Even though Pinker has long been a fan of Strunk & White's and other style guides, he realized that they are often larded with proscriptions that exist simply because they have been passed down from earlier, different eras, rather than because they are based on any sound grammatical logic or understanding of linguistics (Strunk & White, he notes early on, “misdefined terms such as phrase, participle, and relative clause”). For example, Pinker argues, in many cases it’s perfectly finein fact, desirableto dangle participles, split infinitives, describe things in the passive voice, and engage in various other practices frequently frowned upon by our most authoritative style sources.
As a classically-trained translator of Latin, I have to disagree with this. Latin teaches and encourages proper English grammar better than most other languages. If you can read, understand, and compose Latin, you have a pretty damn good understanding of the "proper" rules of English composition.
Tradition is as important as modern usage when it comes to grammar. Just because I hear people all the time say, "Hey, ax d'man yer quest," doesn't mean any of that is correct. Readability and understandability are key, I agree. But, getting rid of certain rules leaves it open to discussion about the others. And, the day subject and verb agreement become inconsequential, I am giving up on the world...
However, with that said, I do think that dangling both participles and prepositions useful at certain times, as well as splitting infinitives and using the passive voice (which, honestly, I can't figure out the rule against in the first place... a voice is a voice and all have their use).
And his advice about how to avoid such writing is couched in cognitive-science theories that help him advise readers not just on how to write better, but on why certain decisions lead to smoother, easier-to-parse prose.
But, then he goes on to state that about "easier-to-parse prose." Wrong! The proper, traditional use of English makes the language far more able to be parsed than current language use. I know, I've done a ton of it and modern speech and writing is nearly impossible to cohesively parse without quite a bit of trouble.
Anyways, those are my thoughts, but whether you agree or not, please take the time to read the article. It is very interesting and worth at least one read through. p.s. I read it three times and look forward to a fourth perusal.