I’ve always been perplexed by this claim by some teachers: Teaching grammar doesn’t improve students’ writing. A problem with it is that it doesn’t define what is meant by “teaching grammar” nor what is meant by “improve students’ writing.” It seems to imply that they have looked at every conceivable way that grammar could be taught and worked with, and they found that none were effective.
When I questioned their basis for this belief, I was often directed to some studies in the 1970s and 80s. Typically, these studies started with students writing a paper. Then for a period of time, they worked on diagramming sentences, doing sentence-combination exercises, identifying parts of speech and completing some grammar worksheets. After this, they wrote another paper, and surprise, surprise, the writing in their essays hadn’t improved. From this, they concluded: teaching grammar doesn’t improve students’ writing.
On top of that, one researcher claimed that his students’ writing got worse, and somehow, he even knew that it was because the students had become obsessed “with avoiding error at all costs, to the point where fluency, content, and reasoning lost their importance.” *
Some teachers have pointed to this “research” as support for their justification to not work with grammar in their writing courses. Their philosophy tends to be: students improve their writing by writing.