I think the whole point of that part of the article is that the only answer that could satisfice O'Reilly and viewers is that of religion, and Norvig says Chomsky has a philosophy "(some would say religious belief)" i.e. some unscientific belief that "language should be simple and understandable", which is balderdash to claim that is a religious viewpoint, in my opinion.
There are several ways one could model language, from a top down purely statistical approach that Norvig likes, something in the middle which Chomsky proposes, to a bottom up neural model of chemical interactions. There are advantages and disadvantages to each method for many different reasons.
> There are several ways one could model language, from a top down purely statistical approach that Norvig likes, something in the middle which Chomsky proposes, to a bottom up neural model of chemical interactions.
Yeah, I was just trying to take a step back (and maybe I was too OT, I agree), but at some point we should start asking ourselves more fundamental questions. Anyway, this discussion is way over my head, I'm just glad that HN users think there's an answer for everything, is like Godel or Kant have never written anything in their entire lives.
There are several ways one could model language, from a top down purely statistical approach that Norvig likes, something in the middle which Chomsky proposes, to a bottom up neural model of chemical interactions. There are advantages and disadvantages to each method for many different reasons.