How Norvig refutes Chomsky

In a well-written and elaborated post, Peter Norvig reputed Noam Chomsky's arguments against statistical models. (Links: I, II)

"Then, to get language from this abstract, eternal, mathematical realm into the heads of people, he [Chomsky] must fabricate a mystical facility that is exactly tuned to the eternal realm. This may be very interesting from a mathematical point of view, but it misses the point about what language is, and how it works."

On speech recognition, I cannot help but think Norvig is right.  There are many times introduction of linguistic knowledge proven to be not as helpful as we hope.   If you are interested in the topic, you should also read Jelinek's piece on Some of my Best Friends are Linguists.

Arthur