1
$\begingroup$

I had asked this question on math.stackexchange.com, but I have not received a response. Hoping to get some help here.

I am trying to prove this result and I am stuck at one step. Let $(R,m)$ be a Noetherian local ring. Let $P$ be a prime ideal. Define the n-th symbolic power of $P$ as $P^{(n)}={{r\in R: \exists y \in R\setminus P}}$ s.t. $yr \in P^n$.

I want to show $P^{(n)}=P^n:m^{\infty}$.

Let $r\in P^n:m^{\infty}$. Then, $\exists t$ s.t. $rm^t\in P^n$. Now, $m^t \subsetneq P^n$, else by taking radicals, we have $m=P$. Let $y\in m^t \setminus P$. Then, $yr\in P^n$, so that $r\in P^{(n)}$.

For the converse, let $r\in P^{(n)}$. Suppose, $y\in R\setminus P$ s.t. $ry\in P^n$. If $y\in R\setminus m$, then $y$ is a unit and $r\in P^n\subseteq P^{(n)}$.

I am not sure how to handle the case, where $y\in m\setminus P$. Is it true $yR$ is $m$-primary? This would imply $yR$ contains a power of $m$ and thus completing the proof.

$\endgroup$
8
  • $\begingroup$ No, $yR$ is not $m$-primary, since it has height at most one by the Principal Ideal Theorem. I'm pretty sure the result you're trying to prove is false, but can't see why immediately. $\endgroup$ Jul 6, 2011 at 18:00
  • $\begingroup$ Something similar is true, replacing the maximal ideal by the intersection of the non-minimal associated primes of $P^n$. See Examples 4.3 of these lecture notes of Herzog: cel.archives-ouvertes.fr/docs/00/37/46/20/PDF/CIMPAHerzog.pdf. In particular I think what you want is true in dimension $\leq 2$. $\endgroup$ Jul 6, 2011 at 18:13
  • $\begingroup$ @Graham: I guess I am missing some hypothesis then. In section 4, in the following paper, this is the formula used for calculating symbolic powers.people.reed.edu/~iswanson/topology.pdf Is the statement true if $R$ is regular, local? $\endgroup$ Jul 6, 2011 at 21:13
  • 1
    $\begingroup$ This is false for the following reason. $P^n:m^{\infty}/P^n$ is a module of finite length, since all elements are annihilated by a power of the maximal ideal. In general $P^{(n)}/P^n$ is not. To make an example, take an example of some prime ideal in a power series ring in $k$ variables such that $P^n\neq P^{(n)}$. Then consider the same inside a power series ring with one more variable. $\endgroup$
    – Mohan
    Jul 6, 2011 at 22:40
  • $\begingroup$ @Mohan: I am not sure I understand your comment. The author uses this formula in section 4 of this paper that I referenced above people.reed.edu/~iswanson/topology.pdf in a power series ring with 3 variables. $\endgroup$ Jul 7, 2011 at 0:11

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.