[davidivadavid]

Uniqueness of names is definitely important — the question is can we do better?

If all we want is uniqueness, we could just number theorems and concepts using GUIDs.

Might as well abdicate and embrace the fact math is going to be automated, and all mathematical objects are just abstract constructs devoid of meaning that can be referred to by arbitrary labels.

But it seems like there's more to it.

The names of mathematicians are interesting for the genealogy of theorems... but they're also completely opaque about their semantics.

Is it not possible to think a system could make math more intuitive by relying on a more structured nomenclature?