Every Korean computer, phone, and PC bang keyboard runs the same layout: Dubeolsik (두벌식, “two-set”), the national standard formalized in South Korea in 1969 and carried into the KS X 5002 keyboard standard. Its design principle is elegant: hangul syllables are always consonant-then-vowel, so Dubeolsik puts all consonants under the left hand and all vowels under the right — typing Korean becomes a natural left-right alternation. This reference maps every QWERTY key to its jamo and explains the composition rules that turn keystrokes into syllable blocks.
The full key map
Left hand — consonants:
| QWERTY key | Base jamo | Shift jamo |
|---|---|---|
| Q | ㅂ | ㅃ (tense) |
| W | ㅈ | ㅉ (tense) |
| E | ㄷ | ㄸ (tense) |
| R | ㄱ | ㄲ (tense) |
| T | ㅅ | ㅆ (tense) |
| A | ㅁ | — |
| S | ㄴ | — |
| D | ㅇ | — |
| F | ㄹ | — |
| G | ㅎ | — |
| Z | ㅋ | — |
| X | ㅌ | — |
| C | ㅊ | — |
| V | ㅍ | — |
Right hand — vowels:
| QWERTY key | Base jamo | Shift jamo |
|---|---|---|
| Y | ㅛ | — |
| U | ㅕ | — |
| I | ㅑ | — |
| O | ㅐ | ㅒ |
| P | ㅔ | ㅖ |
| H | ㅗ | — |
| J | ㅓ | — |
| K | ㅏ | — |
| L | ㅣ | — |
| B | ㅠ | — |
| N | ㅜ | — |
| M | ㅡ | — |
Only seven keys carry shifted jamo: the five tense consonants ㅃㅉㄸㄲㅆ on Q W E R T, and the two “y-diphthongs” ㅒ ㅖ on O and P. Everything else — including all compound vowels and consonant clusters — is composed from sequences, not Shift.
How keystrokes become syllable blocks
Korean text is written in syllable blocks (날, 라), but you type individual jamo; the OS input method engine (IME) assembles them in real time. The interesting part is the ambiguity rule: when you type consonant + vowel + consonant, the final consonant might be the coda of the current syllable or the onset of the next — and the IME cannot know until the next keypress.
Type S K F: ㄴ + ㅏ + ㄹ → the IME shows 날. Now type another vowel,
K: the ㄹ detaches from 날’s coda and becomes the onset of a new syllable —
the display reflows to 나라. This “stolen coda” reflow (called dogig
resyllabification in IME implementations) is why Korean text seems to
mutate as you type; it is deterministic, and every IME implements the same
rule.
Compound vowels are typed as two vowel keys in sequence: ㅗ+ㅏ→ㅘ, ㅜ+ㅓ→ㅝ, ㅡ+ㅣ→ㅢ, and so on — the IME composes them automatically. Consonant clusters in codas (as in 닭, 값) are likewise two consonant keys typed in sequence in coda position: ㄹ+ㄱ, ㅂ+ㅅ.
Why “two-set”, and what the alternative was
The name distinguishes the layout from Sebeolsik (세벌식, “three-set”), the rival design by Dr. Kong Byung-woo. Sebeolsik assigns different keys to the same consonant depending on whether it is an onset or a coda — three sets: initial consonants, vowels, finals. That eliminates the IME ambiguity entirely and its devotees type very fast, but it needs more keys (including the number row) and a steeper learning curve. The government standard went to the simpler two-set layout, and network effects did the rest: today Sebeolsik survives only among enthusiasts, and Dubeolsik is effectively universal. The trade-off Dubeolsik accepted — coda/onset ambiguity — is exactly the reflow behaviour described above, solved in software rather than on the keycaps.
Setup and practical notes
A physical US QWERTY keyboard works perfectly — Korean input is an OS toggle, not hardware:
- macOS: System Settings → Keyboard → Input Sources → add “Korean, 2-Set”. Switch with the Fn/globe key or Ctrl-Space.
- Windows: Add the Korean language pack; toggle Hangul/English with the right Alt key (한/영 key on Korean hardware) or Win-Space.
- Linux: ibus-hangul or fcitx5-hangul, both defaulting to Dubeolsik.
Keyboards sold in Korea print the jamo on the keycaps and add two dedicated keys — 한/영 (Hangul/English toggle) and 한자 (hanja conversion) — but the letter positions are identical to this reference.
Learning tips for new Korean typists
Memorise the five Shift-tense pairs first (Q W E R T → ㅃㅉㄸㄲㅆ) — they are the only shifted consonants in standard Korean. Then internalise the hand-split rather than individual keys: knowing “consonants left, vowels right” halves the search space for every character. Frequency helps too — ㅇ(D), ㄴ(S), ㄱ(R), ㅏ(K), ㅣ(L), ㅡ(M) dominate real Korean text, and all sit on or near the home row by design. Expect the syllable-reflow behaviour and trust it; deleting mid-block with Backspace removes one jamo at a time on most IMEs, not the whole block, which is usually what you want.
Everything on this page runs locally in your browser.
References
- National Institute of Korean Language (국립국어원) — hangul and orthography resources
- Unicode Standard, Hangul Jamo block (U+1100–U+11FF) and Hangul Compatibility Jamo (U+3130–U+318F)