6
$\begingroup$

Fix a universe $\mathcal{U}$. Call a category $\mathcal{U}$-complete if every diagram indexed by a $\mathcal{U}$-small category has a limit, and a functor $\mathcal{U}$-continuous if it preserves $\mathcal{U}$-small limits. Usually, when one fixes a universe, one calls this simply complete and continuous.

Now assume we are given $\mathcal{U}$-complete categories $C,D$ and a $\mathcal{U}$-continuous functor $F : C \to D$. Does then $F$ also preserve limits (which exist in $C$) which are not necessarily $\mathcal{U}$-small?

$\endgroup$

2 Answers 2

4
$\begingroup$

No. I'll answer for the case of colimits as follows: Consider for example the category $O$ of ordinals (as a poset), and adjoin a terminal object $T$, making a larger category $C$. Then this terminal object is a (large) colimit over the diagram $O\to C$. However, a cocontinuous functor $C\to D$ can send $T$ anywhere that admits a cone under the diagram $O\to D$, not necessarily its colimit (if the colimit even exists).

$\endgroup$
1
  • $\begingroup$ Right! Thank you $\endgroup$ Feb 17, 2011 at 16:07
4
$\begingroup$

The answer is yes if you put in some additional smallness/accessibility/presentability assumptions. For example, if $C$ is the $U$-cocompletion of a $U$-small subcategory, each of whose objects are ($U$-small)-presentable, then any $U$-cocontinuous functor is a left adjoint, and hence absolutely cocontinuous.

$\endgroup$
3
  • $\begingroup$ Interesting! Actually this situation seems to arise very often ... $\endgroup$ Feb 17, 2011 at 18:40
  • $\begingroup$ @Martin: Yes, I think it's the standard situation, and Owen's answer is the standard counterexample to the more general claim. In other news, I'm only recently a big fan of presentable/accessible categories, but I have become a big fan of them, so I hope you don't mind that I answer most of your questions with reference to them. $\endgroup$ Feb 17, 2011 at 23:15
  • $\begingroup$ Au contraire, I like your answers. $\endgroup$ Feb 18, 2011 at 8:56

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.