📻 Checking a claim — including one that suits us
This article doubts an explanation that happens to work in our favour. “English sits higher in frequency, so an ear trained on another language cannot pick it up” is a convenient thing for anyone selling listening training to believe, because it leads straight to “then listen to high frequencies”. The trouble is that the explanation has a hole in it.
The claim is usually drawn as two bars. Korean occupies roughly 500–2,000 Hz, English roughly 1,000–5,000 Hz; the 2,000–5,000 Hz stretch is therefore unfamiliar ground, and what is unfamiliar cannot be heard. The same picture, with the bands redrawn for whichever language the reader happens to speak, turns up in coursebooks and lecture slides in every market — this is not a local superstition.
Take the chart at its word and do the arithmetic. On its own numbers the gap between Korean and English works out at something like 60 percent, and the gap between Korean and Japanese at around 50 percent. If band mismatch were the cause, Japanese ought to be nearly as hard to catch. Yet the particular complaint of “I know this word and still cannot hear it” is far rarer among learners of Japanese than among learners of English. The explanation predicts one thing and the world does another.
There is an obvious objection, and it is a good one. Japanese comes easily to Korean speakers largely because the word order matches and a great deal of vocabulary overlaps — reasons that have nothing to do with hertz. The objection generalises: a Spanish speaker finds Italian easy for reasons no frequency chart contains. So one counter-example does not demolish the frequency story outright. What it does demolish is the stronger version of it — that band mismatch alone explains the difficulty. More than one thing is going on.
Infants start out able to tell apart the sounds of very nearly every language. At six months a baby distinguishes English ra from la; somewhere past twelve months that distinction fades. Ten months in the womb and twelve months after birth — twenty-two months of listening — set the standard, and sounds outside that standard are filtered out more and more. It is an efficient design for acquiring a first language quickly. The problem is what ended up in the discard pile: for a child raised on Korean, several English consonant distinctions were in it, and for a speaker of any other language a different set was.
The point is this. The sound is not failing to reach you physically; you are failing to divide what arrives. The cue that separates b from d, or ra from la, passes in a very short moment. That is a question of timing resolution and category boundaries, not of bandwidth.
The two explanations lead to different training. If the cause were the frequency band, the remedy would be to listen to a great deal of high-frequency material. If the cause is that the brain cannot divide phonemes, the remedy lies elsewhere — the speed and precision with which the brain processes sound has to come up. That is why our own training begins not with English but with the language the learner already knows, played fast. When no effort is going into working out what the words mean, all of the effort goes into processing speed.
Why speed listening in your first language should help with English listening is not settled science. A neurologist who has lectured alongside us put it as an estimate: “presumably the neural connections for fine distinctions in pronunciation develop, and that appears to help in separating English sounds.” An estimate, not a conclusion. We designed the training on changes we have watched repeat in the room; we do not claim to have established the mechanism. That part belongs to researchers.
The frequency story favours us. Left alone it would have made the marketing easier. But arriving at a right conclusion by a wrong route means that when the route collapses, the conclusion goes with it. We would rather the ground we are standing on were accurate.
Measuring is quicker than reading about it.
Source
Reconstructed from the ra/la discrimination experiments shown in EBS, “Koreans and English”, and KBS, “The Real Reason You Cannot Hear English”, together with Oh Seung-jong’s 2021 blog post “English frequencies versus Korean frequencies”. On resolution rather than bandwidth see [[fast-forword-004sec]]; on the ear a first language builds see [[why-english-doesnt-click]]; on measuring the top of your own range see [[audible-frequency-age]]. ※ A self-measurement in an app does not replace a clinical hearing test.
Read the original →※ This article summarizes and restructures the key points of the source above.
🧭 How the method works · 11 more on this