When I was first learning to type Hangul, I quickly found a lot of logic in the 두벌식 layout. Consonants on the left, jamo that can be doubled in the top row, nasals in the middle, aspirated in the bottom, vowels on the right, ㅏ ㅑ pairs above each other, ㅗ above ㅜ like arrows, and so on.

What has been always bothering me and making my life a bit difficult is that the order of the aspirated consonants is different from the unaspirated, without any obvious reason. For example, the layout of QWERTY is structured as to prevent from arm jamming on a mechanical typewriter. The 두벌식 layout seems to be built on easy memorization and some logical patterns, however I wasn't able to find much on the 두벌식 layout in English.

두벌식 keyboard layout

(The 두벌식 keyboard layout by Cyp on Wikipedia)

My first idea is that it is because of ergonomics – that would make the more common letters right to left, with the most common letter under the index finger. Unfortunately, I wasn't able to find a frequency list of individual jamo. I coded up this JSFiddle, that counts the jamo frequency in any given text. After feeding it a few Wikipedia articles (mostly about countries, which may have caused some bias, especially with 국), I arrived at a result with about 100,000 letters total (full result in percentages on pastebin). The order of the letters from most frequent to least frequent is not matching with the hypothesis: ㅊ ㅌ ㅍ ㅋ. Maybe more and better data would show a different result? (You can try the tool with more text, or if you have some long text or a Korean language corpus, you can send it to me.) Either way, the aspirated letters are not even very frequent in the language, the benefit would not be huge. However the ordering of the upper row seemed to match, if we assume the data was skewed by the 국 in countries name mentioned earlier.

My second idea was that it is because of alphabetical ordering, but the current order in South Korea is ㅊㅋㅌㅍ.

Any other ideas as to why the change in the order? Why not just leave it as it is in the upper column for consistency and clarity, since the rest of keyboard seems to be made for easy memorization?

  • 1
    The standard 두벌식 layout partly follows a 네벌식 layout. According to this, 황해용 said "기계 구조의 간편 및 실용성, 국제 규격 및 타 기종의 적용성, 글자의 균형성, 타자 동작의 효율성, 타자 학습의 용이성을 고려하였다"; however, these reasons seem to be lies because there are more efficient 두벌식 layouts. It is highly likely that the government forced people to use it, as this says "네벌식 자판은 박정희 정권 때에 과학기술처가 비전문가들을 동원하여 졸속으로 공포한 표준 글자판이었다."
    – Klmo
    Commented Jul 25, 2020 at 18:01
  • 1
    I have just found that this page has the official document for the 네벌식 and 두벌식 layouts. It is titled "한글기계화 표준자판(안) 확정" (about 100 MB in total) and it has the frequency list they used. It seems to me that except for some consonants, they focused more on the ease of learning than on the frequency. I have also noticed that ㅂㅈㄷㄱ pairs with ㅋㅌㅊㅍ in a reversed order; this relates to the figure ("그림 4" in the document) on the typing efficiency they provided but it does not say how they estimated it and where they used it.
    – Klmo
    Commented Jul 28, 2020 at 9:30

1 Answer 1


In a document provided by @Klmo in a comments, I found frequency analysis they made for the letters of Korean alphabet on page 564 of internal numeration. For the letters ㄱㄷㅈㅂㅋㅌㅊㅍ they found the frequency distribution goes like this:

ㄱ > ㄷ > ㅈ > ㅂ > ㅊ > ㅍ > ㅌ > ㅋ

(In your data it's almost the same, with only single difference: ㄱ > ㄷ > ㅈ > ㅂ > ㅊ > ㅌ > ㅍ > ㅋ)

So in my understanding they saw the pattern of reversed frequencies for those letters and it made sense to put ㄱ under pointer finger as the most frequent and ㅋ under pinky finger as the least frequent and everything else mirrored for consistency. (Because ㅍ is slightly out of order.)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.