Talk:Proto-Min

"The Proto-Min finals"

@Kanguole: Hi! I really love your work on this article. I was wondering where I could find a copy of Norman's 1981 "The Proto-Min finals" article, in Proceedings of the International Conference on Sinology (Section on Linguistics and Paleography). I'm interested in this article as there seem to be many Proto-Min forms proposed in that article, but not much luck with Internet search it seems. Thank you! Wyang (talk) 10:59, 20 October 2016 (UTC) (from Wiktionary)

Thanks. I don't think it's online – I've added the OCLC number in case that helps find it in a library. The article doesn't give full protoforms for any words, but it does list a few example words under each final. PM initials and tones of various words are given in other papers. Kanguole 12:09, 20 October 2016 (UTC)

Tones

It is established that Old Chinese didn't have tones. How are both Proto-Min and Middle Chinese independently derived from Old Chinese, yet the tones in both systems map each other so neatly? --188.99.142.6 (talk) 13:18, 29 March 2017 (UTC)

Not quite established, but certainly widely held. The theory is that the same categories existed in Old Chinese, but they were marked by different final consonants rather than pitch contours. The shift between these is believed to have happened some time between the Western Zhou and the Sui periods, but the precise timing is uncertain, and there may have been a lengthy period when both features were present. Proto-Min is believed to have branched off in Han times. Note also that it's believed that the same shift also occurred in neighbouring non-Chinese languages (see Mainland Southeast Asia linguistic area). Kanguole 14:08, 29 March 2017 (UTC)

Old Wu–Min

Wu Chinese § History mentions an ancient Chinese variety called "Old Wu–Min" that was spoken around Jiangsu and northern Zhejiang before Wu Chinese (and is therefore not a direct ancestor of Wu Chinese, but rather a substrate to it). The only way I am able to make sense of the account in that article is that this Old Wu–Min was an ancestor of Min Chinese, and temporally, the fit with Proto-Min appears close. Therefore I have assumed that they are essentially the same thing, even though Old Wu–Min was supposedly spoken considerably to the north of Fujian, and linked the term "Old Wu-Min" to this article, and later (in the section "Substrate influences") equated the two terms explicitly. In particular, I wonder if the dialectal source of the Go-on readings can be equated with Old Wu–Min or Proto-Min judging from the temporal and geographical fit. Please inform me if my deduction was incorrect. The account in Wu Chinese is fairly confused. --Florian Blaschke (talk) 14:20, 2 August 2017 (UTC)

It was Norman's view was that Min was descended from an Old Wu dialect, as the earliest Chinese settlers of Fujian probably came from the lower Yangtze area, and that the Old Wu remaining in the homeland was described by Guo Pu in the early 4th century, but was largely replaced with nothern varieties after the fall of Jin. I have seen statements that modern Wu can be almost entirely derived from Middle Chinese, but can't locate them now. It's certainly the conventional assumption that Go-on reflects southern Early Middle Chinese, while the Qieyun reflects a composite of northern and southern EMC. I've never seen any use of Go-on in Proto-Min work. I agree that the Wu Chinese article is unreliable on this issue. Kanguole 15:33, 2 August 2017 (UTC)

My understanding is that the situation of Wu is similar to that of Yue: Yue can also basically be derived from Late Middle Chinese (= the Tang koine, reflecting the Tang-era wave of immigration), but there is an older, Min-like stratum that may be interpreted as a substratum (reflecting earlier, Han-period waves of migration – I wonder if this earlier stratum can be identified in Old Sino-Vietnamese, early Chinese loanwords in Vietnamese). Does Wu have such a "southern" substratum as well (even if it is not as extensive)? --Florian Blaschke (talk) 13:37, 4 August 2017 (UTC)

Here is an abstract of a paper by Ting Pang-hsin proposing such a substratum, referring to the fact that some dialects in the south of the Wu area have some instances of the merger of dentals and retroflex dentals also found in Min. Kanguole 18:05, 4 August 2017 (UTC)

@Kanguole: What does "southern Early Middle Chinese" mean in this context? Is that a variety with or without the features typical of Min? --Florian Blaschke (talk) 20:51, 8 July 2020 (UTC)

Northern and southern EMC are thought to reflect the reading traditions of the capitals of the late Northern and Southern dynasties, Luoyang and Nanjing, or more broadly Central Plain vs lower Yangtze. That would be without the Min-specific features. Kanguole 21:52, 8 July 2020 (UTC)

Dating of the divergence from mainstream Chinese

The section "Separation from common Chinese" appears to imply that Min diverged already during the Western Han period, so presumably in the first – or even second? – century BC. It probably happened only after southern China was first settled permanently with Chinese speakers starting in the second century BC (as is also implied more directly in the article with the reference to the defeat of Minyue) – the expansion during the Qin period appears to have been only temporary.

However, it seems possible that Old Southern Chinese (as posited by Norman) in the Han period had already diverged from the rest of (Old) Chinese while still spoken only in the lower Yangtze region (specifically in Wu), before its southward expansion. Then, the divergence might be dated to the second century BC – or possibly even to the third.

In any case, the facts listed strongly suggest that Min / Old Southern Chinese had already diverged by the first century AD, so I'm not sure why Schuessler dates the common ancestor of Min and mainstream Chinese to the first and second centuries AD (or even to c. 200 AD, per here – what Schuessler might have had in mind is the wave of immigration following the collapse of the Han, referred to here), into the Eastern Han period, by which time the two listed changes in mainstream Chinese (merger of the finals *-jaj and *-je, and palatalisation of Old Chinese velars in certain environments) are said to have already happened. This implies that the true common ancestor is the language of the Western Han period (after the conventional end of Old Chinese proper around 200 BC, with the end of the Qin). (However, in Minimal Old Chinese and Later Han Chinese, p. 1, Schuessler concedes that Later Han Chinese "could, with hindsight, be considered Middle Han Chinese of the first centuries BC and AD", so he did see the problem too, and came to the same conclusion that his reconstruction did not strictly reflect Eastern Han speech, and therefore, the name given to the reconstruction is not quite appropriate.)

By the way, I'd like to see details or even examples for the palatalisation referred to, because I'm not sure what phenomenon exactly the change in question actually is – is it this one?

It would certainly be helpful for the reader to be more explicit about these points in the article rather than expecting the reader to figure them all out on their own (like I just did, as a non-Sinologist) if the reader is interested in them.

As an aside: I do understand that the evidence for especially the Sui (because of the Qieyun, although I understand it contains older traditions) and Zhou/Qin periods (because of the Old Chinese literature, especially the rhymes of the Yijing) is more accessible (and more directly dateable) and was researched into earlier, but de-emphasising for now the effects of the history of research and the tradition following from it, being trained in conventional (especially Indo-European) historical linguistics, it would make more sense to me to focus first on Common Dialectal Chinese and compare it with the evidence that can actually be dated to the Tang period, to which the reconstruction seems to correspond most closely, and then – with all the Middle Chinese evidence assembled – focus on the reconstruction of the common ancestor of Min and mainstream Chinese ("Proto-Sinitic" or simply "Proto-Chinese"), and compare it with the evidence datable to the Han period, especially from the first century BC, to arrive at a maximally reliable picture of the Chinese of the Han period, as inferred from all the available evidence. Only then I'd address Old Chinese proper of the Zhou and Qin periods, systematically building up from the evidence for the more recent stages. It seems to me that research in historical Chinese phonology has put the cart before the horse to some extent, by trying to reconstruct Old Chinese proper before the Chinese of the Han, and Middle Chinese before Common Dialectal Chinese and the Chinese of the Tang, due to the tradition of neglecting conventional historical linguistic reconstruction in favour of the philological approach. In fact, with the Chinese of the Han, we are on firmer ground than we are for earlier times, and the time around the first century BC is, after all, a key period in Chinese history and literature, so it makes sense to focus one's efforts on figuring out on what the language was like at the time, for now at least, and citing the results as standard reconstructions of Classical Chinese pronunciation in reference works. --Florian Blaschke (talk) 08:55, 11 November 2019 (UTC)

For the first part, you seem to be asking to correlate the linguistic observations of Ting, Scheussler and Baxter & Sagart with historical events and migrations. It's difficult to find sourcing for that. But note that Chinese settlement in the south was quite sparse for a long time, and also concentrated in less mountainous areas.

I've added an example of palatalization of velar initials.

On the last point, that was of course the programme advocated by Jerry Norman. But even for his Common Dialectal Chinese he began with the categories of the Qieyun and combined those not distinguished in modern varieties. I guess the Qieyun is data is irresistable because it's so convenient and the historical interaction of Chinese dialects is so complicated. They didn't keep out of each other's way like Indic, Germanic, Tocharian, etc. Kanguole 12:32, 11 November 2019 (UTC)

Uhh, I can't say that this is the case at all in Indic and Germanic, and many other language families I'm informed about; there's plenty of cross-dialectal interaction going on, wave-model style. Sure enough, directly attested rather than reconstructed historical material is invaluable, but that's why combining evidence is so crucial. When multiple lines of evidence converge, confidence in our result increases accordingly. That's a truism, and no doubt you agree.

I'd say it's methodically fair to start out from the Qieyun and "reconstruct forward" to some extent, in order to minimise later interference. I never advocated pure external reconstruction of proto-stages; I suggested trying to correlate them with directly dateable evidence and in this way correct and enrich them.

That brings me to yet another point that I wanted to make. As I have read, and already mentioned, Middle Chinese reconstructions based on the evidence from the Qieyun and further, more recent sources (the rime tables, especially; supplemented by Sino-Xenic, primarily) reflect a stage that is not contemporary with the sources, but continues even older traditions; therefore, our Middle Chinese reconstructions do not directly reflect a spoken stage, but, apparently, a synthesis between two different traditions from the Northern and Southern dynasties, a northern and a southern one. However, since these traditions likely diverged from an even older common stage, but do not fit Min, it appears that they are more recent, but nevertheless substantially older than the sources. It appears quite possible and even likely to me that their common basis might date back as far back as the Eastern Han itself, shortly after the divergence of Min, in light of the above. Thus, indirectly at least, Middle Chinese reconstructions may well reflect a stage that is not much more recent than Schuessler's (de facto) "Proto-Chinese" (or Late Old Chinese) at all, only a bit. So it seems rather questionable to rely on them so much as standard reconstructions, rather than synthesising the Middle Chinese evidence with the evidence from Min and the other available evidence for the Han period and use the resulting "Proto-Chinese" as the new go-to reconstruction of Classical Chinese original pronunciation instead, the standard citation form. (Without denying that for certain purposes, Middle Chinese will be more suited – especially when the differences between Min and mainstream Chinese are discussed, of course.) Basically, the pronunciation of Old Chinese proper (pre-Han) is too uncertain, and the reconstructions too speculative and full of question marks; and Middle Chinese is too ahistorical; so "Proto-Chinese" seems to me a good compromise, as encapsulating all dialectal and other evidence, whenever the later dialectal developments are irrelevant (such as when discussing pre-Han Chinese, and comparing Chinese to its relatives). Does my argument make sense?

Thanks for your addition. It does help!

Oh, and a further issue, much more directly relevant to the topic of the article: The sentence "Baxter and Sagart suggest that the later part of the Proto-Min period may have overlapped with Early Middle Chinese" makes the reader (well, at least me) wonder about absolute datings again. As discussed, Early Middle Chinese would ultimately seem to reflect a stage spoken in the early first millennium AD, but maybe what is intended here is the middle of the millennium, rather, chronologically speaking. What can be offered in the way of an absolute chronology of the divergence of Proto-Min into (presumably) Coastal and Inland Min (or whatever the primary branching is)? I'd put, as the terminus post quem, the divergence of Min from mainstream Chinese, and given the high internal diversity, a very recent date for the divergence of Proto-Min seems unlikely, but I cannot see a more precise terminus ante quem, save for the Tale of the Lychee Mirror; but a date somewhere in the first millennium AD seems plausible (the schematic diagram in the infobox agrees with this conclusion).

Presumably, "may have overlapped" refers to the palatalisation of dental stop initials (again, examples would be handy, because I can't figure out from Historical Chinese phonology what change is meant), and insinuates that this sound change postdates the divergence of Min from mainstream Chinese, and was shared because of later contact? This seems to imply that speakers of Proto-Min were bilingual in (Early) Middle Chinese; the layer of loanwords from Early Middle Chinese identified by Norman is consistent with this (the subsequent, literary, layer is presumably contemporary with, and from the same source as, Kan-on, Sino-Koreanic, Sino-Vietnamese, and also Common Dialectal Chinese). Therefore, the end of the Proto-Min period in the form of incipient divergence into modern dialect groups would seem to not postdate the Qieyun a lot – although again, the implication is maddeningly left out (an educated Sinologist can figure this out easily, but that's not our target group). Baxter and Sagart would probably put the divergence in question into the mid-first millennium AD, and the palatalisation of dental stop initials might have happened even somewhat earlier than that. A rough absolute chronology at least for the Middle Chinese side should be feasible. At least preliminary estimates for datings do not seem too much to ask for to me.

I know that scholars and also Wikipedia like being non-committal, to be on the safe side, but all the tantalising hints and clues scattered throughout the article, asking to be followed to conclusions, drive me up the wall. Would it really be that outrageous if the murmurs were replaced by more concrete assertions, however strongly qualified? Is it just me who perceives it as an (even if unintended) intellectual scavenger hunt? --Florian Blaschke (talk) 16:39, 11 November 2019 (UTC)

@Kanguole: Can the layers identified by Norman and listed in "Strata" all be reconstructed back to Proto-Min, including the layer from the Northern and Southern dynasties and the literary layer from the Tang? --Florian Blaschke (talk) 15:44, 13 November 2019 (UTC)

Just the earliest two. The others correlate with Middle Chinese. Kanguole 17:09, 13 November 2019 (UTC)

Thanks! I had just noticed the example given in the article, 葛, and it piqued my attention, because the gloss dolichos mystified me. (Apparently Dolichos is meant; however, Wiktionary gives the translation kudzu, i. e., a group of plants in the genus Pueraria. Maybe the everyday sense is different though from the literary meaning in Min – perhaps due to regional flora divergent from the north, and Dolichos being more common in Fujian, if that is actually the case, I don't know; still confusing.) On Wiktionary, I noticed two things: First, literary readings are given only for Southern Min; second, more importantly, they all end in /t/. In contrast, the everyday terms end in a glottal stop. From our article, I learn that Proto-Min final stops have merged into a glottal stop after open vowels in Southern Min. (By the way, too bad the tables don't show any examples from Pu–Xian dialects.) Therefore, it is evident that the literary lexemes were borrowed into Southern Min only after Proto-Min had already split up. This implies that by the Tang period, Proto-Min had already diverged into the modern dialect groups, and therefore gives a more useful terminus ante quem. The terminus post quem is arguably the defeat of the Minyue by 110 BC, and it's so early that it can be treated as certain. (By the beginning of the Eastern Han 25 AD, Min was certainly already distinct, judging from the "two changes" mentioned in the article, and it seems unlikely that Proto-Min had diverged this early into the modern dialect groups, but it's not completely certain.) More useful termini post quem are probably the end of the Han, and the description of Guo Pu, although I'm less certain about them.

Bottom line, dating the divergence to around the publication of the Qieyun, as suggested above, makes sense, but it's only a rough guess. If Norman's third, Qieyun-type layer can be reconstructed to Proto-Min, the divergence probably postdates the end of the Northern and Southern dynasties 589 AD; but you say it cannot. Otherwise, my bet would have been on the Sui period, at least not much earlier, and hardly much later. But so, the divergence could have predated the Northern and Southern dynasties. This could have been a point pretty early in the first millennium, maybe around when Guo Pu wrote.

A terminus post quem specifically for the divergence of Southern Min can probably be given, too (I'm not sure if it can be said to have diverged from Proto-Min directly, because of the Coastal Min proposal), but I'm out of my depth here and can't find anything helpful. --Florian Blaschke (talk) 19:10, 13 November 2019 (UTC)

"But note that Chinese settlement in the south was quite sparse for a long time, and also concentrated in less mountainous areas."

Of course. A most plausible candidate for the origin of Min-type Chinese would be coastal settlements somewhere between the mouths of the Yangtze and the Pearl, possibly beginning as early as the late 2nd century BC (after the conquest of the Minyue in 135 BC, though probably only after 111 BC). Therefore, a beginning of the divergence in the first century BC seems most likely to me. (Ba-Shu might already have diverged around the same time.) However, Fangyan might possibly reflect an older layer of dialects – gradually overlain by dialects descended from Western Han Chinese, a process which according to my hypothesis began just around this time. --Florian Blaschke (talk) 02:19, 22 July 2020 (UTC)

To quote what I wrote above: "Thus, indirectly at least, Middle Chinese reconstructions may well reflect a stage that is not much more recent than Schuessler's (de facto) "Proto-Chinese" (or Late Old Chinese) at all, only a bit."

I just saw that Pittayaporn makes the same point that Qieyun-based Early Middle Chinese is very similar to Later Han Chinese and distinguished from it only by a few innovations here on pp. 56f. (in the sentence spanning both pages). This reinforces my point that when talking about ancient China and especially the Han period, and specifically when giving renderings of foreign names into Chinese, it makes most sense to use Later Han Chinese reconstructions of the pronunciation of the characters – more so than Middle Chinese ones, given that any kind of post-Han pronunciation can easily be derived from them; but either is better than the incredibly annoying practice of using modern Mandarin-based readings for this purpose. Just as a random example, the article Funan completely glosses over the fact that the pronunciation of Chinese has changed radically since the time when 扶南 is attested in Chinese as the name of the polity in question (a fact doubtless not clear to most lay readers), instead of mentioning that the pronunciation of 扶南 in antiquity has been reconstructed as having been something like [po nam], as opposed to modern Standard Chinese Fúnán. --Florian Blaschke (talk) 13:50, 30 August 2020 (UTC)

Schuessler (2007) has relevant remarks too:

On p. 119f.: "The nature and validity of M[iddle Chinese] has been much debated [...]. MC forms are widely quoted, even by critics, for reference and identification of traditional phonological categories. MC [the Qièyùn system] does not reflect a natural language, as many scholars emphasize; [...] The L[ater Han] forms provide a transliteration which is probably closer to some actual language."

(It might be ideal in historical contexts to quote both Later Han and Middle Chinese reconstructions. This would also mitigate the potential risk of circularity in case Later Han reconstructions are partly based on exactly those borrowings the author intends to illuminate, for example.)

On p. 125: "Mĭn dialects apparently split off from the rest of the language, starting with the Qin and Han dynasties (second, even third centuries BC)."

On p. 131: "The Chinese (or Sinitic) branch has today evolved into seven major "dialect" groups, actually "Sinitic languages", which began to diverge during the Han period (ca. 200 BC and after), but most dialects can be traced back to the more recent Tang Dynasty (ca. 600–900) [...]"

Here, Schuessler remarks that the "Old Southern dialect" has left traces in Wu as well as in Gan, Xiang, Hakka and Yue.

By the way, a friend who is trained and still working in an Indo-Europeanist context but has also taken an interest in Sinology lately agrees that it is strange that the attempt to straightforwardly reconstruct "Proto-Chinese" (with or without the Min evidence respectively) has not been made earlier in the history of Sinology. --Florian Blaschke (talk) 03:29, 14 October 2020 (UTC)

Austronesian

I have again reverted the insertion of Austronesian, because it is not mentioned by the cited sources, and there is no evidence of interaction between proto-Min and Austronesian. (There have been proposals of a relationship between Austronesian and Tai–Kadai or Sino-Tibetan, but these are for a much earlier date.) Kanguole 14:03, 24 September 2021 (UTC)

Yeah but there's proof there is still an Austronesian language spoken in Fujian though, and Fujian is close to Taiwan right? And Taiwan has many aboriginal people that still speak Austronesian languages so yeah there is still some proof on that though. I read that on Minyue. Jjdawikieditor103 14:35, 25 September 2021 (UTC)

Taiwan certainly has Austronesian speakers, whose ancestors are believed to have migrated there from the Chinese coast 6,000 years ago. Fujian does not, and there are no linguistic traces of Austronesian there.

As for the specific places you inserted Austronesian:

• The first is a list of neighbouring languages with a Middle Chinese-like syllable structure, with a three-way division of initial consonants and four tonal categories, including a checked tone characterized by a final stop. Austronesian languages have none of that.
• The second was a list of languages mentioned in Baxter and Sagart (2014) as recipients of early loans from Chinese, and thus preserving evidence of Chinese from that period. Baxter and Sagart do not mention Austronesian, and indeed there are no such loans in Austronesian. (Sagart had earlier proposed a common ancestry of Sino-Tibetan, Austronesian and Tai–Kadai, but any such split would be several thousand years earlier than the proto-Min period.)

Kanguole 15:44, 25 September 2021 (UTC)

Yeah but read very carefully at the Minyue culture, then you'll know that it is possibly Austronesian and that the Minyue people share this culture with the aborigines of Taiwan. Jjdawikieditor103 16:11, 2 September 2021 (UTC)

Wikipedia articles are not reliable sources. And this article is about a specific linguistic topic, not cross-strait contacts in the Neolithic. I have pointed out above the specific problems with the two places you have inserted Austronesian, to which you have not responded. Kanguole 12:22, 28 September 2021 (UTC)

The references added (Chang&Goodenough and Chen) are archaeologists rather than linguists, and do not addrees the above objections: they do not claim that Austronesian languages have a Middle Chinese-like syllable structure, nor that they contain early loans from Chinese. Kanguole 15:11, 29 September 2021 (UTC)