Hacker News new | past | comments | ask | show | jobs | submit login

> No Mandarin goes way further back than Old Mandarin

Can you provide credible references that support your statement as historical studies of phonological and tonal changes in Old (Ancient) and, later, Middle Chinese assertively point to emergence of various Chinese languages that we identify today as Wu, Yue, Mandarin etc to Late Middle Chinese (LMC) (with only exception being Min that had split off from the Old Chinese before Middle Chinese emerged)? In Western linguistics, Late Middle Chinese is identified with late Tang-early Song periods.

Interestingly, Tang imperial annals do, in fact, note already emerging regional pronunociation differences in Middle Chinese across the empire, and, ironically, it was perhaps the first and the last point in the history of all Chinese languages where there existed true dialects of the Middle Chinese language – in the Western sense, anyway.

Even the earliest occurence of what we identify as Mandarin today already indicates the beginning of profound phonological changes, the process that took the next few centuries, namely: a complete loss of checked syllables/contour tones, arbitrary tone changes owing to a loss of nearly all Middle Chinese finals which has also led to an emergence of polysyllabic words in Mandarin: Old Chinese words were nearly entirely monosyllabic and atonic, Middle Chinese words were mostly monosylabic and the tonogenesis had yielded four tones. Conservative varieties of Chinese, e.g. Yue, still have a largely monosyllabic core vocabulary, and Mandarin words, owing to the reduction of phonemic and tone pallettes, are remarkable for having two or three syllables on average. Later influence of Mongolian and Jurchen languages has also had profound effects on Mandarin as we know it today, phonetically and morphologically.

> For example this is how the Song philosopher 朱熹, a Fujian native, spoke in public:

> That's pretty recognizably Mandarin. Even earlier from the beginning of the Tang Dynasty, we have poems that are recognizably vernacular Mandarin.

What distinguishing qualities of the two cited texts make you state your assertions? My paleolinguistic knowledge of Chinese languages is insufficient to discern differences, therefore it is a genuine enquiry on my behalf.

Also, with respect to the Song era, the work «THE NATURE OF THE MIDDLE CHINESE TONES AND THEIR DEVELOPMENT TO EARLY MANDARIN / 中古汉语声调的本质和到早期官话的演变» (https://www.jstor.org/stable/23752830) specifically notes the following (page 187):

3. NORTHERN SONG. Somewhat paradoxically, independent sources for the synchronic state of the language in Song are less satisfactory than for either the preceding Tang or the following Yuan periods. The great Song rhyme dictionaries, such as the Guang-yun and Ji-yun, were revisions and enlargements of the Qie-yun, preserving many distinctions which had long been obsolete. The existing versions of the earliest rhyme table, the Yun-jing, date from Southern Song but the work itself must have originated in late Tang times and its categories are those of LMC rather than the more evolved standard language of Song. Later rhyme tables such as the Qie-yun zhi-zhang tu show some contemporary influence but it is not always easy to distinguish what is merely traditional and what is strictly synchronic. Rhyming in poetry, especially «ci», as opposed to «shi» which continued to be dominated by Tang norms, gives important information but only on limited parts of the phonology.

It would be interesting if you could shed the light on the available evidence (with references) that has shaped your point of view.




So first off I'm explicitly downplaying the phonological side of things here. This entire thread has been about the fundamental grammatical and vocabulary differences rather than phonological differences, i.e. realizing that different varieties of Chinese are not just the same language with different pronunciations sprinkled on top. I'll get to the phonological side later because the story there is more complicated than you're making it out to be.

So leaving aside the phonology, I offer a rough idea here of why the Song vernacular is essentially Mandarin: https://news.ycombinator.com/item?id=29528008

> Conservative varieties of Chinese, e.g. Yue, still have a largely monosyllabic core vocabulary, and Mandarin words, owing to the reduction of phonemic and tone pallettes, are remarkable for having two or three syllables on average.

I don't believe this, at least not without more research. The core vocabulary of modern Mandarin is about as monosyllabic as Yue (pronouns, core verbs, core adjectives, etc. are all primarily monosyllabic). The reason Mandarin is primarily polysyllabic is because of non-core vocabulary, but that is shared with Yue as well.

It is by no means settled in modern Western scholarship or in Chinese scholarship that all varieties other than Min are descended from Middle Chinese. It is best to think of the concept of Middle Chinese (as exemplified by the Qieyun System) as a useful fiction, which works to provide a framework to analyze a lot of historical and modern Chinese varieties, but which does not exist either as a synchronic language ever actually spoken by anybody nor is accurately thought of as a true linguistic ancestor to modern varieties.

I unfortunately am shooting from the hip here and don't have time at the moment to lay out sources and thoughts in more detail. Please let me know if you'd like them and I'll lay them out more thoroughly when I next can get to a computer.


> So first off I'm explicitly downplaying the phonological side of things here. This entire thread has been about the fundamental grammatical and vocabulary differences rather than phonological differences

You are single-handedly dismissing the whole body of research that has allowed us to arrive at the realisation and evidence of the historical progression of Chinese languages (please do note the use of plural) in the first place. Owing to the isolating nature of all Sino-Tibetic languages and, critically, a millenia long use of the logographic script that has no other known precedents, the phonological and comparative regional linguistic studies have been the single main instrument that has made the reconstruction of Old (Ancient) and Middle Chinese languages possible. Which also includes the understanding of the grammar of Old Chinese and which was vastly different even from that of Middle Chinese.

> So leaving aside the phonology, I offer a rough idea here of why the Song vernacular is essentially Mandarin: https://news.ycombinator.com/item?id=29528008

No. In the context of this particular conversation, leaving out the phonology is not possible due to the peculiarities I have mentioned above. It is also not possible to provide a self-reference as a reputable reference or a proof of your own claims.

I have collated a couple of responses you are self-referencing to in another thread below, anyway.

> 2. The near demonstrative is 這 (also spelled 者, 遮, or otherwise) and the far demonstrative is 那. While other demonstratives exist (as is the case even in modern Mandarin) these are far and away the most popular, regardless of region.

The source of https://humanum.arts.cuhk.edu.hk/Lexis/lexi-mf/search.php?wo... does not support your assertion concerning 那 and states that the use of 那 as a distal demonstrative had been established by the Tang era, i.e. before the emergence of Middle Chinese.

> 4. The copula is always 是.

The source of https://humanum.arts.cuhk.edu.hk/Lexis/lexi-mf/search.php?wo... contradicts your statement and states that 是 had already been used as a copula during Han dynasty times. Which also predates the emergence of Middle Chinese:

至於「繫詞」意義的「是」,最早出現於戰國的《孟子.告子上》:「公都子問曰:『鈞是人也,或為大人,或為小人,何也?』」趙岐注:「鈞,同也。」意謂公都子問道:「同樣是人,有些是君子,有些是小人,甚麼緣故?」(楊伯峻)。東漢漸多見,如《論衡.死偽》:「余是所嫁婦人之父也」。佛經傳入以後,才全面流行。

> I don't believe this, at least not without more research. The core vocabulary of…

Please do provide references to peer reviewed scholar views and research that can substantiate your beliefs. I am eager to read through them and contemplate.

Summing it up, so far you have presented an alternative timeline of the history of Mandarin that has had a supposedly continuous history of existence starting from (or even predating) early Middle Chinese. Mainstream scholar views and scientic studies as well as existing evidence do not support your theory which is a fringe theory at best. You have also eschewed any reputable references that would add credibility to statements you have made so far. Therefore, I propose we settle on this: you have provided opinions that are of your own, and that is that. Which is not to say that your knowledge of Mandarin appears to be remarkable and is commendable irrespective of however you choose to use it to suit your narrative.


PART 3:

> Please do provide references to peer reviewed scholar views and research that can substantiate your beliefs. I am eager to read through them and contemplate.

I have to go so I don't have time at the moment to give you a thorough set of citations for this, but the easiest off-hand reference for "core vocabulary" I can think of is to just look at the 207 word Swadesh list for Chinese. I'd try to give you more authoritative links, but I don't have time at the moment. Maybe I'll circle back if there's interest.

There's two versions a quick search brought up. Either the 207 word list here https://zh.wikipedia.org/zh-hans/%E6%96%AF%E7%93%A6%E8%BF%AA... or the one here https://en.wiktionary.org/wiki/Appendix:Mandarin_Swadesh_lis.... The former in my opinion undercounts the number of polysyllabic words, and the latter overcounts it (by using pretty formal language such as 恒星 for star), but between the two you're looking at something like 65-80% of the core Swadesh vocabulary being monosyllabic in modern Mandarin.


Going off votes, it appears that there are other folks interested in this thread so I'll add an addendum here.

Let's begin with

> historical studies of phonological and tonal changes in Old (Ancient) and, later, Middle Chinese assertively point to emergence of various Chinese languages that we identify today as Wu, Yue, Mandarin etc to Late Middle Chinese (LMC) (with only exception being Min that had split off from the Old Chinese before Middle Chinese emerged)

I'm going to need to see citations from you in turn for this. In particular I'd be shocked to see any scholar in the last 50 years postulate that Wu split from LMC (again keeping in mind that "Middle Chinese" is basically more of an analytical shorthand than a real language).

To refresh my memory I decided to look at Chappell's introduction to Sinitic grammar: Synchronic and diachronic perspectives (2001), "Synchrony and Diachrony of Sinitic Languages: A Brief History of Chinese Dialects" (which conveniently is included in many course syllabi and hence the introduction can be found online in many places such as here: http://bartos.web.elte.hu/ma/chappell-dial-intro.pdf) and the situation is as crazy as I remember.

This is her summary of Xiang:

> The language of the Chu kingdom, an important dialect before the Jin dynasty was established in the third century CE, was spoken in the area of modern-day Hunan and Hubei. It appears not to have been mutually intelligible with the court language of the Zhou, Xià 夏, the goal of reconstruction for Archaic Chinese. You (1992) regards Old Chu as the basis for Old Xiang which probably split off from Medieval Chinese some time prior to the Tang dynasty. [pg. 10]

This is her summary of Northern Gan:

> Northern Gan was probably formed during the period between the end of the Eastern Han and the beginning of the Tang dynasty (third to seventh centuries CE) [pg. 11]

This is her summary of Wu and Min:

> The less contentious standpoint would be to posit a common ancestor for Wu and Min, formed by the first century CE, from which both split off. [pg. 12]

This is her summary of Mandarin (which she uses interchangeably with 北方話)

> The concept of běifānghuà 北方話 ‘northern speech’ is first mentioned in the dialect work by Guo Pu 郭璞 known as Fāngyán Zhù 方言注 (Commentary on the Fāngyán) and compiled during the Eastern Jin dynasty (317-420 CE). Certain dialect words, listed in earlier works such as the first century Fāngyán 方言(Dialects) by Yang Xiong, are given in this later work by Guo Pu as the common term in northern speech (You 1992: 94). This is interesting in that it suggests some unification of the northern dialects of Chinese had already taken place by this time - the period of Early Medieval Chinese in the fourth and fifth centuries CE. [pg. 10]

This is her summary of Hakka:

> In the first view, espoused by Lo (1933), Hakka was already formed in the north in the area of present-day Henan province, before the first southwards migration took place in the fourth century CE following the fall of the Western Jin dynasty in 313... In the second more plausible view, Hakka evolved in the south in the period of the Song dynasty, specifically in southeastern Jiangxi and western Fujian, after migration had taken place [pg. 15 - 16]

This is her summary of Yue:

> The Yue dialects were thus formed in a language contact situation with non-Han peoples in a frontier area, first annexed as Chinese territory in the Qin dynasty (late 3rd century BCE), as outlined above. [pg. 16]

The language situation is both far older and less well-ordered than you are making it out to be. Also just to underscore how "Middle Chinese" is really just an analytical tool than a real language, note that Chappell also almost never talks of "Middle Chinese" other than the "Middle Chinese phonological system" and instead is careful to talk about "Medieval Chinese" instead.

Also profuse apologies to Sagart! I apparently have been misremembering his position and have been confusing it with authors he was arguing against!

But to sum it all up yes, I will stand by what you are asserting is a "fringe theory," but which ample evidence suggests is fairly mainstream, or as you put it

> Summing it up, so far you have presented an alternative timeline of the history of Mandarin that has had a supposedly continuous history of existence starting from (or even predating) early Middle Chinese.

The only contention is to what degree what we call Mandarin existed pre-Tang (as Chappell suggests, there is evidence, but it gets very fragmentary very quickly). But Norman provides clear evidence by the 8th century (which he hedges with 8th or 9th, but a lot of the texts that he is transitively citing are 8th century texts) there is clear evidence of a mature and distinct Mandarin family and there is fragmentary evidence that would push the beginnings of the Mandarin family even earlier (Norman believes no earlier than the beginning of the Tang, which I can accept, but again other authors will find glimmers of early Mandarin even earlier).

And I stand by my original comment to hker, which is that Mandarin has been the spoken standard (or different varieties have been simultaneous standards depending on specific times) for literate speakers across the Chinese empire for > 1000 years (certainly by the Northern Song).

P.S.

There is one clarification I'll make to Chappell's summary here:

> By the time of the Southern Song, early Mandarin extended only as far south as the middle and lower reaches of the Yangtze, apart from one small pocket where Northern speakers had crossed the Yangtze into the Northwest Xiang area after the Ānshǐ 安史 turmoil of the mid-Tang dynasty [pg. 11]

Apart from the implication here that Mandarin was distinct and already existed by the mid-Tang dynasty, the reason why I brought up the fact that 朱熹 is a Fujian native (and spent the overwhelming majority of his life there) is that he is a good demonstration that while Mandarin may not have been the first language of a lot of people in China during the Song, it was the lingua franca. Indeed while Fujian does lie outside the region Chappell is talking about, and 朱熹 does make reference to a distinction between proper speech and the local language, the fact that his vernacular writing is effectively the same as elsewhere in the empire is strong evidence for a Mandarin-based lingua franca among at least literate speakers.

And finally for a more authoritative source that the core of Mandarin vocabulary remains mostly monosyllabic than "look at Swadesh word lists", there's 论汉语的单音词 (2009) by 李如龙 who states in no uncertain terms:

> 越是常用的词 ,单音词所占比例越大。在500 个高频词中 ,单音词的总数比双音词多出一倍。一千个高频词之后双音词才大幅度增加 , 到 1500号 ,双音词才和单音词平分秋色 ,数量相当。


PART 2:

So let's return to my assertion that often times Mandarin is studied as the standard Song vernacular and has roots that date back much further.

Norman explicitly calls out Mandarin as the predominant language of the Song Dynasty and dating to the Tang.

"The first type includes the Jianghuai dialects and Southwest Mandarin and may be traced to the southern type of Mandarin which flourished during the Song-Yuan Period. The second type includes the rest and may be traced to the northern type of Mandarin of earlier times.... It is also shown that Mandarin as a dialect group was already in existence in the eighth or ninth century, that is, in the middle of the Tang dynasty." [Norman, "Some Thoughts on the Early Development of Mandarin", In Memory of Mantaro Hashimoto, 1997, pg. 21-28]

But I believe you are also missing my point in linking to my other comment.

Everything I've listed there indeed has a pre-Song history (most from Sui/Tang, some from the Northern/Southern Dynasties, 是 as a copula Mair will argue comes even from the Warring States period). They are also why some studies of Mandarin will take Mandarin to have started in the Tang (or earlier!). But taken as an ensemble they are the DNA of what structurally (rather than just phonologically) distinguishes Mandarin from the other Sinitic families and the fact that they are ubiquitous across Song vernacular from all parts of China make it pretty overwhelming that the standard Song vernacular is structurally a Mandarin language, rather than any other Sinitic family. This is a harder case to make pre-Song (but still can be and has been made), but it's very clear by the Song.

yorwba's question is a very reasonable one precisely because the things you've raised are not uniform across all Sinitic families. 那 is not the usual demonstrative in a lot of these families (rather we have in various cases including Min and Wu, where we have demonstratives such as 彼 or 許 alongside some varieties of 爾). Likewise 是 is not the usual copula in some Sinitic families (e.g. Yue where the usual copula is 係). Another hallmark is pronouns where 你 and especially 他 are not the usual second and third person pronouns (the third person is especially different, where it is 佢 in Yue or other various versions of 渠 or 伊 in other families).

This is what we mean when we say that the different Sinitic families differ in grammar and structure rather than just phonology, and why it is possible on the basis of this grammar and structure alone to identify something as Mandarin.

More to the point we can and do analyze the grammar and structure of various medieval Sinitic families separate from their phonology. Open up any post-Tang period-specific textbook and that is exactly how it is laid out (usually the first chapter is phonology and then the rest, vast majority of the book is grammar and structure). We can do this as soon as we have a flourishing vernacular body of writing, which starts roughly in the Tang (with fits and starts before that starting in the Three Kingdoms), but really hits its stride in the Song. The only reason we have to resort to phonological-specific measures for OC is because it is unclear whether we have any vernacular writing from that time period and thus we have to use phonological hints to try to deduce to what extent a diglossia existed.

And what of the particulars of my other comment? My assertions about the pronouns, demonstratives, etc? Well you can find those in any intro Song vernacular textbook. For example I happen to have 宋代語言研究 (2001) by 李文澤 on my bookshelf at the moment and I can back up everything I wrote off the cuff there in his 語法編 chapter starting on page 214. You could also cross-check against the 近代汉语概论 by 袁宾 (which regretfully I do not have on-hand at the moment so I can't give you specific page citations). It's also blatantly obvious as soon as you start reading various Song Buddhist dialogues, 理學 dialogues (not their Classical Chinese prose works but their vernacular dialogue collections), 話本, particularly vernacular 宋詞, or any other Song vernacular works that what I'm saying there is borne out.

> the phonological and comparative regional linguistic studies have been the single main instrument that has made the reconstruction of Old (Ancient) and Middle Chinese languages possible. Which also includes the understanding of the grammar of Old Chinese and which was vastly different even from that of Middle Chinese.

First a correction: The knowledge we have of OC as a living language is extremely limited. Hence it is debatable whether we have any good understanding of the grammar of OC (the Chinese side of scholarship usually rejects this and says we do have a good understanding of OC through Zhou texts) in the same we have for MC. What Wikipedia often calls OC is debated whether it was actually ever spoken by anyone, or is just an artifact of a Classical Chinese that was in fact never a living language. Mair makes this argument forcefully in his seminal 1994 article "Buddhism and the Rise of Written Vernacular in East Asia:"

> Linguistic data that LS [Literary Sinitic, i.e. Old Chinese as represented by Warring States and Sprint-Autumn texts] and VS [Vernacular Sinitic, i.e. the true living, spoken language] have been distinct systems as far back as they can be traced. This is certainly true from the Warring States period on, but I suspect that eventually we will be able to demonstrate conclusively that LS, starting with its earliest stage in the oracle shell and bone inscriptions (around 1200 B.C.E.), was so obviously abbreviated and so replete with obligatory nonvernacular conventions used only in writing that it never came close to reflecting any contemporary living variety of Sinitic speech. [Mair, "Buddhism and the Rise of the Written Vernacular in East Asia," 1994]

Now I'm personally not sure how much I believe that (and e.g. the Chinese side of scholarship generally rejects this argument), but I bring it up to underscore again the reason we cannot do a phonologically independent analysis of living (at the time) Sinitic varieties before the Tang era (well kind of, we have some vernacular before then, but certainly by the Han) is because it's not clear we have any direct written evidence of them, whereas by the Song it's extremely obvious that we have a fairly true-to-life structural reflection in writing of how people actually spoke, and hence we do not need to rely solely, or even mostly, on phonological analyses.


> You are single-handedly dismissing the whole body of research that has allowed us to arrive at the realisation and evidence of the historical progression of Chinese languages (please do note the use of plural) in the first place.

PART 1:

Oof. Fighting words. Nothing I've said here is outside the bounds of mainstream Sinology. Now that I'm in front of my computer and bookshelf I can break out the citations.

Let's restate my original assertion about MC. MC (as traditionally represented by the Qieyun System/QYS) is a convenient analytic fiction, which provides a useful and comprehensive set of categories to study the history of various Sinitic families. However, MC itself was unlikely to ever exist as a real, synchronic language spoken by anyone and likewise any talk of a genetic linguistic relationship in the classical root-stem linguistic sense between non-Min Sinitic families and MC is also a convenient analytic fiction. The idea that MC was a real spoken language and represented a true linguistic ancestor of today's non-Min Sinitic families was an idea championed by Karlgen in the 19th century but has fallen out of favor in the last 50 years.

So let's begin with the citations from various giants in the field of Sinology.

First, Baxter in his treatment of Middle Chinese explicitly calls this out:

> I emphasize again that the Middle Chinese transcription proposed here is not intended as a reconstruction of any synchronic state of the Chinese language. A number of its notations are merely representations, more or less arbitrary, of distinctions which are preserved in the Chinese phonological tradition. Indeed, given the fact that the Qieyun probably represented more distinctions than were preserved in any single dialect, it may be that no true linguistic reconstruction should include all of its distinctions. [Baxter, "A Handbook of Old Chinese Phonology", 2010, pg. 30]

What of the notion that there exists a true, traditional root-stem genetic relationship between non-Min Sinitic families and MC in a similar sense of Latin and the Romance languages? Pulleyblank and Hashimoto basically reject the idea that thinking of a genetic relationship is productive at all, or that at least it should be tempered with extreme caution. Even authors who try to study genetic relationships will acknowledge that it is an analytic convenience to think of some entity known as MC as the genetic ancestor and that on closer inspection, the genetic relationship often goes back to OC or further.

From Pulleyblank:

> This [various influences of Old Chinese] suggests that the effort to delimit clear boundaries between proto-Min, proto-Wu, proto-Yue, etc., to which several of the conference papers address themselves may be misguided. I agree with Mantaro Hashimoto that a strict Stammbaum model [i.e. genetic model] is quite inappropriate for studying the history of Chinese dialects. Some kind of network model, with provincial and regional centers of influence as well as successive national centers of influence in the form of standard languages based on imperial capitals, seems to be called for. [Pulleyblank, "Chinese Dialect Studies," 1991]

Ballard goes further and says to the extent that you can apply the genetic model, the various Sinitic families are best thought of as descending from proto-Sinitic, not anything at all in the vicinity of Middle Chinese.

> Thus, Wu, Cantonese, Chu, and Min, traditionally regarded as being divergent dialects derived from Ancient Chinese [Middle Chinese] (Archaic [Old Chinese] in the case of Min) actually represent separate linguistic traditions that have incorporated much Chinese material. [Ballard, "Aspects of the Linguistic History of South China", 1981]

Indeed to the extent that people are willing to talk about genetic relationships, it is a much more confused and crazy story than "non-Min languages are descended from MC." Sagart claims that Wu, Xiang, Yue, and Gan predate Min. Chappell claims that Wu and Min split in the Han Dynasty, but the other families split in the Sui-Tang. (Going off memory, I don't have the time to look them up at the moment, but I can if you're interested) I'm inclined to agree with Pulleyblank and Hashimoto here and say that a genetic model is simply inadequate and obscures more than it clarifies, all the more so because we don't believe anyone actually spoke MC!

I'm going to stop there because I hit the comment length and then return to the question of whether the standard Song vernacular is justifiably called Mandarin.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: