The 90-Second Sentence Expand That Gets Kids Writing

Welcome to the show -- Colin, I wanna start with two words I could put on my board tomorrow at 8:05 a.m.: “The cat purrs.” [questioning tone] Two words and a verb -- well, three words, technically -- and that’s the whole routine? [warmly] Pretty much, yes. It’s called Sentence Expanding, and it is a 90-second writing micro-routine where you put a kernel sentence on the board and students grow it by answering which, what, where, when, and why. So “The cat purrs” becomes something like, “The orange cat is sleeping on the couch in the afternoon because he is tired.” Not precious. Not a whole writing workshop. Just... go. “Which, what, where, when, and why” is the bit I like there. Five handles. It gives students somewhere to put their attention instead of just saying, vaguely, “add details.” Exactly. And I want to name the source right away so people can actually find it later: Jennifer Gonzalez wrote about this in Cult of Pedagogy on March 30, 2026, in a piece called “8 Ways to Squeeze Writing Instruction Into a Few Minutes.” She interviews Melanie Meehan and Maggie Roberts, and this comes out of their 2026 book Foundational Skills for Writing: A Brain-Based Guide to Strengthen Executive Functions, Language, and Other Cornerstones for Writers. March 30, 2026 -- lovely, specific, searchable. And Meehan and Roberts are doing something clever there: they’re shrinking writing instruction down to a unit small enough that a teacher can actually use it on a Tuesday, not just admire it in a PD slideshow. [laughs] Yes, because teachers do not need one more “transform your classroom” pitch. We need something that survives 3rd period. And this one might, because the hidden win is that it gets reluctant writers moving without freezing them on grammar terms. If I say, “Identify the subject and predicate, then expand with modifiers,” I lose kids immediately. If I say, “Who’s the doer? What are they doing?” -- which is Meehan’s plainer language -- now I’ve got them. “Doer” and “doing” instead of “subject” and “predicate” -- that’s not dumbing it down, that’s reducing the entry cost. You’re delaying jargon until it’s useful. Right. And that matters with middle schoolers, because the second they feel they might be wrong in public, some of them just... stop. I had a kid this year -- sweet kid, smart kid -- who could tell me out loud exactly what was happening in a chapter, but if I asked him to “write a sentence with an introductory phrase,” he looked like I’d handed him a tax form. But if the board says “The army marched,” and I say, “Okay, which army? Where? When? Why?” he can do that. [deadpan] Few twelve-year-olds wake up eager to produce an introductory phrase before first break Shocking, I know. [chuckles] But that’s the tension for me: from the outside, Sentence Expanding can look like cute little sentence play. Like, aw, they’re adding adjectives. But in a real classroom, it is often a launch ramp. It gets the pencil moving. It gives students a way in before the big scary stuff like drafting an argument or writing a paragraph about symbolism. And because the kernel is so stripped down -- “The cat purrs,” “The plant grows,” “The army marched” -- there’s almost no decision fatigue at the start. The student isn’t choosing a topic, an audience, a structure, a claim, and a tone all at once. They’re just answering one of five questions. Yes. One move at a time. Put a visible timer up for 30 seconds and the room changes, because now it feels brief and doable. You’re not saying, “Write for ten minutes.” You’re saying, “Can you make this sentence a little bigger before the timer hits zero?” Most kids will at least TRY that. The visible timer is a useful detail. Thirty seconds feels like a sprint, not an ordeal. And if they only get halfway there? Fine. You still got language on the page. That is not nothing. Especially with students who’ve started to think of themselves as “bad at writing,” which, honestly, is often code for “I get overwhelmed fast.” So here’s the mechanism underneath it. Meehan and Roberts argue that the real issue is cognitive load, not content. In other words, many students do have ideas, but sentence production itself is consuming so much working memory that there’s not much left for argument, organization, or idea development. If you automate sentence-level moves through brief, repeated practice, you free up mental bandwidth for the larger thinking. “Cognitive load” can sound abstract, so let me translate that into classroom reality: the kid is not empty-headed. The kid is busy. Busy trying to remember how to start, where the period goes, whether this sounds dumb, what word comes next, whether they’re allowed to say it that way... Exactly. Working memory is crowded. And when it’s crowded, higher-order tasks get squeezed out first. So the promise of a micro-routine is not “write a masterpiece in 90 seconds.” It’s “make one tiny part more automatic.” Which is why I like the actual routine because it is so concrete. Thirty seconds to expand the kernel. Then 60 seconds to hear three volunteer versions out loud. Then 30 seconds to rearrange one sentence two different ways so students hear that meaning holds even when word order shifts. That last 30 seconds -- the rearranging -- is the sneaky bit. Because now you’re teaching flexibility, not just elaboration. Students hear, oh, “In the afternoon, the orange cat is sleeping on the couch because he is tired” still carries the same basic meaning as the original expanded version. Yes! And that matters so much for kids who think there is one correct school sentence and if they don’t hit it exactly, they’re done. Hearing two versions breaks that spell a little. Let me try to say it back. The aim isn’t grammar drills for their own sake. It’s repeated sentence rehearsal so that later, when students need to write an explanation or an argument, sentence construction isn’t eating the whole engine. [responds quickly] That’s it. And tomorrow-morning practical? In my 3rd-period middle school ELA class, I’d pull the kernel from the text we’re already reading. Not an extra thing. If the chapter gives me a clean little action, I use that. Three students share, they sit back down, we move on. Ninety seconds, maybe two minutes if someone gets chatty. “Three students share” is important. Three, not thirteen. Otherwise your 90-second routine becomes a 12-minute detour with opinions. [laughs] Correct. Teachers know that trap in our bones. Keep it tight. If you teach science, use “The plant grows.” If you teach history, use “The army marched.” Same structure, different content. And because the kernel comes from your lesson, the writing practice is attached to knowledge students are already building. That content link matters. “The plant grows” can become roots, sunlight, soil, season, purpose. “The army marched” can pull in location, time period, weather, motive, destination. You’re not pausing the curriculum; you’re compressing writing into it. And I would really underline this: use it daily for two weeks before you judge it. Not once on a Friday when everyone’s weird, then declare it didn’t work. Two weeks. Enough repetitions for the routine itself to become familiar. Fourteen days is a fair test. Day one measures novelty. Day fourteen starts to show whether fluency is changing. And if you’re listening and thinking, “Okay but I do not have time to invent 10 good kernel sentences tonight,” fine -- outsource the boring part. Paste this prompt into ChatGPT, Claude, MagicSchool, or Brisk: “Generate 10 kernel sentences for a 7th grade [subject] class on [topic]. Each kernel should be three to five words and use one concrete subject and one concrete verb. After each kernel, list five guiding questions in plain student language -- which, what, where, when, why -- a student can answer to expand the sentence to about 15 to 20 words. Format as a two-column table I can paste into Google Slides.” “Three to five words,” “one concrete subject,” “one concrete verb,” “two-column table” -- that level of specificity is why the prompt will actually produce something usable instead of AI soup. Exactly. Don’t ask the robot for inspiration. Ask it for prep help. Then tomorrow morning, put one tiny sentence on the board and see what happens. [reflective] Because sometimes better writing instruction doesn’t begin with a grand unit plan. It begins with three words, a timer, and a student realizing they can make a sentence bigger than they thought.