
Systems of parameters and parameter ideals
Definition
Let be a finite-dimensional Noetherian local ring of dimension . A tuple
is a system of parameters when
Equivalently, the ideal generated by the tuple has maximal radical. That ideal is called a parameter ideal.
When , the empty tuple is a system of parameters exactly when . In a zero-dimensional Noetherian local ring this is the correct convention, so systems of parameters still have length equal to the dimension.
Parameter ideals are exactly the m-primary d-generated ideals
Statement
Let be a -dimensional Noetherian local ring and let . Then is a system of parameters if and only if is -primary.
Facts & Assumptions
Given: A -dimensional Noetherian local ring and the ideal .
By definition, is a system of parameters exactly when (Systems of parameters and parameter ideals).
For a finite module over a Noetherian ring, a proper ideal is -primary exactly when the quotient has associated-prime set ; equivalently, when some power of kills the quotient and every element outside acts injectively (Primary submodules of finite modules are characterized by a singleton associated-prime set).
Proof
Suppose . Since every element outside is a unit in a local ring, it acts injectively on . Also means some power lies in , hence kills . Therefore [L2] makes a -primary ideal.
Conversely, if is -primary, [L2] applied to the finite module gives . Then [L1] says is a system of parameters.
Thus parameter ideals are exactly the -generated -primary ideals.
Choose a parameter that misses the top-dimensional minimal components
Statement
Let be a Noetherian local ring of positive dimension . Then there exists outside every minimal prime of . For every such one has
In particular misses the top-dimensional minimal components.
Facts & Assumptions
Given: A Noetherian local ring with .
A Noetherian ring has finitely many minimal primes (A Noetherian ring has finitely many minimal prime ideals).
Finite prime avoidance chooses an element of outside finitely many proper prime ideals (An ideal contained in a finite union of prime ideals lies in one of them).
If a prime is minimal over a principal ideal generated by a nonzerodivisor, then it has height (A minimal prime over a principal nonzerodivisor has height one).
Prime ideals of a quotient correspond to primes upstairs containing the quotient ideal (Prime ideals of a quotient ring are exactly the prime ideals containing the ideal).
Proof
By [L1], the minimal primes of form a finite set; because , none equals . Hence [L2] provides outside every minimal prime.
Let be a minimal prime of . By [L4], the prime is minimal over in . If is a minimal prime of , then step 1.1 gives , so the image of in the domain is a nonzerodivisor. Therefore [L3] shows that has height , and every chain in extends upward to a chain in longer by one step.
Since , step 2.1 implies . Taking the supremum over all minimal primes of yields .
Thus one can choose a first parameter outside the minimal components, and every such choice lowers dimension by at least one.
A first parameter lowers local dimension by exactly one
Statement
Let be a Noetherian local ring of positive dimension , and let be a system of parameters. Then
Facts & Assumptions
Given: A Noetherian local ring with and a system of parameters .
The image of in is a system of parameters there, equivalently an -primary ideal generated by elements (Systems of parameters and parameter ideals, Parameter ideals are exactly the m-primary d-generated ideals).
A prime minimal over an ideal generated by elements has height at most , and conversely a prime of finite height is locally minimal over generators (Krull's height theorem, Converse to Krull's height theorem in localised form).
Prime ideals of correspond to prime ideals of that contain (Prime ideals of a quotient ring are exactly the prime ideals containing the ideal).
Proof
By [L1], the maximal ideal of is minimal over an ideal generated by elements. Applying the height theorem in the quotient ring, and then reading dimension as the height of its maximal ideal, gives .
Suppose instead that . Then [L2] applied in the local ring provides an ideal generated by at most elements whose radical is the maximal ideal . Lifting those generators to and adjoining , we obtain an ideal of generated by at most elements with radical . Applying the height theorem to the maximal ideal of would then give , contradiction.
Therefore .
Every finite-dimensional Noetherian local ring has a system of parameters
Statement
Let be a finite-dimensional Noetherian local ring, with . Then has a system of parameters.
Facts & Assumptions
Given: A finite-dimensional Noetherian local ring of dimension .
A first parameter can be chosen in outside every minimal prime, and for such a choice the quotient has dimension at most (Choose a parameter that misses the top-dimensional minimal components).
A system of parameters is a -tuple whose generated ideal has radical (Systems of parameters and parameter ideals).
Prime ideals of a quotient correspond to primes upstairs containing the quotient ideal (Prime ideals of a quotient ring are exactly the prime ideals containing the ideal).
Proof
If , the empty tuple is a system of parameters by [L2].
Assume and that the theorem is known in dimensions . By [L1], choose outside every minimal prime. Then is a Noetherian local ring of dimension at most . By the induction hypothesis, choose a system of parameters in of length . Lift those elements to .
The ideal generated by has radical in , so [L3] says the ideal has radical in . By [L2], the tuple is a system of parameters.
Therefore every finite-dimensional Noetherian local ring has a system of parameters.
Local dimension is the minimal number of generators of an ideal with maximal radical
Statement
Let be a finite-dimensional Noetherian local ring of dimension . Then is the least integer for which there exists an -generated ideal with .
Facts & Assumptions
Given: A finite-dimensional Noetherian local ring with .
Systems of parameters exist, and by definition a system of parameters gives a -generated ideal with radical (Every finite-dimensional Noetherian local ring has a system of parameters, Systems of parameters and parameter ideals).
If an ideal generated by elements has radical , then the maximal ideal is minimal over it; the height theorem therefore bounds by (Krull's height theorem).
The local converse to the height theorem produces generators whose radical is the maximal ideal (Converse to Krull's height theorem in localised form, Parameter ideals are exactly the m-primary d-generated ideals).
Proof
By [L1], there exists a -generated ideal with radical . So the least such number is at most .
Conversely, let be an -generated ideal with . Then is minimal over , so [L2] gives . Hence every such generating number is at least .
Steps 1.1 and 1.2 show that the least number is exactly . The same conclusion may also be read from [L3] as the local radical form of the converse height theorem.
Quotienting by a first parameter lowers local dimension by one
Statement
Let be a Noetherian local ring of positive dimension, and let be a system of parameters. Then
Facts & Assumptions
Given: A Noetherian local ring of positive dimension and a system of parameters .
By definition, is the first member of a system of parameters (Systems of parameters and parameter ideals).
A first parameter lowers the local dimension by exactly one (A first parameter lowers local dimension by exactly one).
Proof
The tuple satisfies the hypothesis of [L2] by [L1].
Therefore .
Localisation does not increase Krull dimension
Statement
Let be a nonzero commutative ring and let be a multiplicative subset with . Then
Facts & Assumptions
Given: A nonzero commutative ring and a multiplicative subset with .
Prime ideals of correspond to prime ideals of disjoint from , with strict inclusions preserved (Prime ideals of a localization are exactly the primes disjoint from the denominator set).
Krull dimension is the supremum of the lengths of strict prime chains (Krull dimension of a nonzero ring).
Proof
By [L1], every strict prime chain in contracts to a strict prime chain in of the same length.
Taking suprema and using [L2] gives .
Hence localization does not increase Krull dimension.
Passing to a quotient does not increase Krull dimension
Statement
Let be a nonzero commutative ring and let be a proper ideal. Then
Facts & Assumptions
Given: A nonzero commutative ring and a proper ideal .
Prime ideals of correspond to prime ideals of containing , with strict inclusions preserved (Prime ideals of a quotient ring are exactly the prime ideals containing the ideal).
Krull dimension is the supremum of the lengths of strict prime chains (Krull dimension of a nonzero ring).
Proof
By [L1], every strict prime chain in lifts to a strict prime chain in of the same length.
Therefore [L2] gives .
So passing to a quotient never increases Krull dimension.
A prime chain in R extends to a longer chain in R[x]
Statement
Let be a commutative ring. If
is a strict prime chain in , then
is a strict prime chain in . Consequently whenever is finite.
Facts & Assumptions
Given: A commutative ring and a strict prime chain in .
For a prime ideal , the quotient is an integral domain ( is an integral domain if and only if is a prime ideal).
Prime ideals of a quotient correspond to prime ideals containing the quotient ideal (Prime ideals of a quotient ring are exactly the prime ideals containing the ideal).
Krull dimension is computed by strict prime chains (Krull dimension of a nonzero ring).
Proof
For each , the quotient is a polynomial ring over the domain , so [L1] and [L2] show that is prime. Strictness of the original chain makes the extended chain strict.
The quotient by is again , a domain, so [L1] and [L2] show that is prime and strictly contains .
The displayed chain in therefore has length , and [L3] yields whenever is finite.
Only one saturated step can lie over a fixed contracted prime in R[x]
Statement
Let be a commutative ring and let be prime ideals of with the same contraction . Then , and there is no prime ideal strictly between and .
Facts & Assumptions
Given: A commutative ring and prime ideals of with common contraction .
The quotient is a domain, so it has a field of fractions ( is an integral domain if and only if is a prime ideal, The field of fractions of an integral domain).
Prime ideals of a quotient and a localization correspond to prime ideals upstairs containing the kernel and avoiding the denominator set (Prime ideals of a quotient ring are exactly the prime ideals containing the ideal, Prime ideals of a localization are exactly the primes disjoint from the denominator set).
Over a field, every ideal of is principal, and every nonzero prime ideal of is maximal (For every field , is a principal ideal domain).
Proof
Passing to the quotient by and then localizing away from the nonzero elements of , [L1] and [L2] identify the fiber over with the prime spectrum of . Under this identification, the images of and are comparable prime ideals of , with the image of properly contained in the image of .
By [L3], the only way two comparable primes in can be strictly nested is for the smaller one to be and the larger one to be a nonzero maximal prime. Therefore the image of in is zero. Contracting back through [L2], this means . The same description also shows that no third prime can lie strictly between and , because no third prime lies strictly between and a nonzero prime in the PID .
Hence over a fixed contracted prime in there is at most one extra strict step, and the lower prime in such a pair is exactly the extended prime .
A prime chain in R[x] has length at most one more than its contraction chain
Statement
Let be a Noetherian commutative ring and let
be a strict prime chain in . Then the length is at most .
Facts & Assumptions
Given: A Noetherian commutative ring and a strict prime chain in .
In the local ring , the maximal ideal can be made minimal over generators, where (Converse to Krull's height theorem in localised form, The height of a prime ideal, is local with unique maximal ideal , Every quotient and every localisation of a Noetherian ring is Noetherian).
A prime minimal over an ideal generated by elements has height at most (Krull's height theorem).
If is prime, then is a polynomial ring over the domain , and after localizing at the nonzero elements of that domain its nonzero prime ideals become nonzero prime ideals of a PID and hence are minimal over one generator (Prime ideals of a quotient ring are exactly the prime ideals containing the ideal, For every field , is a principal ideal domain).
Proof
Let . Localizing the given chain at preserves its length, so it is enough to bound the height of the prime . By [L1], after relabelling and choosing , the maximal ideal is minimal over .
If , then is minimal over : any prime of containing contracts to a prime of containing , hence to itself. Therefore [L2] gives .
Suppose now that . Then is a nonzero prime ideal of , where . By [L3], choose a lift whose image generates that nonzero prime of . Any prime containing and has contraction containing , hence equal to by step 1.1; modulo , the prime contains the generator of , so it equals that prime. Therefore , and is minimal over the ideal generated by elements.
By [L2], step 2.2 gives . Since the localized chain still has length , we have . Steps 2.1 and 3.1 cover both cases.
Therefore every strict prime chain in has length at most .
A Noetherian polynomial ring has dimension one larger
Statement
Let be a Noetherian commutative ring of finite Krull dimension. Then
Facts & Assumptions
Given: A Noetherian commutative ring of finite dimension.
Every prime chain in extends to a prime chain in that is longer by one step (A prime chain in R extends to a longer chain in R[x]).
Every prime chain in has length at most (A prime chain in R[x] has length at most one more than its contraction chain).
Proof
Fact [L1] gives .
Fact [L2] gives .
Therefore .
A polynomial ring in n variables over a field has dimension n
Statement
Let be a field and let . Then
Facts & Assumptions
Given: A field and an integer .
Adjoining one polynomial variable to a finite-dimensional Noetherian ring raises dimension by one (A Noetherian polynomial ring has dimension one larger).
Proof
For , the ring is the field , whose only prime ideal is , so its dimension is .
If , then , so [L1] gives dimension .
Therefore for every .
A finite affine extension of a polynomial ring has dimension at most the number of variables
Statement
Let be a field, let be a finite-type -domain, and suppose is module-finite over a polynomial subring . Then .
Facts & Assumptions
Given: A field , a finite-type -domain , and a module-finite inclusion .
The polynomial ring has dimension (A polynomial ring in n variables over a field has dimension n).
In an integral extension, comparable primes with the same contraction are equal (Comparable primes with the same contraction are equal under an integral map).
Proof
A module-finite extension is integral, so any strict prime chain in contracts to a strict prime chain in by [L2].
The base ring has dimension by [L1], so no strict prime chain there has length greater than . Therefore no strict prime chain in has length greater than .
Hence .
A finite affine extension of a polynomial ring has dimension at least the number of variables
Statement
Let be a field, let be a finite-type -domain, and suppose is module-finite over a polynomial subring . Then .
Facts & Assumptions
Given: A field , a finite-type -domain , and a module-finite inclusion .
The coordinate-prime chain in has length (A polynomial ring in n variables over a field has dimension n).
Finite prime chains lift through integral extensions once the first prime upstairs is chosen (Integral extensions lift finite prime chains from the base).
Proof
The coordinate-prime chain in has length by [L1].
Because the extension is module-finite and hence integral, [L2] lifts that chain to a strict prime chain in of the same length .
Therefore .
Affine-domain dimension equals transcendence degree
Statement
Let be a field, let be a finite-type -domain, and let . Then
Facts & Assumptions
Given: A field , a finite-type -domain , and its fraction field .
Noether normalization provides algebraically independent elements such that is module-finite over (Noether normalisation yields module finiteness over a polynomial subring).
A finite affine extension of a polynomial ring has dimension at most, and at least, the number of polynomial variables (A finite affine extension of a polynomial ring has dimension at most the number of variables, A finite affine extension of a polynomial ring has dimension at least the number of variables).
Algebraicity is transitive in towers of fields (Algebraicity is transitive in towers of field extensions).
Proof
By [L1], choose algebraically independent elements such that is module-finite over . Applying [L2] to the inclusion yields .
Because is integral over , every element of is algebraic over the rational function field . Thus [L3] shows that is algebraic over a purely transcendental extension of degree , so .
Steps 1.1 and 2.1 give .
The dimension formula for affine domains
Statement
Let be a field, let be a finite-type -domain, and let . Then
Facts & Assumptions
Given: A field , a finite-type -domain , and a prime ideal .
For a finite-type -domain and a prime ideal , the local dimension formula gives
Because is prime, the quotient is a domain, so its residue field at the generic point is ( is an integral domain if and only if is a prime ideal).
The height of is the dimension of the local ring (Height equals local dimension).
For affine domains, dimension equals transcendence degree of the fraction field (Affine-domain dimension equals transcendence degree).
Proof
Applying [F1] to and gives , where [L1] identifies the residue field term with .
By [L2] and [L3], one has and . Substituting these into step 1.1 yields .
Transcendence degrees along affine prime quotients add correctly
Statement
Let be a field, let be a finite-type -domain, and let be prime ideals of . Then
Facts & Assumptions
Given: A field , a finite-type -domain , and prime ideals .
The affine-domain dimension formula applies to the quotient domain and the prime (The dimension formula for affine domains).
Proof
The quotient is a finite-type -domain, and is a prime ideal of it. Applying [L1] to that quotient domain gives .
This is exactly the displayed identity.
Height plus quotient dimension equals ambient dimension in an affine domain
Statement
Let be a field, let be a finite-type -domain, and let . Then
Facts & Assumptions
Given: A field , a finite-type -domain , and a prime ideal .
The affine-domain dimension formula identifies (The dimension formula for affine domains).
For affine domains, dimension equals transcendence degree of the fraction field (Affine-domain dimension equals transcendence degree).
Proof
Replace the two transcendence degrees in [L1] using [L2], once for and once for the quotient domain .
The result is exactly .
Maximal ideals of an affine domain have full height
Statement
Let be a field, let be a finite-type -domain, and let be a maximal ideal of . Then
Facts & Assumptions
Given: A field , a finite-type -domain , and a maximal ideal .
The residue field is a finite extension of (A maximal ideal of an affine algebra has finite residue field over the base field).
Proof
By [L1], the quotient is a field. Hence .
Applying [L2] now yields .
Thus every maximal ideal of an affine domain has full height.
Maximal chains in an affine domain all have the same length
Statement
Let be a field and let be a finite-type -domain. Every saturated prime chain from to a maximal ideal of has length .
Facts & Assumptions
Given: A field , a finite-type -domain , and a saturated prime chain
In a domain, the minimal prime has height zero (Minimal primes are exactly the primes of height zero).
Every maximal ideal of an affine domain has height equal to the full dimension (Maximal ideals of an affine domain have full height).
For primes in an affine domain, (Transcendence degrees along affine prime quotients add correctly).
For every prime of an affine domain, (The dimension formula for affine domains).
Proof
By [L1], .
For each , the quotient is a domain and the chain is saturated, so there is no prime strictly between and in . Hence . Applying [L3] to and comparing [L4] at and gives .
Starting from step 1.1 and iterating step 2.1, we obtain . By [L2], . Therefore the saturated chain has length , and every such chain has the same length.
Why the equal-chain statement stops at affine domains
Remark
The preceding corollary is deliberately stated only for finite-type domains over a field. That is the catenary range supplied by the page's affine-dimension package.
Outside that range, arbitrary Noetherian rings need not be catenary. In particular, one cannot promote the equal-length conclusion for saturated prime chains to a general theorem on this page without adding genuinely new hypotheses and proofs.
5 · Examples, counterexamples and false statements
None yet.
Sources
- J. S. Milne, A Primer of Commutative Algebra, v4.03, §21
- The Stacks Project, Section 10.60: Dimension
- Melvin Hochster, Dimension theory and systems of parameters
- Allen B. Altman and Steven L. Kleiman, A Term of Commutative Algebra, 13th ed., §21
- J. S. Milne, A Primer of Commutative Algebra, v4.03, §§18, 21
- The Stacks Project, Section 10.116: Dimension of finite type algebras over fields, reprise
- The Stacks Project, Section 10.105: Catenary rings