对 (duì)

« previous post | next post »

Listening to people around Beijing over the past few days, I've noticed a couple of things about a common Chinese word. The Wiktionary gloss for 对 (dui4) suggests the pattern:

Yes! Correct! I agree!; The word is used often in spoken language. It is common to repeat the word three times when you want to make clear that you understand and agree.

My impression is that a single duì is common, and three-fold repetition is also common, and sometimes even five in a row (grouped 3+2?), but not two or four. (I think I heard a double duì once, but it was more like two phrases "duì, duì".)

I also got the impression of a gender association: triple duì seemed common in women's speech, but I don't think I heard a man say it. Of course, this could be a random consequence of the small sample particular people that I was with — colleagues, their students, taxi drivers, passers-by talking on cell phones, etc.

I might have been tempted to infer something about the metrical organization of vernacular Mandarin, but  是 shi4, which the  Wiktionary glosses (among other things) as "yes (as answer to a question)", seems very often to be doubled "是是", and more rarely triple.

One other interesting thing about 对 dui4 is variation in its pronunciation. The Wikipedia article on pinyin give the pronunciation of ui as [u̯eɪ̯], but the Wikipedia pronunciation

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

and the Mandarintools pronunciation

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

both hint that the high back vowel is the nucleus of the syllable, with the [eɪ] part being an off-glide. I got that impression even more strongly in hearing it used, with nucleus a bit lowered and the off-glide fairly short, something like [doɪ]. One speaker even seemed to omit the high front off-glide entirely in producing a five-fold string like [dododododo].

On the other hand, single repetitions of 对 dui4 sometimes seem to shift the nucleus to the second part of the syllable, something like [dwe:], with the vocal part sounding quite a lot like the French vernacular pronunciation of oui often written as "ouais".

I have the feeling that these variations in the vocalic part of 对 dui4 are the sort of thing where the IPA forces us to categorize what is really a more gradient sort of variation.

Anyhow, I don't have a lot of confidence in the validity of these impressions, except as reinforcement for the obvious idea there are a lot of things to be learned about sociolinguistic variation in Chinese.

 



25 Comments

  1. Victor Mair said,

    June 28, 2014 @ 4:51 pm

    I knew an American woman who was fairly fluent in Mandarin, and whose favorite expression was duì pronounced five or more times.

    Just the other day, I wrote this note to some friends:

    =====

    I often hear Chinese say something like this: hao3 hao3, hao3 hao3 hao3 ("good, good; good, good, good, good") or even more times.

    =====

  2. Eric said,

    June 28, 2014 @ 10:40 pm

    Before I learned any Mandarin, I used to imitate my (male) Chinese friend's phone conversations as "hao… hao… dui… dui… dui dui dui……" (hang up). Later, I found that to be a pretty complete summary of many phone conversations.

  3. Randy Alexander said,

    June 29, 2014 @ 2:13 am

    Would be fun to look at these in the phone corpora. Also exchanges of "bye bye" over and over (each softer and lower than the preceding) at the ends of calls.

  4. Jacob said,

    June 29, 2014 @ 3:46 am

    No phone conversation is complete without multiple 嗯 嗯 嗯, quite possibly a 好嘞 or an O儿了 if you're in the Northeast.

  5. languagehat said,

    June 29, 2014 @ 9:25 am

    This usage is very familiar to me from my days teaching English and linguistics in Taiwan (if anyone reading this attended Tamkang University in 1977-78: I miss you all!), but as I recall it the nucleus was definitely the second part of the syllable. Of course, that was many years ago, and I don't vouch for my memory; I'll be interested to hear from those more recently familiar with Taiwanese Mandarin.

  6. Michael Watts said,

    June 29, 2014 @ 10:06 am

    It was my impression that, for lack of a more fluent phrasing, dui 对 and wei 为 are traditionally viewed as having the same rime and different onsets (d- for dui, zero for wei). Do you have an opinion on what forms the nucleus of wei? How possible is it to draw a distinction between -ui with zero onset and -ei with w- onset?

  7. Michael Watts said,

    June 29, 2014 @ 10:09 am

    One thought that immediately (well, slightly too late) occurs to me is that wei can be realized with a [v] for the [w], but I don't think I've ever heard "dvei" or the like.

  8. Max Pinton said,

    June 29, 2014 @ 1:49 pm

    Pardon the newbie question, but how did “-uei” get truncated to “-ui” in Pinyin? If it had to be two letters, wouldn’t “-ue” be closer to how it sounds? Consider all the poor souls mispronouncing feng shwee.

  9. Michael Watts said,

    June 29, 2014 @ 2:38 pm

    I'd kind of like to know the answer to that question myself, but as a first step I note that Wikipedia indicates all three rimes that have a missing vowel in pinyin omit the same vowel (indeed, are represented the same way) in Wade-Giles. Wade-Giles seems to have made some questionable choices itself, though; I can't imagine any reason for transcribing pinyin shuo as "shuo" but pinyin zhuo as "cho". They're perfect rhymes.

    For reference, the vowel-omitting rimes are -ui (wei), -un (wen), and -iu (you).

  10. Jongseong Park said,

    June 30, 2014 @ 8:07 am

    I would guess that the constraint on the number of repetitions of duì is for purely rhythmical reasons.

    The french interjections oh là là and ah là là can be lengthened by tacking on any even-numbered repetitions of so that the total number of syllables is odd, e.g. oh là là là là là là !

    My impression is that the Spanish interjection ¡Ay, ay, ay! also favours an odd number of repetitions if lengthened.

  11. Jongseong Park said,

    June 30, 2014 @ 8:47 am

    @Max Pinton: I would think a lot of English speakers would have trouble interpreting "ue" as anything close to [weɪ̯]. The last syllable of "Feng shue" would probably be interpreted to sound like "shoe" for many. Then again, at least for consonants many of Hanyu Pinyin's choices are incomprehensible for English speakers to begin with…

    In my personal set of notes (probably from Wikipedia at some point), there is a phonemic analysis that use only /a, ə, i, u, y/ to describe most Standard Mandarin vowels. Here, -ui (wei) is underlying /uəi/, pronounced [weɪ̯]; -un (wen) is /uən/ [wən]; -iu (you) is /iəu/ [joʊ̯]. So Pinyin omits the vowel of a rime when it is underlying /ə/ surrounded by two other sounds (a glide and a consonant or two glides). It is not enough for the /ə/ to be preceded by a glide, because /uə/ [wɔ] is written -uo (wo) in Pinyin (except before labials b, p, m, f, where it is realized simply as [o] and also written as -o in Pinyin).

    So this could be at least a systematic feature of Pinyin, or I might be giving it too much credit.

  12. hanmeng said,

    June 30, 2014 @ 11:13 am

    something like [dwe:], with the vocal part sounding quite a lot like the French vernacular pronunciation of oui often written as "ouais"

    I have the impression women do this more with dui4 than men (or am I listening to too many women?).

  13. KevinM said,

    June 30, 2014 @ 4:33 pm

    @Jongseong Park
    Yes, my experience with Spanish is that an emphatic "no" is always an odd number. "No, no, no" or "No, no, no, no, no". Never "No, no" or "No, no, no, no".

  14. Thomas Bartlett said,

    June 30, 2014 @ 6:34 pm

    The simplified form of the character 对 consists of 又, which originally means ""hand", and 寸, which originally means "thumb". That is, it graphically represents the "thumbs up" gesture of approval, made with clenched fingers and raised thumb. This was not invented by the 20th century savants who designed the PRCs system of officially promulgated simplified characters. Rather, it was used informally in times long before the modern movement for language reforms. Was this graph formed to represent a customary popular gesture among Chinese in times long past?r

  15. maidhc said,

    June 30, 2014 @ 11:05 pm

    I remember hearing someone with a theory that New Yorkers always said things in threes:

    "No! No! No! Not like that! Not like that! Not like that!"

    I can't remember who it was now.

  16. John Swindle said,

    June 30, 2014 @ 11:13 pm

    Volga German in Nebraska in my mother's day also had something that sounded like "Ei jei jei!", in threes like that or occasionally in fives. I don't know how to spell it. Maybe it's normal German? It's an expression of shock rather than an affirmation. The preacher, overcome by his own evocation of the glories of heaven, could only say "Ei jei jei wie schön, ei jei jei wie schön, ei jei jei wie schön."

  17. Simon P said,

    June 30, 2014 @ 11:52 pm

    But what about Cielito lindo? "¡Ay, ay, ay , ay! Canta y no llores." Though the lyrics to that song are pretty old, I believe. Maybe the ood-number-of-ays-thing is more recent? Or maybe it should be parsed "¡Ay! ¡Ay ay ay!"?

  18. Michael Watts said,

    July 1, 2014 @ 9:44 am

    In my personal set of notes (probably from Wikipedia at some point), there is a phonemic analysis that use only /a, ə, i, u, y/ to describe most Standard Mandarin vowels.

    I guess pinyin -e is schwa? Wade-Giles represents the pinyin -ong ending as -ung, which isn't comprehensible to me if it's pronounced the way I've learned it; I always put it down to Wade-Giles being based off a different pronunciation system. But… how does this 5-Mandarin-vowels system explain the difference between shi and she? To my poor english-speaking mind, they're both schwa. Sh- followed by -i should turn into xi, right?

    Look:
    sh- + -a: sha 杀
    sh- + -ə: shi 师 (I'm calling this one the -ə syllable because it is traditionally viewed as having an empty rime, although the same is true of xi)
    sh- + -i: xi 西
    sh- + -u: shu 书
    sh- + -y: xu 需
    sh- + -e: she 舌 (this one isn't the same tone as the other examples, but shi 十 makes a perfect minimal pair).

    The same distinction occurs with the s- onset; si 四 contrasts with se 色. s- isn't supposed to license a following -i, but we can observe that chinese people pronouncing the syllable [si] in foreign words tend to say xi.

  19. Rodger C said,

    July 1, 2014 @ 11:24 am

    @Michael Watts: si and shi are +- pronounced (and in some systems written) sz and shr. The tongue position of the consonant is extended into the vowel. Sh is a retroflex, x is a palatal.

  20. Michael Watts said,

    July 1, 2014 @ 5:37 pm

    I know that (well, I know shi can be pronounced shr; I also know it doesn't have to be — rhotic shi is not standard), but they have a perfect correspondence where the palatal is obligatorily followed by front vowels and the retroflex is followed by non-front vowels. You can also view that as a system where they're allophones of each other, with which appears being governed by the frontness of the vowel (perhaps you'd say "the tongue position of the vowel is extended into the consonant"). Hence the Wade-Giles syllables of, say, chih (pinyin zhi) vs chi (pinyin ji).

    And you haven't addressed my question at all: if we view Mandarin as having five phonemic vowels, [a], [ə], [i], [u], and [y], how can we account for the contrast between shi and she? -a, -i, -u, and -y rimes all either already occur and contrast with those syllables, or are precluded by the phonology from occurring, depending on how you want to view the sh-/x- situation. Either way there's only one phoneme left to occur in a pair of contrasting words.

  21. JS said,

    July 2, 2014 @ 7:02 am

    @Michael Watts
    There are analyses of Mandarin (Hartman 1944; later picked up by Edwin Pulleyblank?) that provide for only two vowel phonemes, so I guess the answer is the analyst can do as s/he pleases — by positing "vowelless" syllables (in the cases of zi ci si zhi chi shi), underlying diphthongs, etc.

    @Rodger C
    The terms "retroflex" (for zh ch sh) and "palatal" (j q x) remain common in introductory textbooks but don't appear in serious analyses I don't think; the consensus seems to be that both are technically inaccurate.

  22. Rodger C said,

    July 2, 2014 @ 11:48 am

    @JS: Thanks. Another sign that I studied a lot of things a long time ago.

  23. julie lee said,

    July 2, 2014 @ 12:54 pm

    Many Mandarin speakers answer the telephone with the word "Wei?" which means "Hello?" on the telephone. "Wei" here is in the second tone. My married name Wei is fourth tone. Years ago, when I was a grad student, I was seeing an American grad student (who didn't know Mandarin) and on his visits to me he'd hear me answer the phone with "Wei (hello)?" and thought I was talking to another student with the surname Wei whom we both knew (I was not talking to Wei but just saying "Hello"). He thought I was talking to Wei and was extremely jealous, until the misunderstanding was cleared up.

  24. Jongseong said,

    July 2, 2014 @ 7:19 pm

    @Michael Watts: But… how does this 5-Mandarin-vowels system explain the difference between shi and she? To my poor english-speaking mind, they're both schwa. Sh- followed by -i should turn into xi, right?

    Well, the system in my notes posits two sibillant phonemes, s /s/ and sh /ʂ/, and explains x [ɕ] as arising from underlying s /s/ or h /x/ as a regular allophone before /i/ (reflecting early Mandarin distinctions that have since merged) rather than from sh /ʂ/. Underlying sh /ʂ/ does not occur before /i/ or /y/.

    I'm not sure exactly how shi or si are supposed to be represented in this system, but I'm guessing they are treated as "vowelless syllables" as JS said:

    sha /ʂa/ [ʂa]
    she /ʂə/ [ʂɰʌ]
    shu /ʂu/ [ʂu]
    shi /ʂ/ [ʂɻ̍]

    sa /sa/ [sa]
    se /sə/ [sɰʌ]
    xi /si, xi/ [ɕi]
    su /su/ [su]
    xu /sy, xy/ [ɕy]
    si /s/ [sɯ]

  25. Jongseong said,

    July 2, 2014 @ 7:32 pm

    Note that I said that the aforementioned five-vowel analysis explains "most" vowels of Standard Mandarin. I think the only cases that this system can't deal with are the rare syllables o, lo, and yo (bo, po, mo, and fo can of course be treated as having -uo rimes with underlying /uə/). So we could add /ɔ/ as a marginal vowel in this system to plug the gap.

RSS feed for comments on this post