Contact between Chinese and Vietnamese: an over-researched field?
[Blogmaster’s note: I’m overjoyed by Alexis Michaud’s proposal to write a guest post for this blog. A skilful phonetician and determined fieldworker, Alexis will give an important legacy for the generations to come in terms of documentation and has a breadth of knowledge never failing to surprise, which the readers can see for themselves in this wonderful piece.]
- English
- 汉语与越南语的接触史:依然大有可为
- Sự tiếp xúc giữa tiếng Trung và tiếng Việt: một lĩnh vực nghiên cứu còn rộng mở
This post is about a research topic that appears to belong squarely in the “Indo-Sinica” sphere: contact between Chinese and Vietnamese. That field could appear as over-researched and not at all a priority. Many highly endangered languages in Vietnam are likely to disappear before a good linguistic documentation is recorded; by contrast, the conspicuous Chinese component in Vietnamese has been the object of studies by Nguyễn Tài Cẩn, Wang Li 王力 and many others. There exist several dictionaries of Vietnamese readings of Chinese characters (the “Sino-Vietnamese” layer in Vietnamese: Hán-Việt 汉越), such as the wonderful 1932 dictionary by Đào Duy-Anh. But Đào Duy-Anh and Nguyễn Tài Cẩn belonged to the last generations of Vietnamese scholars who had excellent command of Classical Chinese. Following the demise of Chinese-language schooling in the early 20th century, Chinese philology has become much less accessible to Vietnamese scholars. This, together with general difficulties of the Vietnamese higher education system in recent decades, explains why there are relatively few specialists of Sino-Vietnamese within Vietnam.
On the Chinese side, relatively few linguists have followed in the footsteps of Wang Li 王力. My impression from exchanges with Chinese friends and acquaintances is that history textbooks, media and propaganda (宣传) depict Vietnam as a rebellious and secessionist Chinese province; and the Vietnamese as ungrateful neighbours, who received support from the PRC during their fight against Western imperialism and reciprocated by turning their weapons against the Chinese. Maybe this carries over to some extent into scholarly circles, explaining why few colleagues are attracted to the study of Vietnamese and Sino-Vietnamese?
Finally, in the West, funding agencies such as the Fulbright Foundation apparently place less emphasis on Vietnam, Laos, Cambodia, Burma… than on China (again for non-linguistic reasons), with the result that career opportunities for young scholars appear fewer.
What are the issues in Sino-Vietnamese studies that await further investigation? I’ll try to provide a few pointers concerning loanwords, which are still far from fully studied and exploited.
Did you know about the hypothesis that ‘wax’ là 蜡 is a borrowing from Vietnamese?
Borrowings from Chinese dialects into Vietnamese are far more numerous, and still far from being fully described and analyzed. As part of a team endeavour (led by Martine Mazaudon) to translate papers by Haudricourt , Guillaume Jacques and Marc Brunelle translated Haudricourt’s seminal 1954 articles about the reconstruction of Old Chinese and the origin of tones in Vietnamese, respectively. These papers adduce key evidence from Vietnamese. Haudricourt sorts Chinese loans in Vietnamese into various layers based on the sound correspondences (following the classical method of comparative linguistics). Strikingly, many of the examples are not commonly considered as loanwords. Others are generally considered as belonging to a layer of borrowing other than that to which Haudricourt assigns them. Here are two examples.
First example: an early loan into Vietnamese, not commonly considered as such
In his usual elliptic style, Haudricourt does not spell out a complete argument about the history of each word. Did he facetiously pack his reasoning in the choice of gloss for the word: providing a gloss that is strictly correct neither for the Chinese original nor for the Vietnamese — a gloss that constitutes a hint about the hypothesized path of evolution? This seems to be the case in this example:
number | char. | K-OC | K-MC | [B-MC] | Vietnamese | [gloss] |
GS 1109j | 務 wù | mi̯og | mi̯uʾ | mjuH | mua [muəA1] | to get |
In both Old Chinese and Modern Chinese, 務 wù is associated with two sets of meanings: ‘strive, exert oneself, be necessary’ and ‘task, affair’, but not ‘to get’. The meaning in Vietnamese is ‘to buy’. The best guess of the editors of the forthcoming volume is… that ‘to get’ is Haudricourt’s best guess on the semantics of this early borrowing: from ‘to exert oneself’ to ‘to get’, and later from ‘to get’ to ‘to buy’.
To my knowledge, sixty years after the publication of this article, no further research has been made into this, e.g. examining the words for ‘to buy’ in the entire Viet-Muong (a.k.a. Vietic or Việt-Chứt) language subgroup. This question is not hard to examine, building on the evidence collected on this subgroup in the last 60 years. For ‘to buy’, Arem has /mʊː/, close to Vietnamese /muəA1/. But comparative evidence from the entire Vietic group leads Ferlus to a reconstruction of ‘to buy’ as *caːk. Seen in this light, the Vietnamese word is clearly a case of lexical replacement, further borrowed from Vietnamese into some other languages of the subgroup, such as Arem and Rục. This confirms the plausibility of Haudricourt’s interpretation.
Second example: an early loanword commonly considered as belonging to the Sino-Vietnamese layer
In his analysis of the history of vowel changes in Chinese, Haudricourt lists correspondences between words reconstructed by Karlgren with the rhyme *-ək and words with two different rhymes in two different layers of Vietnamese borrowings. Haudricourt’s argument is that the vowel in Sino-Vietnamese readings is ă whereas an earlier stratum of loanwords has ɯ. This is confirmed by examining the Sino-Vietnamese readings of characters in the table below, as compared to the pronunciation of early loanwords. In Sino-Vietnamese, the character 特 tè reads đặc [ɗăkD2], and 墨 mò ‘ink’ reads mặc [măkD2], both with the vowel ă.
GS | char. | K-OC | K-MC | [B-MC] | Vietnamese | [gloss] |
905 | 得 dé | tək | tək | tok | Sino-Vietnamese đắc [ɗăkD1] | to obtain |
919k | 德 dé | tək | tək | tok | đức [ɗɯkD1] | virtue |
961h’ | 特 tè | dhək | dhək | dok | đực [ɗɯkD2] | male (animal) |
904c | 墨 mò | mək | mək | mok | mực [mɯkD2] | ink |
The gloss that Haudricourt cites for 特 tè represents the oldest sense of this etymon, i.e. ‘bull, male (animal)’, attested both in Old Chinese and as an early Vietnamese loanword. In Modern Chinese, the meaning of 特 tè is ‘special, exceptional’ (as also in Sino-Vietnamese, where 特 reads đặc [ɗăkD2]). The sense development from ‘male (animal)’ to ‘special’ is well documented from the end of the Old Chinese period onward. In this example, the early loanword and the Sino-Vietnamese loanword are well distinguished, and all is for the best.
But the word đức [ɗɯkD1] ‘virtue’ (德 dé) is less straightforward. The word apparently did not undergo any considerable semantic change in Chinese, unlike 特 tè (from ‘male’ to ‘special’); it was borrowed only once by Vietnamese, as đức [ɗɯkD1]. The form đức [ɗɯkD1] serves as the reading of the corresponding character (德 dé) in the system referred to as ‘Sino-Vietnamese’, and is therefore listed in dictionaries of Sino-Vietnamese. Since Sino-Vietnamese is defined as the Tang period reading of characters, it is tempting (but mistaken) to consider all Sino-Vietnamese pronunciations to date back to the Tang period. Haudricourt based himself on the sound correspondences; on the basis of this word’s ɯ vowel, he concluded that it must be an early loan: had it belonged to the Sino-Vietnamese layer proper (words borrowed during the Tang dynasty), one would expect it to have the vowel ă. In light of Haudricourt’s argument, it can be hypothesized that the early Chinese loanword for ‘virtue’ was carried over into the new (Sino-Vietnamese) system for reading Chinese characters.
Such interferences between different layers of Chinese borrowings in Vietnamese, which create exceptional readings within the Sino-Vietnamese system, are mentioned by Nguyễn Tài Cẩn 1979, who provides some wonderfully conclusive examples, and opens perspectives for fresh work. Were I to specialize in Vietnamese language history, I would love to collaborate with Dr. Trần Trọng Dương, from the Hán-Nôm Research Institute, who is working on an etymological dictionary of Vietnamese; also with SHIMIZU Masaaki, Mark Alves and John Phan, who would also be terrific participants in a collective endeavour. But I guess that, rather than delving deep into Vietnamese historical phonology, I shall continue to work mostly on the documentation and description of little-studied language varieties in the decades which hopefully remain before me. I’ll die with my fieldworker’s boots on — or rather, with 1 ‘phonetics’ boot and 1 ‘fieldwork’ boot on! So I thought I’d bring this topic to the attention of “a new generation of Sino-Tibetanists and SEA linguists” — to cite from this blog’s inaugural argument.
在一个名叫“Indo-Sinica”的blog上,我就来写篇文章,谈谈一个“Indo-Sinica”范围内的问题:汉语与越南语之间的接触。这个领域的研究进展看起来已经相当成熟,并不是一个值得优先关注的话题。越南有大量极度濒危语言,在得到严肃认真的调查描写之前,很可能就会消失。相比起来,越南语当中引人注目的汉语成分已经有阮才谨、王力等多位学者研究过。有好几本词典记录汉字在越南语当中的读音(即汉越音,Sino-Vietnamese),其中就有陶维英在1932年出版的力作《汉越词典》(Hán-Việt từ điển)。陶维英和阮才谨都属于最后一代精通汉语文言的越南学者。在20世纪初,越南的汉文教育逐渐消亡之后,汉文与汉学对越南学者来说变得难上加难。这一历史变迁,加上越南的高等教育在近几十年中碰到的一些普遍的困难,就能够说明越南相对缺乏汉越音专家的现况。
而在中国这边,王力之后,从者寥寥。就我与中国朋友与熟人的交流所得到的印象来说,历史课本、媒体和宣传当中,越南是一个整天闹分家的叛省,而越南人是一群忘恩负义的邻居——中国帮助他们打西方帝国主义,他们却调转枪口来打中国人。是不是这种印象对学术界产生了一些影响,结果是很少有同行受到越南语和汉越音的吸引?
在西方,同样是出于无关乎语言学的原因,富布赖特项目等资助计划对越南、老挝、柬埔寨、缅甸等国家的注意程度远低于中国,所以青年学者在就业方面的机会就很要少上一些。
汉越研究中有哪些有待进一步研究的问题呢?这篇文章试举要说明,借词远不是一个得到了完整的研究和利用的领域。
大家有没有听说过,有一个假说认为,“蜡”là字是从越南语中借来的?
从汉语各方言借进越南语的词就更多了,并且远未得到完整的描述和分析。由Martine Mazaudon牵头进行的对奥德里库尔(Haudricourt)著作的翻译计划当中,向柏霖(Guillaume Jacques)和Marc Brunelle翻译了奥德里库尔1954年的两篇影响深远的文章,内容是分别上古汉语的构拟和越南语声调的起源。这些文章中引证了一些越南语中的关键证据。奥德里库尔使用比较语言学的经典方法,按照语音对应将越南语中的汉语借词分成几个层次。奥氏引证的不少例子通常并不认为是汉语借词,这一点相当引人注目。而其他不少的例子,借词层次的分派,奥氏往往与主流意见不同。下面试举两例进行说明。
第一个例子:通常不被认为是借汉成分的早期汉语借词
奥德里库尔文风简练晦涩,往往对每个词具体的源流嬗变并不详细讲明。是不是有的时候,他对词条给出的释义既不符合汉语原义,也不符合越南语今义,而是通过释义,来用一种简练、幽默的方式,给出自己对词义演变的假设?这个例子似乎就符合这种情况:
字号 | 字 | 高上古 | 高中古 | 白中古 | 越南语 | 奥释义 |
汉文典 1109j | 務 | mi̯og | mi̯uʾ | mjuH | mua [muə阴平] | to get |
在古汉语和现代汉语当中,“務”字都有两组意思:“勉也,求也,专力而必为之”,“事务也,事业也”,但是从来没有奥氏笔下“to get”的意思。而越南语里面,mua字是“买”的意思。我们即将出版的新书的编者的猜测是……奥氏写的这个“to get”,表达了他对这个早期借词的语义演变的猜测:先是“专力而必为之”,然后是“得到”,然后才变成“买”的意思。
这篇文章发表六十年以后,在我知道的范围内,这个词并没有什么更加深入的研究,比如说,看看整个越语支的其他语言里“买”是怎么说的。通过最近六十年中对这些语言的资料的收集,这个问题的研究没有过去那么难了。“买”,Arem语说/mʊː/,跟越南语的/muə阴平/类似。不过,Ferlus根据整个越语支的比较语言学材料,把“买”的原始形式构拟为*caːk。这样看来,越南语的mua字显然属于词语替换的情况,这个词又转而从越南语借入同组的其他语言,如Arem语和Rục语当中。这就验证了,奥德里库尔的解释的确可能有道理。
第二个例子:一个通常被认为是汉越音的早期借词
奥德里库尔在分析汉语的元音演变的过程中,对于汉语德韵字(高本汉拟为*-ək),举出了越南语两个不同的借词层次中,与两组不同的韵母的对应关系。奥德里库尔认为,汉越音中对应的元音是ă,而早期借词中对应的元音是ɯ。我们将下表例字的汉越音与早期借词音做做对比,就能看出这一点来。汉越音当中,“特”字读đặc [ɗăk阳入],墨字读mặc [măk阳入],两字元音都是ă。
汉文典号 | 字 | 高上古 | 高中古 | 白中古 | 越南 | 释义 |
905 | 得 | tək | tək | tok | 汉越 đắc [ɗăk阴入] | to obtain |
919k | 德 | tək | tək | tok | đức [ɗɯk阴入] | virtue |
961h’ | 特 | dhək | dhək | dok | đực [ɗɯk阳入] | male (animal) |
904c | 墨 | mək | mək | mok | mực [mɯk阳入] | ink |
奥德里库尔对特字给出的释义,是这个字的本义,“牡也,雄也”。上古汉语里,以及越南语的早期汉语借词的词义,都是这么解的。而在现代汉语当中,特字是“特别、特殊”的意思(汉越语里也是一样,这么解的特念đặc [ɗăk阳入]音)。从“牡也,雄也”到“特别、特殊”的词义演变,从上古汉语晚期开始就斑斑可考。就这个例子而言,早期借词与汉越音泾渭分明,没有什么问题。
但是đức [ɗɯk阴入] “德”的情况就没那么简单了。德字与特字不同,在汉语当中,从来没有发生过大的词义转换,而且在越南语当中只被借了一次,借过来的形式就是đức [ɗɯk阴入]。在所谓“汉越音”的体系中,“德”字的读音是đức [ɗɯk阴入],所以这个形式也列进了汉越音词典中。汉越音的定义就是唐代字音,所以我们很容易把所有汉越音的字音都看作唐代的东西。这种想法虽然很有吸引力,但是是错的。奥德里库尔自己则用语音对应来做出判断:因为这个字的元音是ɯ,他就认定这个字一定属于早期借词——如果真是属于狭义汉越层(唐代借来的字词),则元音应当为ă。按照奥德里库尔的论证,我们可以做出这样一个假设:汉语“德”字的早期借词形式在汉越音这一新的汉字字音体系中被沿用下来。也许这种关乎世道人心的词汇,比日常用词更加保守一些。
越南语中的这种不同借汉成分层次之间的互相干扰,在汉越音体系中创造出了一些例外读音。阮才谨(1979)提到了这一现象,并给出了一些有说服力的妙例,开启了新的研究方向。如果我自己专研越南语史的话,就会很希望与汉喃研究院正在编写越南语的词源词典的陈仲洋博士合作,还有清水政明、Mark Alves以及John Phan——这几位在合作的事业当中一定会是好伙伴。不过我估计,自己这辈子大概不会深钻越南语的历史音韵学了。我希望在未来的几十年里,继续专心对研究不足的语言做描写和典藏。我希望到死的那天,脚上都穿着田野工作者的靴子——不,要一只脚穿着“语音学”靴子,一只脚穿着“田野”靴子!所以我就在这里写篇文章,好向——引句这个blog的开篇——“a new generation of Sino-Tibetanists and SEA linguists”提起对这个问题的注意。
(文/米可,译/龚勋)
số | ký tự Hán | tiếng Hán cổ theo Karlgren | tiếng Hán trung cổ theo Karlgren | tiếng Hán trung cổ theo Baxter | tiếng Việt | Haudricourt dịch ra |
GS 1109j | 務 wù | mi̯og | mi̯uʾ | mjuH | mua [muəA1] | to get |
Ỹ nghĩa của ký tự 務 trong tiếng Hán cổ là ‘cố gắng; cần’ chứ không phải nghĩa mà ông Haudricourt đưa ra (‘to get’, tức là ‘được’). Ý nghĩa ‘được’ cũng khác với ý nghĩa của từ mua trong tiếng Việt hiện nay. Có lẽ Haudricourt đã giả sử rằng khi người Việt mượn từ này từ tiếng Hán cổ, họ đã thay đổi ý nghĩa, từ ‘cần’ thành ‘được’. Sau đó, ý nghĩa này lại được chuyển từ ‘được’ thành ‘mua’.
Thật ra, nguồn gốc của từ mua trong tiếng Việt là gì? Để xác định vấn đề này, ta cần xem xét những ngôn ngữ khác trong nhóm Việt-Chứt (Vietic) có phải cũng có một từ cùng nguồn gốc với mua không. Trong tiếng Arem, ‘mua’ là /mʊː/, gần tiếng Việt /muəA1/. Nhưng từ cổ hơn trong Việt-Chứt là *caːk (theo ông Michel Ferlus). Điều này có thể chứng minh rằng từ mua đã thay cho một từ gốc Môn-Khơ Me cổ hơn. Như thế, có lẽ mua đúng là từ vay mượn từ tiếng Hán như Haudricourt đã giả sử.
Một ví dụ khác là từ đức (trong đạo đức), là một từ thường được coi là từ Hán-Việt (tức là vay mượn trong thời Đường) nhưng Haudricourt lại coi là từ vay mượn cổ (trước Hán-Việt). Cụ thể như sau:
Haudricourt cho rằng có những từ được vay mượn hai lần từ tiếng Hán, một lần trước Hán-Việt, và một lần trong lớp Hán-Việt. Như ký tự 特 đã được vay mượn lần một, thành từ đực của tiếng Việt, rồi được vay mượn lần thứ hai với phát âm khác là từ đặc (trong đặc biệt). Như vậy, ý nghĩa cũng khác, vì ý nghĩa của ký tự 特 đã thay đổi trong chính tiếng Hán từ thời cổ đến thời trung cổ. Như thế, từ đức (trong đạo đức), thuộc cùng một lớp vay mượn với từ đực (ở lớp cổ) chứ không phải cùng một lớp với từ đặc (lớp Hán-Việt).
số GS | ký tự Hán | tiếng Hán cổ theo Karlgren | tiếng Hán trung cổ theo Karlgren | tiếng Hán trung cổ theo Baxter | tiếng Việt | [dịch ra] |
905 | 得 dé | tək | tək | tok | Hán-Việt đắc [ɗăkD1] | to obtain |
919k | 德 dé | tək | tək | tok | đức [ɗɯkD1] | virtue |
961h’ | 特 tè | dhək | dhək | dok | đực [ɗɯkD2] | male (animal) |
904c | 墨 mò | mək | mək | mok | mực [mɯkD2] | ink |
Vậy vì sao mà tự 德 không được đọc là đắc, theo cách phát âm thời Đường (tiếng Hán trung cổ)? Có thể giả sử rằng người Việt thời Đường không chấp nhận phát âm mới (đắc) vì từ vựng tôn giáo bảo thủ hơn các từ “thường”, nên phát âm đã có từ trước (tức là đức) được lưu giữ và bảo vệ. Từ đó, cách đọc đức được coi là thuộc cách đọc Hán-Việt (cách đọc chữ Nho đã hình thành ở nước ta sau thời Đường).
Hy vọng rằng mối giao thoa giữa tiếng Việt và tiếng Hán sẽ tiếp tục được nghiên cứu sâu hơn. Theo hướng đi của TS. Trần Trọng Dương (Viện nghiên cứu Hán-Nôm) trong cuốn Nguyễn Trãi – Quốc âm Từ điển (2014), hy vọng rằng ông và các chuyên gia khác sẽ xây dựng cuốn Từ điển từ nguyên tiếng Việt phân biệt những lớp vay mượn, phản ánh sự phong phú của lịch sử tiếng Việt.
(Alexis Michaud và Đỗ Thị Ngọc Diệp)
OpenEdition vous propose de citer ce billet de la manière suivante :
Alexis Michaud (20 décembre 2015). Contact between Chinese and Vietnamese: an over-researched field? Indo-Sinica · 震南古聲隨記. Consulté le 17 février 2025 à l’adresse https://doi.org/10.58079/q69g
A propos de l’hypothèse de Michel Ferlus sur 蜡 là “cire”:
http://panchr.hypotheses.org/754
What is “[ă]” – is it [ɐ]?
The macron indicates a short vowel. In closed syllables, Vietnamese contrasts long and short a, and long and short ɤ.
There’s a scintillating paper by Haudricourt on this topic (an exercise in ‘dynamic synchrony’). An English translation of that paper is available here, with a note on the state of the art.
Je précise que 務 wù mjuH <*mo-s est très probablement cognat avec le khaling |mu| 'faire', même si j'ignore comment reconstruire le sens de cet étymon.