Where did Chinese tones come from and where are they going?

« previous post | next post »

Recently we've had several discussions about how tones in Sinitic languages aren't as uncomplicated or inflexible as one might imagine or as is often claimed:

"When intonation overrides tone"

"Mandarin by the numbers"

In these posts and in the comments to them, we have seen how stress and musical tune / melody often override or distort the canonical tones for given morphosyllables in sung or spoken context.  This is a completely different matter than tone sandhi, where tones are modified according to their position within a sequence of syllables (I believe that most instances of tone sandhi occur for simple physiological reasons, e.g., in normal speech it is virtually impossible to pronounce two full third tones in a row because they both dip so low in an individual's register that one needs a means for readying oneself for the onset of the utterance of the second third tone and does this by changing the first third tone utterance to a rising second tone).

Recently (2012), University of Oslo student Øystein Krogh Visted has written a very interesting M.A thesis entitled "Nuances of Pronunciation in Chinese:  Lexical Stress in Beijing Mandarin."  Here's a brief description of the thesis:

The pronunciation of Beijing Mandarin, which is the basis for Modern Standard Mandarin, is in reality not as straightforward as it is usually presented. General books on the language and common textbooks in English on the subject usually only give very basic, prescriptive (though supposedly descriptive) analyses of the basic features of pronunciation. Finer points are generally not discussed in any detail. The treatment of amongst other things the aspect of word stress (the parts of words that are emphasized in speech) in mastering and indeed properly understanding Chinese is thus neglected. It has not yet acquired the position in Chinese language-teaching it arguably needs, so that the language may begin to be taught and indeed learned in a more comprehensive manner. This book will take a basic analytical approach to the phenomenon of word stress in Beijing Mandarin. It compares and discusses available meta-information on the topic, as well as its theoretical underpinnings and practical applications, and from a pedagogical starting point aims to bring attention to these important nuances in the Chinese language.

Because of the above described phenomena and increasing polysyllabicization, some scholars have even mentioned to me that they think northern Mandarin, at least around Beijing, is evolving into a stress language rather than a tonal language, though I haven't seen any papers or books that document this development.

This has prompted me to reflect that the "older" Sinitic languages (e.g., Cantonese, Min) tend to have more tones, the "younger" ones (e.g., the Mandarin branch) fewer tones, and to pose the following questions:

1. How to account for the greater number of tones in the older languages?

2. How to account for the lesser number of tones in the younger languages?

3. How many tones did Middle Sinitic have?

4. How many tones did Old Sinitic have?

I ask these questions because it appears that northern Chinese is gradually shedding its tones and evolving into a stress language, at least in the opinion of some colleagues.  Of course, others have pointed out that stress has always played a role in Chinese, but it just hasn't been described and studied in detail.  Still, it does seem that stress and melodic contours are assuming greater weight now than in the past.

From W. South Coblin

Questions 1 and 2 are really different sides of the same coin. The diminution in number of tones in the younger languages is due to phonological mergers, merger and split being two fundamental processes in sound change. The medieval lexica of the Qieyun family of texts [VHM:  medieval rhyme books] indicate four tones, i.e., ping, shang, qu,and ru [VHM:  "even, rising, leaving / falling, and entering"]. It is assumed that the first three differed in level and/or contour. The forth was characterized by syllable final stops, i.e., checked finals.

It is widely assumed that this four-tone system had two phonetic registers, upper and lower, conditioned by the presence or absence of initial voicing/murmur on the syllables. This would have been a phonetic rather than phonemic difference, and many native speakers would presumably not have been aware of it. Thus, the lexica indicate only four tones. Later loss of voicing/murmur, i.e., loss of the conditioning factor, would have brought the register distinction into full phonemic prominence. This process is thought to have begun sometime in the Tang period in at least some dialect families. The result would then have been an eight tone system consisting of paired upper and lower (i.e., yin and yang) ping, shang, qu, and ru tones. This eight-tone system, which is actually rather rare today, would then have been reduced in various ways, due to mergers in most dialect groups. It is these mergers that have reduced the number of tones. Standard Cantonese is an exception, in that the upper yinru tone underwent a phonetic split, conditioned by vowel quality. That yielded nine phonetic (as opposed to phonemic) tones.

The tonal situation in Early (now usually called "Old") Chinese has been a source of controversy. Karlgren and Li Fang-kuei, among others, assumed that the four "classical" tones were already present at that stage. Li definitely believed that tones were present at the time of the composition of Shijing odes, but he was noncommittal about their ultimate origins. The most common newer view is that tones are of secondary origin in Chinese and arose through a process called "tonogenesis", a term coined by James A. Matisoff. Tonogenesis became a hot topic in the sixties and seventies of the last century, when it began to be thought that most or perhaps all tonal systems in East Asian languages were ultimately secondary. Briefly, East Asian tonogenesis is conceived of as a process in which phonemic tones arose when certain early syllable final consonants were lost. In Chinese, as well as in various other languages, it is widely thought that the lost consonants in question were the glottal stop, laryngeals such as -h, and sibilants such as -s.

The classic view for Chinese is that the loss of a final glottal stop yielded the medieval shang tone, while loss of final -s and/or -h yielded the qu tone. The ping tone would have arisen when neither consonant was present, and ru would simply comprise syllables having various obstruent finals, usually posited as -p, -t, or -k, though some authorities envisage other such stops. This view, with individual variations, is rather widely held by people who work on Old Chinese today.

You have asked what I think. The answer is that I do not work on Chinese of that period at all and do not concern myself with the question, except in those rare cases where it may have some bearing on the historical and comparative dialectology I do nowadays. For pre-medieval Chinese, I use Jerry Norman's Early Chinese system, which is specifically designed for use in the historical study of dialects and is much simpler than the Old Chinese systems developed by people who specialize in that field. In this system, ping tones are unmarked, shang tones are marked with an -x, and qu tones are indicated by an -h. These symbols are purely algebraic and are borrowed from Li. They are non-committal regarding tonal origins, since no modern Chinese dialect is atonal. Thus, this system of tonal marking is quite adequate for all historical work involving the modern vernaculars. For the medieval period, I use Norman's Common Dialectal Chinese. There, tones are indicated by the 1-8 numbering system of Y. R. Chao that is conventionally used by Chinese dialectologists today.

From Axel Schuessler:

Old Chinese (OC) is assumed to have had no tones. During (before?) Hàn an assumed final glottal stop (still surviving in some southern dialects) got lost and resulted in Middle Chinese (MC) shangsheng 上聲 [VHM:  "rising tone"] (i.e., words with a glottal stop presumably already  had a certain short-stopped intonation which survived after ʔ was lost, hence the tone). An OC final -s ( or h as well as -s) was also lost by MC, so that only the earlier concomitant intonation survived, hence qusheng 去聲 [VHM:  "leaving / falling tone"]. Where there were no OC *ʔ and *s, the syllable either ended in a vowel (> pingsheng 平聲 [VHM:  "even tone"]), or a stop consonant (rùsheng 入聲 [VHM:  "entering tone"]). This adds up to four MC tones.

After MC, voiced initial consonants became voiceless (b>p, d>t, etc). To maintain a distinction, MC tones split according to voicing of initial consonants. In Mandarin, MC pingsheng split: voiceless consonant became today's tone 1 (gē < kâ 'song'), voiced initial became tone 2 (with aspiration:) táo < dâu. MC shangsheng became Mandarin tone 3, after voiced initial it merged with qusheng > Mandarin tone 4 (hence 受 OC *duʔ > MC shangsheng 'to receive' and *dus > MC qusheng 'to give' are both Mandarin shòu). All MC qusheng syllables have become Mandarin tone 4.  Because of these splits in pingsheng, syllables beginning with today's l- (e.g., lí-kai ["leave"]) can only end up in Mandarin tone 2, there is no tone 1 beginning with l- (except for a few rare, odd exceptions, perhaps). MC rùsheng words became very irregular and unpredictable in Mandarin.

As you see, in northern Chinese MC tones reshuffled themselves, merged and split to wind up with 4 tones again; some Shandong dialects of Mandarin are said to have three tones. Because of the splits of the MC four tones, Sinitic languages can have up to 8 tones (2×4) when no mergers occurred. This is what you find in the south.

I seem to recall that Mantaro Hashimoto once wrote a paper (and perhaps someone else did so later as well) showing that Sinitic languages in the north near toneless Altaic languages tend to have few tones, languages in the south (Cantonese, Min, etc.) near Tai languages with their multitudinous tones tend to have preserved more tones.  Perhaps Tai substrates may have something to do with this too (Anne Yue-Hashimoto would know more about it). So there is some north-south gradation due to neighboring languages. There is a Sinitic language, I think a dialect of northern Chinese somewhere in Central Asia, that has lost all tones.

It is evident from the accounts of historical phonologists such as Coblin and Schuessler that the number of tones in Sinitic languages has waxed and waned, that they are not immutable, and that they are indeed in a state of flux at the current time.  How many there will be a century from now is impossible to predict.


  1. dw said,

    June 25, 2013 @ 5:41 pm

    Punjabi would be a great comparison for study of "tonogenesis" , since we know that its tones developed from breathy voiced consonants.

  2. Jim said,

    June 25, 2013 @ 6:01 pm

    Polsyllabicization may have something to do with it.

    In Shanghai loss of consonant contrasts has resulted in compounds of several syllables and the development of tone envelops, and Shanghai has only two phonemic tones. In Manadarin the polysllabicization is developing for another reason, but coincident with this reduction in the one system.

  3. Olivia F said,

    June 25, 2013 @ 6:50 pm

    If I had written this, I would have given it the parenthetical title: "Or, Why I'm Never Attempting to Learn Chinese". Seriously. I have taken classes in six different foreign languages, and I'd like to become advanced in at least 10 or so, but I can't fathom figuring out how to pronounce a language like Chinese. Major props to people who master this as a second language, especially those who don't grow up around it.

  4. Thom said,

    June 25, 2013 @ 7:20 pm

    I struggled greatly with tones for the six months that I attended classes for Chinese Mandarin. Once I became immersed in Beijing, what helped me most was not consciously acknowledging the tones. Instead I listened and just tried to emulate the speech. Tone is rarely that relevant in everyday conversation because collocations and context are usually enough to understand the meaning of the words being uttered. Non-verbal cues and intonation are also usually more useful than tone identification.

  5. Lazar said,

    June 25, 2013 @ 7:33 pm

    Could you explain what you mean by older and younger Sinitic languages?

  6. Mark S. said,

    June 25, 2013 @ 8:07 pm

    The full text of "Nuances of Pronunciation in Chinese: Lexical Stress in Beijing Mandarin" is here:

  7. Therese said,

    June 25, 2013 @ 8:34 pm

    Olivia: Mandarin Chinese is rather easy to pronounce, I'd say, so long as you convince yourself that you can do it and don't become too bogged down with the idea that it's difficult. Many speakers of non-Chinese languages manage to learn Chinese languages, and there's no reason that you cannot do the same. :)

  8. gummy worm said,

    June 25, 2013 @ 9:43 pm

    "There is a Sinitic language, I think a dialect of northern Chinese somewhere in Central Asia, that has lost all tones."
    The Wutun language (五屯话) in the Qinghai province has no tones at all. According to the Wikipedia article, it is a Chinese–Tibetan–Mongolian mixed language. As soon as I hear this bizarre name I became very interested but I could not find a single credible source to support this claim yet.

    By the way, how is tone realized in Shanghainese? On this website, http://wu-chinese.com/zanhei/pitch.html, it says "these contours disappear completely on the polysyllabic level; syllables in Shanghainese words mostly all have leveled (flat) pitches."

  9. gummy worm said,

    June 25, 2013 @ 9:49 pm

    Paper on the Wutun language:

    1. Topic-Prominence And Coordinate Converbal Structures In Wutun (http://facta.junis.ni.ac.rs/lal/lal201101/lal201101-02.pdf)
    2. Writing Chinese in Tibetan: On the alternatives for an Wutun orthography (http://scripta.kr/scripta2010/en/proceedings/proc08v01_009.pdf)

    3. Asian Highlands Perspectives. 1 (2009), 367-371REVIEW

    4. Bonan Grammatical Features in Wutun Mandarin (http://www.sgr.fi/sust/sust264/sust264_sandman.pdf)

    5. Atlas of Languages of Intercultural Communication in the Pacific, Asia, and …

  10. J.Xiao said,

    June 26, 2013 @ 1:05 am

    'Older languages' are conservative language varieties which retain more archaic features than the 'younger' ones. It's a rule of thumb that more geographically remote varieties would usually be more conservative than the 'central' ones. Here, Canton and Minnan lie on the periphery of China, hence the saying. For an easy example in English, look at the distribution of rhoticism in rural England by Trudgill (non-rhoticism is more recent).

  11. John said,

    June 26, 2013 @ 1:17 am

    I wonder how the existence of repositories like youtube and youku will affect the evolution of languages. In 100 years time it will be trivial to find out how people spoke today rather than all the educated guesswork and reconstruction that is occuring with Old Chinese, PIE, etc. Will prescriptivists cause evolution to slow since they can easily point to examples of what is "right"?

  12. Martin J Ball said,

    June 26, 2013 @ 7:43 am

    "I wonder how the existence of repositories like youtube and youku will affect the evolution of languages. In 100 years time it will be trivial to find out how people spoke today"
    You're assuming they'll be able to access all this material in 100 years' time; experience from recent changes in recording technologies is not very reassuring in that regard!

  13. Katie said,

    June 26, 2013 @ 7:58 am

    Thanks for pointing to that thesis. Looking forward to reading it. It was somewhat of a revelation in my Mandarin studies when I realized that Mandarin not only has stress but that a lot of my pronunciation problems were from not paying attention to it. Kind of ironic that as a a native English speaker I'd have this problem, but there it was.

  14. Neil Dolinger said,

    June 26, 2013 @ 12:01 pm

    Two points I took from this post are that
    1) there is a relatively uncontroversial theory that Chinese developed tones to make up for a shrinking inventory of consonants, and
    2) there is some thought among Siniticists that the relative tonality among Chinese languages has been affected by the non-Sinitic languages nearby.

    Has anyone studied the development of tone in these non-Sinitic language families, and are the theoretical origins the same as for Chinese?

  15. JS said,

    June 26, 2013 @ 1:00 pm

    ^ I think (1) is oddly stated… better to say that tonal contour and subsequently "register" are considered to have been new realizations of earlier, segmental contrasts, the former involving syllable-final segments like laryngeals, the latter initial voicing/murmur.

    As regards (2), as it happens, if I'm not mistaken, the seminal works on both processes in fact concerned Vietnamese: Haudricourt's (“De L’origine Des Tons En Vietnamien,” 1954) suggestions about the emergence of contour contrasts were subsequently applied to early Chinese by Pulleyblank and others, while Maspero ("Études sur la phonétique historique de la langue annamite: les initiales," 1912) observed among the tonal categories of Vietnamese a binary distinction he traced to an old voicing contrast.

  16. David B Solnit said,

    June 26, 2013 @ 4:20 pm

    @Neil Dolinger
    Yes, there has been quite a bit of study of the development of tone in Tai-Kadai, Hmong-Mien, and Tibeto-Burman languages. The main thing to mention in that connection is that when it comes to tone splitting and merging, the conditioning factor is not simply initial voicing, but more generally initial laryngeal features: voicing yes, but also glottalization and aspiration. A seminal article on this is “Bipartition Et Tripartition Des Systèmes de Tons Dans Quelques Langues d’Extrême-Orient.” Bulletin de La Société de Linguistique de Paris 56 (1): 163–80.
    Highly recommended is the English translation of this by Christopher Court: “Two-way and Three-way Splitting of Tonal Systems in Some Far Eastern Languages (Translated by Christopher Court).” In Tai Phonetics and Phonology, edited by Jimmy G. Harris and Richard B. Noss, 58–86. Bangkok: Central Institute of English Language, Mahidol University, 1972

  17. Jim said,

    June 26, 2013 @ 4:39 pm

    "but more generally initial laryngeal features: voicing yes, but also glottalization and aspiration."

    Jihn Kingston says the same thing about tonogenesis in Athapaskan:


  18. AntC said,

    June 26, 2013 @ 5:02 pm

    Does the decrease in the number of tones over history parallel the decrease in the number of noun cases or verb person/number in language (families) with declensions? (Say in European languages.)
    (I'll not say decrease in complexity, because I guess there would be an increase in complexity in other areas of the language, to compensate.)

  19. Etienne said,

    June 26, 2013 @ 5:16 pm

    Professor Mair–

    In answer to your points 1 and 2: I cannot help but note that Mandarin is the Northernmost variety of Chinese, and its linguistic neighbors further north (various Mongolic and Tungusic languages) are as a rule non-tonal. This stands in sharp contrast to Southern Chinese varieties, which, if they are in contact with non-Sinitic languages, are as a rule in contact with languages whose tone systems are often quite elaborate (For instance, with various Hmong-Mien and Tai-Kadai languages).

    Considering all the known interaction between Mandarin Chinese speakers and their northern neighbors (the Manchus being the most recent example), I must admit that it is tempting to explain the reduced tone system of Mandarin (when compared to other Sinitic languages) as being a consequence of long-term language contact involving Mongolic and Tungusic languages. If memory serves me correctly Jerry Norman, in his book on Chinese, had mentioned that some Mandarin features do align it with Mongolic and Tungusic languages as opposed to more Southern varieties of Chinese. Has any (further) serious work on the topic been done? Or on related topics, such as, for example, the informally acquired L2 Mandarin spoken by older L1 speakers of Turkic, Tungusic or Mongolic languages within China? If such varieties are tone-free, or even if they exhibit simplified tone systems, it would be indirect evidence in favor of my hypothesis.

  20. David Morris said,

    June 26, 2013 @ 5:48 pm

    When I qualified as an ESL teacher, I researched various countries overseas as places to teach. I decided against China partly because of the tones, but mostly because of the writing system. I also decided against Japan mostly because of the writing system, and decided on Korea partly because of the writing system.

  21. Amos Teo said,

    June 26, 2013 @ 11:41 pm

    John Kingston also has a chapter on tonogenesis in the Blackwell Companion to Phonology that summarises various different processes in languages around the world – a draft copy is available here: http://people.umass.edu/jkingstn/web%20page/research/TBC_097.kingston.1.proofs.pdf

    Graham Thurgood's (2002) paper on Vietnamese tonogenesis updates Haudricourt's original model (from "De L’origine Des Tons En Vietnamien”) by proposing a laryngeal theory of tone development, as opposed to one purely based on consonants.

    Of special mention, Mark Post once directed me to work by Donegan & Stampe who proposed that differences in rhythm within the Austroasiatic family are responsible for structural differences between the Munda languages and the Mon-Khmer languages. (http://www.ling.hawaii.edu/faculty/donegan/Papers/2004rhythm.pdf) Specifically, they find that a falling rhythm, characterised by phrase and word-initial accent (i.e. a trochaic rhythm) correlates with such features such as atonality, polysyllabic words (typical of a synthetic language), etc. In contrast, Mon-Khmer languages that have maintained a rising rhythm, characterised by phrase and word-final accent (i.e. an iambic rhythm) correlates with tonogenesis, monosyllabic / sesquisyllabic words (typical of an isolating language) etc.

    This model has interesting implications for the development of Sinitic languages too. Northern Sinitic languages with fewer tones like Mandarin also typically display a falling rhythm, evident in words like 爸爸bàba and 妈妈māma where the second syllable is toneless. Similarly, Shanghainese is said to have developed 'word tone', with only the tone on the first syllable of the word being contrastive. In contrast, many Southern Sinitic languages like Minnan display what looks more like rising rhythm – for instance, in languages like Taiwanese Hokkien (and in my experience Singaporean Teochew) tone sandhi rules apply to all syllables in a word / phrase, except the last one. This suggests that the final syllable in these languages is the most prominent one. Arguably, it is Northern Sinitic that has shifted to a falling rhythm under influence of Altatic in the North, while Southern Sinitic languages have maintained a rising rhythm, and also a larger number of tone contrasts.

  22. Victor Mair said,

    June 27, 2013 @ 4:18 pm

    From Brian Spooner:

    Although I have no specialist knowledge of China I find this post particularly interesting. But in my opinion if we want to understand any fading of tonality in the pronunciation of Mandarin in northern China we should examine it in one or more comparative contexts. I would offer two possibilities:

    1. Since I spent some time studying Thai, I first thought of Emeneau's work on "India as a linguistic area," and, like Etienne, above, that it would not be surprising to find a tone gradient from northern China to south/southeastern China, irrespective of the language spoken. Tonality could well be a Sprachbund feature that varies and changes in relation to the larger social context, which includes speakers of other languages, rather than exclusively as a feature of any particular language. How long has China been a single language community anyway? As far as spoken language is concerned, it may have made a lot of progress towards becoming one since the Cultural Revolution. Before that it may have been a single language community for the small number of individuals in the cities who communicated in classical written Chinese. But what proportion of the total population of China were they? And whatever the original relationship between classical Chinese and the language of oral communication may have been, it has been weakening and widening over the past century, and the relative status of each has changed. So the factors that originally generated tonality are presumably no longer active.

    2. More importantly, I would suggest this is one variety of a general process that often kicks in when the amount of public oral interaction in a language increases significantly. For example, both Persian and Greek lost almost all their inflections after Alexander the Great, at a time when the societies of Middle Persian and the Greek koine were becoming very different from those of the Achaemenian Empire and the Greek city states, in terms of the sizes of the arenas in which people interacted. English changed in comparable ways after the Tudors, under the Stuarts (the century when public nonconformist Protestantism broke out from under the control of Tudor controlled Protestantism). I would expect Panjabi to lose its tones if it replaced Hindi as the main language of public interaction in the northwest of the Subcontinent (as could easily have happened over the past two generations under different political conditions).

    Every language I know something about (that currently comes to mind) has undergone significant structural change of one sort or another over the last generation or so at the same time as its speakers have experienced increasing rates of social change and the expansion and intensification of the arenas of interaction in which they participate.

    I have never read anything about the origin of tones in Chinese, but what I have read about the origin of tones in Thai (which is presumably speculative) would fit this, i.e. it is related to a change in the relationship between (implicit) social control of correct speech and the social universe of public language use during a period of history when correct use of language was closely related to correct social and political behaviour. The opposite process kicks in now when the relationship between language use and public behaviour is becoming very different.

    What has happened over the past fifty years, in China as in most of the rest of the world, is that as social arenas have expanded and merged, within countries as well as between countries, the hierarchy of correct language use has faded.

  23. Dave Cragin said,

    June 27, 2013 @ 9:49 pm

    Olivia: Chinese has both easy and hard aspects. The easy aspects are: no verb conjugation, no articles, no grammatical gender, and virtually no word endings. Most words are just 1 or 2 syllables and 2 syllable words are just 1 syllable ones put together, so if you can say the 1 syllable words, you can say the 2 syllable ones. Numbers, days of the week, and months are simple & logical.

    The difficult aspects are the writing (as noted by David Morris), the tones and lots of homophones, the latter 2 making comprehension difficult. I also find remembering tones difficult, i.e., when I began to forget a word, I forget the tones 1st, the phonetics 2nd. (eg. I told friends "I ate a scorpion," but because I got the tone wrong, it sounded like I ate a shoe. They didn't expect me to eat a scorpion, so it just confused them).

    Chinese grammar can be remarkably similar to English. My favorite: 我老婆会杀死我的 (Wo laopo hui shasi wo de). A word-for-word translation: "My wife (old lady) will kill me." Idioms never translate, but this one does perfectly – with the same word order and meaning.

    My major Professor in graduate school was Japanese and he used to say that the thinking in Chinese is much more similar to English than Japanese. Now that I know much Chinese, I better understand his perspective (but I'd be interested to hear from others if they share this perspective).

  24. Keith said,

    June 27, 2013 @ 10:36 pm

    VHM quotes Alex Schuessler as saying, "syllables beginning with today's l- (e.g., lí-kai ["leave"]) can only end up in Mandarin tone 2, there is no tone 1 beginning with l- (except for a few rare, odd exceptions, perhaps)."

    I can think of one first-tone l-initial Mandarin word: 拉. It's the only one that comes to mind, but it's hardly a rare or odd word.

    What's the story behind 拉?

  25. Daniel Trambaiolo said,

    June 28, 2013 @ 3:01 am

    Schuessler is talking about characters with initial l- that were originally pingsheng. The character lā 拉 was originally rusheng, with a final -p (like its phonetic element lì 立), so this rule does not apply.

  26. Daniel Trambaiolo said,

    June 28, 2013 @ 6:17 am

    One of the things that surprised me in the process of raising a daughter whose first language is Chinese is that tones were one of the easiest aspects of the language for an infant to start learning. At twelve months, my daughter wasn't particularly good at consonants, but she could imitate tone contours pretty well. I am not particularly familiar with the research on this issue, but from what I have seen there have been similar observations on Mandarin and Cantonese-learning children; correct pronunciation of Mandarin tones 1 and 4 generally comes earlier than tones 2 and 3, which was also generally the case for my daughter. For adult learners whose L1 is non-tonal, of course, the tones are much more difficult to master, whereas the consonants and vowels are relatively straightforward.

  27. leoboiko said,

    June 28, 2013 @ 6:26 am

    > Idioms never translate

    They seem to do way more often than I expected, e.g. (Sino-)Japanese 一石二鳥 isseki nichō one-stone two-birds. I suppose they are easy to borrow by calquing, and also some metaphors come naturally (lots of cultures seem to have the extended meaning of "aspiration" for "dream", or good/bad for up/down).

    Re experience—I know only the most basic fundamentals of Mandarin grammar, but not being a native speaker of either, I was immediately struck by how much English-like it felt.

  28. julie lee said,

    June 28, 2013 @ 11:21 am

    @Olivia, @Dave Cragin, @Daniel Trambaiolo, and others:

    Yes, Chinese has its easy and difficult aspects. I am Chinese and have studied French and German. I always thought the hardest part of these European languages were the gender of the nouns. It's not hard for me to remember words in German or French, but I have difficulty remembering when a noun takes le or la or die, der, or das. So reading French or German is much easier for me than speaking those languages.
    I used to think the Mandarin tones were very difficult for a foreigner to learn because I hear so many foreigners pronounce the tones incorrectly. But when I taught my half-German grandson, whose first language was German, the four tones, he picked them up very quickly and correctly. He was five. I did treat them as melodies (it's true that Chinese is a sing-song language) and drilled him many times for a few days to sing the tones, which are pitches or melodies, and to sing the four tones as a melody one after another. I also used the method with an adult Caucasian and it worked. However, another grandson, now 11, whose is Chinese-American, still utters the 3rd tone as if it were the 2nd tone. He can't get the 3rd tone. It's in a lower pitch. He doesn't drop low enough. He also learned Mandarin at 5, from another teacher, not from me. He's in a class, and didn't get individual instruction. I haven't had a chance to drill him myself yet, but I'm convinced if I drilled him for a few days, he'd get it right. As I see it, the four tones are melodies, whether a one-note melody or a multi-note melody. And insofar as anyone can pick up a melody, anyone can memorize the tones of Mandarin, Cantonese, or any other Chinese dialect or regional speech, or the melody of a Chinese phrase or sentence. I learned Cantonese as a second language, but the tones were never a problem. Much easier for me than remembering when to use le or la or der, die, das in French or German.

  29. Victor Mair said,

    June 28, 2013 @ 5:30 pm

    From Tsu-Lin Mei:

    (1) I am somewhat surprised that the commentators on Thai tones failed to mention Fang-kuei Li’s classical articles: “The hypothesis of a Pre-glottalized Series of Consonants in Primitive Tai” Bulletin of the Institute of History and Philology, 11 (1943) :177-188; “Laryngeal Features and Tonal Development” BIHP 51(1980) : 1-13. Li was the Herman Colitz Professor at the Summer Institute of Linguistics of 1977 sponsored by the Linguistic Society of America.

    (2) Tonogenesis is intimately connected with our present view of the composition of the Sino-Tibetan family. If you read the 19th-century writers on Southeast and East Asian languages, e.g., people like John Leyden and Henri Maspero, you get the view that Chinese is a monosyllabic, tonal language without morphology (hence, “analytic”).

    These three features (monosyllabic, tonal, analytic) are also shared by Thai and Vietnamese. Hence, these three languages must be related.

    This view is highly attractive because Fang-kui Li has shown that Proto-Tai has four tones, A, B, C, D, each with two registers. This is very much like the Middle Chinese (601) tone system, with ping, shang, qu, ru 4 tones and two registers. So if being tonal is an inherited feature, the conclusion must be that Tai and Chinese are genetically related. This is the view of August Conrady and Fangkui LI.

    Haudricourt’s article, “De l’origin des tons Vietnamien” (1954), upsets the applecart. Haudricourt showed that Vietnamese was originally a Mon-Khmer language, and as such, was polysyllabic and non-tonal. Vietnamese became monosyllabic and tonal through contact with tonal languages such as Thai and Chinese. He also showed that the tones in Vietnamese came about through lost final consonants.

    Haudricourt “Comment reconstruire le chinoise archaique” (1954) suggested that Old Chinese qusheng (departing tone) was the result of losing the final –s. Haudricourt’s theory was developed by Pulleyblank, Jerry Norman, and Tsu-Lin Mei, and the generally accepted view is that Proto-Chinese was a toneless language, and the rising tone was due to the lost glottal stop ( -? ) consonant, while the departing tone was due to the lost –s consonant.

    This solves another puzzle in comparative Sino-Tibetan. Written Tibetan (of the 9th century) or Old Tibetan (of the 7th or 8th century in the Dunhuang manuscripts ) evidently had no tones. So under the old view that Chinese was eternally tonal and Tibetan was non-tonal, the question is how can a tonal language be genetically related to a non-toal language. (How can a cat be related to an eagle?) With tonegenesis, the problem also disappeared.

    So in The Sino-Tibetan Languages (2003), the view is that Tai, Vietnamese, Miao-Yao are not part of the Sino-Tibetan family, that Chinese and Tibeto-Burman are related, and that the Sino-Tibetan family also includes many non-tonal languages, e.g., Tangut, Qiang, Kiranti, Dulong, Jinghpo. Many of these languages have been in contact with Chinese for thousands of years, yet they did not develop tones. Spoken Lhasa Tibetan has tones which were of late origin and thousands of years of contact with Altaic languages did not seem to have wore down Tibetan tones.

    So we should be a bit cautious in coming to sweeping conclusions.

  30. Daniel Trambaiolo said,

    June 28, 2013 @ 7:29 pm

    As far as I am aware, "one stone, two birds" 一石二鳥 is a recent calque from English (or some other European language). A better example of an idiom that works in both English and Japanese might be "to lend [one's] ears" 耳を貸す, which is attested at least as early as the late eighteenth century—well before the Japanese had become familiar with the works of Shakespeare.

    @julie lee
    When learning Cantonese, to what extent did you feel that you were applying tone-conversion rules from Mandarin? (I assume Mandarin was your first language.) Did it take you longer to remember the Cantonese tones for characters that don't follow a predictable pattern of correspondence with Mandarin tones (e.g. 市 vs. 是)? What about the three different versions of the entering tone? Or did you feel that you were acquiring Cantonese tones independently of your knowledge of Mandarin tones?

    I am English L1, Mandarin L2, and at a very early stage of learning Cantonese. I find that my habit of trying to figure out pronunciations based on Mandarin equivalents can cause these sorts of issues, and I would be curious to learn whether Mandarin L1 speakers learning Cantonese have similar problems.

  31. julie lee said,

    June 28, 2013 @ 9:19 pm

    @Daniel Trambaiolo

    To answer your question, I learned Mandarin as a child without being conscious of tones. People spoke Mandarin and I absorbed it. I didn't know there were tones, or that there were four tones. Then I was conscious of picking up Cantonese at age five and six, and again at ten and eleven because we were then in Hong Kong and the people spoke Cantonese, and I picked it up, without knowing anything about tones or that there were seven tones in Cantonese. So I was not applying "tone-conversion rules from Mandarin", as you put it. Even now, if I learn a new word in Mandarin or Cantonese, I will learn it with the tones, and have no trouble remembering the tones that come with the words (what I call the melody, or lilt) of the words. Perhaps that's because I first learned Cantonese as a child.
    Later, when I tried to learn the Shandong dialect spoken by my in-laws, which was very hard for me to understand, I was much older and more analytical, and I figured out some tone-conversion rules from Mandarin to assist my memory—for instance, that words with 2nd tone in Mandarin might be, say, 4th tone in that Shandong dialect. It was helpful, but just listening to the lilt of the Shandong dialect is also helpful.

  32. Daniel Trambaiolo said,

    June 28, 2013 @ 9:38 pm

    @julie lee
    Thanks for your reply. I incorrectly assumed you had learned Cantonese as an adult. Since you started learning as a child, picking up the tones automatically would seem more natural than figuring out the correspondences with Mandarin. Your experience of learning Shandong dialect as an adult is much closer to the sort of thing I am curious about, but I would expect the patterns of correspondence moving from Mandarin to Shandonghua to be more predictable than they would be moving from Mandarin to Cantonese.

  33. julie lee said,

    June 28, 2013 @ 11:03 pm

    My mom learned Cantonese as an adult, in her late 20s. She moved from Hankow to Beijing to Shanghai, to Hong Kong, picking up the local speech as she moved from place to place. I'm sure she didn't learn the local speech by analyzing sound-correspondences with Mandarin. She just learned Cantonese by ear, at first by interacting with the household help, then with shopkeepers and so forth. I think a lot of Chinese people, adults, pick up the local speech by ear if they live in a place long enough, and if they are interested enough. She did tell me that she noticed that many of the words and constructions in Cantonese were from archaic Chinese, The Shandong speech I heard was just as different from Mandarin as Cantonese, much more different than the Sichuan or Hunan speech I have heard, but I feel sure I would have been able to speak it if I had stayed longer among Shandong people.

  34. Daniel Trambaiolo said,

    June 29, 2013 @ 1:37 am

    The case of your mother is an interesting example. Can you really be so sure that she wasn't analyzing sound correspondences, at least at some level? I am thinking of cases when she would have wanted to pronounce a word that she was familiar with from one dialect but which she would not have heard spoken in the language she wanted to speak. For example, I have never heard the second character of the word 上漲 spoken in Cantonese, but if I ever needed to pronounce it in Cantonese
    I am pretty sure I would be able to guess. (I just looked it up online, and my guess was correct.) Surely your mother would have been able to do the same sort of thing?

  35. julie lee said,

    June 29, 2013 @ 1:19 pm

    Presumptuous of me to say "I am sure". But that's how I feel. For instance, I can speak the Sichuan dialect or speech, or the kind I heard my friends at Taiwan University speak, those who grew up in Sichuan province during WWII. I never consciously thought of correspondences between Sichuanese and Mandarin. I picked it up by ear. No doubt deep inside, my mind made those correspondences and conversions. It's funny how the mind works. Cantonese is of course much more different from Mandarin than that particular variety of Sichuanese speech, but I believe the same kind of subconscious conversion can be made. We who've been educated in the West tend to be very analytical. But I think someone like my mom, who was educated in old China (and only one year of college) was not so consciously analytical about the sounds of a language. She would have "picked it up by ear".

  36. Victor Mair said,

    June 29, 2013 @ 3:18 pm

    From Bill Baxter:

    "Just a note about 拉 la1 'pull': It is true that this character is given a rusheng (entering tone) reading in the Guangyun and other Middle Chinese written sources; in the transcription Laurent Sagart and I use it is MC lop.

    "But this is a case that demonstrates how excessive reliance on MC written sources can lead us astray in dialectology. In Mandarin dialects like Nanjing that preserve rusheng, 拉 is not rusheng; in Nanjing it is [lɑ 1], tone 1 as in standard Mandarin; if it were rusheng it would have a final glottal stop.

    "In fact there are quite a few words in Mandarin that have voiced resonant initials in tone 1, which have no plausible origin in the Middle Chinese system. Many of them are verbs referring to bodily actions, and some of them cannot have come from rusheng words because they end in a nasal, like 扔 rēng 'throw'. Other examples are 摸 mō 'feel, stroke'; 捋 luō 'to smooth with the palm'; 捞 lāo 'to fish out of (water, one's pocket, etc.)'; 撩 liāo 'to raise (one's dress, a curtain, etc.)', also 'to sprinkle'; 拎 līn 'carry, lift'; 溜 liū 'slip away', 熘 liū 'deep-fry with starch', 搂 lōu 'hold in the arms'… You could reconstruct these for "proto-Mandarin" with glottalized initials like *ʔl-, *ʔr-, *ʔm-, etc. to account for the high-register tone; or perhaps the high tone reflects the earlier presence of some kind of verbal prefix that has now been lost."

  37. Victor Mair said,

    June 29, 2013 @ 10:52 pm

    From Tsu-Lin Mei:

    Just think of it, tonegenesis was unknown until Haudricourt proposed it in 1954, and I learned about Haudricourt’s articles from Jerry Norman in 1968 at Princeton. Matisoff has a book on this topic, I think, and this was a topic of lively discussion in Sino-Tibetan conferences in the 70s and 80s. Too bad Fang-kuei Li to the very end refused to believe it. Jerry Norman often pointed out to me Chinese linguistics as practiced by the Institute of History and Philology has a deeply ingrained conservative streak.

    I met Haudricourt once, in Paris, through the good office of Alain Peyraube, and I admired him greatly. Haudricourt’s Chinese scholarship is so-so; he also works in Vietnamese, Thai, and Austroasiatic, and Austronesian languages. This convinced me that foreign devils have much to contribute to Chinese studies — it is not their learning, but their insight. Think of Max Weber. His knowledge of Chinese history is so-so. But what he said about Chinese religion and Chinese bureaucracy stands the test of time.

    Haudricourt was the keynote speaker at the Berkeley conference on Sino-Tibetan languages and linguistics in 1992. I was at that meeting. The place was packed and I couldn’t hear him.

  38. Victor Mair said,

    June 30, 2013 @ 4:13 am

    From Stephan Stiller:

    This post raises the possibility of gradual tonothanatos* :-) in some northern Sinitic languages (such as Mandarin).

    I just skimmed the thesis "Nuances of Pronunciation in Chinese: Lexical Stress in Beijing Mandarin" by Øystein Krogh Visted referred to in Victor's post.

    I find his "minimal pair" examples unconvincing. Yes, there is morphologically and syntactically conditioned prosodic emphasis of certain elements [not sure whether this is related to what I have elsewhere mentioned about present-day Mandarin neutral tone patterns not being entirely captured by dictionaries], but it's not entirely surprising. (For example in an ordinary VO-compound, the second part carries emphasis, exactly like in any language with a verb followed by an object the object carries phrasal stress.) I didn't locate any true minimal pairs of expressions of an identical part of speech contrasting (supposedly) only in "stress", and the bits of the thesis that I've read don't seem to suggest that -– beyond ordinary neutral tone annotations –- there is an additional phonological feature that needs to be lexicographically captured because, say, morphology and syntax wouldn't be able to predict some of the variation.

    There is a question about where a neutral tone can morphologically and syntactically occur (someone has pointers? there might be some in the thesis, actually). There is also a possibility that, when morphologically assigned phrase-level emphasis gets out of sync with the present tendencies through new lexicalizations, true stress could come into being, but for now I am not convinced that stress is phonemic, as he postulates.

    (If I am misreading his data and somebody would like to correct me, please go ahead.)
    * the opposite of tonogenesis

  39. julie lee said,

    June 30, 2013 @ 10:36 am

    Professor Mei Tsu-lin's remark– that although Haudricourt's Chinese scholarship was so-so, and Max Weber's knowledge of Chinese history was so-so, they nonetheless contributed important insights to Chinese studies– made me think of the Chinese philosopher Mou Zongsan's remarks about Hegel. Mou said that although Hegel had little knowledge of Chinese history he contributed valuable insights on Chinese history with his theory of freedom, for instance, Hegel's insight that China was a society in which only one man was free, the emperor.

  40. Dave Cragin said,

    June 30, 2013 @ 8:05 pm

    As another example of English/Mandarin similarity in sentence construction: "I am learning Chinese" = "我在学中文" (wo zai xue Zhongwen), Although Mandarin uses "zai", in this usage it is the equivalent of "am," so word order is the same.

    Julie: You mentioned remembering the gender of nouns is hard (I agree!). I’ve wondered whether it is harder for a Mandarin speaker to learn German than it is for a Mandarin speaker to learn English. What do you think?

    My Chinese friends often struggle on when to use “the” and I would think trying to also remember which “the” to use would make it even harder. Also, word order in English seems much closer to Mandarin, than German (It’s been years since I took German & my 1st language is English.)

    (*In my post above, I wrote “Chinese” while I should have been specific, because I was referring to Mandarin).

  41. Daniel Trambaiolo said,

    June 30, 2013 @ 9:23 pm

    {I wanted to add a further comment on 拉, but it isn't showing up. Is there a trick involved?}

  42. Victor Mair said,

    July 1, 2013 @ 2:01 am

    @Daniel Trambaiolo

    Since you just sent in this short note and have often commented upon previous occasions, I don't know why you had trouble commenting about 拉 this time. I just checked the comments pending section of the LLog dashboard, and there was nothing there. If you happen to have a draft of what you wrote, please try again.

  43. Victor Mair said,

    July 1, 2013 @ 2:07 am

    From David Solnit:

    Haudricourt was certainly one of a kind. To Prof Mei's list of languages should be added Tibeto-Burman: Haudricourt brilliantly (Paul Benedict's term) discovered the basic outlines of Proto-Karen using just one trilingual dictionary (English, Pho Karen and Sgaw Karen).

  44. Victor Mair said,

    July 1, 2013 @ 2:11 am

    From Boyd Michailovsky:

    Martine Mazaudon (who tried to translate for Haudricourt at the 1992
    meeting mentioned by Mei) and Alexis Michaud are editing a volume of
    translations of Haudricourt's articles (including the 1954 articles on
    Chinese and Vietnamese) to be published by Mouton.

  45. julie lee said,

    July 1, 2013 @ 2:43 am

    @David Cragin
    David: Yes, English "the" is hard for a Chinese to master. I learned English as a second language at age 7 and found it very difficult, mostly the pronouns (Chinese has much fewer pronouns) and words like "will" "would" "shall" "should". But after a year I was speaking English like the other children. I guess children learn fast. I began to learn German in college, and then went back to it many years later. I find German grammar more difficult than English, though it's hard to compare learning a language as a adult with learning one as a child. Yes, Chinese word order is a lot like English. However, I've been struck with at least one similarity between German and Chinese: the adjectival clause preceding the noun-being-qualified in German and Chinese, whereas in English the same adjectival clause would follow the noun:
    (Below an example from a book, but I apologize for not having the umlaut symbol):
    "eine egalitare, auf gegenseitiger Hilfe und Hingabe fur das Wohl der Gesamtheit grundete Gesellschaft" (here the adjectival clause describing the noun "Gesellschaft (society)" precedes the noun).
    In English: "an equal society that is based on mutual assistance and devotion to the common good" (here the adjectival clause, unlike the German, follows the noun "society").
    In Chinese: "一個 平等,以互助而追求共同利益為基礎的社會" ( Translated into English, the Chinese here says: "an equal, on mutual assistance and devotion to the common good based society," like the German, with adjectival clause preceding the noun "society". )

  46. Daniel Trambaiolo said,

    July 1, 2013 @ 3:09 am

    Trying again with the comment…

    Thanks to Bill Baxter for his clarification concerning 拉. My own explanation was clearly too hasty and based on thinking about too little of the available evidence. 思而不學則殆, indeed.

    So how, then, do we explain the discrepancy between written sources for MC that give 拉 as rusheng and the absence of the glottal stop in Nanjing? Is it relevant that although the most common modern meaning for this character is "to pull," judging from a quick database search this meaning seems to have been rare in pre-Tang sources; the more usual meaning in pre-Tang texts was something like "to break by twisting or bending" (e.g. 摧枯拉朽), while "to pull" started to become more common from the Song dynasty onwards. (In Ming dynasty vernacular literature, "to pull" seems to have become entrenched as the most common meaning.)

    Could we explain the discrepancy between MC and Nanjing by proposing that these are not etymologically the same word at all, but that the (originally rusheng) character 拉 meaning "to break by twisting or bending" was borrowed to transcribe a (non-rusheng) northern vernacular or dialect word meaning "to pull" due to vaguely similar sound and meaning? Or would the correct explanation be something else entirely?

    Moving beyond Chinese pronunciations of the character, I have for some time been curious about the Japanese and Korean pronunciations of the word 拉致 (“capture, kidnap”). This word is quite common in Japanese, but as far as I know it doesn't exist in Chinese and seems to have come into use in Japanese at some point in the early twentieth century. I suspect that it may have come into being as an alternative way of writing 羅致, which is homophonous in Japanese and has a long history of use in Chinese with a closely related meaning (“to capture or gather in with nets,” often used metaphorically). However, to write this word as 拉致 , it is necessary to assume a non-rusheng pronunciation of 拉. (Some dictionaries give “ratchi” らっち or even “rōchi” ろうち as alternative pronunciations, either of which could correspond to the rusheng pronunciation of 拉, but I have never heard anyone pronounce it as anything other than “rachi.”)

    However, the usual pronunciation in Korean is “napchi” 납치, corresponding to the rusheng pronunciation. How did this happen? Did Koreans at some point look at the Japanese characters and decide how to pronounce them based on dictionaries that were in turn based on MC rhymes?

  47. Jongseong Park said,

    July 1, 2013 @ 8:35 am

    Daniel Trambaiolo:
    If 拉致 is indeed a relatively recent coinage of Japanese origin, then it is just one more example of the "neo-Sinitic" vocabulary (parallel to neo-Latin vocabulary in the West) that has enriched the Korean language in recent times. Because these words were transmitted through writing in translated works, they are pronounced in Korean according to the established, regular readings of Chinese characters in Korean without regard for the pronunciation in Japanese. Hence 拉致 becomes 납치 napchi, where 拉 takes the usual reading 랍 rap with initial r converted to n due to the initial sound rule. The Japanese reading of 拉致 has no bearing whatsoever on how it is pronounced in Korean. The difference between the pronunciation of Chinese characters in the two languages is too great for Korean speakers even to pick up on anything strange going on in Japanese, let alone figure out how to change the Korean pronunciation accordingly. There is certainly no awareness at all among the average Korean speaker that 拉致 is a Japanese coinage anyway. Isn't the case similar for the hundreds of Japanese neologisms that entered Modern Chinese?

  48. Daniel Trambaiolo said,

    July 1, 2013 @ 10:05 am

    @Jongseong Park
    My question concerns the reason why the "usual" reading of 拉 in Korean should be 랍/납 rap/nap, rather than 라/나 ra/na. The former is what we would expect based on the MC written sources, while the latter is what we would expect based on the pronunciation in modern Chinese dialects, including those that preserve final -p.

    There don't seem to be any other commonly used Korean compounds containing the element 拉 랍/납, aside from a few derivatives of napchi拉致 납치 (e.g. 被拉 피랍, 拉北 납북, etc.; please correct me if I am wrong); so presumably the first Koreans to start using this word would have seen it in Japanese text and had to look it up in a dictionary to find the pronunciation. What is surprising is that these dictionaries must have given the pronunciation 랍, corresponding to the obscure MC word "lop", rather than 라, corresponding to the extremely common Mandarin word "lā", which seems to have been the basis for the Japanese choice of the character 拉 to write this word in the first place. In other words, the Japanese and the Koreans were working from different assumptions what the "usual" pronunciation of this word ought to be.

    That this was not the only possible pathway that might have been followed can be seen from the example of another very common Korean word, in which etymological 拉 is pronounced in Korean as 라: i.e. ramyǒn 라면 "noodles", originally from Chinese lāmiàn 拉麵, probably via Japanese rāmen ラーメン. In this case, it seems to have been important that neither Japanese nor Korean tend to use kanji/hanja to write this word. (Incidentally, would you say that modern Koreans are typically aware of this etymology?) Perhaps if Japan had adopted the habit of writing the word in kanji rather than katakana, modern Koreans might have ended up pronouncing it as nammyǒn 납면, following the same logic that produced napchi 납치 .

    (I found some discussion on the etymology of Korean ramyǒn here:
    However, this discussion does not mention that the word probably came into Korean via Japanese rather than directly from Chinese.)

  49. Jongseong Park said,

    July 1, 2013 @ 12:25 pm

    Daniel Trambaiolo:
    Thanks for clarifying your question. The Korean pronunciation of Chinese characters is usually taken to originate in MC pronunciation, so it is not surprising that 拉 is 랍/납 rap/nap in Korean. The subsequent development of Chinese doesn't seem to have had much influence on the pronunciation of Sino-Korean vocabulary. This is an interesting difference from the situation in Japanese.

    So even if 拉 isn't a commonly used character nowadays except in the derivatives of 拉致 napchi, its Korean pronunciation would still have been established in the numerous Korean character dictionaries produced throughout the centuries, based originally on the Korean approximation of MC and then carried through the subsequent sound changes of Korean of course.

    I emphasized that 拉致 납치 napchi must have been borrowed through writing. There are of course examples of loanwords from Chinese through the other mode of borrowing, speech. Your example of 라면 ramyǒn illustrates this, though it is a hybrid example: the Sino-Korean reading of 麵 면 myǒn, which is a recognizable morpheme meaning 'noodles', was preserved in preference to myen or men which would be closer to the Chinese or Japanese pronunciations. Another frequently cited, older example of a spoken borrowing from Chinese is 붓 put ("brush"), apparently from an older Chinese pronunciation of 筆. This character is read p'il in Sino-Korean pronunciation. Koreans write words with hanja (Chinese characters) only if they follow the canonical Sino-Korean pronunciations, so hanja is not used for cases like 라면 ramyǒn or 붓 put. These cases are definitely not considered Sino-Korean vocabulary even if they are loanwords (ultimately) from Chinese.

    For 라면 ramyǒn, at least some Koreans would be vaguely aware of a Japanese or Chinese etymology, especially if they are familiar with Japanese rāmen. The initial r is a giveaway that this isn't a native Korean word or an established Sino-Korean form, which points to this being a (non-Sino-Korean) loanword. I cannot judge whether the borrowing is directly from Chinese or via Japanese (note that there is only one liquid phoneme in Korean, so initial r could reflect an /l/ in the source language and is frequently pronounced as a lateral in this position despite being romanized as r). Compare the case of 짜장면/자장면 tchajangmyǒn/chajangmyǒn, which is directly taken from Chinese 炸醬麵 zhajiangmian without a doubt.

    As the name for a noodle dish, 라면 ramyǒn seems to be a prime candidate for spoken rather than written borrowing. Whereas Koreans would have mostly encountered 拉致 납치 napchi only in texts, they would have heard the source word for 라면 ramyǒn in the original language (whether Chinese or Japanese), which explains why the former was adopted as a Sino-Korean word and the latter as a simple loanword.

  50. Jongseong Park said,

    July 1, 2013 @ 12:36 pm

    Based on a quick internet search, it looks like the first Korean commercial ramyǒn was produced in 1963 and referred to instant noodles based on the Japanese rāmen invented a few years earlier. Unless we find earlier instances of the word, this seems to favour the theory of the word coming into Korean via Japanese.

  51. Daniel Trambaiolo said,

    July 1, 2013 @ 7:29 pm

    Yes, I think you are right. The borrowing of 拉致 is dependent on the written language; given this pathway for borrowing, it is is not too surprising that the final pronunciation in Korean became napchi. Thanks for clarifying the distinction between genuine Sino-Korean vocabulary and spoken-language borrowings; the situation in Korean is evidently much less chaotic than in Japanese, and it is probably the Japanese reading of these characters as rachi rather than the pronunciation in Korean that requires some sort of special explanation.

    However, there do seem to be some other peculiarities in the pronunciation for Sino-Korean words of Japanese origin. Is there any accepted explanation for why 自動車 is 자동차 while 自轉車 is 자전거? I would have expected both these words to have entered Korean through the same pathway and to have taken the same pronunciation for 車, but evidently something else was at work. Both 차 and 거 are based on possible MC pronunciations, but why would a different selection have been made in each case? (In modern Mandarin, the pronunciation is usually encountered in the wild only when talking about Chinese chess, but that is presumably irrelevant here.)

    The case of "ramyǒn" is, as you say, an interesting hybrid. Presumably the second element miàn/men/myǒn 麵 "noodle" was familiar enough to Korean speakers that it made sense to use the standard Sino-Korean pronunciation, whereas the first element ra 拉 was not only unfamiliar but also semantically inappropriate, since by the 1960s the Japanese word rāmen no longer referred to "pulled noodles"; it could thus have been retained as an unanalyzed prefix, as if we were to refer to these noodles in English as "ra-noodles."

  52. Jongseong Park said,

    July 1, 2013 @ 9:51 pm

    For the different readings used for Sino-Korean 車 in 自動車 자동차 chadongch'a "automobile" and 自轉車 자전거 chajǒngǒ "bicycle", an explanation I've seen is that when it refers to a human-powered vehicle, it is 거 kŏ/gŏ but when it is powered by something else, it is 차 ch'a.

    Hence we have 거 kŏ/gŏ in 自轉車 자전거 chajǒngǒ "bicycle" and 人力車 인력거 illyŏkkŏ "rickshaw" as well as more obscure examples like 曲車 곡거 kokkŏ, which refers to a small hand-driven cart used in pre-modern times. For the vast majority of cases, we have 차 ch'a, as in 自動車 자동차 chadongch'a "automobile", 汽車 기차 kich'a "train", and 風車 풍차 p'ungch'a "windmill".

    I'm not sure what rule there is if any when the word containing 車 doesn't refer to the vehicle specifically. In 停車場 정거장 chŏnggŏjang, which refers to bus or train stops, it is 거 kŏ/gŏ. But 停車 which refers to the action of stopping a vehicle can be either 정거 chŏnggŏ or 정차 chŏngch'a. I'm similarly at a loss to explain why certain readings are used for certain words in other similar cases of multiple hanja readings, such as 茶 차 ch'a or 다 ta "tea".

    I don't speak Japanese or Chinese, but a cursory search of 自轉車 (not used in Chinese?) and 人力車 in these languages shows that the reading sha (Japanese) or che (Mandarin) is used in these cases, suggesting that the semantic distinction in Korean using the alternative reading of the character is not parallelled in these languages. So this distinction may be a Korean innovation.

  53. Jongseong Park said,

    July 1, 2013 @ 9:57 pm

    In modern Mandarin, the pronunciation jū is usually encountered in the wild only when talking about Chinese chess, but that is presumably irrelevant here.
    In Korean chess, by the way, 車 is pronounced 차 ch'a, so there is no cross-linguistic correspondence in this case.

  54. Daniel Trambaiolo said,

    July 1, 2013 @ 11:23 pm

    Jongseong, thanks for these fascinating details.

    After browsing the dictionaries a little further, I was surprised to discover a further anomaly in Korean pronunciations of 車. Words like 馬車 "horse-cart" and 牛車 "ox-cart" are pronounced in ordinary contexts as 마차 mach'a, 우차 uch'a. However, when mentioned in the context of the Lotus Sutra's Three Vehicles (삼거, 三車), they are pronounced as as 마거 magŏ, 우거 ugŏ. (The third vehicle, the "goat cart", yanggŏ 양거 羊車, is so rarely encountered in everyday contexts that the dictionaries I have consulted don't even bother giving yangch'a 양차 as a possible pronunciation.)

    To the best of my knowledge, Chinese and Japanese don't make this distinction. Standard dictionaries give the pronunciation of all these compounds with C. chē, J. sha. However, Buddhist vocabulary often preserves unusual and archaic pronunciations, and it would be worth checking in specialized dictionaries (or asking some monks) to confirm.

  55. Daniel Trambaiolo said,

    July 1, 2013 @ 11:39 pm

    Incidentally, you are correct to say that Chinese doesn't use 自轉車. PRC Mandarin uses 自行車, Taiwanese Mandarin uses 腳踏車, and Cantonese uses 單車. I believe 腳踏車 and 單車 are both older terms, while 自行車 is more recent, but I'm not sure of the exact history.

  56. Victor Mair said,

    July 1, 2013 @ 11:55 pm

    From South Coblin:

    Haudricourt was unquestionably an unusual and brilliant man. He marched to a different drummer and tended to be ostracized by “establishements”, both academic and otherwise. He was by training a botanist, and studied Asian languages as a sort of avocation. He had studied in Hanoi after WW II, was passionately opposed to the Vietnam War, and for that reason refused to visit the US until Matisoff prevailed upon him to attend the Berkeley meeting in 1992. He died several years after that. Laurent Sagart knew him well and admired him very much.

  57. Jongseong Park said,

    July 2, 2013 @ 8:25 am

    Fascinating. I'm not familiar with the Three Vehicles of the Lotus Sutra, so this is the first time I'm seeing those terms. By the way, for most Koreans 羊 양 yang by itself only means "sheep" so I didn't expect the rendering "goat-cart" for 羊車, but I see that the original Chinese character can also refer to goats.

    You are right about Buddhist vocabulary often preserving unusual pronunciations. I came across a blog post in Korean about the character 茶 "tea" which also has two canonical readings, 차 ch'a and 다 ta/da. Some highlights:
    • Both readings are attested from Middle Korean sources at least as old as the 15th century (the reading 차 ch'a may even be indicated in an early 12th century source).
    • The general pattern was that 다 ta/da was preferred in the court, in Buddhist temples, and among the Confucian elite, while commoners usually said 차 ch'a.
    • This pattern explains 다 ta/da being used in words like 茶房 다방 tabang "teahouse", which originally referred to the court pharmacy, or 茶毘 다비 tabi, a Buddhist term for cremation from the Pali jhāpita.
    • For this reason, many Koreans think lofty words use 다 ta/da and everyday words use 차 ch'a, although the situation is not that clear-cut.
    • 차 ch'a is used when referring to actual tea, as in 綠茶 녹차 nokch'a "green tea", or in compounds with native Korean words as in 찻그릇 ch'atkŭrŭt "tea plate" (ch'a + kŭrŭt "plate").
    • 茶禮 is usually 다례 tarye when it refers to rites having to do with serving tea, but 차례 ch'arye when it refers to ancestral rites. In contemporary Korea tea is usually not part of 차례 ch'arye.
    • 茶甁 is frequently 다병 tabyŏng when it refers to the kettle containing tea in liquid form, but 차병 ch'abyŏng when it refers to a container of tea leaves. I'm not familiar with these terms though and these don't appear in the dictionaries I've consulted.

    It looks like slightly different traditions of reading Chinese characters among different circles are responsible for these distinctions in the modern language. Something similar may be going on for 馬車 and 牛車 being pronounced differently in Buddhist as opposed to everyday contexts.

  58. Jongseong Park said,

    July 2, 2013 @ 8:26 am

    The blog post I referred to above: http://blog.daum.net/haharbang/6217463

  59. JS said,

    July 2, 2013 @ 1:37 pm

    Daniel Trambaiolo, Re: 拉, your suggestion of two separate words 'break' (from final -p) and 'pull' (open syllable) seems most likely… I just noticed that Axel Schuessler has suggested, speculatively, that lā 'pull' could be an "archaic colloquialism" related to the word tuō 拖 'pull', this seemingly from an old lateral (GSR Companion, p. 214). Don't know how this would relate to the Sinoxenic items tho…

  60. Brian said,

    July 23, 2013 @ 12:42 pm

    On the two readings of 車: does anyone know what the deal is in Cantonese? I know it is normally pronounced ce1 (analogous to Mandarin che1, Japanese しゃ, etc.) but I have heard it pronounced geoi1 (Mandarin ju1, etc.) in the idiom 螳臂当车. This online dictionary…
    … says that ce1 is preferred in vernacular speech and geoi1 is a literary reading, with additional examples. This distinction (literary / vernacular) doesn't seem to correspond to the situation in Korean (human / non-human).

    Kangxi also has both pronunciations but I don't understand Classical Chinese so I can't figure out the distinction in there.

  61. David Eddyshaw said,

    September 4, 2014 @ 3:15 pm

    @Brian Spooner via VM:

    "For example, both Persian and Greek lost almost all their inflections after Alexander the Great, at a time when the societies of Middle Persian and the Greek koine were becoming very different from those of the Achaemenian Empire and the Greek city states, in terms of the sizes of the arenas in which people interacted."

    By no stretch of the imagination has Greek "lost almost all" of its inflections, even today, much less "after Alexander the Great."

RSS feed for comments on this post