The Plain Series 평음: ㄱ ㄷ ㅂ ㅅ ㅈ and the Sonorants

Korean sorts its consonants into three parallel columns — plain (평음, also called lax or lenis), aspirated (격음), and tense (경음) — and the plain series is the base that the other two are built on top of. This page covers the five plain obstruents ㄱ ㄷ ㅂ ㅅ ㅈ, the four sonorants ㄴ ㅁ ㅇ ㄹ, and the breathy ㅎ. Get these right and you can already read most of a Korean page; the aspirated and tense series are then just controlled modifications of what you learn here.

The plain series also hides the single most useful pronunciation fact in the whole alphabet — automatic voicing — which trips up nearly every English speaker for a very specific reason. We will build up to it.

The five plain obstruents and where they are made

Each plain stop and the fricative ㅅ has a fixed place of articulation. Learn the place (where your tongue or lips go) once, and the aspirated and tense versions reuse the exact same place — only the airflow changes.

Word-initially the plain stops are voiceless and only lightly aspirated — much softer than the strong puff of ㅋㅌㅍㅊ, and softer even than an English "k" in "key." That is why they sit in the middle: not tense, not aspirated, just plain.

The big rule: plain stops voice automatically

Here is the fact that reorganizes everything. The plain stops ㄱ ㄷ ㅂ ㅈ are voiceless at the start of a word, but they voice — turn on the vocal cords — whenever they sit between two voiced sounds (between vowels, or after a nasal ㄴ/ㅁ/ㅇ or the liquid ㄹ). This is not optional and not regional; it is automatic and total.

The classic demonstration is 부부, "a married couple":

The word 부부 has two identical ㅂ's, yet they come out differently: the first ㅂ is voiceless [p] (word-initial), the second ㅂ is voiced [b] (between vowels). If you listen, it sounds roughly like "pubu." Now look at the romanization line above: Revised Romanization writes both ㅂ's as "b" — "bubu." That is deliberate, and it is the whole point.

In 고기 (gogi, "meat") the first ㄱ is [k] and the second is [g]. In 아버지 (abeoji, "father") the ㅂ is voiced [b] and the ㅈ is voiced [dʑ]. In 바보 (babo, "fool") the first ㅂ is [p], the second [b]. Every time: voiceless at the edge, voiced in the middle.

This is exactly why 비빔밥 is romanized "bibimbap" on menus worldwide: three ㅂ's, and the reading walks through the whole rule — the first (word-initial) is [p], the one in -bap (after the nasal ㅁ) is voiced [b], and the final batchim ㅂ is an unreleased [p]. Same letter, three surface sounds, zero choices.

The one plain obstruent that does not join the party is ㅅ. Korean has no [z] sound, so ㅅ stays voiceless [s] even between vowels: 부산 is "Busan" [pusan], not [puzan]; 이사 (moving house) is "isa," not "iza."

ㅅ before ㅣ and y-vowels: the "sh" shift

ㅅ has one more quirk. Before the vowel ㅣ (i) or any y-glide vowel (ㅑ ㅕ ㅛ ㅠ), the tongue pulls back and ㅅ is pronounced [ɕ] — the "sh" of English "she." So 시 sounds like "shee," not "see."

Note the mismatch you have to hold in your head: Revised Romanization always writes ㅅ as "s," so 시 is spelled "si" — but it is pronounced [ɕi], "shee." The romanization tells you the letter; your ear has to supply the "sh."

The sonorants ㄴ ㅁ ㅇ ㄹ

The sonorants have no three-way contrast — each is just one consonant — so they are the easy part.

• ㄴ = [n], like English "n."
• ㅁ = [m], like English "m."
• ㅇ as a syllable onset is silent (a placeholder), and as a batchim it is [ŋ], the "ng" of "sing." That double life gets its own treatment on the ㅇ page.
• ㄹ is the tricky one, because it changes shape by position (below).

ㄹ: a flap between vowels, an [l] at the bottom

Korean has one liquid letter doing two jobs. Between vowels, ㄹ is a flap [ɾ] — the tongue taps the ridge once, exactly like the "tt" in American English "butter" or the single "r" in Spanish "pero." It is not the heavy, bunched American "r." As a batchim (final), ㄹ is a clear [l] — 달 is "dal" (moon), 물 is "mul" (water), 서울 is "Seoul."

So 우리 (we) is "uri" with a light tap, and 나라 (country) is "nara" with a tap — but 달 ends in a solid "l."

ㅎ: the breath letter

ㅎ is [h], a plain glottal breath, as in English "hat." 하다 (to do), 호수 (lake), 형 (older brother, said by a male). Between voiced sounds ㅎ often weakens or drops in fast speech (좋아 → [조아], 전화 → [저놔]), but that is a sound-change topic for later; treat ㅎ as a clean [h] for now.

ㅎ also matters because it is the source of aspiration: when ㅎ meets a plain stop across a boundary, they fuse into an aspirated sound (좋다 → [조타]), which is where the aspirated series ㅋㅌㅍㅊ comes from.

Common Mistakes

1. Over-aspirating a word-initial plain stop until it becomes another word. English "t," "k," "p" come out with a strong puff at the start of a word, and if you carry that habit into Korean you can flip a plain consonant into its aspirated twin. Aspirate 자다 (to sleep) and it becomes 차다 (to be cold / to kick) — a different word.

2. Aspirating (or keeping voiceless) the medial stop instead of voicing it. Between vowels the plain stop must switch on the vocal cords. Say 고기 with a hard, puffy middle consonant and it sounds like a non-word "gokhi" instead of "gogi."

3. Pronouncing 시 like English "see." Before ㅣ and y-vowels, ㅅ is [ɕ] ("sh"). 시 is "shee," 셔츠 is "shyeocheu."

4. Using a heavy American "r" for intervocalic ㄹ. ㄹ between vowels is a single light tap, not the retroflex English "r." 사랑 is "sa-ɾang" (a quick flick), and 우리 is "u-ɾi."

Key Takeaways

• The plain series ㄱ ㄷ ㅂ ㅅ ㅈ is voiceless and only lightly breathed at the start of a word — the neutral baseline the other two series modify.
• Plain stops voice automatically between voiced sounds: 부부 = "bubu" ([p]…[b]), 고기 = "gogi," 아버지 = "abeoji." This is one phoneme with two sounds, never a choice — so never agonize over g/k, d/t, b/p.
• ㅅ never voices (no Korean [z]), and before ㅣ/y it becomes [ɕ] — 시 is "shee."
• ㄹ is a light tap [ɾ] between vowels and a clear [l] as a batchim; ㅁ ㄴ are plain [m] [n]; ㅎ is [h].