Grammar Instruction That Actually Improves Writing

This is not what most language arts teachers want to hear: fifty years of research on isolated grammar instruction — diagramming sentences, worksheet exercises, fill-in-the-blank drills — shows that it does not improve student writing. Not a little. Not for some students. Across studies and methodologies, isolated grammar instruction produces near-zero transfer to writing quality.

This finding has been known since George Hillocks' 1986 meta-analysis and confirmed by subsequent research. Yet isolated grammar instruction remains ubiquitous because we don't have a good alternative pedagogy and because grammar instruction feels productive — students complete worksheets, teachers can grade them, skills appear to be learned.

The skills are learned. They just don't transfer to actual writing.

Why Isolated Grammar Fails

The transfer problem is fundamental: grammar knowledge (knowing rules) and grammar fluency (applying those rules in writing) are different cognitive skills. A student who can correctly identify the subject and verb in a test sentence and a student who instinctively constructs clear subject-verb relationships in their own writing have learned different things.

Isolated grammar instruction teaches the first skill. Writing instruction requires the second. The research shows that teaching the first skill doesn't automatically produce the second.

What Actually Works

Three instructional approaches have solid research support for improving writing quality through grammar instruction:

Sentence combining. Students take short, choppy sentences and combine them into one effective sentence. This is different from fixing sentences — it's generating multiple options and evaluating effectiveness. Research dating to the 1970s shows consistent improvements in writing syntactic complexity and quality.

Example: "The dog ran. The dog was brown. The dog was running fast. It was toward the fence." → "The brown dog sprinted toward the fence." Students discover subordination, modification, and coordination through the act of combining, rather than through rule explanation.

Mentor sentence study. Students examine well-crafted sentences from published authors, identify what makes them effective, and imitate the structure with new content. Craft study — looking at how professional writers use grammar — builds tacit knowledge of effective constructions.

Jeff Anderson's "Everyday Editing" approach uses this method systematically: students find, discuss, imitate, and apply specific craft moves from published writing.

Grammar in the context of revision. When students revise their own writing, addressing specific grammatical choices in their own text — not in exercises — produces transfer. The key: the grammar instruction is tied to real writing decisions in real pieces.

"Look at your third paragraph. You have four sentences that all start the same way. What could you do to vary them?" This works. "Complete exercises 1-10 on sentence variety" doesn't.

A Practical Framework

Mini-lessons on specific craft moves. Rather than covering grammar comprehensively, focus on moves that appear in student writing. If students are struggling with run-ons, teach sentence boundaries in the context of revision, using their own writing. If students are writing in monotonous short sentences, teach sentence combining.

Mentor text analysis. Regularly share excellent sentences from literature, journalism, and student writing. Analyze them: what choices did the author make? What effect do those choices have? Imitate the structure.

Sentence combining practice. Regular sentence combining exercises (10 minutes, a few times per week) develop syntactic flexibility. The key is requiring multiple versions and discussing which works best and why.

Grammar in conference, not just in writing. When conferencing with students on their writing, address specific grammatical choices in their specific text. "What were you trying to do here? Could you try it this way?" makes grammar concrete and purposeful.

What to Do About Conventions

Conventions — spelling, punctuation, capitalization — are different from grammar in the sense above. They're rule-governed and more directly teachable, though still most effectively taught in the context of real writing.

Proofreading and editing are genuine skills worth teaching. The difference is:

• Grammar (sentence structure, syntax) → teach through sentence combining and mentor text, apply in revision
• Conventions (spelling, punctuation) → teach and practice explicitly, apply in editing

Both matter. Neither should dominate writing instruction.

The Tension With Mandated Curricula

Many teachers are required to use grammar curricula that take the isolated approach. If that's your situation, the research suggests:

• Spend the minimum time required on isolated grammar
• Add sentence combining and mentor text study
• Make sure revision conferences address grammar in the context of real writing
• Assess writing quality rather than grammar exercise performance

Students who write well will perform adequately on grammar assessments without extensive isolated instruction. Students who perform well on grammar assessments often don't write well.

The goal is writers who make sophisticated grammatical choices instinctively — not students who can identify parts of speech in sentences that aren't theirs.