Why Do People Think ㅿ was /z/ in Middle Korean?

[Moose-tache]

So I've been reading about the history of Korean, and I always knew that the triangle consonant ㅿ, or "half J," was a voiced fricative /z/. That's just a truism; everybody knows it. But recently when I started looking into transcription methods for pre-Hangeul Korean scripts and early Hangeul, I noticed that ㅿ is primarily used to transcribe Chinese roots beginning in ɲ, and a few in j. Later ㅿ merged with ㅇ, just like initial ŋ did. The evidence that ㅿ = ɲ seems so obvious that I am honestly wondering how the idea that ㅿ = z ever came about. I haven't been able to find any literature giving coherent reasons for the standard etymology, just side comments that take the z pronunciation as a given.

[Raholeun]

Are you sure it ㅿ is stated to be the phoneme *z? Studies of proto-phonemes often do not have a logical correlation between the original sound and its transcription.

[Ketsuban]

Is ㅿ used to write the Chinese sounds or the Korean pronunciation of Chinese loanwords? Early Middle Chinese /ɲ/ developed into /ʐ/ in northern dialects, so in the former case a Korean /z/ being used to represent this sound makes some amount of sense to my mind. (The small number of cases of /j/ could be the result of borrowing between Chinese dialects.)

[Kuchigakatai]

I don't know anything about the topic, but I notice the Wikipedia article on Hangeul has an unsourced description of ㅿ that might be useful:

• ㅿ z (bansiot 반시옷 "half s", banchieum 반치음): An unusual sound, perhaps IPA [ʝ̃] (a nasalized palatal fricative). Modern Korean words previously spelled with ㅿ substitute ㅅ or ㅇ.

If it was a nasalized [ʝ], that would allow it to render Chinese [ɲ] and [j]. If it's true it sometimes merges with ㅅ [s], then that makes it more likely it was some kind of voiced coronal fricative too. The name ban.si.ot "half s" is also significant, as well as ban.chi.eum (半齒音 half-tooth-sound) where chi.eum seems to refer to coronal affricates and fricatives (in modern Hangeul, ㅈ ㅊ ㅉ ㅅ ㅆ).

[Pabappa]

Are we sure of the pronunciation of the contemporary Chinese? Reconsturctions of Chinese are difficult because there's no external representation, and phoneme inventories get harder and harder to believe (for me at least) the further back we go. THis is common also to https://en.wikipedia.org/wiki/Proto-Hmo ... #Phonology and some others. Contemporary Chinese may not have been unified to begin with, as, for example, the invewntory listed at https://en.wikipedia.org/wiki/Old_Mandarin#Phonology has no postalveolars, unless they are being analyzed as /nj/, etc ... but https://en.wikipedia.org/wiki/Middle_Chinese#Phonology (which despite its name is older) does have them and has a lot more consonants in general, and also includes an explicit /ʐ/.

[Kuchigakatai]

Are we sure of the pronunciation of the contemporary Chinese? Reconsturctions of Chinese are difficult because there's no external representation, and phoneme inventories get harder and harder to believe (for me at least) the further back we go. THis is common also to https://en.wikipedia.org/wiki/Proto-Hmo ... #Phonology and some others. Contemporary Chinese may not have been unified to begin with, as, for example, the invewntory listed at https://en.wikipedia.org/wiki/Old_Mandarin#Phonology has no postalveolars, unless they are being analyzed as /nj/, etc ... but https://en.wikipedia.org/wiki/Middle_Chinese#Phonology (which despite its name is older) does have them and has a lot more consonants in general, and also includes an explicit /ʐ/.

Chinese medieval rimebooks are a better resource than you probably think, so yes, scholars are pretty confident in their understanding of the development of Early Middle Chinese down to Old Mandarin and further so to Beijing and Nanjing Mandarin.

What happened is that Middle Chinese underwent a beautiful chain of continuous mergers among onset consonants, greatly reducing the number of distinguished onsets. The only new consonants in Old Mandarin are /f/ and /ʋ/, and in some sense, /ɻ/. The phoneme /f/ developed from Early Middle Chinese /pʰj pj bj/, /ʋ/ from /mj/, and /ɻ/ is quite simply the continuation of /ɲ/. All other Middle Chinese consonants that you feel are "missing" in Old Mandarin (including /ʐ/) aren't there because they merged into some phoneme or other.

I could go into more detail on the great consonant mergers of Middle Chinese over time, but such a discussion is not going to help Moose-tache regarding the nature of Middle Korean ㅿ.