Building a Syllable Block 음절

Korean letters do something no European alphabet does: they refuse to stand in a straight line. Instead they gather into little squares — one square per spoken syllable — and it is these squares, not the individual letters, that fill a line of Korean text. This looks like the feature that makes Korean "a character script," but it is exactly the opposite. Each square is a transparent package of ordered letters, and the entire skill of reading Hangul is learning to unpack it. Master that, and you can read blocks you have never seen before.

One block = one syllable

A 음절 (syllable block) is the square unit Korean writes in. Every block corresponds to exactly one spoken syllable and occupies one equal-sized square, no matter how many letters are crammed inside. The word 음절 itself is two blocks, two syllables:

Inside each block there are up to three slots, always in this order:

• 초성 (onset) — the starting consonant, top or top-left.
• 중성 (nucleus) — the vowel, to the right of or below the onset.
• 종성 (coda), better known as the 받침 ("supporting floor") — an optional final consonant, sitting underneath.

The minimum block is just onset + vowel (a CV syllable). Add a final consonant and you get onset + vowel + coda (a CVC syllable):

That is the whole architecture. 가 is two letters in one square; 강 is three. Both are one syllable, one block, one square.

Every block needs an onset — even vowel-first syllables

Here is the rule that surprises everyone. A block must have an onset slot filled, visually. So what happens when a syllable begins with a vowel sound, with no real consonant? Korean fills the onset slot with a placeholder: the letter ㅇ (이응, ieung), which in that position is completely silent. It is a spacer that keeps the block square, nothing more.

So a lone vowel never appears in real text; ㅏ by itself is not a word, it is just a letter waiting for its block. Write 아, not ㅏ, for the sound "a." (Beware: this silent ㅇ is silent only in the onset. Down in the batchim slot the very same letter is pronounced [ŋ] — that is why 강 is gang, not ga. The two jobs of ㅇ get their own ieung page.)

Decomposition is the skill — not memorization

Coming from English, the instinct is to see 한 as a single dense "character" and to try to memorize it whole, the way one memorizes a Chinese hanzi. That instinct is the one thing that will make Hangul feel impossible. There is nothing to look up. There is a block to build — and, when reading, to take apart — from about forty letters you already know.

Train the decomposition explicitly. Read each block onset → nucleus → coda, in that order:

Put two of these together and you have read a word. 한 + 글 = 한글, the name of the script itself:

The payoff is enormous and immediate: because you assemble blocks rather than memorize them, a block you have never encountered is no harder than one you have. See its three letters, read them in order, done.

You never memorize the 11,172 blocks

The arithmetic makes the point. Korean has 19 possible onset consonants, 21 possible vowels, and 27 possible final consonants — plus the option of no final. That is 19 × 21 × 28 = 11,172 theoretically possible blocks. Only a couple of thousand actually occur in the language, but here is the thing: you memorize none of them. You memorize roughly forty letters and the rule for stacking them, then compute any block on demand. An eleven-thousand-item "character set" collapses into a forty-item alphabet plus a construction rule. That is the entire gift of the alphabet Sejong designed — see the featural-alphabet page for why even those forty letters are systematic.

Real text, then, is just a stream of these decodable squares. Nothing in it is a picture to recognize:

Where the letters physically go

You may have noticed that ㅏ sat to the right of its onset (한) while ㅜ sat below it (국). That is not random and not a matter of taste — it is a fixed rule keyed to whether the vowel is "vertical" or "horizontal." Because that geometry is a topic of its own, it gets its own page on block shapes. For now, hold onto the structure (onset → vowel → optional batchim, all in one square); the next page handles the arrangement.

Common Mistakes

1. Memorizing a block as one unanalyzable picture. There is no "han symbol." Pull every block apart into ordered letters.

2. Writing a vowel with no onset. A vowel-initial syllable still needs the silent ㅇ to fill the onset slot.

3. Thinking the batchim ㅇ is also silent. In the onset ㅇ is silent; as a batchim it is [ŋ]. So 강 is gang, not ga.

4. Reading a block bottom-up or scrambling the slots. The order is always onset → vowel → batchim. The batchim is read last, never first.

5. Stringing letters out in a line. Korean packs each syllable into a square; it does not write ㅎㅏㄴ across the line like h-a-n.

Key Takeaways

• Korean writes in square blocks (음절), one block per spoken syllable, each an equal square.
• A block has up to three ordered slots: 초성 (onset) + 중성 (vowel) + optional 받침 (final consonant). Minimum is consonant + vowel (가); add a coda for CVC (강).
• A vowel-initial syllable fills the onset with the silent ㅇ (아 = ㅇ+ㅏ). The same ㅇ as a batchim is [ŋ], not silent.
• The real skill is decomposition — read every block onset → vowel → coda — because that lets you read any of the ~11,172 possible blocks from just ~40 letters.
• How the letters arrange inside the square depends on the vowel's orientation — the next page.