Im a newbie and I have superficial knowledge regarding Chinese History.

I wish to ask, when did dialects originate?

In the first place I guessed that it was originated from the Period of Warring States, when there were many different languages that vary from state to state. And currently the scripts of these languages we now call 'dialect' are probably burnt during Qin Shihuang's reign. But since Qin Shihuang standardised the language throughout his empire, how did the languages of the different warring states manage to survive for a very long period of time, from Qin Dynasty up till the current PRC?

That was my mind-set in the beginning.

I think your premise is a little off.

Dialects didn't suddenly appear during the Warring States, they are formed regionally and over a long period of time. Language is adaptive and doesn't originate form a single source or fully conform into only one standard form.

Also, Chinese dialects and Chinese script are not-related as phonetic languages are. Qin Shihuang standardized the SCRIPT but not the language--that would be extremely difficult. So, there was no instant dialect shift to a standard Mandarin language. Instead, large geographic areas (the south) and ethnic pockets maintained their distint dialects, though *Standard* (that is the Beijing dialect) Putonghua has been slowly diffusing into these areas over time via population migrations, trade, and the need to adapt to the court dialect to attain a better position. Recently, modern media and mandatory Putonghua in schools have accellerated this process causing the spoken language to merge a little quicker towards one dialect.

Dialects did not 'suddenly' appear. Languages are influenced by the local people's language. The Han chinese conquered and intermarried with the local people over few thousand years.

By the time of the warring state era, there seems to be 4 main branches of chinese languages - the Qi, Qin, Jin, Chu.

http://dictionary.laborlawtalk.com/Chinese_language

There was a standard written form used by Qin ShiHuang but no standard speech. Mandarin was not even formed during his time. I would think the imperial language during his time is called 'Qin'. However, the Qin language evolved and split into today's Mandarin, Cantonese , hakka etc.

There wasn't a need for a standard speech because all communication at that time between provinces were done in written form. There was no TV or long distance phone during that time.

By the warring states period the Zhou language split into 3 dialects Chu to the south (along the middle and lower yangzte), Qi was spoken in the central plains and Jin was spoken in the northern end of the central plains. Eventually Qi split into Qin to the west central plains and Qi to the east central plains. The Jin and Qi eventually got replaced by Qin. However, the Qi dialect was preserved in the south in Fujian and is todays Min.

Over a series of migrations from the Central plains, central plains people (which spoke Qin) settled in Jiangxi and in Guangdong. The Qin dialect spoken in Jiangxi evolved into Gan whereas the Qin dialect in Guangdong evolved in Cantonese and Hakka. The Qin spoken in Central Plains evolved into todays Mandarin.

The northern migrations caused the Chu language to be cut into 2 at Jiangxi, Xiang to the west of Jiangxi and Wu to the east of Jiangxi.

Mandarin, Wu, Xiang and Gan have been under influences from Mongolic, Tungusic peoples. This can be seen with their loss of final consonants and simplification of tones. Mandarin and Wu also suffer from heavy palatalisation of initials. As Cantonese and Hakka are far away from the north, extreme simplification of tones, palatalisation and losing of finals have not affected them and they preserve features of Tang/Song Chinese which have been lost in other Chinese languages.

Even without much influence from other, languages do have tendency to change over themselves.
More or less I'm referring to the "lazy sounds".

How do they group those modern dialects into "Qin" language? Any specific evidence? Or plain speculation?

The loss of final stops and nasal consonants to glottal stop and vowel nasalization are also found in MinNan, which of course are not found by "new Chinese" loanwords.
Fuzhounese only have -k, glottal and -ng ending.

Suzhounese and Minnan both have 5+2 tones. Nothing less conservative in Suzhounese for the tones.

Simplification of tones do affect Cantonese.
斷 and 近 have 2 sounds in Cantonese, one of them is older and the other is influence from Yang-Qu and Yang-Shang confusion from some older version of Mandarin.
Most Hakka even only have 4+2 tones which would be very similar to tone system of Beijing Mandarin except that they have 2 more Ru tone.
And I thought Hakka are not affected by palatalization, but I heard 香港 "Siong Kong" instead of "Hiong Kong".

I'm quite sure that palatalization are not from Mongolic people as the velar sound + i-like sound are still there in Yuan Chinese, e.g. from document of Baijiaxing (however ng- sound is gone).
I also doubt they are from Manchu because many of their older surnames have a syllable of "giya".

I'm thinking more of Chinese-own lazy sounds (the place of articulation shifted to palatal under influence of [j]).

Example?

yes, i have heard of that too... there are people (especially meixian) that speak like that...eg. hiao tet (know) becomes 'siao tet' (s is palatalised)... i have been to shanghang, a hakka area in fujian and it seem palatization was pretty heavy over there, seems like affricatisation and fricativisation of velars is not a uniquely mandarin phenomenon then?




i m not sure if jin was wholly absorbed by the qin dialect or the central plains speech, even if it received influences from the standard speeches of various times, it certainly retains a separate character of it s own, especially the more rural areas.. (one has to remember that jin was 'attacked' by mandarin from 3 sides, from north, west and south)

Jin language was quite isolated from the mandarin population due to its high mountains surrounding the area.
That is why it still retains features such as k,t,p endings and extensive tone sandhi.


Shanxi is located on a plateau, which is in turn made up of higher ground to the east (Taihang Mountains) and the west (Lüliang Mountains), and a series of valleys in the center through which the Fen River runs. The highest peak is Mount Wutai (Wutai Shan) in northeastern Shanxi at an altitude of 3058 m. The Great Wall of China forms most of the northern border of Shanxi with Inner Mongolia

Fuzhounese is not Minnan but Minbei. The Minbei dialects have less endings compared with Minnan which has all the middle chinese endings - n, m, k, t, p, ng.

Although Hakka and Gan did evolve from Qin which is their grandfather language but their father language is still cantonese. The Hakka and Gan migrated later than the Cantonese from the central plains.

That's why hakka can learn to speak cantonese quite easily as there are many similarities.

No -p, -t, -k only glottal stop.
I doubt extensive tone sandhi is preserved from Middle Chinese.
There were only 4 tones, anyway.


Minnan 文讀 and 白讀 has somehow different ending phonology.
For 文讀 p t k m n ng, no [glottal stop] and [nasalization]
For 白讀 [glottal stop] ng [nasalization], maybe p and t in some cases but I'm not sure when they do not turn into [glottal stop].

The pattern of change is also different so that Wendu is much more similar to Hakka/Cantonese while Baidu is farther.
e.g. In Baidu, I see voiced sh->yang ts change quite regularly (as in 食 cia', 十 cap, 蛇 coa) while in Wendu I see voiced sh->s (十sip 蛇sia)...
Wendu looks more like Middle Chinese loans.

To a lesser extent, later Cantonese are also affected by central language.
After Cantonese split out, there was a major change in the North: YangShang->YangQu.
Example: 斷 近 上 下 淡
The characters I list above have 2 readingsin Cantonese, one is Yangqu and one is Yangshang. Yangqu one is considered 文讀 whereas the Yangshang one is considered 白讀.
In Cantonese, Yangshang stops are aspirated but Yangqu stops are unaspirated.

But some of the Yangshang characters, e.g. 似 and 市 do not seem to be affected by above change.

Similarity to each other between 2 languages does not show who is more original and who is less.

I don't see anything that makes the whole Middle Chinese can be classified as a Cantonese dialect, even outside phonology.

BTW, the word order "畀啲菜我" "Give some vegetable me" feels strange to me because I am more used to "Give me some vegetable" (as in Mandarin and English).

Dialects develop as languages do. Some dialects may have the chance to develop into a language while others may die.

I find palatalisation of Hakka initials really weird since it went through a process of de-palatalisation with alveo-palatal initials before... How much is palatalisation affecting Hakka? Is it just a minority? BTW how does Huizhou differ from Meixian Hakka? I heard it has an older set of tones.

Maybe a new effect of Putonghua because s+i vs xi, z+i vs ji, c+i vs qi is not distinguished?
Actually (compared to Guangyun), most Hakka merges 3 consonants to 1, Mandarin to 2... Just a merger...

Maybe Meixian Hakka and a few surrounding dialects?


I'm not sure how...
What I have heard is that some Hakka dialects have 5+2 instead of 7+2 tones.
And also in Fuichew[?] dialect there are words like "fun gaau" [瞓?]覺, "tei" 睇 (those 2 looks Cantonese) and 我 "nguoi". The question words are also quite different from Meixian (well, I think Dabu Hakka is also different).
If I'm not wrong, 我 in most Hakka is "ngai", and in Fuzhounese "nguai".

there are a quite a number of 'cantonism' in the hakka dialects (especially 'non standard hakka' i.e. non-meixian) for example: 'fun gaau' for 'sui muk' (like) , 't'ai' for 'k'on' (see), vun for ts'im (search), and even m guooi for kam chia (thanks), i do wonder sometimes, if they should all be attributed to influences from yue dialects, as not all of these dialects are located adjacent to yue dialects

btw, i believe that the cantonese of guangzhou or hong kong is probably not the oldest or most conservative yue dialect. which dialect would that honour go to? it is a little tragic that, cantonese similar to putonghua s position in mandarin, becomes so important that the other dialects are overlooked.

You mean 'fun gaau' for [something] (sleep) and 'jung yi' for 'sui muk' (like)?
BTW, as far as I see, k'->h in Cantonese affects most of Hakka dialects, and one exception is Dabu.

I see that 尋 is read ts'im and 謝 is read ts'ia there, so /z/-> ts' (somehow similar to Cantonese except the aspiration)?

Maybe they absorbed some Cantonese as they displaced Cantonese speakers from that area?


I'm not sure which one is the most conservative.

What I heard, Zhongshan is more preservative in single-vowel rimes than Guangzhou Cantonese, but I'm not sure if they also have other more/less preservative features.

Taishanese is better in vowels before a consonant e.g. 一, and the "originally flat s" has become the voiceless lateral fricative "ls" [ɬ]. Words with "sh-" has become s. And"Nh"->ngi like Hakka.
飲 im, 人 ngin, 二 ngei
三 lsam 四 lsei 手 siu 七 thit [well, seems that the Taishan dialect is not as good at preserving the consonant]
and also like well-known voiced (aspirated?) stop -> h
If I'm not wrong, the tone is less preservative, 三 and 四 the same tone...
edit:
(copied from http://www.zompist.com/sino.htm; not in IPA. "lh" is the lateral fricative:
jit 55 ngei 31 lham 33 lhei 33 ng 55 luk 31 thit 5 pat 3 kiu 55 sip 11
The correspondence of Rusheng pitches to normal tones is apparently different. And I doubt that 31 and 11 represent different tones in Cantonese)

Zhongshan Cantonese is probably most similar to Guangzhou Cantonese of the 1800s. Are there border areas between Cantonese and Hakka which preserve the vowels and medials of Hakka and the tones and finals of Cantonese?

my apologies!! what was i typing? i meant 'sleeping' for both words... somehow i really fell asleep ...'sui' or 'gau' is a element for 'sleep' in most hakka dialects..... 'fan' looks more yue though





i have often heard of a 'sam yap' and 'sei yap' division of cantonese, where would zhongshan lie lingustically and culturally? i think toishan, together with hoipeng, onpeng sanwui seemed to be sei yap, would the rest be sam yap? please correct me, thanks

The "fan" 瞓 looks like a reduction of initial from something like "khwan" (related to 睏), as we have kh -> h + hw->f in the Yin register.
You may notice that this "睏" is the word for "sleep" in Fujianese. The word in Wu is something like khungaw...

Some Hakka dialect pronounce 苦 as not k'u but fu because of the same effect (Meixian should be one of those dialects).

I'm not sure... but the languages among Sei Yap agree more to each other, and Sam Yap should also be like that, therefore Zhongshan is surely not Sei Yap...
And I'm not sure if Guangxi Cantonese would be different.

Zhongshan goes into the "Yue Hai" category or SamYap which is in the Guangzhou, Hongkong, Shunde and Zhuhai bracket. Jiangmen, Taishan and Kaiping are the SeiYap. The Cantonese spoken around Jiangmen is known as GaoYang dialects.

The SamYap dialects have more vowel shifts and dipthongisation then SeiYap and Gaoyang. Zhongshan dialect was immune to the dipthongisation which Guangzhou and Hongkong speakers have. eg. 3rd person is 'Ki' in Zhongshan while it has changed to 'Keui' in Guangzhou dialect. This pronounciation is exactly the same as 3rd person in Middle Chinese. Also, the n- to l- shift, gwo- to go- shift, kwo- to ko- shift and the dropping of ng- is seen in the Guangzhou dialect but it has been resisted in Zhongshan dialect.

The Seiyap dialects generally preserve simple vowels better but its initial consonants have changed more than Samyap.

SeiYap is not immune to diphthongization... look at the '4' of Taishan dialect.

Cannot be the same because it was voiced in Middle Chinese.
And, the "eui" diphthong points to "ü" rather than "i".

Are you sure? I only see that in Hong Kong. The n-/l- confusion and ng- droppage should not the case in Guangzhou. In fact that's the source of jokes about Hong Kong Cantonese. I've heard some cases of lazy "kwo", though.
There are also split of tone in Yinping category for Guangzhou dialect, but those are apparently re-merged in Hong Kong. "Flatter" 1st tone have taken over the falling 1st tone in Hong Kong.

About tones, I've heard some people from Shunde having difficulty distinguishing Yangqu and Yinqu. How about Zhongshan?

Example?
Diphthongs in Guangzhou/HK Cantonese can only occur if there is no ending.

Middle Chinese 3rd person was 其, 'ki' in Zhongshan Cantonese.


Pretty sure, I know people from Guangzhou and they don't speak 'correct' Cantonese but I havn't been to Guangzhou before. Have you been to Guangzhou and listen to the way people speak? There was a 'theory' to explain the occurence of lazy sounds and it states that places with high number of immigrants would simplify the language so its easier to communicate. It uses that to justify the simplification of Hong Kong Cantonese, Taiwan Mandarin and American English. Guangzhou also has high number of immigrants and are exposed to HK media so I'm assuming they have these lazy sounds as well. That theory can also explain the oversimplifying of Mandarin since northern China was constantly exposed to and governed by 'foreigners'.

Neither have I.
Some Hong Kongers can "detect" Guangzhou accent: when one of my friends from Guangzhou speak his language to a shopkeeper, often the shopkeeper will switch to Putonghua.
I don't know what difference is important, though.

Non-native speakers tend to speak the language wrong, but not neccessarily too simplified, e.g. Taiwan Mandarin.
Don't you think British English phonology is more simplified than American English, by dropping the -r endings?
Some (not all) American accent have vowel merger though. Well, but Bristol in English also have more consonant mergers. It's simply a generalization to say all American English accents are more simplified.

Well, the simplification by exposure would tend to fit into the interacting languages instead of just making it "as simple as possible".
As far as I know, Manchu do not have glottal stop and -m ending.

The main difference between Guangzhou and HK Cantonese is the tones. Guangzhou Cantonese tends to stress and emphasise tones more than HK Cantonese.

This is what explains HK lazy sounds and other lazy sounds in general. Its in Cantonese Chinese not Mandarin Chinese.

1949年後，香港方言開始出現大量懶音，當中以鼻音消失（即n/l不分）同埋w拗音嘅消失最為顯著。新一代年青人普遍把「你」[nei]同「我」[ngo] 唸成 [lei]和 [o] 。把「國」[gwok] 誤讀成「角」[gok]，「過」[gwo] 讀成「個」[goh] 。呢個現象似乎同大量外地移民有關，對佢哋嚟講，n/l 嘅發音差異唔大，喺大部分情況下混淆兩者亦唔會帶嚟嚴重嘅溝通障礙，於是佢哋來港學習這種新方言嗰陣，往往捨難取易，淡化一啲難以分辨嘅發音差異。呢個亦即係所謂「移民理論」，呢種現象喺台式國語、美式英語嘅演化過程中，亦曾經出現。

Are Korean and Manchu language from the same roots? How come Korean is so much more diverse than Manchu?

Guangzhou Cantonese one is falling in general, flat it some cases, but in new HK Cantonese all becomes flat. Some older people in HK do pronounce falling tone, but I'm not sure about their origin.

Well, it seems to me that most Mandarin Chinese dropped the ng- or "畫蛇添足" on ng- since a long time ago.
The n-/l- confusion also exists in some dialects of Mandarin, for example Nanjing.

One of my Cantonese friends come from Malaysia and his father is a Guangzhou Cantonese. Other than his tone is rather flattened because of Malaysian Chinese influence, I don't really spot lazy sounds like ng- dropping and n/l confusion. (Hakka influence on the 呢 'this' -> 'li')
But then, after spending months in Hong Kong, he developed a "bad habit": when he is talking to Hong Kongers he "switches" to the lazy reading. (maybe intentionally because when I test him he still can pronounce those sounds naturally)

One other Malaysian speak lazy sounds exactly like Hong Kongers but I don't know how come. Dropping the ng- looks like a very recent feature near Hong Kong.

How can you be sure that it's really 1949?
Interacting languages tend to fit their phonology to each other, but most languages distinguish n- and l-. Actually I am suspecting the n/l confusion come from Fujianese.

The simplification in Taiwanese Mandarin are clearly effect of Fujianese speakers (i.e. they are non-native speakers). But sometimes the change in Beijing Mandarin is larger and irregular.

How are American accent more simplified than British accent apart from merging the 'a' as in farm and the 'o' as in long? But the -r dropping is more significant to me than merging those 2 sounds (only distinguished by rounding). Without -r ending, English just sounds unconfortable for me...
And I remember that there are some words that lost distinction in British English because of this...
How about American English?

{
EDIT: in fact there are so many: http://en.wiktionary.org/wiki/Special:Sear...ccent+homophone
Example, "awe": homophone with "ore", "or" and "oar"
Well, maybe I have misjudged British English, but surely in London the -r had been dropped since long time ago: http://en.wikipedia.org/wiki/Rhotic_and_non-rhotic_accents
}

New effects in Hong Kong that might not be too dominant is the confusion of -k/-t (more common) and -n/-ng (less common).
I see that the -k and -t are glottalized for the speakers who are confusing them and that is one explanation.

Some linguists theorize that Korean are Altaic, while seemingly most disagree.
Korean are lazy enough for the endings (pronounced more simply than what is written) but they still distinguish quite many.

Do you know if theres any government action taken against lazy sounds in HK? They proposed before to make a pronounciation exam as part of education in government schools. But I noticed the reporting of HK news are very accurate in pronounciation.

I find it pretty ridiculous with the -n/-ng and -t/-k confusion and also when people say 'heui' for 3rd person.

I don't know. I have to ask locals.
Some people seemed to have learned something about n/l difference.
Strangely, some of them still leave out the case for ng- while it seems a later effect.

Actually I think "lazy sound" is not too accurate for most cases because they are confused rather than somebody preferring the easier sound.
These can be truly considered as lazy sounds:
characters "Yale"
七十九 "cha-a-gau"
即係話 "je-e-wa"

HK news are most of the time accurate in pronunciation, except for that they regularly 畫蛇添足 when it should be "null" consonant, e.g. I've never heard the standard version of "安全" in news.

Btw, in Cantonese and most other languages, the "null" consonant actually is realised as a glottal stop. From fanqie, it seemed that ancient Chinese considered the glottal stop to be a consonant, e.g. 於 哀都切, 影 於而切, 安 於寒切.

Glottal stop is voiceless, but ng- in Middle Chinese was voiced and that is why words starting with glottal stop always go to Yin category, while word of ng- most of the time goes to Yang category, which make it difficult to find a minimal pair. 勾 changed its initial from voiceless unaspirated k ("g" in Yale) to ng that is how a minimal pair can be found with glottal stop, 勾 vs 歐.

Yeah....
And even in television there was an advertisement of -t/-k confusion. Usually when -t/-k are confused they are glottalized somehow. Without that glottal stop they won't sound similar. But some do not pronounce with glottal stop but still do it wrong...
And one question one may ask is, why "heui" when "k" and "h" are not confused...
Though we do see some change from k to h and h to k in words like 吸 and 去. (well, the latter case is quite common for "Yin" kh. I wonder why 區 and 啟 are not affected...).

barely anyone in HK differentiates between n- and l- except for some songs and the news


the n- and l- really did start off as a lazy sound and the later generations are the ones that are confused. its jus lik ng- which was in Mandarin, it startd off as a lazy sound and current Mandarin speakers don't say ng- not cos of laziness, but they don't know that it 'exists'. now the lazy sounds become the standard, 'I' in Mandarin is 'wo' instead of 'ngo'.

people actually say 'chat a gau' not 'cha a gau'
n they say 'jek hai wa' not 'je e wa'


I never noticed that but most of the time they are accurate. In the end its better to 'overuse' the 'ng-' then for it to completely dissapear lik in Mandarin.


So you mean the 'vowel initial' actually has a glottal stop before the vowel?


*sigh* ive heard people saying 'heui' in HK drama...
why does h become k since people chose k -> h as h was easier to pronounce.

i actually think cantonese initials arent 'fully' depatalised
次 is ts'i and 字 is tsi
but for 蠢 and 准 the initials arent really the same as 次 and 字.
蠢 is more lik ch'eun not ts'eun, while 准 is more lik cheun and not tseun.
so maybe cantonese has 21 initials and not 19??

but sh- and s- differentiation is really lost (at least in HK) as I hear noone saying 'sha tin' they all say 'sa tin'.

"Ngo" should be "o" but it did not become "e" as usual pattern (鵝, 餓), but rather "wo".
Ng- was already not there since Yuan dynasty Chinese.

btw, how do you prove that it was a lazy sound?
If lazy sound can be blamed for every change... then we can say that all modern languages are products of lazy sounds, unintelligible because of different lazy sounds...

Well, those examples are what I heard from my surroundings.
I think many people speak this "je e wa" anyway. Some people speak "jik hai" as "je hai" anyway (I mean when there is no "wa"), notice Uncle Bus' language. If I recall correctly, in his language there are sometimes "jik hai" as well, but "je hai" is more common.

---

For fast speech there can be linking, like 49 "sei ya gau". That is one reason to doubt the "chat a gau".
Well, no alveolar stop t; not even glottal stop is heard between "cha" and "a" to my experience.

There can be merger of "i"s like 第二間 "daii gaan" (I "invented" a new symbol "ii").
Then 二十九 will be "yia gau". "ia" is not a valid rhyme in slow Cantonese, so it will be either "e" or "a".
廿 should sound more like "yap" 入 if you follow the classical Chinese, 二十切 == 人汁切 ~ 入.
I don't know where "nian4" of Mandarin came from.

I think the overusing of the "ng" will lead people to confuse it with the "zero initial" because they really become indistinct.

Almost all of the time I really hear glottal stop.
In a few language it matters if you add a glottal stop before a vowel... for example Hawaiian.
In English I think there should also be a glottal stop initial, which is lost in the consonant linking, e.g. an aim -> a name.

Well, they try to immitate usual speech as much as possible.

Well, you have to find *direct* minimal pair in order to show that they are really different instead of being "allophonic" (esp. conditional allophones).
If it happen to be true, then tsh tsh' s would be phonemes. And I think the glottal stop should be an initial phonemes as well... (though many just consider it "null"/"zero" initial)

Well, Chinese dialects are somewhat like French, Italian and Spanish in Europe, which are not mutually intelligible but belong to the same language family. The only point is that China is a still a unified country while the Roman Empire has fallen for good.

Can you explain in more details why you said Jin is replaced by Qin ? I thought there is a chinese language called Jin. How different is it from the ancient Jin language ?

Modern French, Italian and Spanish are derived from vulgar Latin language,and they are not considered dialect to one another.
Dialect is also not perculiar to Chinese language but also to other languages as well,i once read a thread by a French stating that dialect from other French regions are rather unintelligible to one another, while German language also contains dialects which could be understood in written form but not orally.

Yes. I am a Fui Chew 惠州 Hakka. Some words peculiar to this Hakka sub-dialect that differ from Standard Hakka include:

睏覺 fūn-gău (vs. 睡目 sôi-mūk "sleep")
睇 t'ǽ (vs. 看 k'ôn "look")
閒嘢 hăn-yă (vs. 東西 tūng-si "thing")
畀 bí (vs. 分 pūn "give")
哪誰 năi-shü (vs. 物-e-人 mat-e-ngin "who")
哪裡 năi-ní (vs. 哪位 lā-wûi "where")

There is a debate on whether it should be classified as a Hakka sub-dialect, or as a Cantonese sub-dialect (commonly known as 惠城話).