When a small child says a word wrong, the world falls over itself to help. Parents repeat the word slowly. Teachers make a game of it. Other children — less kindly — make sure the error is never forgotten. It is a loud, constant, sometimes brutal feedback loop, and it works: by six or seven, the child sounds like everyone around them. Then adulthood arrives, and the corrections stop. Completely. You can live fifteen years in a second language, raise children in it, run meetings in it — and not one person will ever mention the vowel.
The silence is not neglect. Correcting an adult’s speech is a social transgression: it interrupts the conversation to grade the speaker, and it points at the very thing polite people have agreed not to notice. So listeners do something quieter instead — they adapt. They lean on context, reconstruct the intended word, and keep the conversation moving. Linguists call it accommodation; socially it is kindness. Acoustically it is a conspiracy: the better your friends and colleagues get at understanding you, the less evidence ever reaches you that anything needs fixing.
There is a cruel twist underneath. You cannot supply the missing feedback yourself, because you hear your own speech through the sound categories of whatever language you grew up with. If two English sounds share one category in your ear — the “ih” of ship and the “ee” of sheep, say — then your own recording sounds perfectly fine to you. The substitution is inaudible precisely to the one person who could correct it. Feedback has to come from outside, and from outside, nobody is talking.
One signal does survive the politeness filter: the repeat request. But “say that again?” names nothing. You know something failed; you do not know which word, let alone which sound inside it. String enough of those moments together and people draw the natural, wrong conclusion — that their “accent” is a single enormous thing, too big and too late to change. It almost never is. Listen closely to almost any fluent second-language speaker and the distance between them and the room is a handful of specific substitutions, repeated everywhere, doing nearly all of the damage.
“Fifteen years, and nobody said anything? Madonna santa! One little vowel, wearing the wrong costume every single day — somebody should have told you on day one!” — Nunzio, of course, has opinions.
What broke at adulthood was never your ear or your mouth — it was the supply of specific, immediate, unembarrassed correction. That is restorable. Read one sentence aloud, have every individual sound compared against what the sentence called for, and let something with no social stakes whatsoever tell you plainly: this sound, in these words, came out as that sound. It is the childhood loop again, minus the playground. Each of the sounds involved has a page in the sound library, with an anchor word to hang it on — because a sound you can name is a sound you can practise.
The Nunzio app is exactly that loop: five sentences, every sound scored on your own phone, and one warm, dramatic Italian uncle who mentions the vowel — the way somebody should have years ago.