This is Gustixa & Rhianne’s cover of Keane’s “Somewhere Only We Know.” Rhianne’s pronunciation is close to RP — the “BBC English” of southern England. Below, every line of the song is written twice: on the left in English spelling, and on the right in 正音, sound by sound, as she actually sings it.
正音 (“correct sounds”) is the original form of the Korean alphabet, created by King Sejong in 1443. Modern Korean writing, Hangul, still uses most of its letters. But the original set was designed as a phonetic system: each consonant letter is a small picture of the mouth making that sound. That design is what lets 正音 spell sounds that the Korean language itself does not use.
“Somewhere Only We Know” — English lyrics (left) and their 正音 transcription (right), part 1. ‘Somewhere Only We Know’ 영어 가사(왼쪽)와 正音 표기(오른쪽) — 1면.“Somewhere Only We Know” — English lyrics (left) and their 正音 transcription (right), part 2. ‘Somewhere Only We Know’ 영어 가사(왼쪽)와 正音 표기(오른쪽) — 2면.
(In the brackets, 英 marks a British pronunciation and 美 an American one. Some letters here — ㅿ, ㆆ, ㅸ, ㆄ — are 15th-century letters that modern Korean no longer uses.)
How to read the letters
You don’t need to know Korean to follow the logic. Here is the whole consonant set at a glance, with its sound values:
The 正音 consonant chart: each column is one family of sounds, with the sound value of each letter. 正音 자음표 — 조음 위치별 자음 가족과 각 글자의 음가.
Each column is one family of sounds made at the same place in the mouth; the names on top — wind, tree, fire, earth, metal, water — are the book’s element labels for those families. Read a column downward: the top row is the soft base shape (ng, n, m…), the next row adds a stroke and the air stops (g, d, b…), and the bottom row adds one more stroke for a puff of air (kʰ, tʰ, pʰ…). In the wind column — the s-family — the extra letter is written small underneath the ㅅ instead.
A few shapes carry most of this song.
ㄴ is a side view of the tongue, its tip touching the ridge behind the upper teeth. Add a horizontal stroke on top — and the stroke touches — and you get ㄷ, then ㅌ. The shape is now sealed. Sealed shape, stopped air: these are the plosives [t, d], sounds made by blocking the airflow and releasing it.
ㅅ is a picture of a tooth. ㅅ and ㅿ wear no hat, so nothing blocks the path: the air keeps flowing through a narrow gap. These are the fricatives [s, z].
ㅈ and ㅊ are the tooth shape ㅅ wearing a gat — the black horsehair hat of old Korea, the one you may recognize from Korean period dramas. Here the gat sits touching the tooth. The air is blocked first, then squeezed out through the narrow gap. Stop plus friction: these are the affricates — like the ch in branches.
ㅎ follows the same rule from the other side: its strokes float above the circle without touching it, so the air is never blocked — a fricative, [h].
One rule of thumb covers it all: a stroke that touches means the air is stopped; an open shape means the air keeps flowing.
Plosives, fricatives and affricates: IPA symbols and their 正音 letters. 파열음·마찰음·파찰음의 IPA 기호와 그에 대응하는 正音 글자.
In 正音, ㅅ and ㅿ picture the teeth — and the affricates ㅊ and ㅈ contain the same tooth shape inside them.
In the plosives ㅌ ㄷ and the affricates ㅊ ㅈ, the added stroke sits closed: it touches.
In the fricatives — ㅅ ㅿ, and below — nothing touches the tooth shape. It stays open.
The two “th” sounds
English has two sounds that most alphabets can’t spell cleanly: [θ] as in thing and [ð] as in the. Korean has neither. Yet 正音 writes them: and — a t or d block with the tooth letter ㅿ tucked underneath. The horizontal stroke never touches ㅿ, so by the system’s own logic the block reads as a fricative — which is exactly what [θ] and [ð] are.
Now scroll back up and find the, pathway, something, this in the lyrics. You’ll see the small triangle ㅿ sitting under those blocks, every time.
The transcription above is one page from 《正音으로 듣고쓰다》 (Hearing and Writing in 正音), the book behind this site.
Read more:
Hangul, Upside Down — how each consonant letter pictures the speech organs, with a 3D model.
The International Phonetic Alphabet (IPA) has a famous chart for consonants. Each column is a place in the mouth, from the lips on the left to the throat on the right. The book 《正音으로 듣고쓰다》 puts a 正音 letter into each cell of that chart. This post reads the chart column by column and shows the letters inside real words, from Spanish and French to Japanese and Hindi.
Some letters are old Hangul letters that modern Korean no longer uses (ㆆ, ㅿ, ㆁ, ㅸ, ㆄ and others). When a spelling uses them, we show the book’s own picture instead of typing it. Vowels have their own post, A Compass Made of Vowels; here we look at consonants.
The IPA consonants in 正音. Columns: BL bilabial (both lips), LD labiodental (lip and teeth), D dental, AL alveolar, P-AL post-alveolar, R-flex retroflex, PL palatal, Velar, U uvular, PH pharyngeal, G glottal. IPA 자음표 전체를 正音으로 적은 표 — 왼쪽 두입술소리(BL)부터 오른쪽 목구멍소리(G)까지 조음 위치별 세로줄.
Two keys help in reading it. First, each column is one place in the mouth, and the columns run from the front of the mouth to the back. Second, the characters 木 火 土 金 水 mark the five consonant families of 正音 – wood ㆁ ㄱ ㅋ, fire ㄴ ㄹ ㄷ ㅌ, earth ㅁ ㅂ ㅍ, metal ㅅ ㅈ ㅊ ㅿ, water ㅇ ㆆ ㅎ – the families described in A Year in a Rice Field. The tinted cells hold letters built from two consonants; they are the subject of section 8.
1. Why is there a vowel in a consonant chart?
Look at three columns side by side. In the middle is the velar column (the back of the tongue touches the soft palate): ㄲ k, ㄱ g, ㅋ kʰ. On its left is the palatal column: 끼 c, 기 ɟ, 키 cʰ. On its right is the uvular column: 끄 q, 그 ɢ, 크 qʰ.
So palatal = velar + ㅣ, and uvular = velar + ㅡ. The ㅣ and the ㅡ are not extra sounds. They tell the tongue where to go. ㅣ is the front vowel: the tongue moves forward and up, toward the hard palate. ㅡ is the back vowel: the tongue pulls back, toward the uvula. In the vowel compass, ㅣ is east (front) and ㅡ is west (back).
In a real word a vowel follows the consonant, and the ㅣ or ㅡ joins that vowel in the same block. For example, c + a = 끼 + ㅏ. Since ㅣ followed by ㅏ is ㅑ, the result is 꺄. In the same way q + a = 끄 + ㅏ: the ㅡ and the ㅏ sit together in one block, as in the book’s line below. Only when nothing follows, at the end of a word, is the chart form written as it is. The sections below show how this works inside real words.
From the book: palatal = velar + ㆆ/ㅣ, uvular = velar + ㅎ/ㅡ. With the vowel a, velar 까 가 카 become palatal 꺄 갸 캬, and uvular puts ㅡ and ㅏ in one block. 책 본문 — 경구개음은 연구개음+ㆆ/ㅣ, 구개수음은 연구개음+ㅎ/ㅡ. 모음 a가 붙으면 까·가·카는 꺄·갸·캬이 된다.
Greek shows the same thing. Several Greek letters change their sound before e and i, and the new sounds are exactly the palatal ones – the 正音 letters that contain ㅣ.
Greek letters before e and i (人 = before e, i). Κ [k] ㄲ becomes [c] 끼, Ν [n] ㄴ becomes [ɲ] 니, Λ becomes [ʎ] 리. 유성 = voiced, 무성 = voiceless. e·i 앞에서 소리가 바뀌는 그리스 문자들의 正音 표기 (人 = e·i 앞).
2. Palatal column: 끼 기 키 니 리
Spanish ll is the palatal [ʎ], written 리. In llamar, 리 + ㅏ becomes 랴, so the word is 랴마르. galleta becomes 갈예따 or 가례따.
French gn is the palatal nasal [ɲ], written 니. In agneau, 니 + ㅗ becomes 뇨: 아뇨. At the end of a syllable there is no vowel to join, so the book writes ㄴ with ㆆ at the bottom of the block (Czech kůň). In the palatal column ㅣ and ㆆ go together: ㅣ joins a vowel, and ㆆ takes its place at the bottom of a block. French words like Champagne, Bourgogne and Bretagne end in this sound.
[ɲ]: French agneau and Czech kůň (final ㄴㆆ). [ɲ] — 프랑스어 agneau, 체코어 kůň(받침 ㄴㆆ)의 正音 표기.
3. Velar and uvular: ㄲ ㄱ ㅋ and 끄 그 크
The velar nasal [ŋ], the ng in sing, is ㆁ. The velar fricative [x] is the double letter ㆅ.
The uvular sounds are made even further back. The uvular nasal [ɴ] is ㄶ. Japanese ん at the end of a word is this sound, so in みかん the last block ends in ㄶ.
Japanese ん [ɴ] at the end of みかん is written with a final ㄶ. 일본어 ん [ɴ] — みかん의 끝소리는 받침 ㄶ으로 적는다.
The uvular fricative [χ] is 흐, and the same rule as before applies: 흐 + ㅏ puts ㅡ and ㅏ in one block. The French r lives in this column too. A dictionary writes it [ʁ], 正音 ㅎㄹ, but in real speech it comes out as a trill (ㅀ), as [χ] (흐) or as [x] (ㆅ), depending on the sounds around it. Paris can be written 빻~리, where the ㅎ at the bottom is a voiced h (more on this ㅎ in section 7).
The French r: Paris and bonjour. In real words it is heard as ㅀ [ʀ], 흐 [χ] or ㆅ [x]. 프랑스어 r — Paris와 bonjour의 正音 표기. 실제 말소리에서는 ㅀ [ʀ]·흐 [χ]·ㆅ [x]로 들린다.
The voiced velar fricative [ɣ], the Greek gamma, is 긔: it holds both ㅡ and ㅣ. Inside a word only one half stays. Greek agora keeps the ㅡ, which joins the next vowel; Spanish Higuera keeps the ㅣ: 이계라.
ㅃ is p without a puff of air (as in spin), ㅂ is b, ㅍ is pʰ with a puff (as in pin), and ㅁ is m.
Korean has no f or v today, but 正音 had them in the 15th century: put a small ㅇ under a lip letter and the lips become “light.” ㆄ is f and ㅸ is v. ㅱ is the sound between w and v, [ʋ], as in Dutch wang.
[ʋ] = ㅱ in Dutch, Tamil, Slovak, Ukrainian, Catalan and Khmer words. [ʋ] = ㅱ — 네덜란드어·타밀어·슬로바키아어·우크라이나어·카탈루냐어·크메르어 낱말의 正音 표기.
5. Tongue tip: dental, alveolar, post-alveolar
ㄸ t, ㄷ d, ㅌ tʰ, ㄴ n, ㅅ s and ㅿ z. ㄹ at the start of a block is the quick tap [ɾ].
English th [θ ð] is made with the tongue on the teeth. 正音 writes ㅌ or ㄷ with ㅿ at the bottom of the block. ㅿ is shaped like a tooth, so the picture reads “tongue touching the teeth.”
English think and thank you. 영어 think와 thank you의 正音 표기 — 첫 글자 블록 아래에 ㅿ.
One column to the right sit [ʃ] and [ʒ], the sh of she and the s of vision. The chart writes them as ㅅ and ㅿ with one extra dot. In 正音, one more dot on a vowel adds a y-glide (ㅏ → ㅑ, ㅓ → ㅕ), and the dot in this letter does the same work: s with a y-colour becomes sh. It is the letter used for Sanskrit ś in the Sanskrit post: शक्ति śakti = 샥띠, शब्द śabda = 셔브더.
English has two kinds of l. The clear l is a final ㄹ. The dark l at the end of milk, fall and pull is ㅭ: ㄹ closed by ㆆ.
Dark l [ɫ] = final ㅭ: milk, fall, fool, pull. 어두운 l [ɫ] = 받침 ㅭ — milk, fall, fool, pull의 正音 표기.
The English r [ɹ] is ㆆㄹ, ㆆ before ㄹ. About a hundred years ago, missionaries such as Homer Hulbert already used this pair to write the English r in Hangul. Here is the start of an English song in 正音:
“Somewhere Only We Know” – Gustixa & Rhianne
“Somewhere Only We Know” (Gustixa & Rhianne), first verse. ‘Somewhere Only We Know’(Gustixa & Rhianne) 1절의 正音 표기.
6. Retroflex: adding ㅿ or ㆁ
For a retroflex sound, the tongue tip curls back behind the ridge of the gums. The chart builds these letters out of familiar parts too: retroflex t, d and n take ㅿ at the bottom of the block, and retroflex s, z and ch are ㅅ ㅈ ㅊ with a small ㆁ. (Why ㅿ and ㆁ – see section 8.) Hindi, Swedish and Mandarin all use retroflex sounds.
[ɳ]: Hindi ṭhaṇḍā, Telugu goṇugu, Swedish garn. [ɳ] — 힌디어 ṭhaṇḍā, 텔루구어 goṇugu, 스웨덴어 garn의 正音 표기.
Mandarin zh, ch and sh are written 즤, 츼, 싀, and the Mandarin r [ɻ] is ᄛ, ㄹ with a small ㅇ under it – a letter already described in the 1446 book that explained 正音.
7. Throat: ㆆ and ㅎ
ㆆ is the glottal stop [ʔ]. At the bottom of a block it cuts the syllable short, like the small っ in Japanese あっ!. ㅎ is h.
Is that h voiced or voiceless? Most alphabets do not say. The Roman letter h covers both the voiceless [h] and the voiced [ɦ], and nobody minds. ㅎ is the same kind of letter. At the top of a block it is voiceless: the chart’s [ç], [x], [χ], [ħ] and [h] are all built from ㅎ and ㆆ. But in two kinds of spelling the book puts ㅎ at the bottom of a block for a voiced sound, and there it is read as the voiced h [ɦ]: the French r of section 3 (Paris 빻~리, bonjour), and the deep throat sounds of Arabic in the Pharyngeal column – the voiced [ʕ] and the “emphatic” consonants (tˤ, dˤ, sˤ, zˤ) are all written with a final ㅎ.
A voiced final ㅎ looks strange to Korean eyes at first: in modern Hangul, a final ㅎ on its own is usually read like an unreleased [t̚]. But Korean already voices ㅎ next to voiced sounds. Before ㄱ, ㄷ, ㅈ a final ㅎ makes ㅋ, ㅌ, ㅊ (놓고 [노코]); next to nasals and vowels it melts away (놓는 [논는], 좋아 [조아]). Old Korean even had nouns that ended in ㅎ: 살ㅎ + 고기 became 살코기 (lean meat), and 수ㅎ + 닭 became 수탉 (rooster). A letter that changes with its position is nothing new in Hangul either: ㅇ is silent at the top of a block and [ŋ] at the bottom. So the book adds one small agreement: in these spellings, the ㅎ at the bottom of the block is voiced.
How Korean treats a final ㅎ. Before ㄱ ㄷ ㅈ it makes ㅋ ㅌ ㅊ (놓고 [노코]); with ㅅ it makes ㅆ (닿소 [다쏘]); next to ㄴ and vowels it melts away (놓는 [논는], 좋아 [조아]). 한국어 받침 ㅎ의 변화 — ㄱ·ㄷ·ㅈ 앞에서는 ㅋ·ㅌ·ㅊ(놓고 [노코]), ㅅ과 만나면 ㅆ(닿소 [다쏘]), ㄴ·모음 옆에서는 탈락(놓는 [논는], 좋아 [조아]).
8. Two consonants together: only five pairs
正音 can join two different consonants: side by side, like ㄶ, or one under the other, like ㅸ. The book uses only five pairs.
The pairs come from the five families marked on the chart. Each family has a strong half (yang, +) and a soft half (yin, −), ten parts in all, called the ten heavenly stems (甲 乙 丙 丁 戊 己 庚 辛 壬 癸). On the star, each element controls the next one along a line. A yang stem pairs with the yin stem of the element it controls.
The five pairs are not equally strong. The book gives them a strength from 1 to 5:
Pair strength: 甲己 1, 乙庚 2, 丙辛 3, 丁壬 4, 戊癸 5. 다섯 천간합의 세기 — 甲己 1부터 戊癸 5까지.
戊 + 癸 (5, strongest): ㅂㅍ + ㅇ → ㅸ ㆄ (v, f) and ㅱ (ʋ). These letters were already used in 15th-century books.
丁 + 壬 (4): ㄴㄹ + ㆆㅎ → ㄶ ㅀ ㅭ ㆆㄹ ㄴㆆ ㅎㄹ. ㄶ and ㅀ are still final clusters in Korean today. In this pair the order matters, and each order has its own place in the chart: ㄹ + ㆆ (ㅭ) is the dark l, ㆆ + ㄹ is the English r; ㄹ + ㅎ (ㅀ) is the trilled [ʀ], ㅎ + ㄹ is the French [ʁ].
丙 + 辛 (3): ㄷㅌ + ㅿ → retroflex t and d, and the th of English.
乙 + 庚 (2): ㅅㅈㅊ + ㆁ → retroflex s, z and ch. In Mandarin the ㆁ turns into the vowel ㅢ: 즤 츼 싀.
甲 + 己 (1, weakest): ㄱ + ㅁ → 긔 [ɣ]. It is rare and unsteady: inside a word only 그 or 기 stays, as in agora and Higuera.
The stronger the pair, the more firmly it already lived in written Korean. The weaker pairs use ㅿ and ㆁ, letters that dropped out of Korean, so the book had to search for their sounds.
The two letters of a pair rarely stand together at the start of a block. ㅸ and ㆄ are stacked; 丁壬 pairs sit at the bottom of the block; 乙庚 becomes the vowel ㅢ; 丙辛 and 甲己 split between the top and the bottom of the block. So a 正音 block keeps the shape of an ordinary Hangul block. The bottom row of the full chart names the pair behind each column:
The full chart. Bottom row (천간합 종류 = stem pair): 무계合 戊癸 for the lips, 병신合 丙辛 dental, 정임合 丁壬 alveolar and post-alveolar, 을경合 乙庚 retroflex, 정임合 (ㄴㆆ) palatal, 갑기合 甲己 velar, 정임合 (ㄶㅀ) uvular. 조음 방법별 전체 자음표 — 맨 아릫줄은 각 세로줄에 대응하는 천간합의 종류.
正音, the alphabet King Sejong of Korea made in 1443, sorts its consonants by where they are made in the mouth, and it adds a stroke when a sound gets stronger. If you know the Devanagari consonants in their five rows – क, च, ट, त, प – this will feel familiar. In this post we go through the sounds of Sanskrit and Hindi one group at a time and show how each one is written in 正音. Buddhist words, yoga words and everyday Hindi words can all be written this way.
All tables and examples come from the book 《正音으로 듣고쓰다》. Some of the letters (ㆍ, ㅿ, ㆆ, ㆁ, ㅸ) are old Hangul letters that modern Korean no longer uses, and they need a font with Old Hangul support. When a spelling uses such letters, we show it in the book’s own pictures instead of typing it.
1. The whole system in one table
Devanagari and 正音. Columns, left to right: voiceless unaspirated (무성·무기), voiceless aspirated (무성·유기), voiced (유성·무기), voiced aspirated (유성·유기), nasals (비음, written as final consonants), approximants (접근음), voiceless and voiced fricatives (마찰음). Bottom right: vowels, long and short. 데바나가리와 正音 대응표 — 왼쪽부터 무성·무기, 무성·유기, 유성·무기, 유성·유기, 비음(받침), 접근음, 마찰음. 오른쪽 아래는 장·단모음.
The table keeps the order every Indian school child knows: five places in the mouth (velar, palatal, retroflex, dental, labial) and, for each place, four stops and a nasal. The sections below read it one column at a time.
2. The four stops: voice and breath
Hindi and Sanskrit separate stops in two ways: with or without voice, and with or without a puff of breath (aspiration). 正音 shows both in the shape of the letter.
voiceless, no breath
voiceless, breath
voiced, no breath
voiced, breath
velar
क k = ㄲ
ख kh = ㅋ
ग g = ㄱ
घ gh = 그
palatal
च c = ㅈ + final ㆆ*
छ ch = ㅊ
ज j = ㅈ
झ jh = 즈
dental
त t = ㄸ
थ th = ㅌ
द d = ㄷ
ध dh = 드
labial
प p = ㅃ
फ ph = ㅍ
ब b = ㅂ
भ bh = 브
* च is written with a first ㅈ and a final ㆆ at the bottom of the block (see the table in section 1). The book notes that च can sound like ㅊ or like ㅉ, depending on the speaker’s region and tone.
Voiceless, no breath (क त प) are the doubled letters ㄲ ㄸ ㅃ: a tight, clean sound with no puff of air, like the k in English skin.
Voiceless, with breath (ख थ फ छ) are ㅋ ㅌ ㅍ ㅊ. The breath is one extra stroke inside the letter, so kh is one letter, not two.
Voiced (ग द ब ज) are the plain letters ㄱ ㄷ ㅂ ㅈ.
Voiced, with breath (घ झ ध भ) are the plain letters with a horizontal ㅡ. The ㅡ is the breathy murmur. Before a short vowel it joins the vowel inside the same block, so ध is still one syllable, just as it is one akshara. (What happens before a long vowel is in section 6.)
3. Retroflex sounds and the five nasals
Retroflex sounds (ट ठ ड ढ ण ष) are written with the dental letters plus a final ㅿ at the bottom of the block. ㅿ is an old Hangul letter; here it means “curl the tongue back.” Where IAST puts a dot under the letter (ṭ ḍ ṇ ṣ), 正音 adds ㅿ. So त and ट, or न and ण, never look the same.
The nasals are written as final consonants, at the bottom of the block, one for each place:
ङ ṅ = ㆁ (velar)
ञ ñ = ㄴ + ㆆ (palatal)
ण ṇ = ㄴ + ㅿ (retroflex)
न n = ㄴ (dental)
म m = ㅁ (labial)
4. Semivowels, the three s-sounds and h
य y is the glide in ㅑ ㅕ ㅛ ㅠ (य = 야).
र r is ㄹ. Between two vowels a single ㄹ is a quick tap of the tongue, like the Hindi r.
ल l is marked ㄹㄹ in the table. This is not a new double letter. It means that between vowels, l is written as a final ㄹ in one block plus a first ㄹ in the next block, which is how Korean writes a clear l sound. So धूलि dhūli has only one l, but its spelling 두훌.리 has two ㄹ. With one ㄹ it would read as a tapped r.
व v/w is ㅸ, the old “light b,” or the glide ㅝ when it sounds like w.
The three s-sounds: स s = ㅆ, ष ṣ = ㅅ with the retroflex ㅿ, and श ś = ㅅ with one extra dot. In 正音 one more dot on a vowel adds a y-glide (ㅏ → ㅑ, ㅓ → ㅕ). If ś were plain ㅅ, śa would be 사; but śa sounds like [ʃa], 샤, which has one dot more. That extra dot is what the ś letter stands for. In real words: शक्ति śakti = 샥띠, शब्द śabda = 셔브더.
ह h = ㅎ.
5. Vowels and length: one block, one beat
In Hindi a short vowel (a, i, u) takes one beat (mātrā) and the others (ā, ī, ū, e, o, ai, au) take two. The book follows one simple rule: one 正音 block = one beat.
अ and आ. Short a is written ㅓ and long ā is written ㅏ. So तथागत tathāgata = 떠타거떠, and in Hindi बल bal (strength) = 벌 but बाल bāl (hair) = 발.
Sanskrit grammar since Pāṇini sorts vowels into three grades: the basic vowel, a middle grade (guṇa) and a strong grade (vṛddhi). Modern Hindi, Pali and Bengali have mostly lost this system. One example: in Sanskrit, u and ū differ only in length, but in Hindi short u is [ʊ] and long ū is [uː], so Hindi can tell them apart even without a length mark.
The three vowel grades: basic (Mūla/Hrasva), middle (Guṇa) and strong (Vṛddhi). 산스크리트 모음의 세 등급 — 기본(Mūla/Hrasva)·중간(Guṇa)·강화(Vṛddhi).
6. Breath and long vowels
The beat rule also decides where the breath of a voiced aspirate (gh, jh, ḍh, dh, bh) goes. The rule depends on the vowel that follows:
Short vowel: the ㅡ stays inside the block, joined to the vowel (section 2).
Long ū, ai, au: the breath gets its own beat and moves into the next block as ㅎ. भूमि bhūmi = 부후.미, धूलि dhūli = 두훌.리, भौतिक bhautika = 바후.띠꺼.
e and o are the middle grade (guṇa). Here the breath does not move to a new block; the ㅡ stays inside, as with a short vowel. See vibheda and aghora in the picture.
A voiced aspirate before a long vowel. Before ū, ai and au the breath moves to the next beat as ㅎ (bhūmi, Bhaiṣajyaguru, bhautika); before e and o it stays in the block (vibheda, aghora). 유성 유기음 + 장모음 — ū·ai·au 앞에서는 숨이 다음 박자의 ㅎ로 옮겨 가고(bhūmi, Bhaiṣajyaguru, bhautika), e·o 앞에서는 블록 안에 남는다(vibheda, aghora).
The other way round, a plain consonant plus a two-beat vowel stays in one block. In IAST, ai and au take two letters, but 正音 writes each of them as one vowel, so the syllable fits in one block – just as ā ī ū e o are two-beat vowels written in one Devanagari letter with a mark on top. In modern Hindi, ai and au have become single vowels, [ɛː] and [ɔː], so Hindi ऐ is simply ㅐ, and औ is one vowel block too.
ai and au written in one block: मित्र mitra (friend) → मैत्री maitrī (kindness, 慈). The last line shows how Sanskrit saukhya (happiness) is said in Hindi. ai·au를 한 블록에 적기 — मित्र mitra(친구) → मैत्री maitrī(자애, 慈). 마지막 줄은 산스크리트 saukhya(행복)의 힌디어 발음.
7. Anusvāra, visarga, candrabindu
Anusvāra (ं) takes the place of the next consonant, so it is written as the nasal you actually say. In गांधी Gāndhī the next consonant ध is dental, so the dot becomes the dental nasal ㄴ: ga + ㄴ. Before क-row sounds it becomes ㆁ, before प-row sounds ㅁ. (Japanese ん works the same way.) Because anusvāra also makes the vowel nasal, the book writes the vowel a as ㆍ instead of ㅏ or ㅓ in these syllables – see saṃskṛta in the picture above.
Visarga (ः) is a breath after the vowel. It can be written with ㅎ plus an echo of that vowel: iḥ = 히, uḥ = 후, eḥ = 헤. But it is a short sound at the end of a word, not a full syllable. Visarga means “release,” so the book also uses the wave sign that Koreans like to use in chat: दुःख duḥkha = 두~카.
Candrabindu (ँ), a nasal vowel, is written with a final ㅎ + ㄴ, or with ㆁ as the first letter of the next syllable.
8. Everyday Hindi
Hindi has two sounds that Sanskrit did not write: the flaps ड़ ṛ and ढ़ ṛh. They began as Sanskrit ड ḍ and ढ ḍh between vowels, which grew weaker over time and became a quick retroflex flap. The dot under the letter (nuqta) only became part of Hindi spelling in the late 19th and early 20th centuries. In 正音 the flap is a final ㄹ with the retroflex ㅿ; the breathy one adds a breath mark to it.
Top: the Hindi flaps. Bottom: Sanskrit ऋ r̥ next to र, ड़ and ढ. 위: 힌디어 설탄음 ड़·ढ़. 아래: 산스크리트 ऋ r̥와 र·ड़·ढ의 비교.
Sanskrit ऋ r̥ is an r that works as a vowel. In Vedic chanting it is a trilled r; in classical Sanskrit and fast speech it became a single tap. Today only South Indian chanting keeps the vowel-like r. In Hindi it sounds like ri, so कृष्ण is said like “Krishna,” and in Marathi and Gujarati it sounds like ru. The spellings for kr̥ṣṇa and gr̥ha in the picture show the South Indian chanting sound.
In Hindi, the length of a consonant alone can change the meaning of a word. Double consonants are written by closing one block and opening the next:
Single consonant
Double consonant
पता patā (address) = 뻐따
पत्ता pattā (leaf) = 뻗따
बचा bacā (left over) = 버짜
बच्चा baccā (child) = 벚짜
मल mal (dirt) = 멀
मल्ल malla (wrestler) = 멀르
रस ras (juice) = 라쓰
रस्सी rassī (rope) = 랏씨
Hindi words with long and short vowels and with single and double consonants (meanings in Korean). बाल bāl (hair) = 발, बल bal (strength) = 벌. 장·단모음과 홑·겹자음으로 뜻이 갈리는 힌디어 낱말들 — बाल bāl(머리카락) = 발, बल bal(힘) = 벌.
9. Buddhist and yoga words
With these rules, words that travelled out of India with Buddhism and yoga can be written as they sound in Sanskrit.
Each word in Devanagari, then 正音 in brackets, then IAST and the Korean meaning. Yoga readers will find karma, dharma, dhyāna, saṃskāra and samādhi here. 데바나가리·正音(괄호 안)·IAST·한국어 뜻을 나란히 적은 표 — karma, dharma, dhyāna, saṃskāra, samādhi 등.The three s-sounds ś, ṣ, s and more words, including the Heart Sutra (hṛdaya sūtram). The last line writes pratiṣṭhitaḥ two ways: as it is said in Hindi and Marathi, and as it is said in Nepali, Telugu and Tamil. 세 가지 s 소리 ś·ṣ·s와 반야심경(hṛdaya sūtram) 관련 낱말들 — 마지막 줄은 pratiṣṭhitaḥ의 지역별 발음 두 가지.
A few of them in plain text:
प्रज्ञा prajñā (wisdom) = 쁘러갸 – ज्ञ is a common joined consonant, read as gya or as jña; the picture shows both
हेतु-प्रत्यय hetu-pratyaya (cause and condition) = 헤뚜-쁘러땨야
तथागत tathāgata (the Thus-come One) = 떠타거떠
सुभूति Subhūti = 쑤부후.띠
शक्ति śakti = 샥띠
दुःख duḥkha (suffering) = 두~카
The book adds a note on duḥkha, taken from a Korean study of the Heart Sutra: duḥ means “bad” and kha means “hole,” the hole for a cart’s axle. When the hole does not fit the axle, the wheel creaks, and that creaking is suffering. By chance, the same picture appears in the 1446 book that explained 正音: in its preface, Jeong In-ji describes the pain of a people whose speech did not fit the Chinese letters they had to write it with.
The six chakras. Each petal letter is written in 正音 just outside the petal. 여섯 차크라 — 꽃잎의 데바나가리 글자를 꽃잎 바깥에 正音으로 적었다.
10. Same letters, different regions
Indian languages share almost the same consonants, but their vowels drift as the distance between regions grows. Here is how seven languages say six vowels:
Second column: native speakers (억 = 100 million, 만 = 10,000). “장단구별 약화” = the long–short difference has weakened. 일곱 인도 언어의 여섯 모음 발음 비교 — 둘째 칸은 모어 화자 수.
Because 正音 writes what is heard, the consonants stay the same and only the vowels change to match each region, as the pratiṣṭhitaḥ example above shows. India has also tried to solve this from the other side: the Bharati Script, made by a team at IIT Madras, takes the common structure of Devanagari, Bengali, Gurmukhi, Gujarati, Tamil, Telugu, Kannada, Malayalam, Odia and other Indian scripts and builds one shared script from it. It is still not widely used.
below — nothing touches the tooth shape. It stays open.
