The evolution of tones in Chinese

[Player01]

Help! I have been googling a lot but it doesn't seem there is an answer and the curiosity is killing me. I want to know the origin of tones in Chinese and how it happened. I heard some people say Chinese wasn't a tonal language before. And by the way what are some pros and cons for tonal languages.

[DecafLat]

I am not giving you an ultimate answer as to how tones were "invented" and I doubt anyone can. Tones were well established in Middle Chinese, and were categorized into Four Tones in a similar fashion to what we do today. Lu Fayan (陆法言)'s rime dictionary 切韵 (601 AD) is a testament to the status of tones at that time.

No reliable documentary evidence can be found supporting/debunking the existence of tones in Old Chinese. [Or so I understand.]

Since it is more or less considered a given that the Chinese tongue is tonal, the question of pros and cons are quite irrelevant. Tones have been an integral part of the language for so long that it is hard to imagine a workable version of the Chinese language without them.

[renzhe]

Keep in mind that Chinese is not the only tonal language. Some other SE-Asian languages are too, but also some African and (to a certain extent) European languages have tones which convey semantic information.

So tones are a rather common feature which pops up in some languages from time to time.

[Hugh]

I remember reading somewhere that actually the majority of languages use tones, not the other way round. Obviously the situation isn't quite that simple as this doesn't include the number of native speakers and second language learners etc. Annoyingly I can't find the source again.

The point though is that you'd probably need to look far further back in linguistic history to find the 'evolution' of tones. The relevant developments likely occurred long before anything like 'Chinese' came on to the scene.

[Altair]

I remember reading somewhere that actually the majority of languages use tones, not the other way round. Obviously the situation isn't quite that simple as this doesn't include the number of native speakers and second language learners etc. Annoyingly I can't find the source again.

I have read this as well. I have also read that the only large geography lacking tonal languages is Australia.

Tones have been an integral part of the language for so long that it is hard to imagine a workable version of the Chinese language without them.

This was certainly my impression until I learned that Shanghaiese is only marginally tonal. It can be analyzed as having a pitch accent like Japanese. Also, Tibetan is considered by many to be related to Old Chinese as part of the Sino-Tibetan language family, and Wikipedia says that "Classical Tibetan was not a tonal language, but some varieties such as Central and Khams Tibetan have developed tone."

As for the origin of tones in Mandarin, it is pretty much agreed that they go back to the four "tonal" categories of Middle Chinese: "level" 平, "rising" 上, "departing" 起, and "entering" 入. The phonetic values that correspond to these can only be determined in a general sense, since Chinese fangyan 方言 differ quite a lot in how they realize them and the historical descriptions of how they sounded are not sufficiently precise. There is a lot of confidence however, in tracing the development of tones since Middle Chinese. For more on this, read the "Four Tones" page on Wikipedia.

The origin of the four Middle Chinese "tones" is discussed here, where it says at the beginning of section 3.1 on page 15:

Since Haudricourt (1954)4, most linguists agree that Middle Chinese tones come from lost consonants. By comparing Vietnamese and other mon-khmer languages, Haudricourt first demonstrated that Vietnamese sắc and nạng tones came from final -ʔ and that ngã and hỏi came

from *–s through a *–h stage. He made the conclusion that a similar change occurred in Chinese, and proposed that shang came from a glottal stop, and qu from an –s.

In modern dialects, We can find direct traces of these earlier segments. In 孝义 Xiaoyi of Shanxi 山西 (Sagart 1999: 132-3), rising tones have a glottalization, and departing tones have a slight –h. Mei (1970)5 and Pulleyblank (1962, 1978) have found independent evidences to reconstruct a glottal stop in tone Shang and a –h in one Qu.

For more on the original of tones in general linguistics you can read about tonogenesis on Wikipedia.

[JonBI]

Tones weren't "invented" but were in truth discovered eventually by a Poet, who basically worked the transition into Shi form poetry, as before people were just speaking with tones, but were doing it naturally, rather than consciously - it was just the language. With the emergence of this thought, Tones were subject to become codified, in the sense that poetics forced poetry to be regulated over tonal metrics - the tone functions as an essential quality, so offers a pattern of recording - we see tones shift however, as spoken and written languages aren't the same, and the time since the discovery of Chinese tonality, and the shift in speech is a few centuries short of 2000 years. The question of emergence though is like asking how did stress emerge in English.

[Don_Horhe]

Huh?

Tones developed after certain consonants were dropped but the effect they had on the pitch of their adjacent vowels was retained.

Of the top of my head, vowels after voiced stops are pronounced with a relatively lower pitch, compared to those after voiceless stops. Take for example the nonsense syllables *dak and *tak. Phonemically, it's the same vowel, but if you measure their absolute values, you'll find that they're each pronounced with a different voice pitch. Now, assume that, over time, the initial 'd' and 't' wear away and only somebody's granny retains the old *dak and *tak, while everybody else says *ak and *ak. If this happens to isolated word pairs/groups, it won't cause any trouble, but if it happens to the whole language, then people will need a way to differentiate between all those words that now sound the same. This is where pitch comes in - the 'a' in *ak from *dak keeps its lower pitch, and the 'a' in *ak from *tak keeps its higher pitch, and it is precisely this difference in pitch that makes these two words different.

[Hofmann]

Tones weren't "invented" but were in truth discovered eventually by a Poet....

I find it difficult to detect the sarcasm in this post. Citation or it didn't happen.

[Altair]

Of the top of my head, vowels after voiced stops are pronounced with a relatively lower pitch, compared to those after voiceless stops.

A good example of this is the Mandarin rising second tone, which is apparently not a descendant of the Middle Chinese rising tone. The Middle Chinese rising tone generally gave rise to the Mandarin dipping third tone; whereas Middle Chinese level tone syllables that began with a voiced sonorant (m, r, l, n) or a voiced consonant became the Mandarin rising second tone. The voiced consonants became devoiced in Mandarin, and so evidence of the change is not readily apparent; however, one can still see that Mandarin first-tone syllables beginning with m, r, l, or n are comparatively rare. In fact, in the old Gwoyeou Romatzyh transliteration system, unmarked syllables are designated as first tone, unless they begin with m, r, l, or n, in which case they are designated as second tone.

Another example is what happened in Yue Chinese. The four Middle Chinese tones split fairly neatly into two levels of eight or so tones. With minor exceptions, the lower level mimics the contour of the upper tones, except on a lower pitch. Cantonese has had two more, partly optional tone splits that have only marginal phonemic significance.

[Player01]

What is the relationship between tonal languages and homophones. Do all the other tonal languages have the same problem of homophones as seen in the chinese language? Another thing I want to know about is what other countries with tonal language do with their writing systems. Do most of them use the alphabat?

[skylee]

What are your views? What has your research told you?

[Player01]

My view is that the limitation of sound available in Chinese has allowed the growth of homophones in the Chinese language. As for my research it is telling me that it is almost impossible to fix the Chinese homophone problem unless we depend less on the Chinese writing system (汉字) and develop new ways to read/pronounce. In my opinion this can be solved once China adopts Digraphia. Firstly pinyin is very advantageous when it comes to displaying foreign words borrowed into the Chinese language. Like instead of awkward character translation such as "MacDonald's" to "麦当劳" you can just write the word "MacDonald's" in pinyin. Perhaps altering the reading a bit so it is easier to read in Chinese. But the idea is to reduce chance of foreign words using the same sound as the Chinese language or funny translations. Beside that we can also start defining what a word in Chinese is like to the mass public. Most of the people (even a lot Chinese) think Chinese is a monosyllabic language and it is not easy to explain what is a word to them in a sentence of characters. But in pinyin you can display words between spaces so most homophones won't stand alone but instead it will always exist with another syllable. Lastly we can change the way you pronounce some homophones slowly and introduce some new sounds to the language from other Chinese dialects similar to the Standard Chinese of PRC, or in some cases you can even borrow sounds from China's somewhat culturally related neighbors like Japan, Korea and Vietnam. Yet this rise another problem. Some people love the language so much they won't easily accept borrowing foreign sounds into the language. If you are one of those people please consider this. Languages of the world are becoming more and more global and it is not rare to do this. Some examples are Japanese and English. Japanese has Go-on吴音, kan-on汉音 and even more and I don't even need to mention in details the always evolving English. And everything I said above is of course only the opinions I got from my research.

EDIT: By the way Digraphia can actually allow romanization and characters peacefully coexist and evolve in its own way without overpowering one and another. For example if Digraphia is adopted some characters badly simplified now lost its point and can be revised. Simplified characters are not bad but some are just not good.

[renzhe]

Chinese does not have a small number of sounds available. Even the number of valid syllables (when ignoring tones) is significantly higher than non-tonal, or minimally tonal languages like Japanese or Korean.

I am not sure that Chinese has a problem which needs "fixing". Languages tend to do well if you let them develop naturally. There is a slow, but definite move away from tones in modern spoken Mandarin, with the neutral tone becoming much more common, and this process is even more advanced in dialects like Shanghainese. Other dialects like Minnan or Cantonese have developed more tones than Middle Chinese. Languages are natural, and they tend to take care of themselves, and resolve ambiguity through additional information (tone, vowel length, new sounds), new words (such as several characters forming a word, where a single characters used to be common), or context.

Some pitch-accent languges, like Croatian or Swedish, have removed most of the ambiguity to the point where minimal pairs are rare, and don't even write tones (but do retain tonal information in speech). I think that modern Mandarin is not far from this -- you will be understood even if you totally mangle tones, as long as your sentence structure and grammar are correct. Mandarin has a comparatively large number of words where only tones differentiate, but it's not that bad.

[Player01]

Thanks renzhe for replay however I still have something I don't understand.

1. I couldn't agree with you more that language develop naturally and do well if let alone but this comes back to the question of tones. Why did the Chinese language lose its consonants if doing so disadvantageous the language? Perhaps tones were not developed naturally or it was actually good to have tones before?

2. And how will schools influence the natural development of the language? Won't Chinese pronunciation taught in schools prevent the language from doing so?

[renzhe]

1. You'd need a linguist to answer that. More likely, consonant stops were dropped as tones took over their roles, and they weren't needed anymore. But I don't know if anyone can answer that with any certainty.

2. Standardisation always slows down the development of a language.

[Rongxing]

"Tones weren't "invented" but were in truth discovered eventually by a Poet, who basically worked the transition into Shi form poetry," JonBI, I find this fascinating could you please share the poem that was written, the poet, etc.

[herradw]

And how will schools influence the natural development of the language? Won't Chinese pronunciation taught in schools prevent the language from doing so?

I agree with renzhe - school teaching will only slow down the process ... in Germany/France for example the changes in language use are evaluated and standards for teaching as well as the spelling of words are adapted in regular intervals.

[laurenth]

in Germany/France for example the changes in language use are evaluated and standards for teaching as well as the spelling of words are adapted in regular intervals.

Wow, as a French speaker I can tell you that French spelling is in fact extremely conservative and ossified. French spelling is a hotchpotch of etymological, semi-phonetic and assorted groundless conventions. I say "semi" phonetic, because the parts of the spelling that are supposed to be phonetic often reflect pronunciations that have ceased to exist centuries ago. Finnish, for example, has an almost perfectly phonological spelling, no need for "spelling rules". Spanish is also quite phonological, especially if you are from Castilla.

French spelling is *not* "adapted in regular intervals" - linguists, teachers, etc. regularly attempt to do so, but they almost always fail, because spelling is sacrosanct and defended by a powerful establishment. Learning spelling in the French speaking world is voodoo science inculcated by means of daily dictations and sanctioned by, of all things, national spelling contests.

I can see an interesting parallel with Chinese: a good deal of our school time is devoted to learning a written language that has acquired a cultural life of its own and is really different, phonetically, syntactically, morphologically from the way people actually speak.

The only other language I know whose spelling is similarly disconnected from actual pronunciation - apart from Chinese, of course - is English (with the added bonus that the oddities of English spelling were often inherited from old French, which makes it somewhat easier for us French speakers to learn...).

End of rant mode. Take this with a grain of salt : - )