1
$\begingroup$

I'm reading the proof of Prop. I-18 in Eisenbud-Harris The Geometry of Schemes and I'm unsure of a claim they make. Let me just give you some necessary info:

Let $R$ be a commutative unital ring and let $f_1, \dots, f_n$ be elements of $R$ which generate the whole ring i.e. $(f_1, \dots, f_n) = R$. Let $g\in R$ and suppose there exist nonnegative integers $k_i (1\leq i \leq n)$ with $f_i^{k_i} g = 0$. Let $N$ be the maximum of these integers, so that $f_i^N g = 0$ for all $i$. The authors seem to make the claim that the ideal $(f_1^N, \dots, f_n^N)$ contains some power of the ideal $(f_1, \dots, f_n)=R$, but it is not clear to me why it should. Wouldn't any power of $(f_1, \dots, f_n)$ contain certain mixed products of the $f_i$ not divisible by any $f_i^N$?

Here is the (near) original statement which I have just altered slightly to replace $g=h$ to $g=0$ (you can find the original on page 19 here):

This implies that $g$ is annihilated by a power of the ideal generated by all of the $f_a^N$ for some $N$. But this ideal contains a power of the ideal generated by all the $f_a$, which is the unit ideal. Hence $g=0$ in $R$.

$\endgroup$
0
1
$\begingroup$

The power $P=nN-n+1$ answers the question: indeed if all powers but one, say $k_i$, are $<N$, i.e. $k_j\le N-1$, then $$k_i=P-\!\!\!\sum_{\substack{1\le j\le n\\j\ne i}}k_j\ge nN-n+1-(n-1)(N-1)=N.$$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.