The children can read… now they
Everything below is one idea: reading is a road your children can now drive. Writing is the same road, driven the other way — and that direction is harder, for reasons you can see once someone draws them.
How to read this page
Each stop is one idea in plain words, plus one picture that shows the machinery — not a diagram of names, a diagram of what actually happens. Read the picture; the caption tells you what it is claiming.
Research is not all the same weight. Some of what follows has been measured many times, in many countries, with numbers attached. Some of it is a beautiful practice that has barely been tested. Both are useful. Pretending they are the same is how programmes go wrong. So every claim carries a tag:
Stop 01
To you and me, reading cat and spelling cat feel like one skill. To a five-year-old they are two very different jobs, and one of them is much harder.
When a child reads, the page does half the work. The letters are already there, already in order, already sitting still. The child's job is to turn three visible things into one sound he already knows. It is a collapse: three into one. And English mostly cooperates — a letter usually points at one sound.
When a child spells, the page gives him nothing. He has to hold the whole word in his head while it slowly leaks away, chop it into sounds, decide which letters make each sound, keep them in the right order, and move a pencil. It is a fan-out: one sound opens into several possible spellings, and nothing on the page tells him which one is right. The sound /k/ can be written c, k or ck. He must choose.
That is the whole reason encoding is a separate programme and not a bonus activity. Your readers taught the collapse. Nothing has yet taught the fan-out.
Almost every phonics programme sells the top half of that picture. The bottom half is where the second product lives — and it is not a nice-to-have: teaching the fan-out is one of the best-evidenced ways to make the top half faster too. Solid
Stop 02
Adults do not sound out ship. We see it and the word is simply there. Something happened that turned a word we had to work out into a word we just know. Linnea Ehri named that something orthographic mapping — which is a heavy phrase for a simple picture.
Her definition: it is "the formation of letter–sound connections to bond the spellings, pronunciations, and meanings of specific words in memory." In plain English: each letter-chunk gets tied by a little string to the sound it makes. Once enough strings are tied, the whole word lifts as one piece. It stops being three things and becomes one thing.
Here is the part that should make you sit up. Spelling the word is what ties the strings tight. Reading it helps. Writing it — producing every grapheme yourself, in order, from memory — helps more, because you cannot skip a single letter. Reading lets a child glide over the middle of a word. Spelling does not permit gliding.
And it is fast. Children have remembered letters in a novel word after roughly one to four good exposures, and still remembered them a month later. You are not looking at hundreds of repetitions. You are looking at a handful of correct, effortful ones.
Across 53 studies and about 6,000 students, formal spelling instruction improved spelling (effect size 0.54) — and it also improved reading (0.44) and hearing sounds in words (0.51), and the gains held up later (0.53). Teaching spelling is teaching reading. Solid
Stop 03
Virginia Berninger's model of writing is easier to feel than to read about. A writer needs two hands. One hand does the mechanics — spelling the word, forming the letter. The other hand does the ideas — what to say, which word is better, does this make sense. Both hands have to reach through the same narrow door at the same time: the child's working memory.
The door does not get wider. Nothing you do makes a five-year-old's working memory bigger. So there is only one move available: make the mechanics hand smaller, until it barely takes up room, and the ideas hand gets everything that is left.
This is why a child who can talk in wonderful long sentences writes "I lik cat." He is not short of ideas. His ideas are being crushed by the effort of remembering how to make a lowercase g. Ask him to tell you the story instead and you get four sentences with a joke in them. The bottleneck is the pencil, not the imagination.
It also explains the counter-intuitive thing: drilling handwriting and spelling makes children more creative, not less. That is not a slogan. Handwriting instruction has been measured to improve not just legibility (0.59) and speed (0.63), but the quality of what children write (0.84) and how much they write (1.33). You are not stealing time from creativity. You are buying it.
Berninger's model itself is a model — strongly supported by how well transcription predicts writing quality in young children, and extended by Young-Suk Kim's more recent work showing oral language matters just as much from the very start. The instructional claim (teach handwriting and spelling to fluency) carries a "moderate evidence" rating in the US federal practice guide. Solid for the effects; Promising for the exact architecture.
Stop 04
When a child writes KT for cat, most adults see a mistake. It is not a mistake. It is a measurement. It tells you exactly which sounds he can currently hear.
Children climb a predictable staircase, mapped by Richard Gentry and later by Donald Bear's team. First, letters that mean nothing. Then the first sound only: K. Then first and last, because the ends of words stick out: KT. Then every sound he can hear: KAT. Then, last, the bits that are not sound at all but English's private habits: CAT.
Each step up is one more sound captured. So a child's wrong spelling is a photograph of his phonemic awareness. KT means "I hear the edges." KAT means "I hear all three — I just don't know English uses c here." That second child needs a spelling pattern, not more sound work. You can read that straight off the page.
The evidence goes further than "don't panic." Children deliberately coached to invent spellings — to stretch a word out and put down a plausible letter for each sound, wrong or not — ended up better at real spelling and better at reading than children who got sound-work alone. In one dataset, adding invented spelling to the picture lifted the prediction of Grade 1 reading from 45% to 60% of the variance. Struggling to spell is not a failure state. It is the exercise.
Two activities, two red pens. In dictation, you only ever dictate words made of patterns you have already taught — so the target is near-perfect, and you correct on the spot. In free writing, the child reaches past what he has been taught — so you praise the analysis and never mark the page. Same child, same morning, opposite correction rules. Programmes that blur these two either produce fearful writers or sloppy spellers.
Stop 05
Everything above collapses into one short routine that runs every single day. Four stations, roughly fifteen minutes, and the child hands himself more of the wheel as he goes round.
Sound boxes come first. Elkonin boxes: a row of empty squares, one per sound. The child says the word, pushes a counter into a box for each sound, then goes back and writes a letter (or letter-team) into each box. It splits an impossible question ("how do you spell it?") into two easy ones: how many sounds, and which letter for this one.
Word chains come second, and they are the sharpest drill in the whole programme. cat → cot → cop → chop → shop. Change one thing. Because only one thing changed, the child's whole attention lands on exactly the letter you wanted him to look at. Nothing else in early literacy focuses attention this precisely.
Dictation comes third: you say a sentence, he writes it. Strict rule — every word must be built from patterns already taught, plus the handful of tricky words he already knows. Say it, stretch it, write it, read it back and check. Three to five minutes, daily. Not a Friday test.
Free writing comes last, and it is the point of the other three. His sentence, his idea, invented spelling welcome. You walk round, you say the right version out loud once, you do not touch his page.
Sound boxes and phoneme–grapheme mapping show up repeatedly among the common ingredients of effective early-literacy interventions. Solid Word chains are a beautifully designed drill that follows directly from the theory, but there is no trial of "word ladders" as such. Beloved Daily dictation is standard in every structured-literacy programme and rests on the spelling-instruction evidence rather than on trials of dictation itself. Promising
Stop 06
There is a strong temptation to jump from "he can spell cat" to "write me a story." Skip the rung in between and you get the classic five-year-old output: a drawing, and one exhausted sentence underneath.
The missing rung is the sentence itself. Judith Hochman's method — often called The Writing Revolution — makes it the whole curriculum for a long time. You do not ask for a story. You hand him a stem and he finishes it three ways:
The cat sat on the mat because …
The cat sat on the mat but …
The cat sat on the mat so …
Those three little words are a thinking machine. Because forces a cause. But forces a contrast. So forces a consequence. The child has to build a real thought and a longer sentence at the same time, and neither is optional.
Above the sentence sits Pie Corbett's ladder: imitate, innovate, invent. Imitate — the class learns a short model story completely by heart, with actions, until it lives in their mouths. Innovate — same story, swap the cat for a pig, swap the mat for a hat. Invent — his own story, his own shape. The child never faces a blank page cold; he always has a structure he already owns.
And running alongside all of it, Vivian Paley's practice, which is the cleverest thing on this page. The child tells you a story. You write it down, word for word. Then the class acts it out. His hand cannot write a story yet. His mouth absolutely can. Story dictation lets his imagination live years above his handwriting — and watching an adult turn his own speech into marks on paper is the most persuasive lesson about what writing is that a four-year-old will ever get.
Sentence-level work is the best-evidenced rung: sentence combining beat grammar teaching in adolescent writing (effect size around 0.50), and explicitly teaching grammar rules in isolation shows no reliable benefit at all. Solid The Writing Revolution as a whole programme is widely adopted but thinly trialled — its parts are evidenced, the package is not. Promising Talk for Writing was piloted by the Education Endowment Foundation and the differences were "small or very small," low security, not ready for a full trial. It is a superb classroom craft with weak proof. Beloved Paley's storytelling and story-acting has a real randomised study behind it showing gains in oral language and narrative for preschoolers — but measured on speaking, not on writing. Promising
Stop 07
Here is the pleasant surprise. Montessori worked out the bandwidth problem from Stop 03 about eighty years before anyone measured it — and her whole language shelf is a set of devices for removing one specific load at a time.
She noticed something everyone else had backwards. Children write before they read. Composing a word from letters you already know is easier than decoding a word someone else composed. So her sequence puts the pencil last, not first.
Sandpaper letters take the "which letter makes this sound?" load: the child traces the shape while saying the sound, so the hand stores the link, not just the eye. Metal insets take the pencil-control load, on its own, before letters have to carry any meaning. And the movable alphabet is the masterstroke: it lets a four-year-old spell a word his hand cannot possibly write. No forming, no erasing, no cost to changing his mind. It is pure encoding with the transcription load set to zero — which is exactly what Stop 03 says you want.
Later, the grammar symbols do the same trick one level up. A black triangle for a noun, a red circle for a verb, a small blue triangle for the little word in front. A child physically places shapes over a sentence and feels its architecture years before he could define "verb." Then sentence analysis makes the relationships themselves movable.
Montessori's word for what happens when enough load has been lifted was the explosion into writing — the child who wrote nothing for months and then, one week, writes on everything. Read through Berninger, that is not mysticism. That is what it looks like when the mechanics hand finally gets small enough.
Explicit grammar instruction — naming and defining parts of speech — has repeatedly failed to improve children's writing. Keep the Montessori symbols, but use them for what they are good at: sorting, noticing and building sentences by hand. The moment they become vocabulary to be recited, the evidence stops supporting you. Solid (that grammar drill does not work) Beloved (the symbols themselves are untested, and lovely)
Stop 08
The whole stack has one job: get a child from spelling cat to holding a book with his own name on the cover. Everything else is scaffolding for that moment — because that moment is what makes him write the next one.
One spread per pattern: hero-word tracing, sound boxes, a word chain, one dictated sentence. It is the daily loop turned into paper, and transcription practice is the single highest-leverage thing at this age.
Printed letter tiles including team tiles — sh, ch, th, ck, ee, ai. Lets a four-year-old spell a word his hand cannot write yet. Sets the transcription load to zero.
Wipe-clean, one strip per word length. Turns "how do you spell it?" into two easy questions instead of one impossible one.
cat → cot → cop → chop. Cheapest thing on the list and the most precisely aimed: one change means the child's attention lands on exactly the letter you chose.
Word cards colour-coded to the Montessori symbols, plus because / but / so stem strips. This is the missing rung between word and story, and the rung with the best evidence.
Big picture box on top, ruled lines beneath, "told by ___ · written down by ___". Lets his imagination work years above his handwriting, and shows him what writing actually is.
Weekly pick → hero-page template → eight-page bound book → class library. The loop in the diagram above. This is the brand, not a feature.
Same hook format you already have, aimed at retrieval rather than teaching: the song's job is to make a pattern re-retrievable a hundred times across weeks, not to teach it the first time.
Three timed minutes of letter formation, self-graphed by the child. Unglamorous, and it is the thing that shrinks the orange block in Stop 03.
Hear the word → drag graphemes into boxes → instant correction → chain variant. Must stay encoding. The moment it becomes tap-the-right-picture it has quietly become a reading app again.
Removes the "I don't know what to write" problem by supplying content, so the child's whole effort goes into the sentence rather than into inventing a subject.
Ten tiny stories short enough to learn by heart with actions, each with an obvious swap-one-thing version. Corbett's ladder, packaged.
Appendix · second brain
One line each. Who said it, what it says, and the one thing it changes about what you build.
| Idea | Who & when | In one line | What it changes |
|---|---|---|---|
| Orthographic mapping | Ehri, 2014 · 2023 | Letter–sound connections bond a word's spelling, sound and meaning into one stored unit, after very few correct exposures. | Spelling practice is the fastest route to instant word recognition. Solid |
| Spelling → reading | Graham & Santangelo, 2014 | 53 studies: spelling instruction improves spelling (0.54), reading (0.44) and phoneme awareness (0.51). | Encoding is not a second product line; it is more reading instruction. Solid |
| Simple / Not-So-Simple View of Writing | Berninger et al., 2002 · Berninger & Winn, 2006 | Writing = transcription + text generation + executive function, all squeezed through working memory. | Automate spelling and handwriting to buy creative bandwidth. Promising |
| DIEW model | Kim & Schatschneider, 2014–2020 | Oral language and higher-order thinking feed writing from the very beginning, not after transcription is "done". | Run talk and story alongside spelling from day one, not after it. Promising |
| Handwriting instruction | Santangelo & Graham, 2016 | Improves legibility (0.59), speed (0.63) and, downstream, writing quality (0.84) and length (1.33). | Keep a real handwriting strand. It is a creativity intervention. Solid |
| Elementary writing meta-analysis | Graham, McKeown, Kiuhara & Harris, 2012 | Strategy instruction 1.02, transcription teaching 0.55, prewriting 0.54, creativity/imagery 0.70 — grammar teaching, nothing. | Teach transcription and process; never teach grammar as terminology. Solid |
| Invented spelling | Ouellette & Sénéchal, 2008 · 2017 | Coaching children to invent spellings predicts and improves later conventional spelling and reading. | Welcome wrong spellings in composition; they are the exercise. Solid |
| Spelling stages | Gentry, 1982 · Bear et al., Words Their Way | Pre-communicative → semiphonetic → phonetic → transitional → conventional. | Score a child's errors as a stage, not as a mark. Promising |
| Explicit + invented, together | Shanahan, 2022 | "Explicit spelling versus invented spelling is a false dichotomy. I encourage both." | Two activities, two correction rules, same morning. Solid |
| Sound boxes | Elkonin, 1963 | One box per sound; push a chip, then write a grapheme in each box. | The first station of the daily loop, and a printable product. Solid |
| Word chains / ladders | Rasinski · Sounds-Write "Switch It" | Change one element at a time so attention lands on exactly that element. | The cheapest high-precision drill you can print. Beloved |
| The Writing Rope | Sedita, 2019–2022 | Five strands: critical thinking, syntax, text structure, writing craft, transcription. | A checklist for what a full writing curriculum must eventually contain. Promising |
| The Writing Revolution | Hochman & Wexler, 2017 | Sentences before paragraphs; because / but / so; sentence combining; stems and appositives. | Build the sentence rung before asking for stories. Promising |
| Sentence combining | Graham & Perin, 2007 | Effect size around 0.50; beat traditional grammar teaching outright. | Sentence-building cards, not worksheets about nouns. Solid |
| Talk for Writing | Corbett · EEF pilot, 2015 | Imitate → innovate → invent. EEF found "small or very small" differences, low security. | Use the ladder as a design pattern; do not claim it as evidence. Beloved |
| Storytelling & story-acting | Paley · Nicolopoulou et al., 2015 | Child dictates a story, adult transcribes it, class performs it; gains in oral language and narrative. | Story-dictation books — composing above handwriting level. Promising |
| Drawing as composing | Dyson, 1986–2020 · Mackenzie | Children compose meaning in drawing first and weave writing into it; drawing is not a warm-up. | Keep the picture box on every writing page. Beloved |
| Writing Workshop | Calkins, TCRWP | Mini-lesson, long independent writing, conferring, author's chair, publishing. The process approach scores about 0.40. | Steal the rituals (choice, author's chair, publishing); do not adopt its word-reading model, which is heavily criticised. Promising |
| Writing before reading | Montessori, 1912 | Composing a word from known letters precedes decoding a word someone else composed; the "explosion into writing". | Your existing shelf is already the encoding programme. Beloved |
| Daily writing time | IES Practice Guide, 2012 | At least 30 minutes a day in kindergarten, an hour from Grade 1; teach transcription to fluency (moderate evidence). | Sets the minimum daily footprint your materials must fill. Promising |
| Balanced reading + writing | Graham et al., 2018 | Programmes that deliberately balance reading and writing instruction beat reading-only programmes. | Ship the two products as one curriculum, not two SKUs. Solid |
1. Reading collapses letters into a word; writing fans a sound out into choices. That is why encoding is harder and why it must be taught, not assumed.
2. Spelling a word correctly, a few times, is what glues it into instant-recognition memory — so encoding practice pays out as reading fluency.
3. A child's attention is a fixed-width door. Automatic spelling and handwriting is how you buy room for ideas.
4. Wrong spellings are stage readings, not failures — celebrate them in free writing, correct them in dictation, never both at once.
5. His mouth can already tell stories his hand cannot write. Story dictation lets him be an author now, while the ladder carries his hand up.