I'm asking this and other questions in an attempt to tackle something I'm having a hard time to understand from another point of view.
Let's consider a category $C$ with 4 objects, $a$, $b$, $c$, and $d$, and some morphisms such that
- $|Hom(a,c)| \ge 1$,
- $|Hom(b,d)| \ge 1$,
- there is a morphism $m$ in $Set$ such that $m: Hom(a,c) \to Hom(b,d)$.
What do the 3 hypothesis above imply as regards the other hom-sets in $C$?
My understanding is that, at the minumum, $|Hom(b,a)| \ge 1\ \wedge\ |Hom(c,d)| \ge 1$ must be true, but it's not required that the strict inequality holds, nor that the other hom-sets are not empty.
Furthermore, I believe that the concept of dual/opposite category need not be brought in to answer this question.
Is all the above correct?
Let's pause and observe that the third hypothesis follows from the second hypothesis (so it doesn't add anything). Assuming $\mathrm{Hom}(b,d)$ is non-empty, we can pick some $f\in \mathrm{Hom}(b,d)$ and define $m: \mathrm{Hom}(a,c) \to \mathrm{Hom}(b,d)$ to be the constant function with value $f$, $m(x) = f$.
Absolutely nothing.
Nope, that's not the case. And I'm very curious what your reasoning is - this seems like a total non sequitur!
For example, $C$ could have exactly six arrows: $\mathrm{id}_a$, $\mathrm{id}_b$, $\mathrm{id}_c$, $\mathrm{id}_d$, $f\colon a\to c$, and $g\colon b\to d$. Then $\mathrm{Hom}(b,a)$ and $\mathrm{Hom}(c,d)$ are both empty.