Categories and set-theoretic quantifiers

76 Views Asked by At

I'm asking about the quantifiers such as $\forall, \exists,$ etc. Is it aapropriate to use these concerning categories and objects/morphisms within them? For example, is it common to write this way:

$f$ in $Hom(A,B)$ is a epimorphisms $\Leftrightarrow \forall$ objects $Z \ \forall g_1,g_2$ in $Hom(B,Z) \ \ g_1f = g_2f \Rightarrow g_1 = g_2$ I know it's probably doesn't work for $\in$ if $Hom(X,Y)$ is not a set so I've typed "in" instead of the quantifier. Though, correct me you can actually use $\in$.

1

There are 1 best solutions below

0
On BEST ANSWER

Two points.

(1) You don't need "$\in$" to express the claim here. You don't need "in" either! The use of set- or class- or other collection-talk here is in fact superfluous. Just say

An $f$ which is an arrow with source $A$ and target $B$ is epic if and only if, for any object $Z$, if $g_1$ and $g_2$ are both arrows with source $B$ and target $Z$, etc. etc.

That makes perfectly good sense even if there are two many arrows with source $B$ and target $Z$ to be a kosher set, and so $Hom(B,Z)$ is, as they say, a proper class.

(2) But having made the point, we do in fact conventionally allow ourselves to overload notation and write things of the form $x \in C$ even if $C$ is a proper class so that $\in$ can't still express set-membership, and indeed the occurrences need to be translated away. It's convenient (see e.g. Kunen's famed set theory book) even if the really pernickety-minded might feel uncomfortable.