Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Let $R$ be an Artinian Ring and suppose there exists $a,b\in R$ s.t. $I=aR=Rb$, then prove $I=bR=Ra$. (You may assume that a right Artinian Ring is Right Noetherian).

I've managed to get $Ra$, $bR$ contained in $I$ without using the fact that it's Artinian. The hint confuses me because I'm not sure how to make a useful ascending chain to use the fact that it's right Noetherian.

share|improve this question
    
Does the ring $R$ have unity? –  Giovanni De Gaetano Apr 26 '12 at 10:46
    
yes it does have unity –  Alice Apr 26 '12 at 12:22
    
@Arturo Maybe I´m missing something. But how can you conclude $a=bu$? It seems to me that you only have $a=ub$, and you need somehow to prove that $bR$ is a two-sided ideal to get the conclusion. (In specific the Artinianity hypothesis is missing or well hidden...) –  Giovanni De Gaetano Apr 26 '12 at 15:21
    
@Giovanni: I got some wires crossed. Sorry. –  Arturo Magidin Apr 26 '12 at 16:00
    
Before you go any further looking for ascending and decending chains, see if you can't make use of the maximal condition and/or minimal condition first. –  rschwieb Apr 28 '12 at 2:01

1 Answer 1

By symmetry, it's enough to show that $aR = bR$.

Since $R$ is right Artinian, it is right Noetherian. So $R$ has a finite length as a right $R$-module: it has a finite composition series. By the Jordan-Holder theorem, the length $l(M)$ of any composition series of any finite length module $M$ is well-defined (does not depend on the choice of composition series), and satisfies $l(M) = l(N) + l(M/N)$ whenever $N$ is a submodule of $M$.

Now $I = Rb = aR$ so $a \in Rb$ and $b \in aR$. Thus there are $x,y \in R$ such that $a = xb$ and $b = ay$. Hence $bR = ayR \subseteq aR$. Also $bu = 0$ implies $au = xbu = 0$, so the right annihilator $rann(b)$ of $b$ is contained in $rann(a)$.

Thus $l(bR) \leq l(aR)$ and $l(rann(b)) \leq l(rann(a))$.

By the first isomorphism theorem for modules, we know that $aR \cong R / rann(a)$ and $bR \cong R / rann(b)$, so

$l(aR) + l(rann(a)) = l(R) = l(bR) + l(rann(b))$.

Hence $0 \leq l(aR) - l(bR) = l(rann(b)) - l(rann(a)) \leq 0$.

Therefore $l(aR) = l(bR)$ and $aR = bR$.

share|improve this answer
    
Very nice! I thought about this some time ago and never considered annihilators. –  Manny Reyes Jan 5 at 12:15
    
Thanks. The hint given isn't very helpful! I wonder if it's true if $R$ is only assumed to be Noetherian; do you perhaps know a counterexample? –  Konstantin Ardakov Jan 5 at 12:18
    
Konstantin, I was finally able to think about your question. Much to my surprise, I believe that it may in fact be true for (right) Noetherian rings! In your notation above, notice that left multiplication by $x$ defines an endomorphism of the right module $I=aR$ because it is also a left ideal. This map is surjective since $a = xb = xay$. If $R$ is right Noetherian, then $aR$ Hopfian, whence this surjection is an automorphism. But notice that the submodule $bR \subseteq aR$ also maps onto the image $aR$ as $a = xb$. By injectivity, we must have $bR = aR$. Can it really be true? –  Manny Reyes Feb 19 at 19:03
    
Yes, I think it works. Great! –  Konstantin Ardakov Feb 19 at 23:53

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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