Discounts on Second Sinitic Languages

Since I don’t know the answer, I’d like to forward a question left on an unrelated post. “Deany” asks:

How much “discount” can someone gets if she/he is already able to speak cantonese and want to learn mandarin (pu tong hua)?I’ve noticed that a friend of mine who is able to speak cantonese (raised in a family who is able to speak cantonese), he has a higher speed in learning mandarin. Is it true that mandarin is easy to master when you already fluent in one of chinese dialects?

Then Albert of Laowai Chinese chimes in

I myself would like to find the following resource (though I admit it would be a very VERY small market): A Cantonese instruction book for foreigners who already speak Mandarin. I, like you, would love to see an overview of the differences to get an idea of what kind of “discount” I’d get if I actually tried to learn Cantonese.

I figure there’s gotta be a discount from knowing one Sinitic language to another, at the very least because there are lots of cognates and grammatical similaries. It might also be easier just going from one tonal language to another (here is some experimental data and anecdotal comments from Sinosplice).

The difficulty in answering the “discount” question is that lots of folks in China who know two Sinitic languages learned them both as kids. Usually, they started out with something non-Mandarin as a native tongue then learned Mandarin very early in school. So you can’t judge from their experience whether the adult second language experience is easier or not.

It would be interesting to hear from those who learned a second Sinitic language as an adult about how knowledge from the first helped or hindered. Anyone got stories or data?

35 responses to “Discounts on Second Sinitic Languages”

  1. Not Sinitic but Semitic: After spending a few years studying Arabic I started classes at the local Synagogue in Hebrew. After two classes they kicked me out of beginner and into the upper level class… which was then cancelled due to low enrolment. Point is, the languages are so similar, it took little to figure out the complexities of Hebrew coming from an Arabic background.

    So I’m sure there’s a discount, and I’m sure it’s huge. I know someone who grew up speaking Min as a child. Then when a little older they moved to the north where she learned Mandarin. As an adult, she moved to Shanghai (Wu) and spent all of a month here before she could understand almost everything that was said and only slightly longer before she could take part in the conversations.

  2. flow says:

    i never learned another chinese 方言 after mandarin / guoyu / putonghua or whatever you want to call it, but i did learn some japanese and some korean after that. my experience is somewhat mixed: with japanese, i was spared to relearn kanji; the differences between japanese and chinese are rather minor compared to the total number of characters. however, usage in japanese may make a huge difference in some cases, and there is a fair number of false friends where you read one character as something completely different. also, being in japan was so funny because i sometimes couldn’t say the most simple things but then i used a lot of upscale vocabulary—it just read out the chinese characters for whatever i wanted to say and tried to guess their sino-japanese readings, which is often possible (i guess this should work with chinese dialects as well). similarly, when in korea (while i did not advance beyond good morning and thank you as for korean itself), after some time i could read a great deal of shop signs and so on, since half of that vocabulary is chinese in origin, using a simple orthography to spell out 16th-century or so northern chinese; again, you can guess a lot in both directions. in my experience, knowledge of a related language did open some doors, but it also may have kept me from getting deeper into the other languages. also, there is always a fair amount of interference; a friend of mine who studied both italian and spanish tells me the same. she finds it hard to cleanly switch back and forth between the two because they are so confusingly similar. just do it and see how it works!

  3. Karan says:

    There is a HUGE advantage in learning Cantonese if you already know Mandarin. This is from personal experience. There are of course tonnes of words that are different but even more words that are the same and the technical vocabulary is pretty much identical. Shanghainese is even easier to learn if you know Mandarin. I’m starting to pick it up and I have to say it’s not hard.

  4. 凯文 says:

    There was a woman at the who was in my adult Mandarin class in the United States. She is originally from Hong Kong. The class was geared to those who knew no Chinese (we actually used childish like books but supplemented with more reading). She easily wrote the characters but she seemed to struggle as much as I did in memorizing the vocabulary. She had problems with learning the tones and some of the pronunciation.

    I assume being able to learn Cantonese from someone that knows Mandarin but their childhood language is English would be like someone whose original language is Mandarin and knows English trying to learn German.

  5. Yunxiao says:

    I agree with Karan; it only took me a few weeks of living in Guangzhou to understand a little of what people were saying in Cantonese (with prior knowledge of Mandarin, obviously), without any formal instruction.

  6. Syz says:

    @Kellen: Semitic example is cool. I had no idea
    @flow: occasionally people make the case for the “universality” of chinese characters along the lines that you’re talking, that they make, say, Korean and Japanese much easier. There’s no doubt knowing hanzi already saves some time, with the caveats you point out about false friends and so forth. But the grammars are so radically different that I’m doubtful about much benefit beyond the borrowed words.
    @Karan: you should consider membership in the Club for Sinitic Masochists 😀
    @凯文: I’m surprised she’d struggle with vocabulary, somehow. I wonder if it was an interference problem like @flow was describing betw Italian / Spanish…
    @Yunxiao: That’s a pretty strong discount. Are you a native speaker of Mandarin?

  7. Tom says:

    “I assume being able to learn Cantonese for someone that knows Mandarin but their childhood language is English would be like someone whose original language is Mandarin and knows English trying to learn German.”

    This is about right. My Cantonese grammar book gives examples of parallels (or more often points out non-parallels) in Mandarin, which is very useful to me.

    Yunxiao’s comment is very believable. It’s easy to pick up what people are saying (in general terms) very quickly, but of course you don’t have the ability to reply quite so soon. When stuck for the Cantonese word, I end up using the Mandarin word with a Cantonese pronunciation.
    This works even less often than people would imagine, since even Standard Written Chinese in Hong Kong is not what it is in Beijing (I haven’t been to mainland Cantonese speaking areas much, so couldn’t comment about that).

    For example, in writing, Canto-speakers would recognise “我已经付钱了。” But saying that, or using 給 pronounced as ‘kap’, won’t get you far. You need to use 俾 instead.

  8. Tom says:

    And the grammatical particles are different too, I forgot to mention that in my example.

  9. Karan says:

    @Syz and @Yunxiao Once you have the sound mapping down (like ji –> gei, tiao –> tiu, etc.), then it’s just a matter of getting enough exposure. And currently, I’m doing just that by watching Cantonese TV dramas with Mandarin subtitles *blushes*. It’s amazing how much my Cantonese listening and speaking ability has gone up in a matter of just a few weeks by watching these shows… Now, I can pretty much understand people’s everyday conversation and even hold short conversations. I strongly recommend this method to anyone who is interested in learning Cantonese and knows Mandarin.

    @Tom Yeah, since everyone going up in Hong Kong will learn “書面語”, they will be able to recognize Mandarin words, but as you pointed out, the basic grammar still needs to be in Cantonese. In fact, that is a huge advantage for Mandarin speakers because this does not work the other way around at all! For example, you can go to Hong Kong and say “我好喜歡呢條褲子 Ngóh hóu héifūn nī tiùh fují” in a shop and they would understand what you mean, even though a more natural way of saying it would be to use “鍾意” instead of “喜歡“, and to just say “褲” instead of “褲子”. And Mandarin or “Standard Written Chinese” has a lot of influence in Spoken Cantonese as well; for example, although the word “example” traditionally translates to “例” in Cantonese, you might hear a lot of people say “例子” even though the “子” is superfluous in Cantonese. However, a Cantonese speaker cannot go to Beijing and say “我好鍾意這條褲”, or “我踩單車去“ instead of “我騎自行車去”. So, although you will have an advantage going from Cantonese to Mandarin or from Mandarin to Cantonese, I think that perhaps you have a bigger advantage going from Mandarin to Cantonese because most Cantonese speakers know Mandarin vocabulary and grammar at some level.

  10. dlszho says:

    I’d say that there is a huge discount for sure. I started learning the basics of Mandarin at the age of 15 (overseas), stopped, started again and hadn’t really spent all that much time on it–maybe 100 hours of language exchange over five years–until I came to China and found that I was fairly fluent. I speak Cantonese natively, and I also read and write Chinese–which certainly helps.

    Vocabulary, especially the more formal words, are really close, though there are some tricky faux-amis in the colloquial words. Some differences in grammar, but nothing overly complicated. I do feel frustrated not being able to express in Mandarin all the emotions the final particles of Cantonese carry (eg. ge bo, ge lo bo, or gwa etc, especially when two or more are stringed together).

    Tones in Mandarin were fairly easy, though consistently distinguishing tones 1 and 4 was hard. Still, tones can be mapped fairly accurately, say 80-90% of the time from Cantonese to Mandarin–but probably not the other way around: e.g. Cantonese tones 1, 4, 2, 3&6 map to Mandarin tones 1, 2, 3, 4; Cantonese tone 5 maps to Mandarin tones 3 or 4 depending on whether the initials are obstruents or sonorants, etc (tones 7 & 8 are scrambled in Mandarin). On the other hand, while a lot of the initials map from Cantonese to Mandarin, the alveolars and retroflexs (z, c, s; zh, ch, sh,) have merged in Cantonese and I constantly had to guess the correspondence in Mandarin. And there are always freak cases like 秩序 (Mandarin: zhìxù, Cantonese: dit6zeoi6).

    Right now, I’m trying to pick up a little bit of Shanghainese and I kind of feel it’s phonologically between Cantonese and Mandarin (except for the voiced initials). In fact, if I know what the conversation’s going to be about, I can understand quite a bit.

  11. Syz says:

    dlszho: So before that bit of overseas Mandarin exposure at age 15 is it fair to say you had NO Mandarin exposure? TV, maybe?

    Interesting to hear that tones 1 and 4 were hard. I’ve heard elsewhere as well that Cantonese-Mandarin is easier for tone learning than vice versa.

    Good luck with the Shanghainese. Maybe you can report back in a few months on whether it has been harder/easier than Mandarin!

  12. dlszho says:

    So before that bit of overseas Mandarin exposure at age 15 is it fair to say you had NO Mandarin exposure? TV, maybe?

    Well, really almost no exposure to Mandarin at all (this was in the 80s and early 90s)–I probably had more exposure to animes in Japanese than Mandarin. This was a time when Hong Kongers didn’t have too much interest in the mainland (fear and disdain, yes, but not all that much interest); TV was almost completely in Cantonese (or English).

    I did go to Shanghai at the age of 12 for three days. I understood basically nothing–but I could read, more or less. Needless to say, communication was next to impossible (no one spoke English).

  13. Bruce says:

    I find a number of the examples cited above a bit misleading.

    The relationship between English and German, or Hebrew and Arabic, are much more distant than that between Mandarin and Cantonese.

    I think there are two factors confusing us here: 1) Some of the contributors don’t know both Mandarin and Cantonese well, and thus view one or the other as somewhat “foreign”; 2) The long-standing intellectual debate (often thrashed out on this site) about whether Mandarin and Cantonese are dialects, different languages, etc.

    Re: 2, this is perhaps a topic of intense interest to linguists, but for those of us who have actually learned both of them, I would submit that knowing one means a HUGE “discount” when it comes to mastering the other. I mastered Mandarin before arriving in China, and learned basic Cantonese in HK, mainly by listening to and imitating the baby-talk that my in-laws babbled to our daughter.

    Although my Cantonese is limited, I can think simple thoughts in the language, and I know why; because I mastered Chinese years ago — the deep structure, the grammar, the idioms — and can apply them sub-consciously to both Mandarin and Cantonese with relatively few major screw-ups.

    I learned German before I went to university and at one time read Kafka and Thomas Mann in the language. But I rarely “thought” in English when reading German literature, because the “distance” between the two is quite considerable.

    For the reasons I’ve just mentioned, when it comes to learning one or another version of Chinese — Mandarin, Shanghainese, Cantonese, etc — on the ground, solid knowledge of any one of them is a massive advantage. The main reason that Mandarin speakers in particular have trouble with Cantonese is their prejudiced attitude toward it (“just a dialect,” “not the national language,” “sounds like everyone is arguing”), and not the fact that it represents a “different” language. And I wasn’t immune to that: I lived in HK for many years, secure in my Mandarin cocoon, before I deigned to learn Cantonese…

  14. Syz says:

    Looks to me like the crowd supports “huge discount”. Summarizing from above, everyone who has extensive personal experience knowing one Sinitic language and moving to another (Karan, Yunxiao, dlszho, Bruce) has said the learning is pretty rapid.

    @Bruce: good point about “Cantonese is just a dialect” mentality among Mandarin speakers. For some Chinese, the “let them all speak Mandarin” attitude seems even stronger than the “let them speak English” attitude in the US — and that‘s saying something!

  15. Chris Waugh says:

    @Bruce, this statement confuses me: “But I rarely “thought” in English when reading German literature, because the “distance” between the two is quite considerable. ”

    I rarely think in English when dealing with another language that I have acquired beyond a certain degree of “fluency”- for example, I rarely think in English when processing French or Chinese texts, but I would probably need to revert to English fairly often when dealing with written German now (although not so much 10 years ago).

    In any case, I’m inclined to agree that learning one Sinitic language should buy one a big discount in another. I base that on my experience of having picked up texts in Romance and Germanic languages and having had no major difficulties in getting the gist of the article, and watching Hong Kong films and reading the Chinese subtitles while listening to the soundtrack. One must be careful of faux amis and similar problems, of course, but…

  16. Kellen says:

    Other than false friends in fairly prominent places (Arabic and Hebrew both having the word “lahm” but meaning meat in one and bread the another), I’d say the two are still quite close. Enough that the comparison to Sinitic languages ought not be discounted. Unless we’re talking about the differences between Gan and Mandarin or other cases where the amount of historically recent influence is so major. I do think the German/English comparison is no good given the massive amount of non-Germanic words that have replaces their Romantic counterparts.

    I’ve heard it said many times by people who claim to know that learning one foreign language makes the second easier, and the second in turn helps the third. It may be that the learner actually just figures out how they most effectively learn languages, as opposed to their German helping them learn Thai. But the way I’ve heard it, there’d be a discount even if the languages were in no way related.

  17. Claw says:

    Syz wrote:

    Interesting to hear that tones 1 and 4 were hard.

    That’s because most dialects of Cantonese don’t really distinguish high level and high falling tones. They’re both considered 陰平 in Cantonese.

    I knew Cantonese first and also found that there was a huge discount when learning Mandarin. Even before I found out about the systematic tone and phoneme correspondences between the two, I was able to intuit with fairly good reliability how a word in Mandarin should be pronounced based only on its Cantonese pronunciation.

  18. Albert says:

    Very interesting indeed!

    @Claw, et al: do you have a list of those little tips and tricks such as the “systematic tone and phoneme correspondences”? I would pay my weight in zongzi-s for that resource.

    (I just wanted to try the cool little IPA buttons)

  19. Bruce says:

    No doubt there is a list of “correspondances” between Cantonese and Mandarin somewhere out there.

    Cantonese to Japanese correspondances are perhaps better known and simpler, partly because they don’t touch on tone. For instance, it is virtually certain that any Chinese character pronounced in Cantonese with a sound such as k,m,p,t at the end of the syllable will have certain — quite predictable — characteristics in the Japanese. For instance, 国=Cantonese ‘kwok’=Japanese koku (the Cantonese glottal stop becomes ‘ku’ in Japanese). Or, 一=Cantonese ‘yat’=Japanese ‘ichi’in isolation, which becomes a glottal stop before certain sounds. For instance, 一本 is pronounced ippon, with the two ‘p’ representing a glottal stop between the ‘i’ sound and the actual ‘p’ sound. These changes are perhaps most noticeable when the numbers 1 (yat),3 (saam),6 (luk),8 (baat) and 10 (sup) are combined with measure words in Japanese, and that’s quite predictable, because 1,3,6,8 and 10 all end with k/m/p/t in Cantonese.

    I didn’t study Cantonese-to-Mandarin changes systematically, and have never seen a list per se. But tone-wise, I did notice these phenomena: 1) Cantonese syllables with k/m/p/t at end are always flat tones (not rising or sinking) in Cantonese; 2) Most high, flat tones in Cantonese (referred to as first tone, I think) “convert” exclusively to 1st or 4th tone in Mandarin. For instance, 一张台. ‘Zhang’is a first tone (high, flat) in both Cantonese and Mandarin.

    Please note that I am not a linguistics professional, and I speak here — as always — simply as someone who has noticed certain things through practice, daily speech and a bit of reading.

  20. Karan Misra says:

    @Albert Here’s the mapping I figured out from Mandarin to Cantonese:
    陰平 Tone 1 -> Tone 1 (High Flat to High Flat)
    陽平 Tone 2 -> Tone 4 (High Rising to Low Falling)
    上聲 Tone 3 -> Tone 2/5 (Falling-Rising to High Rising and Low Rising)
    去聲 Tone 4 -> Tone 3/6 (Falling to Middle Flat and Low Flat)
    入聲 (nothing) -> Tone 1/3/6

    So, as you can see Mandarin merges 陰上 and 陽上 into just one tone and 陰去 and 陽去 into one, thus when going from Mandarin to Cantonese, you cannot with 100% accuracy determine whether tone 3 is going to go to tone 2 or 5 in Cantonese. For example, the word 意義 is yìyì in Mandarin but yiyih in Cantonese (tones 3 and 6 resp.) Also, there are some exceptions that occur with this rule (such as 錢 which is tone 2 in both Mandarin and Cantonese) but those are mostly because of Cantonese’s tone change rules.

    Also, words that have the 入聲 or glottal stop (-p, -t, -k) in Cantonese do not have any easily decipherable mapping to Mandarin since Mandarin lost all glottal stops. So, for these, rote memorization is the only way. However, the good thing is that if they are glottal stops, they can only be tones 1, 3 or 6 – only the flat ones. So, you have a 1 in 3 chance of guessing right. :-)

  21. Tom says:

    @Karan: I’d worked out a partial correspondence but this is very handy. Thanks!
    As for Cantonese tone changes… gah. Sandhi as a semantic item (if I’m remembering my grammar book right)? No fun for the learner.

  22. Zev Handel says:

    @Karan: the rules can be refined further depending on how much detail you want to get into. For example, a Mandarin fourth-tone word with an aspirated initial will be 陰去 in Middle Chinese (and thus middle flat in Cantonese). Ru-tone words (glottal stop endings in Wu) that begin with resonants (including all nasals and l-) will always be the lowest ru-tone of Cantonese. (lip “elevator” is an exception because it is a borrowing from English.) Non-resonant initial ru-tone words will be high or mid in Cantonese, but this is predictable from vowel length: only Cantonese long-vowel syllables have the mid tone. Cantonese low-rising (陽上) syllables are almost always equivalent to Mandarin third-tone with resonant initials; the handful of exceptions are colloquial reading Cantonese words with aspirated initials that don’t match Mandarin’s unaspirated initials (like ‘sit’ and ’embrace’), plus the third-person singular pronoun.

    The more familiar you are with the two languages (and with Middle Chinese), the easier it is to find these additional details useful.

  23. Zev Handel says:

    @Tom: The tone changes in Cantonese technically aren’t sandhi, because they aren’t predictably conditioned by the tones of surrounding syllables. They are part of the morphology rather than the phonology of the language. That is to say, the change of tone imparts a new meaning or function to the word. In many cases, the change to a rising tone in Cantonese is somewhat analogous to er-hua in Mandarin.

  24. Claw says:

    FYI, a glottal stop refers to a very specific sound (see, and none of Cantonese’s final stops are glottal. The -p/-t/-k stops in Cantonese (words of which have 入聲 tones) did get merged into a single glottal stop sound (-ʔ) in most of the Wu dialects though, so perhaps that is where the confusion came from.

    @Bruce: the -m final in Cantonese is not part of the -p/-t/-k 入聲 finals. -m in almost all cases got merged with -n in Mandarin and can have 平, 上, or 去 tones.

    I don’t have too much time now, but I’ll post more on sound correspondences later.

  25. Claw says:

    Most everyone else has commented on the tone correspondences, but I’ll summarize them here plus add some more information. It helps if you think about the Middle Chinese tone system and how Mandarin and Cantonese evolved from it. The Middle Chinese tone system consisted of four categories: 平 (level), 上 (rising), 去 (departing), and 入 (entering), each divided into two registers: 陰 (yin/upper) and 陽 (yang/lower), in total comprising 8 tones. Incidentally, each of these names were pronounced in their respective tones (e.g., 入 is pronounced with a 入 tone; 陰 with a 陰 tone, etc.).

    As mentioned previously, all words with 入 tones ended with -p, -t, or -k. The 陰/陽 distinction came about due to a initial voicing distinction. All 陰 words began with an unvoiced initial consonant; all 陽 words began with a voiced or sonorant consonant.

    Cantonese essentially retained the Middle Chinese tone system but split the 陰入 into two separate tones, for a total of 9 tones. They are traditionally numbered as follows in most Cantonese romanization schemes:

    1 陰平; 2 陰上; 3 陰去;
    4 陽平; 5 陽上; 6 陽去;
    7 (or 1) 上陰入; 8 (or 3) 下陰入; 9 (or 6) 陽入

    Because 入 tones 7, 8, and 9 are pronounced with the same tone contours as 1, 3, and 6, respectively, many schemes just list 入 words as 1, 3, and 6 since it’s obvious from their -p/-t/-k finals that they have 入 tones. While Cantonese 入 tones are pronounced with the same contours as the other tones, there’s evidence to suggest that the 入 tones had distinct contours in Middle Chinese.

    I also mentioned in a previous comment that words with 陰平 tones may actually manifest with two different tone contours: high level (1) and high falling (1′). For all intents and purposes, you can ignore this distinction since most Cantonese speakers can’t tell the difference anyway. This is why you’ll see various claims of Cantonese having 6, 7, 9, or 10 tones; it depends on how the 入 and 陰平 tone contour distinctions are counted. For most purposes though, 6 tone contours are all you really need to deal with when learning Cantonese.

    Mandarin, on the other hand, lost many of the Middle Chinese tone distinctions, with many of the tones merging together. Here is how the Middle Chinese tones map to Mandarin tones:

    陰平 – corresponds to tone 1
    陽平 – corresponds to tone 2
    陰上 – corresponds to tone 3
    陽上 – corresponds to tone 3 if the initial is sonorant; corresponds to tone 4 if the initial is voiced
    陰去 – corresponds to tone 4
    陽去 – corresponds to tone 4
    陰入 – redistributed into the other four tones with no obvious pattern
    陽入 – corresponds to tone 4 if the initial is sonorant; corresponds to tone 2 if the initial is voiced

    Because Mandarin lost the final -p/-t/-k endings, it lost the 入 tone. Words in this category were redistributed into the other 4 Mandarin tones. I mentioned previously that the tone category names were pronounced in their respective tone category. Due to tone mergers in Mandarin this is no longer true. 入 had an sonorant initial in Middle Chinese ([njəp]), and so became tone 4, which typically contains words with 去 tone. 上 had a voiced initial in Middle Chinese ([zjaŋ]) and so also became tone 4 (去聲).

    Both Cantonese and Mandarin lost the voiced/unvoiced initial distinction (the b-/d-/g- used in romanization thus do not refer to voiced consonants but rather aspirated consonants), but because Cantonese still kept the Middle Chinese tone system intact, it is still pretty easy to figure out correspondences:

    Cantonese tone 1 陰平 -> Mandarin tone 1
    Cantonese tone 2 陰上 -> Mandarin tone 3
    Cantonese tone 3 陰去 -> Mandarin tone 4
    Cantonese tone 4 陽平 -> Mandarin tone 2
    Cantonese tone 5 陽上 -> Mandarin tone 3 if Cantonese initial is l-/m-/n-/ng-; Mandarin tone 4 in most other cases; somewhat unpredictable if Cantonese initial is w-/y-
    Cantonese tone 6 陽去 -> Mandarin tone 4
    Cantonese tone 7 上陰入 -> unpredictable
    Cantonese tone 8 下陰入 -> unpredictable
    Cantonese tone 9 陽入 -> Mandarin tone 4 if Cantonese initial is l-/m-/n-/ng-; Mandarin tone 2 in most other cases; somewhat unpredictable if Cantonese initial is w-/y-

    Cantonese initials w-/y- are listed as being a bit unpredictable in a couple cases. This is because Cantonese words with those initials evolved from different sets of Middle Chinese initials, some of which are sonorant, some of which are voiced, so it is not as easy to predict the correspondence.

    Of course, there will always be exceptions because languages do not always evolve consistently, but in general these correspondences are pretty reliable. Here are some examples:

    花 has Cantonese tone 1, so it has Mandarin tone 1
    奶 has Cantonese tone 4, so it has Mandarin tone 2
    餃 has Cantonese tone 2, so it has Mandarin tone 3
    免 has Cantonese tone 5 and starts with m-, so it has Mandarin tone 3
    社 has Cantonese tone 5 and starts with s-, so it has Mandarin tone 4
    世 has Cantonese tone 3, so it has Mandarin tone 4
    辣 has Cantonese tone 9 and starts with l-, so it has Mandarin tone 4
    白 has Cantonese tone 9 and starts with b-, so it has Mandarin tone 2

    Going from Mandarin to Cantonese is harder because of the multiple tone mergers, but for most words:

    Mandarin tone 1 generally corresponds to Cantonese tone 1
    Mandarin tone 2 generally corresponds to Cantonese tone 4
    Mandarin tone 3 generally corresponds to Cantonese tones 2 or 5
    Mandarin tone 4 generally corresponds to Cantonese tones 3 or 6

    This completely ignores words that were 入 tones in Middle Chinese, but because it’s impossible to pick them out based on Mandarin pronunciation alone (since -p/-t/-k were dropped in Mandarin), the tone mapping from Mandarin to Cantonese will never be as reliable as the mapping from Cantonese to Mandarin.

  26. Claw says:

    I wrote:

    the b-/d-/g- used in romanization thus do not refer to voiced consonants but rather aspirated consonants

    Sorry, I meant unaspirated consonants.

  27. Claw says:

    Oops, I spotted another mistake:

    奶 has Cantonese tone 4, so it has Mandarin tone 2

    That should be 埋, not 奶.

  28. Claw says:

    As others have mentioned, Cantonese also features changed tones, which works somewhat like erhua (兒化) in Mandarin. This can throw a monkey wrench into tone correspondences if you don’t keep it in mind.

    A changed tone in Cantonese is usually manifested as a mid-to-high rising tone, which is the same tone contour as Cantonese tone 2. For this reason, such words are usually marked as having tone 2. (There are a few cases where changed tones have a high level tone like tone 1, but we’ll ignore those for now.)

    For example, 糖 meaning ‘sugar’ has Cantonese tone 4, but when it means ‘candy’, it changes to tone 2. Karan mentioned 錢 has Cantonese tone 2, but this actually is an instance of a changed tone, because the citation tone for 錢 is tone 4 (and it does continue to be pronounced as tone 4 in certain words, such as 價錢), so the correspondence to Mandarin tone 2 is expected.

    While the changed tone for 糖 occurs with the semantic change, words like 錢 and 魚 have changed tones in the general case. Because changed tones in Cantonese are like erhua in Mandarin, they are rather unpredictable, so it can be difficult to know what tone you may need to use to figure out the tone correspondence.

  29. Zev Handel says:

    Claw, thanks for the great detailed summary of the tone correspondences for Mandarin, Cantonese, and Middle Chinese.

    I’d like to buchong one minor point: “This completely ignores words that were 入 tones in Middle Chinese, … it’s impossible to pick them out based on Mandarin pronunciation alone (since -p/-t/-k were dropped in Mandarin).”

    There are actually a few classes of ru-tone word that can be identified reliably on the basis of Mandarin pronunciation. The easiest one to remember is that all second-tone words with unaspirated stop or affricate initials are ru tone. Examples of this kind of syllable are bó, jué, jí, zá, zhuó, dé, etc.

  30. Claw says:

    @Zev: Interesting. I never realized this. Thanks! Do you know if this is documented in any particular source? One logical conclusion from this is that all Cantonese 陽平 (tone 4) words that have initial stops or affricates are necessarily aspirated.

  31. Zev Handel says:

    @Claw: I don’t know that it’s specifically mentioned anywhere, but it’s a logical consequence of the devoicing rule for Mandarin, namely that MC voiced obstruent initials devoiced to aspirated in 平 tone and to non-aspirated in 上去入 tones. (Another consequence of this same sound change rule is that there are no second-tone words with nasal endings and unaspirated initial obstruents, i.e. no syllables like dán, jián, zóng, bíng, etc. The one exception, béng, developed from a fusion of búyòng at a much later date.) The devoicing rule for Cantonese is similar to that of Mandarin (the only exception being colloquial 上 tone words like ‘sit’, ’embrace’, etc.), so yes, in general initial obstruents of Cantonese Tone 4 words will be aspirated.

  32. […] to this thread at Sinoglot, it’ll be easier for me to learn Cantonese (if I ever decided to) by comparing it to the […]

  33. Lina says:

    Re: Cantonese books for foreigners who speak Mandarin:

    There is one!

    A Short Cut to Cantonese: An innovative approach for speakers of Mandarin, by Yin-Ping Cream Lee at Chinese University of Hong Kong.

  34. Zrv says:

    When I was studying Cantonese in Hong Kong some years ago, I was using a textbook written for mainlanders, which of course presumed knowledge of Mandarin. It was really helpful. All the vocabulary in each lesson was glossed in terms of Mandarin equivalents, and the grammar explanations focused on differences between the two languages. Of course, it was all written in Chinese. It wasn’t designed for native English speakers who already know Mandarin!

    Needless to say, the study went much quicker than if one were starting from scratch with a Cantonese textbook for non-Chinese language speakers.

  35. […] finally, in a slightly related post from sinoglot: Do you get a discount for the second type of Chinese you’re learning? Links: […]

Leave a Reply to Kellen Parker