1
$\begingroup$

As the title says, I want to show that when two principal ideals are equal in $R/\langle p^k\rangle,$ where $R$ is a principal ideal domain and $p\in R$ is a prime element, then their generators are associates.

So I started off by letting $\langle a+\langle p^k\rangle\rangle=\langle b+\langle p^k\rangle\rangle$ for some $a,b\in R.$ Then there exist $\langle r+\langle p^k\rangle\rangle$ and $\langle s+\langle p^k\rangle\rangle$ such that $$ar+\langle p^k\rangle=(r+\langle p^k\rangle)(a+\langle p^k\rangle)=b+\langle p^k\rangle$$ and $$bs+\langle p^k\rangle=a+\langle p^k\rangle.$$ Then $ar-b\in \langle p^k\rangle$ and $bs-a\in \langle p^k\rangle.$ But I have no idea how I should proceed from here, or if this is even the right approach.

I've also considered using the Correspondence Theorem for Rings. $\langle a+\langle p^k\rangle\rangle \lhd R/\langle p^k\rangle$ corresponds to the ideal $q^{-1}(\langle a+\langle p^k\rangle\rangle) \lhd R$ that contains $\langle p^k\rangle.$ ($q$ is the natural quotient map.) This ideal should be principal because $R$ is a PID. An ideal that contains $\langle p^k \rangle$ would be of the form $\langle p^i \rangle$ where $0\leq i \leq k.$ But again, I don't know how to proceed from here.

Any help would be much appreciated.

$\endgroup$
  • $\begingroup$ Do you want them to be associates in the quotient or in $R$? $\endgroup$ – rschwieb Jun 17 '16 at 15:48
  • $\begingroup$ I want them to be associates in the quotient. $\endgroup$ – dhk628 Jun 17 '16 at 16:45
  • $\begingroup$ OK, good then. That's what I used. $\endgroup$ – rschwieb Jun 17 '16 at 18:00
0
$\begingroup$

There is a more general version of this (if my interpretation of your question is right.)

If $a,b$ are nonzero elements in a local ring and generate the same right ideal, then they are right associates$^\ast$.

Proof: Write $a=bx$ and $b=ay$. Then $a=ayx$, and $a(1-yx)=0$. Since $R$ is local, one of $yx$ and $1-yx$ is a unit. In the latter case, that would imply $a=0$ with the earlier equation, so that is ruled out.

Instead, $yx$ is a unit. We would like to conclude that $x$ and $y$ are units, and fortunately that's true in a local ring. If $zyx=yxz=1$, then we can see $zy$ is a left inverse to $x$. We also necessarily have $(xzy)(xzy)=x(zyx)zy=xzy$, so that $xzy$ is idempotent. There are only two idempotents in a local ring: $0$ and $1$. It is clearly not $0$, so it must be $1$, and $x$ is a unit, and $y$ is a unit.

A quotient of a PID by a power of a prime ideal is local, so this argument applies here.

$^\ast$ We say $a$ is a right associate of $b$ if there exists a unit $u$ such that $au=b$.

$\endgroup$

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.