Want to take part in these discussions? Sign in if you have an account, or apply for one below
Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.
I just discovered that, all along, the term “quiver representation” was just redirecting to representation. Have started this dedicated page now, with the bare minimum
If a quiver representation is a functor from then doesn’t this mean it is equivalently a graph map ? I.e. you associate to each node in the quiver a vector space and to each edge a linear map. I couldn’t workout what a morphism of quiver representations would be from this point of view however.
Yes, that’s what it means for to be the free category on a graph. The general mechanism is explained at free functor.
Right. So why do representation theorists present quiver representations using the free category if this point of view is simpler?
The mechanism explained in the “free functor” article explains the 1-categorical case well, but we are working in a 2-category Cat here. We would need a biadjunction (really a strict 2-adjunction) to transfer morphisms across the view point. It’s not immediately obvious to me that the free category is biadjoint to the forgetful functor.
I don’t think the 1-category of quivers has enough info to get morphisms of quiver representations. I would assume there is some 2-category of quivers to do that, but I have no idea what the 2-cells ought to be.
And just in case it wasn’t clear, my point is that there can be multiple morphisms of quiver representations since we are asking for natural transformations not equivalences, but not multiple 2-cells of graph maps. It seems a better 2-category (probably bicategory) of graphs is needed to state this POV correctly.
No, it’s just the 1-category of categories and functors between them that is used. There is nothing subtle going on here.
1 to 6 of 6