1
$\begingroup$

I was reading one of the possible definitions of semisimple modules, which is "for every submodule $N$ of $M$, there exists a submodule $P$ such that $M=N \bigoplus P$. After reading this I immediately thought that a particular case of semisimple modules are simple modules: if $M$ is simple then the only submodules are $0$ and $M$, and $M=0 \bigoplus M$, so the simple modules satisfy the conditions of the definition of semisimple modules.

However, I've also read that not all simple rings are semisimple rings, but a simple ring in particular is a simple module, so how can it be that it satisfied the definition of semisimple module but is not semisimple?

I am very confused by this, I would appreciate if someone could clear up my confusion, thanks in advance.

$\endgroup$
  • $\begingroup$ That definition of semisimple leaves out a rather important part: that the direct factors must be simple modules in itself. $\endgroup$ – Timbuc Oct 27 '14 at 5:24
  • 2
    $\begingroup$ @Timbuc This definition is equivalent to yours. $\endgroup$ – user98602 Oct 27 '14 at 5:34
5
$\begingroup$

Be careful. A simple ring $R$ is "A ring with no proper nontrivial two-sided ideals", but a simple module is "A module with no proper nontrivial submodules". A semisimple ring $R$ is "A ring that is semisimple as an $R$-module". Every ring that's not a division ring has left ideals (and thus is not simple as a module over itself). Simple rings do not need to be simple as modules over themselves. (For instance, consider $R = M_2(\Bbb Q)$.)

You are correct that every simple module is a semisimple module. But the notion of simple rings and simple modules don't coincide.

$\endgroup$
  • $\begingroup$ One last doubt: suppose the ring is simple but also commutative, then is it also semisimple seen as an R-module with the "natural" action on itself? $\endgroup$ – user16924 Oct 27 '14 at 5:38
  • 1
    $\begingroup$ @user16924 Yes; commutative simple rings are fields, and thus simple as modules over themselves. (Proof: let $a \in R$; then $Ra$ is a two-sided ideal, because $R$ is commutative; because $R$ is simple, $Ra = R$, and thus there's some $b$ with $ba=ab = 1$. So all nonzero elements have an inverse.) $\endgroup$ – user98602 Oct 27 '14 at 5:40

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.