Pinker really goes all the way in this, bathing the reader in wonderful language, interesting ideas and good old fun and games. But the sad part is that his premise and conclusion--that language is an instinct--is a total and complete non sequitur.
Being a fan of Chomsky, Pinker submits to the notion (and a notion it is) that language and communication aren't necessarily related (as Chomsky (1975) said, "communication is only one function of language, and by no means an essential one"). Although Chomsky in recent years has done a lot to moderate his position, and a lot of research at least suggest that the world has come out of the post-skinnerian, anti-"blank slate" state in which it was in the seventies, when Chomsky reigned, Pinker upholds the sharp divide between grammar and usage. Why?
Because The Language Instinct isn't really about language. It's about completing Pinker's reductionist trilogy, consisting of this one, The Blank Slate, and How the Mind works. In The Language Instinct, Pinker doesn't analyze the facts and draws a valid conclusion. He simply tells us how convenient to his worldview it would be if language really was an instinct. I believe that makes The Language Instinct theology (or at best, philosophy) and not science.
Still, this book is a fine introduction to chomskyan grammar, X-bars and the like. Plus it's fun. But scientifically, it lacks stringency, humility and honesty. The book is filled with thin case studies that could mean the "instinct hypothesis" is correct or wrong, depending on your interpretation (of course Pinker chooses "correct"), and quote mining (the worst example being one in which Pinker gets the one name he's quoting wrong--twice!--plus, the book he's quoting is really about something else than what Pinker claims. The book in question is Heath (1983), Ways With Words.)
So, this book doesn't really tell us much about language, nor about language acquisition. What it does, however, is to educate us in Pinker's worldview, luckily for us in an entertaining manner. But need you learn of language, you must turn to someone who understands communication. Check out Pinker for laughs and a quick read, but if you want to learn something, I suggest Michael Tomasello, Jerome Bruner or Albert Bandura. They have what Pinker lack: an understanding of how complex human communication really is.