Proof Noetherian Ring: $M^2 \ne M$

  • Context: MHB 
  • Thread starter Thread starter Amer
  • Start date Start date
  • Tags Tags
    Ring
Click For Summary

Discussion Overview

The discussion revolves around proving that for a Noetherian ring \( R \) with a maximal ideal \( M \), it holds that \( M^2 \neq M \). The scope includes mathematical reasoning and proof verification.

Discussion Character

  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • One participant presents a proof attempt, arguing that if \( M^2 = M \), then an element \( a_1 \) in \( M \) can be expressed as a sum of products of other generators of \( M \), leading to a contradiction if \( 1 \in M \).
  • Another participant questions whether \( R \) is assumed to be commutative, which is confirmed by the original poster.
  • A third participant challenges the correctness of the proof, suggesting that if \( M^2 = M \), Nakayama's lemma implies the existence of an idempotent element \( e \in M \), leading to \( M = (e) \).

Areas of Agreement / Disagreement

Participants do not appear to agree on the validity of the initial proof attempt, with one participant asserting its incorrectness and offering an alternative perspective based on Nakayama's lemma.

Contextual Notes

The discussion highlights the need for clarity on the assumptions regarding the commutativity of the ring and the implications of finitely generated ideals in the context of Noetherian rings.

Amer
Messages
259
Reaction score
0
Can you check my Work or mention a link for a proof.
Let $R$ be Noetherian ring. Then if $M$ is a maximal ideal in $R$. Prove that $M^2 \ne M$.

Proof:
Since $R$ is Noetherian ring then $M$ is finitely generated. Thus $M = (a_1, a_2 , \cdots, a_k)$ we can choose the $a_i's $ which are minimal in the sense that $M$ can't be generated if we ignore one of the $a_i's$. Also that means one $a_i$ can be generated by other $a's $.

Suppose on the contrary that $M^2 = M$. Hence $a_1 \in M^2 $. But elements of $M^2$ has the form $\displaystyle \sum_{i=1}^{n} k_i a_j a_h$ for some $n$
Therefore
$\displaystyle a_1 = \sum_{i=1}^{n} k_i a_j a_h $ the right hand side is divisible by $a_1$ since the left hand side is. and since $a_1$ can't be generated by other $a_i's$ that means $a_1$ should appear on terms of the right hand side moving these terms to the left and factoring $a_1$ we will get someone like this

$\displaystyle a_1 ( 1 - \sum_{i=1}^{n} c_i a_i) = \sum_{i=1}^{m} k_i a_j a_h ,~~~ i, j \ne 1$
I think the right hand side is equal to zero if it is then we will get $1 = \sum_{i=1}^{n} c_i a_i $, it follows that $1 \in M$ a contradiction. But is it zero ?. if it is not then how i can prove it.

Thanks
 
Physics news on Phys.org
Hi Amer, is it assumed that $R$ is a commutative?
 
Euge said:
Hi Amer, is it assumed that $R$ is a commutative?

yea commutative with 1
 
Ok. The proof you've provided is incorrect. Here's something to keep in mind. If $M^2 = M$, then since $M$ is finitely generated, Nakayama's lemma gives an $e \in M$ for which $m = em$ for all $m \in M$. In particular, $e^2 = e$, i.e., $e$ is idempotent. Furthermore, $M = (e)$.
 

Similar threads

  • · Replies 19 ·
Replies
19
Views
3K
Replies
21
Views
3K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 10 ·
Replies
10
Views
2K
Replies
2
Views
1K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 8 ·
Replies
8
Views
3K