Is there a Korean font that can display arbitrary jamo combinations?

Question

For every possible syllable in modern Korean, a precomposed Unicode character exists.

These syllables have at most:

• 2 choseong (e.g. ㅃ, arguably composed of ㅂ and ㅂ)
• 3 jungseong (e.g. ㅞ, composed of ㅜ, ㅓ, and ㅣ)
• 2 jongseong (e.g. ㄳ, composed of ㄱ and ㅅ)

For older texts, Unicode encodes several dozen sequences of up to:

• 3 choseong (e.g. ᄢ, composed of ㅂ, ㅅ, ㄱ)
• 3 jungseong (e.g. ㅞ, composed of ㅜ, ㅓ, and ㅣ)
• 3 jongseong (e.g. ퟑ, composed of ㄷ, ㅅ, ㄱ)

These do not have precomposed Unicode characters. Instead, the fonts have to be able to “intelligently” compose them into syllables (e.g. ᄢᅰퟑ, which is composed of the nine jamo shown above).

There seem to be some fonts that are able to display these combinations, notably the relevant Noto, Kurinto, or Nanum fonts. Yet these fonts do not allow arbitrary combinations of jamo. Instead, it seems they work on the basis of the existing Unicode jamo sequences. If a single Unicode character exists for jamo that make up the choseong and for jamo that make up the jungseong and for the jamo that make up the jongseong, then it can be displayed as a syllable block. So instead of the theoretical freedom of composing up to 3 choseong, 3 jungseong, and 3 jongseong, we are down to 1 Unicode character for each.

Take for example a sequence of the two choseong ᄂ and ᄀ followed by the jungseong ᅳ. There is a single Unicode character for the choseong sequence ᄓ, therefore the fonts can display the syllable:

• ᄓ + ᅳ → ᄓᅳ (works in Noto, Kurinto, Nanum)

Reverse the order of the choseong to ᄀ and ᄂ, and the fonts will no longer display a syllalbe block:

• ᄀ + ᄂ + ᅳ → ᄀ느 (does not work in any font I know of)

Or a more complicated example:

• ᄀ + ᄀ + ᄒ + ᅩ + ᅥ + ᅡ + ᆯ + ᆺ + ᆮ → ᄀᄀ호ᅥᅡᆯᆺᆮ

So my question is: Are there any fonts where the jamo can be composed freely into syllables? The reason why I am asking is because this would be great for writing other languages with Hangul.

Technically, I think such fonts should be possible, and my computer is even aware that ᄀ느 or even ᄀᄀ호ᅥᅡᆯᆺᆮ form a single block (I can only select the entire unit, and when I move the cursor, it will jump the sequence like one unit) – but there are no fonts that can display them as such.

(I hope I have not mangled up the terminology. I am sorry I do not speak any Korean, I just love the Hangul.)

Answer 1

ᄓ is not an arbitrary combination of jamos, it is a historically-existed jamo and only used by researchers of classics in modern days. That said, ᄀᄂ is not an unusual sequence, it's just not a legit jamo in both old and modern Korean, and thus Unicode does not support it (or more precisely it was never proposed to the consortium as a "character"), and hence none of the modern fonts support glyphs for it. And lastly, ᄀ느 or ᄀᄀ호ᅥᅡᆯᆺᆮ are not single syllabaries. In fact, they are not syllables at all in Korean and I believe the reason the OP can only select them as a whole is the Unicode decoder trying to be smart and failing to understand what actual human characters are. Unicode is not some kind of statutory definition of writing systems that constitute a big part of human languages. Something being "allowed" or "recognized" by a Unicode decoder doesn't necessarily mean that's the right usage of human language. Languages are defined and re-defined every day by actual usage of them. I think Unicode is merely a failed attempt (most successful in history so far though, I admit!) to describe human languages with limited computational tools.

So, the main source of confusion seems that the OP thinks jamos can be freely (and phonotactically) combined into phonetic clusters, but that is not true. That is, for example, even though a jongseong ㄻ

1. looks like a combination of ㄹ and ㅁ, and
2. is named (리을미음) like a combination of ㄹ(리을) and ㅁ(미음), and
3. is typed like a combination on a wide-spread Korean IME,

but it is not a consonant cluster, and it doesn't sound ㄹㅁ (/lm/). It sounds only /l/ or /m/ depending on in which word it's used and sometimes /l.m/ when a choseong-less morpheme directly follows (then it's realized into two syllables). There is no consonant nor vowel clusters in Korean and neither is in Hangul.

Sidenote: that there are other keyboard layouts (example) that actually lay out "composite" jamos as separate keys.

If it is a cluster of any kind, it is a morphological cluster, which is unpacked under certain morphological conditions. To understand this, one might need to understand the debate between formalist orthographic principle vs. phonemic orthographic principle. Modern Korean is using a mixture of two philosophies, and I think it's more on the formalist side, as opposed to, for example, German which has more phonetically and phonologically intuitive orthography (or so I heard; I don't speak German). The usage of all "composite" Hangul jamos is a reflection of the formalist view, namely spellings that show morphological origination rather than actual phonetic realization. Not so surprisingly, even the inventors of Hangul (or more precisely, Hunminjeongeum) weren't very rigid on their stance on these principles (八終聲可足用 vs. 終聲復用初聲, 竝書, 連書), and early documents show different orthographies. The philosophical debate is still going on, especially regarding the translation of foreign names. A longstanding example is a translation of Karl Marx into 칼 맑스 vs. 카를 마르크스. But the "standard" Korean explicitly rules out using composite jamos for phonetic representation of consonant clusters. And the standard Korean is the basis of all current standard computational Hangul handlers, from code points, input methods, to glyphs in fonts.

The only free combination of jamos allowed in modern Hangul is, given three sets of choseongs (첫소리) and jungseongs (가운데소리) and jongseongs (끝소리), a combination of one from each set, and it is called 첫가끝 combination. As many might already know, there is the "Hangul Syllables" block in the Unicode table, providing 11,172 pre-compiled syllabaries of 첫가끝 combinations of jamos that are used in the modern-day Korean written language. For a larger set of jamos and syllabaries that includes those only found in old written Korean, this wiki page could be helpful (spoiler alert: there are 1,638,750 possible 첫가끝 combinations when all historic jamos are included). Note that the list is not final in that, at any point they can dig out a new piece of old document and find an unseen character.

Although Unicode defines a Hangul syllable block as "A sequence of one or more L followed by a sequence of one or more V and a sequence of zero or more T, or any other sequence that is canonically equivalent", (p.142, D134), considering smaller building blocks previously defined as "In Modern Korean, a [choseong/jungseong/jongseong] consists of a single jamo. In Old Korean, a sequence of more than one ... may occur" (pp.141-142, D123, D126, D129, emphasis mine), I'd interpret the D134 definition as rather a future-proof definition for any additional findings they may make in old Korean documents, not is for scripting other languages than Korean.

In conclusion, I don't think there is an openly or commercially available font that supports glyphs beyond the tables in the second link above (namely, jamos that are actually used in the Korean language, either historically or daily). And there's no reason for designers and developers to pursue a "beyond-Korean" flexibility in Hangul fonts, considering the tight coupling between modern Hangul and the Korean language (not to mention technical difficulties to implement input methods and font renderers of that level of flexibility on modern computer text standards developed for Latin scripts from the beginning). Hangul was originally invented as a writing system for Korean, and Korean is the only major language that uses Hangul script still today. That said, Hangul is essentially a part of the Korean language.

The reality is that most fonts don't even support all 첫가끝 combinations because designing and packing over a million glyphs is not a trivial work; see this question (옛한글 on Safari not working)

And finally, a uber-flexible 3-by-3-by-3 Hangul font is theoretically possible, but there are still some practical issues that need to be resolved for the development of such to avoid hurting the usability and readability of Hangul for its primary and native users, Korean speakers. For example, how does this new font visually differentiate non-Korean consonant clusters from those digraph-y, or ligatur-ish, or composite, or whatever not-so-cluster Korean jamos (e.g., ᆯᆨ as /lk/ vs standard Korean ᆰ)? I have no idea how to implement this without somewhat sacrificing the beauty of the square-like assembled rendering of syllables.

Answer 2

See Unicode's "Korean" FAQ at https://www.unicode.org/faq/korean.html#6 where it says: "An L that is not followed by a V should be displayed as if it were the sequence <L, Vf>. A V that is not preceded by an L should display as if it were the sequence <Lf, V>. A T that is not preceded by <L, V> or LV, should display as if it were the sequence <Lf, Vf, T>."

(Here, L, V, T represent Jamo characters for lead, vowel, and trail; LV is a composed "Hangul syllable" character representing the sequence <L, V>; and Lf and Vf are the invisible "dummy" filler characters U+115F HANGUL CHOSEONG FILLER and U+1160 HANGUL JUNGSEONG FILLER, respectively.)

This means that your "ᄀ+ᄂ+ᅳ" example, an instance of <L, L, V>, should display as if it were an instance of <L, Vf, L, V>, with the "dummy vowel" inserted between "ᄀ" and "ᄂ+ᅳ". It further means that the kind of font you are looking for is not Unicode-compliant, and will probably never be produced, at least not by professional font designers who abide by the rules of the game.