Segmenting is the ability to break a spoken word into its individual sounds. A teacher says "cat," and a child says the three sounds: "/k/, /æ/, /t/". Segmenting is the reverse of blending — you're deconstructing a whole word into its phonemic parts. This skill is crucial for spelling and encoding: if a child can segment a word into sounds, they can match each sound to a letter and write the word phonetically.
Start with short, simple words: "at," "it," "up." Use Elkonin boxes (sound counters) — as the teacher says each sound, the child pushes a token into a box, one sound per box. This makes segmentation visible and concrete. Use stretching and holding: say "ssssit" to stretch the initial consonant, making it easier to isolate. Start with two-sound words (/a/ /t/) before moving to three-sound words (/k/ /a/ /t/). Use hand motions, singing, and exaggeration.
You already know that words are made of individual sounds — that's phonemic awareness. Segmentation takes that knowledge one step further: it's the active skill of pulling a word apart into those individual sounds. When you hear the word "cat," your brain instantly recognizes it as a unified unit. But for a child learning to read and write, consciously breaking "cat" into /c/ /a/ /t/ is a deliberate, effortful operation that must be taught and practiced.
Segmentation is the inverse of blending. If blending is "sound by sound, let's make a word," then segmentation is "word as given, let's find all the sounds." Both operations teach the same insight: words are composed of discrete phonemic units. But segmentation serves a different purpose than blending in the reading curriculum — while blending is the operation you perform when decoding (translating letters to sounds to word), segmentation is the operation you perform when spelling (translating word to sounds to letters). A child who cannot segment has no basis for phonetic spelling because they haven't isolated the sounds that need to be represented.
Here's the typical progression: early segmentation starts with simple, highly salient onsets: "Find the first sound in 'top'" — /t/ is pulled out of the word. Final sounds come next: "What's the last sound you hear in 'cat'?" Then onset-rime segmentation: "Say 'sit.' Now say it again, but this time tell me the first sound and the rest": /s/ + /it/. Finally, phoneme-by-phoneme segmentation: /s/ /i/ /t/ as three separate sounds. This progression is important: early segmentation is harder than later segmentation because the entire word is still a unified auditory event. Breaking it into two chunks (onset-rime) is easier than breaking it into individual phonemes. Only after substantial practice with easier segmentation tasks can children reliably segment words into all their component sounds.
The tool most strongly associated with segmentation instruction is the Elkonin box (also called a sound box or sound counter). The teacher says a word, and the child says each sound while pushing a token (often a chip or button) into a box — one token per sound. For the word "cat," the child would push three tokens into three boxes while saying /c/, /æ/, /t/. This makes the abstract operation of segmentation concrete: the child can literally see and count the sounds. After segmentation practice with Elkonin boxes, children develop the automaticity to segment without the physical boxes, but the boxes are crucial scaffolds early on.
Why does segmentation matter so much for writing? When a child writes a word, they must translate from the spoken/conceived word to a sequence of letters. The translation requires segmentation: "I want to write 'sun.' What sounds are in 'sun'? /s/, /u/, /n/. What letters make those sounds? S, U, N." Without segmentation, a child has no way to know how many letters to write or what they should be. This is why segmentation is taught alongside letter-sound correspondence in a comprehensive phonological awareness and beginning writing program. Together, blending and segmentation are the foundations of the alphabetic principle — the insight that writing is a systematic code for speech.