Wouldn’t we see more regularity in the structure of languages then? English and classical Latin are almost opposites by every measure I can think to apply to a language (complexity of grammar, diversity of vocabulary and idiom, etc.). This doesn’t seem like a good assumption.
English and Latin aren’t even anywhere near as different as two natural languages can be. Take a look at this for a quick example, and take a look at the Language Construction Kit (it’s about constructed languages, but AFAICT most of the things exemplified aren’t completely unheard-of among natural languages) for a lot more.
(There are quite a few linguistic universals, but I’m not entirely convinced that all of them exist because a language flouting one of them would be unlearnable by humans, rather than (say) because they were inherited from a common ancestor.)
Wouldn’t we see more regularity in the structure of languages then? English and classical Latin are almost opposites by every measure I can think to apply to a language (complexity of grammar, diversity of vocabulary and idiom, etc.). This doesn’t seem like a good assumption.
English and Latin aren’t even anywhere near as different as two natural languages can be. Take a look at this for a quick example, and take a look at the Language Construction Kit (it’s about constructed languages, but AFAICT most of the things exemplified aren’t completely unheard-of among natural languages) for a lot more.
(There are quite a few linguistic universals, but I’m not entirely convinced that all of them exist because a language flouting one of them would be unlearnable by humans, rather than (say) because they were inherited from a common ancestor.)
There doesn’t have to be one solution to “memetically selected for learnabillity”