Um, let me note that I don't buy everything Kuhn says. Rather, I think he demonstrated that descriptive theories of reference entail a non-realist view of science. This, anyway, is an interesting topic; as you may know, I've done some work on reference.