4 Equivalent Categories.
Suppose that R and S are two rings. We want to determine what it would mean for them to have the same representation theories — or for these theories to be equivalent in some meaningful way. What we want should be pretty clear — the two rings have equivalent representation theories when their module (= representation) categories RMod and SMod are “the same”. But that just trades one problem for another — what does it mean for two categories “to be the same”? This is a bit dicey. We’d probably like to say that they are the same if they are isomorphic. But that’s too strong. For example, the category of all nite sets, an enormous sucker, has exactly the same categorical structure as the category of all nite subsets of N, but there is no possible way to build a bijective function between these two categories. As we shall see, natural transformations will come to the rescue, and we will be able to determine an appropriate notion of equivalent categories and hence of equivalent representation theories for rings.
Let C and D be two (additive) categories. A covariant (additive) functor F : C →Dis an equivalence in case there exists a covariant (additive) functor G : D →Csuch that GF and FG are naturally isomorphic, respectively, to the identity functors 1C and 1D. If such an equivalence exists, then we say that the categories are equivalent. This relation is symmetric. Indeed, the two categories are equivalent i there exist (additive) covariant functors
F : C →D and G : D →C and natural isomorphisms