The wonder of Cantonese particles

« previous post | next post »

Rosalyn Shih has an entertaining and informative piece called "Let's Go Laaaaaaaa:  And learn Cantonese particles" in LARB China Channel (5/1/18)

Some highlights:

…In Singapore, particles have migrated to English, prompting the Quora thread “Why do Singaporeans say lah at the end of every sentence?”

It seems that the more southern the Chinese-speaker, the more particles he or she might use. Citing various studies from 1924 to 1994, Language Log notes the estimates of Cantonese particles are anywhere from 30 to 206….

…So what’s the purpose of all this variation in Cantonese particles? Why would a language develop dozens of particles, when others seem to get along with just a few?

There is the argument that particles make speech softer, the way that vowel breaking in an Southern drawl might make someone sound more genteel. But this idea doesn’t cut weight in Cantonese. Take, for instance, triad movies, where egregious final particles and lilting intonation make gangsters sound anything from garrulous and informal to threatening and icy cold. (Check out Anthony Wong Chau Sang’s character Fei Gor, one of the most memorable gangsters in Cantonese film, and how his flamboyant endings, sloppy personal hygiene and objectionable morals go hand-in-hand.)….

Perhaps the best way to illustrate how Cantonese sentence-final particles convey different niceties, subtleties, and emotions is to take a single Cantonese sentence and then modify it by tacking on half a dozen or more different particles at the end, as Shih illustrates with this series from a Tumblr post:

keoi5 faan1 zo2 uk1 kei2 佢返咗屋企 ("He went home")

佢返咗屋企 aa3 呀 – (informing the listener) “He went home!”
佢返咗屋企 laa3 啦 – (informing the listener) “He already went home!”
佢返咗屋企 wo3 喎 – “Oh… He already went home though.”
佢返咗屋企 gwaa3 啩 – (with uncertainty) “He went home, I guess”
佢返咗屋企 me1 咩?  – “Oh, he went home?”
佢返咗屋企 lo1 囉 – (with emphasis) “He went home already.”
佢返咗屋企 laa3 maa3 啦嘛 – “He went home already! (so he can’t be here right now)”

I think that sentence-final interjections in English function somewhat similarly to sentence-final particles in Cantonese, though we don't have so many of them, they tend to be interrogatives, and they are not so integrally systematized into the language as a whole:

eh, huh, OK, bazinga, (all) right [N.B. "right" can be used to express an amazing variety of nuances, depending upon how it is intoned and drawn out or clipped off, from reservation, skepticism, and sarcasm to enthusiastic agreement]

But most of our interjections come at the beginning of sentences and are detached from the main utterance (as indicated by the use of an exclamation point):

Ugh! Whew! Uh-oh! Eek! Yahoo! Oh! Yuck!

By no means am I saying that English interjections are equivalent to Sinitic particles, only that there are some overlapping correspondences in the way they function.  A good example of the closeness of the two types of expressions may be seen in Singaporean (probably originally Cantonese) "wah" and English "wow".  Quoting from "New Singaporean and Hong Kong terms in the OED" (5/12/16):

wah — I love this exclamation.  Around 1969 in Seattle, I learned it from Pinky Wu, granddaughter of Wang Ching-wei (1883-1944), head of state of the Reorganized National Government of China based in Nanjing during WWII.  Pinky would use it to express admiration, surprise, amazement, or delight, and it was just wonderful to hear her say it with so many different intonations and nuances.  (Pinky seemed to say "wah" about every tenth word, so tickled was she by almost everything.)  I picked "wah" up from Pinky and still to this day I use it where most Americans would probably say "wow!"  The reason I think "wah" is so effective in expressing the emotions that Pinky used it for is that the syllable does not close at the end; it just lingers on and on:  wahhhhh! — but it can also be brief and terse when called for.  Although I do think that "wow" looks nicer on the page than "wah", the intrinsic nature of "wow" is that, when spoken, it closes off at the end:  wow!  So it's hard to prolong your feeling of astonishment, stupefaction, and so forth.  Wahhhhhhhh!  N.B.:  This is a totally different "wah" from the Dartmouth Indian cheer (also picked up by Virginia):  wa-hoo-wa.  For other aspects of the Dartmouth topolect, see:

"Schlump season" (3/21/15)


"Mud season in Russia: Putin, Rasputin" (3/31/18)

In that original note about "wah" quoted above, I didn't mention that Pinky was a native speaker of Cantonese, but that is relevant, because she used "wah" a lot when she spoke Cantonese, and it transferred over into her Mandarin and her English.

In sum, the genius of Cantonese sentence-final particles is that, through different degrees of prolongation and modulation, they allow for the communication of an enormous range of sentiments and feelings.


"Cantonese sentence-final particles" (3/23/17)

"Cantonese intonation" (4/30/15)

"Mandarin is weirder than Cantonese" (8/14/13)

"Weird languages?" (7/2/13)

[h.t. John Rohsenow]


  1. Chris Button said,

    May 14, 2018 @ 9:18 am

    I wonder if the fact that intonation in many varieties of Mandarin often approximates a language like English more closely than Cantonese (e.g. neutral tone in unstressed syllables, compression of tones outside of focus) accounts for the greater relative use of particles in Cantonese?

  2. Christian Weisgerber said,

    May 14, 2018 @ 10:40 am

    So, modal particles. Compare Wikipedia's entry on German modal particles.

  3. Michael Watts said,

    May 15, 2018 @ 2:53 am

    It doesn't seem right to call 啦 a separate particle from 呀. At least in Mandarin, 啦 "la" is a contracted form that occurs when 啊 "a" is preceded by 了 "le". (And 呀 "ya" is a variation of 啊 that occurs in certain phonological contexts.) The translation given for 啦 in Cantonese ("he already went home") strongly suggests that it is also a contraction with 了 there.

    Calling them different particles is like enumerating the many different forms of "he" in English: he, his, him, he'll, he's (he is), he's (he was), he'd (he had), he'd (he would), ….

    Those aren't different forms. They're the same form participating in contractions.

  4. norman said,

    May 15, 2018 @ 10:28 am

    Michael, that might be the case in Mandarin, but in Cantonese, 啦 laa3 isn't a contraction. The particle is meant to represent the particle indicating "already". While 了 means the same thing, it's not used in spoken Cantonese (though it is used in Standard Written Cantonese/Mandarin as taught in school) to mean "already". In fact, in Cantonese it's pronounced differently from Mandarin as liu5 and at least to my knowledge is used in spoken Cantonese only in phrases like 了解 liu5 gaai5.

  5. zeyao wu said,

    May 15, 2018 @ 12:36 pm

    Hi, Michael, the Cantonese particle system is actually different from the Mandarin particle system. I think Cantonese particles are much richer than that of Mandarin. For example, we have 呀aa3,喇laa3,啵bo3,吖aa1,囖lo1,啫ze6 and so on.

    In Cantonese, 啊/吖 is a really common particle, for example,
    1. How much is it?
    (Mandarin) 這個Zhège多duō少shǎo錢qián?
    (Cantonese) 尼nei4個go3几gei2錢zin2吖aa1?

    2. How are you?
    (Mandarin)過Guò的de怎麽 zěnme樣 yàng?
    (Cantonese) 過gwo3的dik1點dim2吖aa1?

    But we seldom use "了" as a particle. It is common to use 啦 or 喇 (same as 啦, but I think native Cantonese speakers prefer to use this one), for example, 我ngo5知zi3喇laa3 (I have already known it). "啦啊” is also not a common usage.

    我知喇 also has a different meaning from 我知呀.
    我知喇/啦 (ngo5 zi3 laa3): It has two meanings. First, an impatient answer which means "I have already known." Second, be suddenly enlightened.

    我知呀 (ngo5 zi3 aa1): Informing the listener that I know this.

  6. Michael Watts said,

    May 15, 2018 @ 1:08 pm

    In fact, in Cantonese it's pronounced differently from Mandarin as liu5 and at least to my knowledge is used in spoken Cantonese only in phrases like 了解 liu5 gaai5.

    This seems identical with the Mandarin pronunciation 了 liǎo (as in 了解 liǎojiě). I had a vague understanding that 了 le and 了 liao were believed to originate from different historical words, but have become combined into one character in the modern day.

    I am broadly aware of three contexts where 了 is read liao in Mandarin:

    – As part of a lexical entry like 了解.
    – As the dummy resultative suffix, providing 了-support to the construction of ability or inability. (动 dong "move", 动得了 dongdeliao "can move", 动不了 dongbuliao "cannot move")
    – In the song 小苹果, part of the chorus goes 春天又来到了 "spring has come again", and the 了 is pronounced liao. (In contrast, in the earlier lyric 终于长出了果实, it's pronounced le.)

    That third one suggests more or less free variation in independent 了, undercutting the theory that liao and le are two separate morphemes.

  7. Neil Kubler said,

    May 15, 2018 @ 4:00 pm

    Regarding Michael Watts' comment about Cantonese final particles A and LA really being the same particle because in Mandarin LA derives from LE followed by A, there are several problems with that. First and foremost, Cantonese doesn't have LE, so what is Cantonese LA supposed to derive from? (Cantonese employs a verb suffix -JO where Mandarin employs -LE). Second, the functions and meanings of Cantonese A and LA are quite different from Mandarin A, LA, and LE. But the most basic problem is that the Cantonese particle system exists completely independently of the Mandarin particle system and, in terms of the modern Cantonese language and the modern Mandarin language, they have nothing to do with each other. The discussion by other contributors about the character 了 that is the written representation of the first syllable of the word (not phrase) 了解 liugaai is of some interest but, of course, completely unrelated to the question of Cantonese final particles.

  8. Victor Mair said,

    May 15, 2018 @ 6:46 pm

    It is baffling to hear native and fluent speakers of Cantonese being lectured on how their particles work, especially since the particles are the most elusive aspect of the language. Just as Cantonese is famous for its large number of tones, so is it well-known for having a proliferation of particles. It would not be wise to declare that the Cantonese tones are just some variant of Mandarin tones, and it makes even less sense to assert that Cantonese particles are somehow derived from or reflections of Mandarin particles. In comparison to Cantonese, the number of particles in Mandarin is conspicuously small.

    Since the disparity in the number of particles between Cantonese and Mandarin is so great, it simply will not work to attempt to equate the two systems. There are many particles in Cantonese that express subtle nuances that cannot be found in Mandarin.

    The literature on particles in Cantonese and Mandarin is vast. Without looking into the scholarship on particles in that language, it would be rash to declare that Cantonese particles can be equated with Mandarin particles.

    Above all, very interesting work has been done on Cantonese particles that indicates close correspondences with the substrate Austroasiatic languages of the region into which Cantonese moved. Naturally, these would be absent in Mandarin.

    See the latter part of this post:

    "Cantonese sentence-final particles" (3/23/17)

  9. Victor Mair said,

    May 15, 2018 @ 9:34 pm

    More from Zeyao Wu:

    My question here is that for this sentence 佢返咗屋企 aa3 呀 – (informing the listener) “He went home!”, it not only means to inform the listener, but also could be used to express doubt. So this sentence could also mean "Does he go back home?(feel surprised).

  10. Victor Mair said,

    May 15, 2018 @ 11:24 pm

    More from norman:

    I went back to read the the comments- in hindsight I could’ve been more emphatic about 了 liu5 as a word that’s NEVER used as a particle in Cantonese!

  11. Chris Button said,

    May 16, 2018 @ 5:21 am

    @ Christian Weisgerber

    Thanks – I suppose that is why, in spite of their overwhelming similarities, German is sometimes considered slightly less overt than English in terms of its intonation.

  12. Victor Mair said,

    May 16, 2018 @ 7:07 am

    From Bob Bauer:

    I've just read Michael Watts' comment and then the following responses by Norman, Zeyao Wu, and Neil Kubler.

    What Michael Watts has said about Cantonese laa3 indicates he needs to learn more about Cantonese utterance-final particles.

    I completely agree with the responses by Norman, Zeyao Wu, and Neil Kubler which are all fully accurate and correct.

    We may also note there is a different particle laa1. These are NOT contractions.

  13. B.Ma said,

    May 17, 2018 @ 2:32 am

    Michael Watts' comments in the present post annoy me, because the post has nothing to do with Mandarin and was not even making any comparison between the two (although a comparison was made between Cantonese and English).

    In a post about something unique to Danish, would it be appropriate to interject with "well, in German it works like this, so you're all wrong"?

RSS feed for comments on this post