5
$\begingroup$

Let $R$ be a commutative ring with unit. If $M$ and $N_i$ are arbitrary $R$-modules, the module $\operatorname{Hom}_R(M,\bigoplus_{i\in I}N_i)$ is not isomorphic to $\bigoplus_{i\in I}\operatorname{Hom}(M,N_i)$ in general. But if $M$ is finitely generated, does the isomorphism hold?

$\endgroup$
2
$\begingroup$

Define a homomorphism $\phi: \bigoplus_{i \in I} Hom(M,N_i) \rightarrow Hom(M, \bigoplus_{i \in I} N_i)$ by sending $f = (f_k)_{k \in I^*}$ to $\phi(f) = \sum_{k \in I^*} f_k$, where $I^*$ is a finite subset of $I$.

Clearly $\phi$ is injective. For surjectivity, take $f \in Hom(M, \bigoplus_{i \in I} N_i)$ and let $x_1,\dots,x_n$ be the generators of $M$. Then for $j \in \left\{1,\dots,n\right\}$, $f(x_j)$ will be non-zero in finitely many indices $I_j \subset I$. So the image of $f$ will be non-zero for at most finitely many indices $I^* = \bigcup_{j=1}^n I_j$. For each $k \in I^*$, $f$ induces a homomorphism $f_k : M \rightarrow N_k$, from which we can see that $f = \phi((f_k)_{k \in I^*})$.

Hence if $M$ is finitely generated, then $\phi$ is an isomorphism.

$\endgroup$
2
  • $\begingroup$ the proof is clear and right,thank you! $\endgroup$ – nick Dec 21 '14 at 9:03
  • $\begingroup$ How this $f_{k}:M \to N_{k}$ are defined? Im thinking about composing $f$ with a projection into $N_{k}$ but doesnt seems to work. Thanks! @Manos $\endgroup$ – Cos Mar 12 '20 at 1:13

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.