[throwaway1967]

I'm sorry, but that still does not explain why the Romans created more gendered 3rd declension variants when using the neuters from the 2nd and 3rd declensions would have made more sense (since the things they were naming were sexless, like rivers and trees and other geographical things) while they would nevertheless have retained the "checksum" and "free word ordering" benefits from a language that declines its nouns.

For example, Finnish is utterly genderless, but it has more declension cases for nouns than Latin had. And so, Finnish affords itself the ability to have free word order and checksum benefits while doing entirely away with the concept of gender.

So I reject your explanation on the basis that it is clear to me that the declension of nouns into cases is what is responsible for the checksum and word order benefits, and not whether a language has grammatical gender.

[javert]

> For example, Finnish is utterly genderless, but it has more declension cases for nouns than Latin had.

That's consistent with simply making a tradeoff. So a Roman could say:

I don't understand why the Finns created more declension cases when they could have just used gender.

[dllthomas]

The free word order seems to come predominately from declension rather than gender, but I don't see how declension provides much in the way of a checksum.