8
$\begingroup$

Let $\mathcal{C}$ be an abelian category. In $\mathcal{C}$ we consider the diagram \begin{array}{ccc} A&&\\\ \downarrow&&\\\ C&\rightarrow&D \end{array} with arrows being monomorphism.

Is it possible to say when there is a $B$ such that we have a push-out square \begin{array}{ccc} A&\rightarrow&B\\\ \downarrow&&\downarrow\\\ C&\rightarrow&D \end{array} and if possible how to get $B$ (up to isomorphism)?

$\endgroup$

1 Answer 1

11
$\begingroup$

If there is such a pushout, then $B\to D$ is also a monomorphism, i.e. $B$ is a subobject of $D$.

Phrased more concretely, you're asking when there is a subobject $B$ of $D$ such that $B+C = D$ and $B\cap C = A$.

Let us mod out $D$ by $A$: we get a monomorphism $C/A\to D/A$, and a pushout square

$\require{AMScd}\begin{CD} C @>>> D \\ @VVV @VVV \\ C/A @>>> D/A\end{CD}$

which we can glue to the previous one to get a pushout square

$\require{AMScd}\begin{CD} A @>>> B \\ @VVV @VVV \\ C/A @>>> D/A\end{CD}$

This means that $D/A\cong C/A\oplus B/A$

Conversely, assume $C/A \to D/A$ has a summand $X$, and consider its pullback $B$ to $D$. Then clearly $B\cap C = A$ and $B+C = D$.

So there is such a $B$ if and only if $C/A$ has a summand in $D/A$; and in fact if you're more careful with the above constructions, it seems like

$\{$ subobjects $B\to D$ realizing such a pushout $\}\to \{$ summands of $C/A$ in $D/A \}$

is a bijection, with inverse given by the pullback; this can be seen as a relativization of the claim that subjects of $D$ containing $A$ are the same thing as subobjects of $D/A$.

$\endgroup$

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.