I personally agree 99% with the Historical facts and Chinese linguistics that mention that both Cantonese and Hakka dialects are descendants of the original Han Chinese and that both Cantonese and Hakka preserve Middle Chinese [Sui, Tang and Song Dynasty Chinese] the most compared to other Chinese dialects.

When i mention Cantonese, i do not mean Guangzhou Cantonese [Standard Cantonese], but all Cantonese dialects like Taishanese, Zhongshanese, etc. Most Cantonese variants preserve the "i" vowel like the word 你 [Ni], 四 [Si], 十 [Sip], etc, except for Guangzhou Cantonese that evolved from i -> ei Example Ni -> Nei, Si -> Sei, Gwok Ki -> Gwok Kei (國旗).

I will use Taishanese as the representative of Cantonese, because it preserves the vowels better. Hakka and Taishanese both are similar in 3 ways:

1) They preserve alot of the old Chinese vocabulary words that are also found in Sino-Influenced languages like Japanese, Korean and to a lesser extent Vietnamese[there are a few words that are similar from my observation compared to korean and japanese]. Words like Ngit Bon (Japanese say Nip Pon) = 日本, Gai (Japanese say Gai) = 街
, Sai (Japanese say Sai) = 西, Ngi (Japanese say Ni) = 二, Ngin (Japanese say Nin or Jin/Korean say In) = 人, Nam (Korean say Nam) = 南 or 男, Sam (Korean say Sam) = 三, Hee (Korean say Hee) = 氣, Fi Gi (Korean say Fi Gi) = 飛機, Gim (Korean say Gim or Kim) = 金 and so on.

2) Both Taishanese and Hakka preserve the initial and final consonant. Both preserve the initial consonant Ng- (Example: Ngin = Person, Ngowt = Moon, Ngit = Day, Ng = 5, etc) and Final consonant -t, -m, -k, -p [Example: Yit[-t] = 一, Sam[-m] = 三, Luk[-k] = 六, Sip[-p] = 十].

3) Both Taishanese and Hakka preserve the "i" vowel. Example: Ni [你], Si [四], Mi Gok [美國], Sing Ki [星期], Gwok Ki [國旗], Pi Fu [皮膚], etc.

A Sample of Taishanese Cantonese [台山話] ->>> http://www.youtube.com/watch?v=gbeiQYt6kKI

Andy ...

I have 1 example for you,散步 ( stroll ) in Japanese sounds exactly like Mandarin & Hakka dialect.

It's the same in Taishanese too. Both Hakka and Taishanese are almost the same. Except that when Hakka pronounce the word 上 it is different from both Mandarin and Taishanese. Hakka pronounce 上 as Siong. Taishanese pronounces it the original Middle Chinese way -> Syang [Mandarin is Shang].

By the way, Mandarin is not the so called "pure Chinese" language.. it evolved differently from Middle Chinese. Mandarin lost the initial and final consonants. Initial Consonants: Ng- to Wu- (Example: Ng vs Wu = 五) OR Ng- to R- (Example: Ngin vs Ren = 人 and Ngit -> Ri = 日).

What is intersting to observe from Beijing Mandarin's recent change is the word 人 and 的. The word 人 in some Mandarin dialects[not Beijing Mandarin] is pronounced as Yen. It seems like the evolution of Beijing Mandarin started from Middle Chinese Nyin -> Yin -> Yen -> Ren. The word 的 is really interesting as well, as many old chinese singers like Teressa Deng, pronounces 的 as Di [in Cantonese and Hakka dialects it is Dik], but many modern singers and current mandarin speakers pronounce 的 as De. So the evolution of Beijing Mandarin probably went from Middle Chinese Dik -> Di -> De. To conclude, Taishanese and Hakka preserves more of Middle Chinese than Mandarin =)

Knowing the fact that both Taishanese(and other Cantonese dialects) and Hakka are very alike (linguistically and somewhat 99% genetically). By looking at geography and statistics, we know that the Hakkas live in Guangdong, Guangxi, Jiangxi & Fujian border. But what is even more interesting that will prove the fact that Hakka and Cantonese linguistically preserves Middle Chinese [Tang Dynasty Chinese] is that there are Hakkas who live in the North-Eastern part of Sichuan province which is next to Henan - the Birth Place of Chinese Civilization. Many famous Hakka people originated from the Northern part of Sichuan such as Deng Xiaoping(邓小平) and Zhu De(朱德).

What is interesting is that the people of Southern Henan look more like you're typical Southern Chinese, where as the Northern part of Henan look different as they intermix with the so called "Northern Barbarians" such as the Mongols, Jurchens, Turks, Manchus, etc...

The interesting thing between Mandarin and Taishanese Cantonese is that we both use the word 哪儿[Where] in our daily speech & we both pronounce it in the same way: Na R vs. Nai Ah. But In Guangzhou Cantonese daily speech[as oppose to reading], they would say "Bin Dou" to mean "Where".

interesting stuff

I wonder where did the hakka in Taiwan originated from? Fujian or GD or both?

Both,I think majority migrated from Canton because I personally met several Taiwan-Chinese could trace ancestral roots there.

Most of the Hakka in Taiwan originated from Guangdong region, predominantly Meixian region.

Andy, I am really indebted to your detailed coverage of this topic. Could you please be so kind as to outline the principle differences between Zhongshanese and Cantonese. As you may recall, my family is from Zhongshan. I find the differences between what my family speaks and proper Cantonese a little confusing. Many thanks, William.

what i know is that zhongshan is a unique cantonese dialect like Taishanese from Guangzhou Cantonese[Standard Cantonese]. Zhongshanese i know retains the "i" vowel example they say Ni instead of Nei for the word You.

Hey Andy,

If you do make more threads on the Cantonese/Middle Chinese connection, could you keep them to one thread?

Though I must admit, you've improved from before, your logic on here is better than the past where you seemed like you thought Cantonese was spoken in the past.

Why do you always say that?! I'll say this again: In the past i never said Cantonese was the exact language used in the past. I have always said (present and previous posts) that Cantonese dialects and Hakka retains most of Middle Chinese - the Chinese spoken during Tang Dynasty period - compared to the other Chinese dialects. But i also said that IMO Cantonese dialects seem to be more preservative than hakka, but i guess others will say otherwise.