I’ve been hijacking the Does 水 mean “beautiful”? thread with a little discovery of mine.
The discovery being there seems to exist triplet forms of many Hanji that originally had /t/ as it’s initial consonant. It seems like /t/ was the original sound in Old Chinese, which evolved into /ts/ and /s/. However, instead of these new sounds replacing the old sound all together, many of the old sounds are preserved in rebus (phonetic loan) or phono-semantic compound characters.
Before going further, I like to point out this is not a Sinitic specific sound change.
Recently went through a youtube video about The prehistory of English by Martin Hilpert, and he pointed out several Indo-european examples of this same sound change.
Previously I found examples on my own,
In English:
/t/ -> /ts/ (in this case /tʃ/)
Statue /ˈstæ.tʃuː/ , which in its original Latin form was statua /ˈsta.tu.a/
other examples in English, some would pronounce Tuesday as choose-day…
Japanese phonology:
/t/ -> /ts/
/t/ > dental affricate [ts]: /tuɡi/ > [tsɯᵝŋi] tsugi 次 ‘next’
for example, when transcribing the English word “two”, Japanese say ツ /tsu/
Seems to be closely related but not exactly the same sound change in Indo-european languages:
PIE *dyéwm 'sky (accusative singular) ’ > *dyḗm > Sanskrit dyā́m, acc. sg. of dyaús, Latin diem (which served as the basis for Latin dies ‘day’), Greek Zen (reformed after Homeric Greek to Zena), acc. of Zeus
We get a /d/ -> /d͡z/ -> /z/ sound change.
Some examples in Taigi (Taiwanese Holo):
/t/ -> /ts/ -> /s/
圵 tong / 仩 tshong / 上 siōng
倘 thóng / 敞 tshàng / 尚 siāng
特 ti̍k / 持 tshî / 寺 sī
辰 sîn / 振 tsín, tín / 震 tín
墜 tui / 墜 tsui / 隧 sui
惰 tui / 墮 tsui / 隋 sui
濤 tô / 鑄 tsù / 壽 siu
錘 tûi,thûi / 棰 tshuê, tshê / 垂 suê, suî
We can also find this phenomenon in Mandarin:
唾 tuo4 / 垂 cui2 / 睡 shui4
提 ti2 / 匙 chi2 / 是 shi4, 匙 shi3
If we compare with Holo and different Japanese Kanji readings, where we have a rough idea which version came first. It seems like the order of sound change is /t/ -> /ts/ -> /s/. Classical dictionaries seem to support this fact as well.
There are 5 ways this phenomenon manifest itself:
- One single Hanji retaining all 3 initial consonant sound changes in a living Sinitic language.
- Hanji triplet retaining all 3 initial consonant sound changes in a living Sinitic language.
- Hanji triplet with all 3 initial consonant sound changes across different living Sinitic languages.
- Hanji pair with only 2 out of the 3 initial consonant sound changes in a living Sinitic language.
- Hanji is described in classical dictionaries with all 3 initial consonant sound changes.
Many cases 3, 4 and 5 could just be because of my lack of knowledge.
some examples of case 1, single Hanji retaining all 3 initial consonant sound changes:
/t/ -> /ts/ -> /s/
峙 tī / 峙 tshāi / 峙 sī
Here are some examples of case 3, all 3 initial consonant sound changes across different living Sinitic languages:
/t/ -> /ts/ -> /s/
提 the̍h / 匙 chi2 in Mandarin / 是 sī
石 dan4 in Mandarin (historic weight unit) / 石 tsio̍h / 石 si̍k
Here are some examples of case 4, pair with only 2 out of the 3 initial consonant sound changes :
/t/ -> /ts/
田 tiân / 田 tshân
重 tāng,tiōng / 鍾 tsiong
童 tông, tâng / 鐘 tsiong, tsing
/t/ -> /s/
單 tan, tuann, siān / 禪 thòo, siân
Examples of case 5, described in classical dictionaries with all 3 initial consonant sound changes.
荼茶
七
十
矢知智雉
珍
陟
直
丈
自
字
者
囟細
In Japanese, the /t/ -> /ts/ rule requires the vowel /u/ to affect the initial consonant. In Sinitic languages, this group of sound changes would appear to be associated with /u/, /i/, /ɨ/, /ʉ/, /a/ and /ia/.
There are 3 things to watch out for when finding these sound changes. First, simplified characters may mislead one to think a character is rebus (phonetic loan) or phono-semantic compound character, when it is in fact a totally different character originally. With simplified, I don’t just mean PRC’s simplification. I mean things like 巖 -> 岩, where you may think 岩 would fit into the 石 triplet somehow, but in fact it is read just as 嚴, it’s original form. There are many many examples such as these.
Second, compound ideographs often doesn’t carry sound from it’s components. It is sometimes easy to confuse it with a rebus or phono-semantic compound character, and as such lead to a wrong conclusion. For example, 員 originally is a circle “O” on top of the “鼎” glyph, to emphasize it’s taking about the circular shape of a bronze cookware. It carries neither sound of 口 or 鼎, and definitely sounds nothing like 貝.
Second, between each transcription from oracle script to bronze script, to the two seal scripts, to Han official script and finally to the classical script, some of the original character is mis-transcribed with a radical that looks similar but different in meaning and pronunciation. Some of it is a form of simplification, but many are just plain wrong. Using the above 員 exmaple, during the many transcriptions someone mistaken the 鼎 glyph for a 貝. To further emphasize it’s a frigging circle, people later added another circle around “the circle” to make 圓. There are many such examples, and somethings causing the character to not look like what it should sound like. Other examples are 思 should be 恖, 細 sould be 糸囟. So neither really carries the 田 sound. 囟 is a picture of the human skull, where the X depicts the fontanelles of a human skull.
