Vol. I · No. 001a pronunciation quarterlyapril 2026

slaytheaccent

“An instrument for the honest correction of the mouth.”
Feature · The pronunciation problem

A quiet method for teaching the mouth what the ear already knows.

Every second-language speaker of English carries a trace — a shadow of the native tongue in the flap of a T, the pre-rhotic colouring of a vowel, the unvoiced th that arrives as an s. The standard correction is a person: a coach, a tutor, expensive and inconsistent. The standard alternative is a toy: gamified, evasive, too kind to be useful.

slaytheaccent is the third thing. Record a phrase. The acoustic signal is aligned, phoneme-by-phoneme, against a native reference. You do not receive a compliment. You receive a diagnosis — and a drill.

“You can teach a person to hear the difference between water and waɾer in a minute. Teaching the tongue takes longer.”
— Peter Ladefoged, paraphrased

The method, briefly.

§II
  1. 01

    Record without performance.

    A phrase, the phone mic, three seconds. No scripts, no artificial prosody — natural delivery is what we are measuring.

    capture
  2. 02

    Align against a native.

    Three acoustic models vote on every phoneme. The output is a diagnosis — expected ɾ, observed t — not a summary score.

    score
  3. 03

    Drill the wrong ones.

    Your worst phonemes rise to the top of the queue. Shadow the reference, re-record, watch the score move. Repeat.

    iterate

Begin. Quietly, if you prefer.

Enter the study →