3
$\begingroup$

In an additive category, we have a zero object and every Hom-set is given the structure of an abelian group such that composition distributes over addition. That is, given $f \in \operatorname{Hom}_\mathcal{A}(A,B) \quad g,g' \in \operatorname{Hom}_\mathcal{A}(B,C) \;$ we have: $$ f(g+g') = fg + fg'$$

$\textbf{1.}$ Is the identity in each Hom-set the unique zero map? I can't find any axiom saying that it is and no reason as of yet that it should be.

$\textbf{2.}$ When defining an $exact$ functor (or left/right exact functor) we say that $F$ is an additive functor taking an exact sequence $$0 \rightarrow A \xrightarrow{f} B \xrightarrow{g} C \rightarrow 0$$
is taken to the exact sequence $$0 \rightarrow F(A) \xrightarrow{F(f)} F(B) \xrightarrow{F(g)} F(C) \rightarrow 0$$

Is this to say that $F(0)=0?$ I suppose that if the answer to question $\textbf{1.}$ is yes then since $F$ is additive then $F(0:0\rightarrow A) = 0:F(0) \rightarrow F(A)$ so that we can replace these $F(0)$s with $0$ anyway?

$\endgroup$
  • $\begingroup$ Actually, I've just seen this question, but im not sure that it actually answers the question. math.stackexchange.com/questions/335824/… Is it claiming there that composition is an abelian group too? $\endgroup$ – SEWillB May 9 '18 at 9:58
3
$\begingroup$

Nice question.

To answer 1., first notice that in any group (not necessarily abelian) the only idempotent is the identity, i.e. $x^2 = x$ if and only if $x = 1_G$, or in additive notation, $x + x = x$ if and only if $x = 0_G$. Thus, we'd like to show that $0_{AB}+0_{AB} = 0_{AB}$ for the zero morphism $0_{AB}\colon A\to B$.

First consider what exactly $0_{AB}$ is. It is given as the unique morphism that factors through zero object, i.e. it is the composition $A\to 0\to B$. So, we can write $0_{AB} = 0_{0B}\circ 0_{A0}$ and therefore we have $$0_{AB}+0_{AB} = 0_{0B}\circ 0_{A0} + 0_{0B}\circ 0_{A0} = (0_{0B}+0_{0B})\circ 0_{A0} = 0_{0B}\circ 0_{A0} = 0_{AB}$$

where $0_{0B}+0_{0B}=0_{0B}$ because $\operatorname{Hom}(0,B)$ is trivial group by the definition of zero object.

For 2., notice that $X$ is zero object if and only if $0_X = \operatorname{id}_X$. One direction is obvious. To prove that $0_X = \operatorname{id}_X$ implies that $X$ is initial (and analogously it is terminal), consider $f,g\colon X\to A$ for any object $A$. By 1. we know that $$ f = f\circ\operatorname{id}_X = f\circ 0_X = g\circ 0_X = g\circ \operatorname{id}_X = g,$$ and we are done. To conclude that $F0 \cong 0$, observe that additive $F$ preserves both zero and identity morphism.

$\endgroup$
  • $\begingroup$ @SEWillB, just a comment on your last sentence. That argument doesn't work since not all objects of $F$'s codomain are necessarily of the form $FA$ or isomorphic to it. And even if so, that still doesn't mean that there are no other maps between $F0$ and $FA$ besides zero map. $\endgroup$ – Ennar May 9 '18 at 14:15

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.