0
$\begingroup$

I saw the following exercise in the Kaplansky's book that is due to D. Lizard. Where can i find the main text for the proof of this exercise?

Let $P$ be a prime ideal of $R$, $I$ the ideal generated by all the idempotent in $P$. Prove that $R/I$ has no non trivial idempotent.(page 8 of kaplansky'book, exercise 6, commutative ring theory)

$\endgroup$
1
  • 3
    $\begingroup$ Is it "D. Lizard", or "D. Lazard"? $\endgroup$
    – Todd Trimble
    Aug 31, 2015 at 2:09

1 Answer 1

1
$\begingroup$

I think this is a proof if $R$ has $1$. Suppose $e + I$ is an idempotent in $R/I$. This means $e^2 \equiv e \pmod{I}$, so $e(1-e) \in I \subseteq P$. Since $P$ is prime, this implies that either $e \in P$ or $1 - e \in P$. Let's say $e \in P$. Since $e(1-e) \in I$, we may write $e(1-e) = \sum_{i=1}^n r_i e_i$, where each $e_i$ is an idempotent that is in $P$. Now, let $f = e \cdot \prod_{i=1}^n (1 - e_i)$. Then, since each $1 - e_i$ is an idempotent, we have $$f^2 = e^2 \cdot \prod_{i=1}^n (1 - e_i) = \left( e - \sum_{i=1}^n r_i e_i\right)\cdot \prod_{i=1}^n (1 - e_i) = e\cdot \prod_{i=1}^n (1 - e_i)= f, $$ so $f$ is an idempotent (and $f \in P$, because $e \in P$), so $f \in I$. Since $f \equiv e \pmod{I}$, this implies $e \in I$. So $e + I$ is zero in $R/I$.

$\endgroup$
1
  • $\begingroup$ More conceptually: if $i$ is an idempotent, any idempotent in $R/(i)$ lifts to an idempotent in $R$ (this is obvious when you think of idempotents as decompositions of the ring as a product). Any idempotent in $R/I$ is already idempotent in some $R/I_0$ where $I_0\subseteq I$ is generated by finitely many idempotents. Since the ideal generated by finitely many idempotents is just the ideal generated by their join, it follows that any idempotent in $R/I$ lifts to an idempotent in $R$. $\endgroup$ Aug 31, 2015 at 2:41

Your Answer

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

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