Category talk:Proto-Japonic terms borrowed from Old Chinese

Question

@Eirikr Hi. Are there any sources for Proto-Japonic borrowings from Old Chinese? I think this is unlikely, due to lack of contact prior to the Han dynasty in 57 AD. Perhaps you might want to consider Category:Old Japanese terms borrowed from Old Chinese. KevinUp (talk) 18:52, 30 November 2019 (UTC)

@KevinUp: I don't have any explicit sources, and I've been careful to make sure that any Proto-Japonic etymologies tracing to OCH are worded to be clearly speculative. That said, there's a growing body of work suggesting that, prior to the Yayoi Period invasion of / migration to the Japanese archipelago, the pre-Japanese people -- or at least some branch of their ancestors -- were living in proximity to Han peoples. Some authors have mentioned archaeology findings around Dalian and elsewhere on the Liaodong Peninsula (for instance, see Box 11.1 here), others point to Japanese agricultural and other terms with apparent cognates from Tai-Kradai languages further south (for instance, see Vovin's downloadable PPT). While the 57 AD date is the first time we have a clear record of Han contact with people on the Japanese archipelago, we do have strong archaeological evidence that the modern Japanese people derive partially from neolithic migrants (the Jōmon peoples, hunter-gatherers and likely ancestors to the Ainu, estimated arrival 10,000 BCE), and partially just-barely-prehistoric migrants from the mainland (the Yayoi people, farmers with horses and metalwork, estimated arrival 200 BC–200 CE). I've long thought that the Yayoi would have been in contact with OCH speakers in the time before migration. It's been a while since I've dug into my texts; your query is a good prompt to get back to it.  :)

FWIW, here's a somewhat related thread with Wyang: User_talk:Wyang/Archive9#Ancient_nativized_borrowings_from_Sino-Tibetan, regarding a possible OCH resolution for the odd morphophonemics of the OJP term ipi, modern JA 飯 (ii, “cooked grain”). ‑‑ Eiríkr Útlendi │^{Tala við mig} 17:39, 2 December 2019 (UTC)

@Eirikr: So far, there doesn't seem to be any evidence of native Old Chinese words, such as words from Shijing (Classic of Poetry, an important source for OC reconstructions) being related to Japonic. Based on Vovin's downloadable PPT, the source listed for possible Japonic words from Old Chinese is from Fangyan (方言), a 1st century CE dictionary of Chinese regionalisms. Fangyan records regional words that differ from the standard lect of Eastern Han Chinese (an intermediate stage between Old Chinese and Middle Chinese) by recording the phonetics using Han characters. Some of the words in Fangyan have been identified as Austroasiatic, Tai-Kadai, Hmong-Mien (see Behr (2009) [1] - Page 4 shows a hypothetical map of bronze age contact languages of OC based on typological clusters; Page 6 lists possible substrates for Chǔ terms in Fangyan identified by other linguists). Since some of the terms in Fangyan have already been identified as Austroasiatic, Tai-Kadai, Hmong-Mien by other linguists, I think the "Japonic" terms identified by Vovin in Fangyan are likely to be derived from one of these languages. In page 43 of Vovin's PPT, Vovin concludes that a contact (non-genetic) relationship probably exists between Tai-K(r)adai and Japonic (he had rejected this idea 20 years ago in 1994). This is why I think Category:Proto-Japonic terms borrowed from Proto-Kra-Dai is more likely to exist, and any similarities between Old Chinese and Japonic is more likely to be caused by some contact between Sino-Tibetan people and non-Sinitic people from southern China (e.g. Austroasiatic, Kra-Dai, Hmong-Mien) who eventually had some contact with Japonic people (as suggested by Vovin's southern China urheimat for Japonic), rather than direct contact between Sino-Tibetan and Japonic people in ancient times. KevinUp (talk) 19:35, 3 December 2019 (UTC)

@KevinUp: another couple examples of nativized terms occurred to me last night that may be more straightforward: 馬 and 梅. The kun'yomi appear to be essentially alterations of the LTC (or older) readings. Horses in particular were brought over by the Yayoi migration, definitely during the pre-Old-Japanese stage. There is some thought that various /ma-/ sound shapes in reference to horse are evidence of a general Wanderwort (c.f. even English mare), but I haven't run across any similar theories for 梅 (“plum”).

There are aspects of Japonic that occasionally remind me strikingly of Polynesian languages, so the idea of an urheimat closer to Taiwan -- where the Austronesian migration is thought to have begun, c.f. w:Austronesian_languages#History -- has some appeal.

Given what I've understood of the archaeological record, the Yayoi migration was via the Korean peninsula into south-central Honshu. As such, the Japonic predecessors must have traveled from any such southern urheimat, and would have moved through Sino-Tibetan-speaking areas. I don't refute the possibility of Tai-Kradai cognates in the Japanese lexicon, but I also don't see a clear reason to rule out the possibility of Sino-Tibetan cognates that arrived during the Proto or pre-Proto stages of the language. ‑‑ Eiríkr Útlendi │^{Tala við mig} 21:00, 3 December 2019 (UTC)

Oh yes, the relationship between Japonic and Austronesian has been discussed by:

1. Vovin, Alexander (1994) “Is Japanese Related to Austronesian?”, in Oceanic Linguistics, volume 33, number 2, →DOI, pages 369-390
2. Robbeets, Martine (2017) “Austronesian influence and Transeurasian ancestry in Japanese: A case of farming/language dispersal”, in Language Dynamics and Change, volume 7, →DOI, pages 210-251

On the other hand, there is also phylogenetic evidence between mainland Japanese people and Tibetan people from Yunnan:

1. Ya Ping Qian, Zheng-Tao Chu, Qin Dai, Can-Dong Wei, Jia You Chu, Atsushi Tajima, Satoshi Horai (2001) “Mitochondrial DNA polymorphisms in Yunnan nationalities in China”, in Journal of Human Genetics, volume 46, →DOI, pages 211-220

Based on analysis of the mtDNA D-loop region, it is shown that Japanese and Tibetan populations formed a single cluster with a very high bootstrap probability (99%) that is distinct from Han Chinese, Wa, Dai, Lahu populations from Yunnan, so the notion that the Japonic predecessors may have once been present at Sino-Tibetan speaking areas in not unfounded. (Note that the genetic distance between Tibetan and Japanese is very long, even though they are part of the same cluster). KevinUp (talk) 21:58, 3 December 2019 (UTC)

@KevinUp: Forgive me, what's a ""bootstrap possibility"? ‑‑ Eiríkr Útlendi │^{Tala við mig} 22:30, 3 December 2019 (UTC)

@Eirikr: Bootstrapping is a statistical test where the original data (e.g. height from 500 individuals) is resampled by random replacement to estimate a certain property (e.g. mean, variance, etc.). The process involves shuffling the original data and randomly replacing some of the items (e.g. by replacing height of person no. 42, 57 with that of person no. 88, 26) to obtain a new set of data. The purpose of bootstrapping is to check for the stability or consistency of the results. In layman's terms, a bootstrap probability of 99% means that the subject of the study, e.g. height of sample A and B are identical. KevinUp (talk) 04:00, 4 December 2019 (UTC)

Regarding Japanese 梅 (ume), according to Axel Schuessler (2007) ABC Etymological Dictionary of Old Chinese, University of Hawaii Press, →ISBN, page 377, 梅 (méi, “Prunus mume”) may be related with Old Japanese ume₂. There are also other tree names with unknown sources that may have a possible Japanese connection, e.g. 松 (OC *sɢloŋ, “pine”), 柰 (OC *naːds, “some kind of pear”), 柘 (OC *tjaɡs, “some kind of mulberry tree”).

Axel Schuessler cites Marc Miyake (1997) “Pre-Sino-Korean and Pre-Sino-Japanese: reexamining an old problem from a modern perspective”, in Japanese/Korean Linguistics, volume 6, pages 179-211, but I am not able to access that paper. KevinUp (talk) 05:58, 4 December 2019 (UTC)

• @KevinUp: To confirm, you mean that some authors propose that the ZH terms were borrowed from JA? Hmmm. For 松 in particular, I find that hard to accept, as the native kun'yomi of matsu has nothing to do with the (ostensibly) ZH-derived on'yomi of shō. I might accept ZH 柰 (nài) as some kind of shift from JA nashi, although the elision of the medial consonant seems odd. The OJP free variation in む～ vs. う～ in both 梅 and 馬 is unique to these two terms, however, AFAICT, and is very ... unusual, seeming non-Japonic.

Interesting food for thought. Must crash for now.  :) ‑‑ Eiríkr Útlendi │^{Tala við mig} 07:29, 4 December 2019 (UTC)

• One last thought -- I just checked the pitch accent pattern for Japanese 梨 (nashi, “pear”), and it's either 0 or 2, i.e. /nàɕí/ or /nàɕí↓/, from low to high with an optional downstep at the end. Unless there's a record of this term having shifted its pitch accent pattern, this would appear to point away from a match for the ZH reading nài, at least in terms of tonality. That said, I'm aware that /naɪ̯⁵¹/ is the modern CMN reading; I have little detailed knowledge of tonal development. ‑‑ Eiríkr Útlendi │^{Tala við mig} 08:18, 4 December 2019 (UTC)

@Eirikr: Schuessler (2007) suggests a possible Japanese connection, but did not explicitly mention that these ZH terms were borrowed from JA.

There are other tree names which have a possible Japanese connection […] The source of these words is unknown.

I have converted the usage of {{zh-l}} in the examples above to {{och-l}} for clarity because modern Mandarin readings are used by Schuessler for indexing purposes.

• For 梅 (OC *mɯː), Miyake (1997) suggests it is related with Old Japanese 梅 (ume₂).
• For 松 (OC *sɢloŋ), Schuessler (2007) cites Miyake (1997) and claims that it is associated with Old Japanese 杉 (sugi₂, “cryptomeria”) and has superficial similarity with Tai son² ("pine").
• For 柰 (OC *naːds), Ulrich Unger (1983) proposed in Hao-ku. Sinologische Rundbriefe, Vol. 22 that OC *nas could be analogous to Japanese 梨 (nashi).
• For 柘 (OC *tjaɡs), Miyake (1997) compared it with Middle Korean 닥 (Yale: tak). I'm not sure why this was mentioned as having a Japanese connection by Schuessler under the entry for 梅 (OC *mɯː).

I haven't been able to access Miyake's 1997 paper, so I'm unsure whether the suggested relationship is Japonic to Chinese or the other way round. KevinUp (talk) 10:15, 4 December 2019 (UTC)

Thank you for the clarifications. My initial thoughts:

• 梅 (OC *mɯː): OJP ⟨e₂⟩ is reconstructed by some as either /we/ or /əj/, essentially as a glide from back-to-front or open-to-close. The OCH reconstructions both seem to be monophthongs with no glide, but the LTC reconstructions then do have a glide. The OCH vowel of /ɯ/ has an analog in Japonic, so any direct borrowing of a word with that vowel -- in either direction -- would presumably maintain that vowel value. And, I cannot think of any mechanism within Japonic whereby /ɯː/ would shift to /we/ or /əj/, or indeed the monophthongic /e/ in modern Japanese 梅 (ume). This all leads me to think that this may be either a coincidental phonetic / semantic match between Sinitic and Japonic, or a prehistoric borrowing into Japonic after the Sinitic term's vowel value had shifted more towards the LTC reconstruction.
• 松 (OC *sɢloŋ): OJP ⟨i₂⟩ is reconstructed by some as either /wi/ or /ij/, again as a diphthong. I'm unsure how OCH /-l-/ in consonant clusters correlates with OJP pronunciations; if there's a correlation with such diphthongs, the case for these two being related could be stronger. That said, I'm unaware of any mechanism by which Japonic /wi/, /ij/, or /i/ could arise from Sinitic /o/, nor does the reverse from Japonic into Sinitic seem to make any more sense.
• 柰 (OC *naːds): OCH *nas ↔ OJP nasi does look a lot more plausible.
• 柘 (OC *tjaɡs): Hmm. There is one rare reading tsuge, OJP tuge, that has roughly the right shape to match OCH /tjaɡs/. However, the tsuge reading indicates boxwood, not mulberry or indeed any other fruiting plant. See also w:ja:ツゲ. The only fruiting plant that Japanese 柘 refers to with a vaguely-similar pronunciation is the pomegranate, in the irregular-reading term 柘榴 (zakuro). For any mulberry sense, Japanese 柘 appears in the Wamyō Ruijushō with the reading tsumi, now regarded as an archaic/obsolete synonym for 山桑 (yama-guwa, “mountain mulberry”), and likely deriving from verb 摘む (tsumu, “to pick, to gather”). (Interestingly, the MS IME on my computer resolves やまぐわ to either 柘 or 檿, but not the canonical 山桑 spelling listed in my dictionaries.) Nor do any of these JA terms have anything to do with plums, leading me to think that Schuessler's mention in relation to 梅 must have been confused somehow...

Granted, these may have come from, or via, some other separate language(s), which may provide the necessary mechanisms for better sound correlations. Interesting stuff, thank you for this thread!  :) ‑‑ Eiríkr Útlendi │^{Tala við mig} 18:04, 4 December 2019 (UTC)