0
$\begingroup$

This seems fairly similar to density. Suppose I have three categories $A,B,C$, and a functor $L: B \to C$ so that every natural transformation $f: L.F \Rightarrow L.G$, for a parallel pair $F,G: A \to B$ belonging to some class of functors $\mathcal{F}$, factors as a natural transformation $g:F \Rightarrow G$ so that $L.g = f$. This feels like a thing that would come up in 2-category theory, but I'm struggling to find a good reference.

Here is a diagram using the quiver app

Improve this question
$\endgroup$
5
  • 1
    $\begingroup$ Could you clarify what your question is? It sounds like you're asking for $L$ to be fully faithful, so that postcomposition by $L$ induces an isomorphism between 2-cells $L \circ F \Rightarrow L \circ G$ and 2-cells $F \Rightarrow G$. $\endgroup$
    – varkor
    Apr 20, 2021 at 23:13
  • $\begingroup$ @varkor full faithfulness would mean a unique factorisation $\endgroup$
    – David Roberts ♦
    Apr 21, 2021 at 2:38
  • $\begingroup$ Yeah it’s a bit weaker than that - I have a 2-monad and three strict algebras of it. The morphism $B$ to $C$ acts like an inserter for every parallel pair of strict algebra homomorphisms. My example certainly isn’t a fully faithful functor - it’s actually a reflector. $\endgroup$
    – Ben MacAdam
    Apr 21, 2021 at 3:39
  • $\begingroup$ I think it is a particular case of an inserter, but without uniqueness of the 1-cell; the non-uniqueness instead suggests some sort of (weak) orthogonality condition. $\endgroup$
    – fosco
    Apr 21, 2021 at 5:58
  • $\begingroup$ If you don't want uniqueness of the factorisation, you can ask for $L$ just to be full. If it arises via a universal property (without uniqueness), then this is suggestive of a weak (2-)limit as @fosco points out. $\endgroup$
    – varkor
    Apr 21, 2021 at 10:58

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.