Does language research invalidate grammar instruction? Why can't students transfer what they have learned to their speaking and writing?

I hear the same two comments at English-language arts conferences all the time: 1. “I’ve heard that research has proven grammar, usage, mechanics, spelling, and vocabulary instruction doesn’t work.” 2. “I teach grammar and they seem to get it. They pass my tests and do okay on the standardized tests, but they don’t transfer the learning to their writing or speaking. And they just don’t retain what we’ve covered. Their next-year teacher always asks why I don’t teach grammar, usage, mechanics, and spelling.”

So, should we bother teaching grammar, usage, mechanics, and spelling? Some would say “No.” This is what Dr. Stephen Krashen recommends, at least until high school. Dr. Krashen finds that students learn grammar, usage, mechanics, spelling, and vocabulary most efficiently through free voluntary reading, not explicit instruction or even writing, as my old National Writing Project colleagues would advocate. Now, to be fair, Dr. Krashen does see the value of teaching some usage issues and grammatical terminology. And he advocates teaching students how to use language resources, such as language handbooks, to correct errors and improve writing style. But he, and others of his ilk, certainly support the overall position described in the first comment listed above. My view is that the collective jury is still out on this research question.

Irrespective of the research into the effectiveness of explicit grammar, usage, mechanics, and spelling instruction, the writers of the new Common Core State Standards (CCSS) certainly affirm the need for instruction in these skill and content areas. In fact, grammar, usage, mechanics, spelling, and vocabulary now have their own CCSS Language Strand in the English Language Arts Standards. Apparently, language instruction is back in style.