Chomsky’s universal grammar: There’s not enough language data for children to learn languages in the absence of inductive biases.
I think there’s more recent work in computational linguistics that challenges this. Unfortunately I can’t summarize it since I only took an overview course a long time ago. I’ve been wondering whether I should read up on language evolution at some point. Mostly because it seems really interesting, but also because it’s a field I haven’t seen being discussed in EA circles, and it seems potentially useful to have this background when it comes to evaluating/interpreting AI milestones and so on. In any case, if someone understands computational linguistics, language evolution and how it relates to the nativism debate, I’d be extremely interested in a summary!
I think there’s more recent work in computational linguistics that challenges this. Unfortunately I can’t summarize it since I only took an overview course a long time ago. I’ve been wondering whether I should read up on language evolution at some point. Mostly because it seems really interesting, but also because it’s a field I haven’t seen being discussed in EA circles, and it seems potentially useful to have this background when it comes to evaluating/interpreting AI milestones and so on. In any case, if someone understands computational linguistics, language evolution and how it relates to the nativism debate, I’d be extremely interested in a summary!