Bourbaki

Lemma 1

Tag 0CSO, permanent. Cite it as [Bourbaki, Tag 0CSO].

Commutative Algebra, chapter VIII, § 7. Multiplicités

Let ρ:A→B\rho : A \to B be a ring homomorphism, and let pp be a prime ideal of A. Consider the following two properties: (i) the canonical homomorphism ρ~\tilde{\rho} of ApA_p into Ap⊗ABA_p \otimes_A B is bijective; (ii) there exists a single prime ideal rr of B lying over pp and the canonical homomorphism ρp\rho_p of ApA_p into BrB_r is bijective. One has (i) ⇒\Rightarrow (ii). If pp is minimal, or if B is integral over A, one has (i) ⇔\Leftrightarrow (ii) 1^1.

1 This lemma remains valid when the ring A is not noetherian.

The ring Ap⊗ABA_p \otimes_A B is identified with the ring of fractions S−1BS^{-1}B of BB defined by the multiplicative subset S=ρ(A−p)S = \rho(A - p) of BB. The prime ideals of S−1BS^{-1}B are therefore the S−1qS^{-1}q, where qq is a prime ideal of BB such that ρ−1(q)⊂p\rho^{-1}(q) \subset p; if qq is such an ideal, (S−1B)S−1q(S^{-1}B)_{S^{-1}q} is identified with BqB_q (II, § 2, No. 5, prop. 11).

If condition (i) is satisfied, there exists (V, § 2, No. 1, Lemma 1) a unique prime ideal rr of BB such that ρ−1(r)=p\rho^{-1}(r) = p. Moreover, BrB_r is identified with the ring of fractions (S−1B)S−1r(S^{-1}B)_{S^{-1}r}, hence also with (Ap)s(A_p)_s, where ss is the inverse image of S−1rS^{-1}r by the isomorphism ρ~:Ap→S−1B\tilde{\rho}: A_p \to S^{-1}B; now ρ~−1(S−1r)=(A−p)−1p=pAp\tilde{\rho}^{-1}(S^{-1}r) = (A - p)^{-1}p = pA_p, whence (ii).

Conversely, suppose that (ii) is satisfied, and let rr be the unique prime ideal of BB lying over pp. Since (S−1B)S−1r(S^{-1}B)_{S^{-1}r} is identified with BrB_r, it is enough to prove that S−1BS^{-1}B is local with maximal ideal S−1rS^{-1}r, that is to say, that every prime ideal qq of BB such that ρ−1(q)⊂p\rho^{-1}(q) \subset p is contained in rr. If pp is minimal, one has ρ−1(q)=p\rho^{-1}(q) = p, hence q=rq = r. If BB is integral over AA, there exists, by V, § 2, no. 1, cor. 2 to th. 1, a prime ideal r′r' of BB such that q⊂r′q \subset r' and ρ−1(r′)=p\rho^{-1}(r') = p; necessarily one has r′=rr' = r, hence q⊂rq \subset r.

Bourbaki puts the proof directly after the statement, so what is above is the statement and its proof together.

Cited by

Nothing.

Cites

Nothing.

Translated with machine help, not yet reviewed by a person.