“See-aye-tee… cat?” Your three-year-old looks up, proud. You smile, but inside you’re calculating the damage. The ABC song was supposed to be a gentle on-ramp. Now it’s the wall they keep crashing into when they try to actually read a word.
A phonics program for this exact situation has to do something most curricula skip — formally undo the letter-name-first habit before stacking new skills on top. This post walks through where it goes wrong, four repair steps that work, and what you should see within a few weeks.
What goes wrong when a child learns letter names before letter sounds?
The brain stored the wrong retrieval cue. When your child sees the letter C, the first thing fired is “see” — the name — not /k/, the sound. Decoding requires sounds. So every attempt to read becomes a translation: name first, then sound, then blend. That’s three steps where there should be one.
This isn’t a sign of a bad learner. It’s a sign that the input came in the wrong order. Most kids who memorized the alphabet song before age three carry this pattern, and most reading apps make it worse by quizzing letter names with cheerful jingles.
A serious phonics program treats this as a repair job, not a fresh start — your child already knows the alphabet, you just need to overwrite the retrieval cue.
How do you replace letter-name recall with sound recall?
Four steps, run in sequence, two minutes a day. No drilling, no flashcards, no pressure.
Step 1: Stop saying letter names out loud for two weeks
When you point at a letter, say only the sound. /m/ — never “em.” If your child says the name, gently echo back the sound. Do not correct, do not lecture. The brain is listening for which cue you reward with attention.
Step 2: Switch to lowercase-first visual exposure
Capital letters live in the ABC song. Lowercase letters live in books. Posters and writing pages that lead with lowercase break the mental link to the song and rebuild it as “text on a page.” This single shift is doing more work than parents realize.
Step 3: Use guided writing to force sound-first retrieval
Tracing a letter while saying its sound is the encoding moment. The hand, the eye, and the voice all fire on the same cue. Names cannot survive this drill — the writing page does not care what something is called, only what it sounds like.
Step 4: Run sound-only blends in 90-second reps
After two weeks of sound-only exposure, point at three letters in a row and ask for sounds, not names. /c/ /a/ /t/. Then say “blend it.” If they say “see-aye-tee,” tap the page and prompt only the first sound. They will self-correct within five reps.
A well-built english phonics course is sequenced exactly this way — sound input first, blends second, names tucked into the background where they belong.
Before and after: what the repair actually looks like
Before. Your child sees CAT and says “see-aye-tee, cat” — using the name route, then guessing the word from memory. Anything outside the memorized list stalls. Outcome: pseudo-reading that collapses on new words.
After three weeks of sound-first repair. Your child sees CAT and says /c/ /a/ /t/ — pause — “cat.” The pause is the win. That’s real decoding, even when slow. Within another month, the pause shrinks to nothing and they’re reading words they’ve never seen. Outcome: durable decoding that scales.
The shift is not about effort. It’s about which cue the brain reaches for first. Once sound becomes the default, names quietly take a back seat and the alphabet song becomes a fun trivia trick rather than a roadblock.
Frequently Asked Questions
Is it bad to teach the alphabet song before phonics?
It’s not catastrophic, but it does create extra work. The song teaches names, and decoding requires sounds. If you’ve already taught it, you’re not behind — you just need a phonics-first sequence that explicitly overwrites the name-first habit, like the approach used by Lessons by Lucia and other classroom-tested programs.
How long does the letter-name-to-letter-sound repair take?
Most children need two to four weeks of consistent sound-only exposure for sounds to become the default retrieval cue. The first week feels slow. The third week looks like a different kid.
Should I correct my child every time they say the letter name?
No. Correction creates resistance and slows the rewrite. Instead, echo back the sound without commenting. The brain follows the cue you reward, and silent modeling is far more powerful than verbal correction.
What’s the difference between a letter-names-first program and a sounds-first program?
A names-first program teaches “A is for apple” and assumes sounds will follow. A sounds-first program teaches /a/ as the primary association and treats the name as bonus trivia. For a child who already knows names, sounds-first is the only direction that resolves the bottleneck.
What happens if you skip the repair
The bottleneck doesn’t fix itself. By kindergarten, the child reads memorized words at speed and unfamiliar words at a crawl — a profile that often gets misread as “advanced” by teachers and “stuck” by parents. The longer the name-first cue stays primary, the more reading energy gets burned on translation rather than comprehension. A two-week sound-first reset before age five resolves what otherwise becomes a years-long fluency gap.
