Writing the Sounds You Hear
To spell a spoken word, break it into its sounds and write a letter for each one in the order you hear them. · 9 min
Yesterday you turned letters into sounds: you looked at map and read /m/ /a/ /p/. Today you turn the move around. Someone says a word out loud, and you write it down. You cannot see the letters yet — but you can hear the sounds. Catch each sound, and write the letter that spells it.
Guess before you learn
Say the word sun out loud, slowly: /s/ /u/ /n/. To spell it, how many letters will you write?
Three. The word sun has three sounds — /s/ /u/ /n/ — and you write one letter for each: s, u, n. If you said 2, you may have skipped the middle sound. That /u/ in the middle is the easiest sound to miss, but it needs its own letter too.
K–2
3–5
Pulling a spoken word apart into its separate sounds is called segmenting. It is blending run backwards. To spell, say the word slowly and catch each sound in order, then write the letter that spells it. The middle sound is a vowel, and it is the one spellers drop most — stretch the word until you hear it, and give it a letter too.
Write the sounds left to right, in the order you say them, because the order of the letters is the word. The sounds /t/ /a/ /p/ spell tap, but the same three sounds in a new order, /p/ /a/ /t/, spell pat. Say the word, catch the first sound first, and let the rest follow in line.
6–8
The sounds you hear are phonemes; the letters you write are graphemes. Spelling matches one to the other. English has roughly 44 phonemes but only 26 letters, so one sound can be spelled several ways — yet in these short CVC words (consonant–vowel–consonant) the match is clean: one phoneme, one grapheme, left to right. Pulling a spoken word into its phonemes is a skill called phonemic awareness, one of the strongest early predictors of who learns to read easily.
Phonemic awareness is an ear skill, not an eye skill — you can practise it in the dark. Stretching a word forces the continuous stream of speech to break at its seams. The middle vowel is the hardest phoneme to isolate, because vowels carry the voice with no clear stop — exactly why beginning spellers so often drop it.
9–12
Each time a child sounds a word out correctly, the brain does something lasting: it maps the word's spelling onto its pronunciation and meaning, filing it away for instant recognition later. This is orthographic mapping, and spelling drives it as hard as reading does. Writing /d/ /o/ /g/ as d-o-g forces attention onto every phoneme in order, welding the letter string to the sound — which is why spelling practice makes later reading faster, not merely neater.
David Share's self-teaching hypothesis names the engine: decoding a word phonemically is itself the act by which a reader teaches themselves that word's spelling — no flashcards required. A handful of accurate sound-outs can fix a word for life. CVC words are the training wheels: a transparent, one-to-one stretch of English where the code never lies, met before the language's messier spellings arrive.
K–2
Say the word slowly, like it is stretching: mmm, aaa, p. That is three sounds. Now write one letter for each sound: m, then a, then p. You spelled the word map!
Reading turns letters into sounds. Spelling turns sounds into letters. It is the same word, walked the other way.
Undergrad
Cognitive models formalise this. Dual-route theory posits a lexical route for familiar words and a sublexical route that assembles pronunciation from grapheme–phoneme correspondences; CVC spelling is that sublexical route run in reverse, from phonology out to orthography. Connectionist (triangle) models drop the separate routes, instead learning statistical mappings among orthography, phonology, and meaning across many exposures. In both accounts the regular, high-frequency CVC mappings are acquired first and anchor everything built on top of them.
One point deserves stress: the phoneme is a linguistic abstraction, not an acoustic slice. In the speech stream /k/ /a/ /t/ are coarticulated — smeared together, with no silent gaps between them. Segmenting is therefore a real cognitive achievement, the learner imposing discrete units on a continuous signal. Alphabetic writing rewards exactly this analysis, which is why phonemic awareness and alphabetic literacy bootstrap each other rather than developing apart.
Postgrad
Neuroimaging locates skilled word recognition in the left occipitotemporal cortex — Dehaene's visual word form area — a region that is not innate but recycled from object and face recognition through the act of becoming literate. Learning to bind graphemes to phonemes physically reorganises this circuitry; longitudinal studies show its response sharpening as children move from laborious decoding to automatic mapping. The humble CVC drill is, at the neural level, the first tuning of a cultural circuit.
Zoom out and English orthography stops looking broken. Chomsky and Halle argued it is morphophonemic — it preserves morphemes (sign inside signal, heal inside health) at the cost of strict sound spelling, a near-optimal design for a reader who trades letters for meaning. CVC words sit at the transparent floor of that system; the deeper regularities met later are features, not defects.
segmenting
Breaking a spoken word into its separate sounds so you can write a letter for each. It is the opposite of blending: /s/ /u/ /n/ pulled apart, ready to spell.
Here is the move, every time you spell. Say the word slowly and stretch it out. Catch the first sound and write its letter. Catch the next sound and write its letter. Catch the last sound and write its letter. Then read back what you wrote, sweeping your finger under the letters, to make sure it says the word.
Why is this true?
Why do you write the sounds in the order you hear them, instead of any order?
Because the order of the letters is the word. The sounds /t/ /a/ /p/ in that order spell tap, but /p/ /a/ /t/ spell pat — the same sounds in a new order make a new word. So you write the first sound first and follow the word in order.
You can spell a small word now: say it slowly, catch each sound, and write a letter for each, in the order you hear them. Reading and spelling are the same word walked two ways. Next you will spell whole families of words at once — cat, hat, sat, mat — by keeping the ending and changing just the first sound.