How do I know when to stop drilling a sound and just start speaking?

When the sound holds up in an unrehearsed sentence you haven't specifically practiced, not when it holds up in the drill sentence you've said fifty times — those are two genuinely different milestones, and mistaking the first for the second is the most common reason a "fixed" sound quietly reappears later.

30-second version: A drill sentence you've repeated many times tests motor memory of that specific sentence as much as it tests the sound itself. The real test of whether a fix has generalized is a sentence with the same target sound in new words, said without rehearsal, ideally in a context closer to real speech — describing something unscripted, or a sentence you're seeing for the first time. If that unrehearsed test comes back clean across a couple of different attempts, the sound has graduated from drill to habit, and continuing to drill it further has rapidly diminishing returns.

Why the drill sentence can be misleading

Repeating the same sentence dozens of times genuinely improves your performance on that sentence — but some of that improvement is memorization of the specific sequence of words and mouth movements, not a general fix to the underlying sound. This is a well-documented pattern in motor learning generally: performance on a specific, repeated task improves faster than the underlying skill actually transfers to new situations. Trusting a rehearsed drill sentence alone as proof of a fix risks declaring victory on a sound that's really only fixed in one specific sentence.

This doesn't mean drilling was wasted time — it's a necessary stage, and skipping straight to unrehearsed speech before the articulation is even available at all just produces the same old error, unrehearsed. The sequence matters: drill until the sound is reliably available under focused attention, then test it under conditions closer to real use.

The specific test

Take the sound you've been drilling and find or improvise a new sentence containing it — one you haven't said before in this practice cycle — and record it once, without rehearsing it first. If that comes back clean, try a second, different unrehearsed sentence a day or two later. Two clean, independent, unrehearsed results are much stronger evidence of a real fix than ten clean results on the same drill sentence.

An even stronger version of this test: use the sound in genuinely freeform speech — answering a question unscripted — rather than reading any sentence at all, since freeform speech surfaces problems that reading aloud can hide. A sound that survives freeform, unscripted use under real cognitive load has cleared the highest bar available.

What "stop drilling" actually means

It doesn't mean never checking the sound again — it means shifting from dedicated, isolated practice sessions to just using the language normally and trusting the sound to hold up, checking back in occasionally rather than daily. Continuing to drill a sound well past the point it's reliably automatic mostly just takes time that could go toward the next sound on your list, covered in tracking a sound with the 44-sound map — most learners have several sounds worth working through, and each one only needs dedicated attention until it clears this test, not indefinitely.

Drill it, then stop drilling it

If /θ/ has been your focus, the drill sentence might be Record: "Think about this thing for three months." The graduation test is a different sentence with /θ/ in new words, said cold, without having practiced that exact sentence first — that's the one that actually tells you whether to move on.

How do I know it's improving?

The clearest signal isn't a rising score on the same drill — it's the gap between your drilled-sentence accuracy and your unrehearsed-sentence accuracy on the same sound closing over time. When that gap is small, the sound has moved from performance to habit, which is the actual definition of "fixed" this whole series has been working toward.

Try it

Open sayit free — no install, no card — and test whatever sound you've been drilling in a brand-new, unrehearsed sentence today. If it holds up clean, that's your answer: stop drilling that one, and pick the next sound from your map.