Beautiful paper, relatively well written and accessible, too. I think everyone _knew_ in some sense that the structure of categorial information about vectors

Beautiful paper, relatively well written and accessible, too. I think everyone _knew_ in some sense that the structure of categorial information about vectors must be hierarchical and have this general kind of structure, but they managed to formalize that intuition into just a few theorems that seem sort of inevitable only in retrospect.