I’m not trying to pick on Mandarin, cuz surely every language has free variation in the pronunciation of certain words. Tomato/tomahto, right?

Well, what if we narrow it down to mean specifically free variation within a pretty homogeneous group of speakers. That would probably eliminate tomato/tomahto, at least among American English speakers from the West like myself, where tomahtos are an affectation. But you’d still have words like “either” [ee-ther / I-ther], with genuine free variation. And maybe I’m not thinking objectively. Maybe there are loads of them. Realtor / re-la-tor? Noocleeur / nookyulur?

But is it possible that free variation in Mandarin (or within Beijing Mandarin) is just higher? Or free variation is more common on high frequency words?

On Sinoglot we’ve mentioned 菠萝 (pineapple) as bōluó or pōluó (although the survey showed great preference for the standard bōluó). Then there was a weird incident of 比, generally bǐ but once pronounced as pǐ quite clearly on this Beijing Sounds recording.

A while back, walking up the stairs to the apartment, I asked my mother-in-law about her xīgài (膝盖 = knees). She responded that her qīgài were just fine as long as it’s not more than one flight. Weird enough, but many of us with pronunciation fetishes have learned to discount datapoints from the older generation. Not to say that they’re illegit, just that China has changed so fast that some pronunciation differences are purely generational. What you might think is free variation within a regional dialect often turns out to be generational.

But not this time. Not when you add in a discussion from the other day, with a forty-something Beijinger acquaintance and her parents. It went something like this, unprompted by me, as we were talking about the right component of 膝 in 膝盖:

Woman [musing]: Wait, is it xīgài or qīgài?

Me: Isn’t it xīgài?

Father: Qīgài, it’s definitely qīgài.

Mother: No, it should be xīgài.

There was some back and forth, but resolution was reached when I showed them xīgài on my dictionary.

Knees. Geez. Has anyone else come across this?

20 responses to “Knee”

  1. M says:

    A friend of mine (Taiwanese parents, Mandarin first language, but lived outside the Sinosphere her whole life) tells me she says shīgài.

  2. And here I thought it was 波罗盖儿.

  3. I suspect the “genuine free variation” examples you claim to find in American English are not as “free” as you might imagine. Although your average speaker couldn’t tell you when either pronunciation of ‘the’ (the or thee) is acceptable, for large swaths of people these issues are largely settled. Try to go a whole week only using the [ I-ther] pronunciation.

    As for the “knee” case, it seems like a pretty weak case, somewhat similar to claiming no one can write 打噴嚏 down, because we can all totally remember the number of times we had to write the word “sneeze” down on paper, with a pen in the last 10 years on at least three fingers.

    Anyway, I think any Mandarin speaker would probably see your “knees” and raise you a “the”, as in “the most common word in your language and you don’t have a definitive pronunciation or rules for pronunciation of it yet?” There’s tons of standardization gaps all across English and Mandarin in all areas (pronunciation, grammar, punctuation, etc.) People can’t agree about what constitutes a sentence or when to use a comma in either language.

    As for M’s comment about Taiwan, many people’s X’s in the South are much closer to S’s, with almost no guttural effect. And of course, in Taiwan and other places in the South, “h’s” are either not present or used with “free variation.”

    • … for large swaths of people these issues are largely settled. Try to go a whole week only using the [ I-ther] pronunciation.

      Confused here. I use “either” with the German “ei” pronunciation, but I know plenty of people who use the “ie” pronunciation 100% of the time.

      Who are the people for whom these issues are largely settled? Are you saying this pronunciation has been standardised in the years since I left the Anglosphere?

  4. Syz says:


    similar to claiming no one can write 打噴嚏 down

    Which in turn is parallel to how female lions do the hunting in a pride and similar, in its own way, to the anaerobic fermentation process used for making kimchi.

    Maybe I just don’t get it: what possible connection is there between “ability to write 嚏” and “variation in the pronunciation of 膝盖”?

    I don’t follow the “the” example either, since it generally follows pretty straightforward rules for where it takes its different pronunciations. I don’t think there’s anything like that for 膝盖, like “hey, we pronounce it /q/ in this situation and /x/ in this.” It’s just that, based on my limited evidence, some people in the same age group in the same city at similar educational levels don’t seem to agree on how to pronounce it. How many words like that are there in English? Maybe a lot, but that’s what I was trying to figure out.

  5. Peter Nelson says:

    Yeah, I think the “free variation” examples Syz is looking for are not register-based variation (e.g., careful “correct” speech vs normal speech pronunciations of “the” and “clothes” in English; zh-ch-sh vs z-c-s variation in speakers who know the distinction but don’t usually make it in colloquial speech). I think something more up the alley of “economics” or “either”.

  6. Kellen says:

    I’d argue that in many cases people are not actually aware of the ‘correct’ pronunciation. 粽子 is a good example of this, as you can instantly cause disagreement about the proper pronunciation by bringing it up in Shanghai.

    Out of curiosity, what’s the ‘economics’ example? Reading it as e·CON·omics? Can’t say I’ve heard that from a native speaker. And I’m still not sure what people are saying about ‘either’ since I was never aware one was ‘correct’ and one was not…

  7. pot says:

    My Webster gives two readings of “economics”: eccoNOmics and eecoNOmics.

    I’d argue that qīgài and xīgài are not free variants. I think the majority of us who were raised in Beijing used to say qīgài until the day we were told the “correct” reading in school. After that some switched to xīgài but others sticked to their native reading. The same can be said for 解剖 jiěpāo/jiěpōu and 脊梁 jíniang/jǐliáng. Those who say qīgài are more likely to say jiěpāo and jíniang.

  8. Sorry, I thought I would receive follow-ups automagically, I guess I forgot to check the box.

    With the “the” example I was trying to say that although asking individual speakers what is “correct” or “correct” in some cases may lead to no good result, there are in fact pretty solid rules for large swaths of people on when to use certain pronunciations of these words. If you wrote them down, or made them say some examples they would quickly realize there is in fact a rule there, even though they can’t come up with it on the top of their head. In Mandarin’s defense, they would say there’s this confusion over usage AND variation in pronunciation of the most common word in the language and you want to talk to about practice. We’re talking about practice, man. Practice.

    Anyway, someone who knew or cared about these things, like our trusty informant “pot” would quickly solve the problem. Most people just don’t know or care.

    I never meant to say the examples were identical.

    As for sneezing,

    -Maybe I just don’t get it: what possible connection is there between “ability to write 嚏” and “variation in the pronunciation of 膝盖”?

    Reading my comment over, I indeed can see how out of left field it appears after a liberal helping of editing and assumptions. I think what I was trying to say there was that “Knees” in your case function like writing down 打噴嚏 in Victor Mair’s example of how his students couldn’t write Chinese characters. Seemingly both common and things everyone should know how to write (or have a standard pronunciation for at least in the same age group/region) these things are in fact nothing of the sort. Straw-men, you might say. Look, they can’t even write “sneeze” down! Look, there’s not even agreement over the word “knees!”

    As for how the number of words in American English with ambiguous/unresolved pronunciations, I’m sure there are many. We’ve had a president for 3 years now and I bet if you walked down the street and asked 3 people his first name you’d get 5 different pronunciations, with different stresses as well, both across region and age. (I know this is a name, not a word like knee. See next line.)

    anti (eye) perspirant or anti (E) perspirant

    Chinese has got tones too and is used to varied pronunciations (and tones) so that context and other factors are all important, just like in english so we don’t think twice about anti-perspirant.

    onvelope meet envelope

    -It’s just that, based on my limited evidence, some people in the same age group in the same city at similar educational levels don’t seem to agree on how to pronounce it. How many words like that are there in English? Maybe a lot, but that’s what I was trying to figure out.

    So, basically, I was saying, you could probably ask the same age group people in the same city in the US about “economics/either/anti-perspirant/(the stress on the letters in TV) or even an actually commonly used word like “the” and they would be unable to tell you either the rules for generally accepted correct use or pronunciation or know if both were right or acceptable or whatever. People don’t know a lot about their own languages, or other ones for that matter.

    P.S. I would have loved to go to school that day when they told me that in fact, the way I and all my classmates and my parents were pronouncing the word “knee” was wrong and that we would have to learn the new correct way from that day forward.

    P.S.S. If the female lions do the hunting, who takes care of the kids? The fermentation process? Maybe it could take them to the gym to do some anaerobics.

  9. Syz says:

    @Peter: “economics” — nice example.
    @Pot: Bingo! That reminds me of what might be the best example yet: teachers in the US used to (hell, maybe still) teach students to pronounce a /t/ in “often”. Pattern as far as I can tell follows exactly what you’re saying for qīgài and xīgài. Some kids gave up and added the /t/. Others stuck with, or eventually went back to, the /t/-less version. The odd thing (to me) in the parallel stories is that at least in the US there was some visible justification. “There’s a T and you’re supposed to pronounce it.” However misguided the logic was, at least it was there. I wonder what the logic is with xīgài, though, if qīgài is, as you say, the default pronunciation in Beijing.

    @Transliterationisms: I think I almost get it now. Maybe you should send me some of whatever you’re smoking there. “Envelope” is a good example too. I notice, though, that most of our English examples have had their variation on the vowels. Any with a consonant? Beyond that, bonus point for anaerobics at the gym.

  10. Consonant ambiguity is probably going to be harder to find in English. I don’t think you’re going to find a bunch of English/Engrish fights or people who say greem instead of green or baper instead of paper. So if you’re looking for a bunch of closer parallels to qi/xigai example you’re probably going to be disappointed. But that’s more a function of the genetic differences between the two languages. Using the consonant/vowel model and exploring everything that sticks out to you from an English-speaking background would kind of be like a Mandarin speaker being shocked by all the stress and tonal/stress variation in English words and sentences. The genes are just different. In Mandarin you have this r/l/n continuum which is pretty much a foreign concept in English. Then you have things like the 東北 rhotacization/droppings of zh/sh/etc, which is also a pretty foreign concept for the most part in American English. The enunciation models and existing words in the language, combined with the interference of non-mandarin native languages just opens up galactic-size bags of worms. I mean, in some of the languages(方言) people just throw a N in front of words sometimes, because you know, that’s just what you do.

    I’m going to Al’s to buy some groceries, you want anything?
    Oh, you’re going to Nal’s! Sure, get me some napple pie.

    Given that, we do have some variation that another commenter pointed out with regards to endings of words in English(z vs. s) that I’m sure are far less interesting and sexy than you might think. So basically, consonants are freer in Mandarin in the ways we typically think about them, assuming we aren’t including things like n/ng and s/z distinctions.

    Finally, forgive in advance the medical terminology below. For readers with soft eyes, you may stop here.

    With a word like clitoris, you have a variation in stress with gives you a variation in the consonant, a T to a D, which isn’t reflected in the spelling (student never gets rewritten sdudent). So, on the big map of sounds, t and d are next to each other, just as Xi/Qi are in Chinese. We never think a thing of it. Yes, my example was clitoris.

    Here’s that chart, you should remember it from those days when you first learned pinyin. Albert seems to have mocked it up for some reason I forget.

  11. pot says:

    @Syz: In Middle Chinese 膝 was pronounced /*sit/, and xi is the deduced Putonghua reading. (In this particular case, the tone cannot be deduced from MC.) Conformity to ancient sources (於古有據) has always been a cherished principle of language authorities of Chinese. This is applied even more rigorously in Taiwan. For example, 賜 (MC /*sje/) is exclusively given the popular cì reading on the Mainland, but Taiwan standards insist on the deduced sì reading in less colloquial words.

    As to the parallel with “often”, I think the xī reading of 膝 is a lot more justifiable than pronouncing a /t/ in “often”. “Often” has always been a frequently used word and the loss of the /t/ was a well attested historical fact. 膝, on the other hand, was probably not even part of the colloquial Beijing vocabulary: the word for knee in the good old Beijing dialect was gēlengbàr. In dialects where 膝 has been preserved, the situation is split: for the word knee, one finds 磕膝/tsʰiʔ/頭 in Nanjing, 咳膝(qī)蓋 in Xi’an, but 咳膝(xī)頭兒 in Chengdu, 波膝(xī)蓋 in Xining, and 波膝(xī)蓋兒 in Ürümchi.

  12. Two points for Ürümchi! Or should I say, four points?

  13. Syz says:

    pot, thanks for the details. Have you been doing knee research around the country? I especially like gēlengbàr, which elicited a chuckle from my mother-in-law. She claims to use it still. I don’t recall having heard her say it, but she was a doctor and might ordinarily stick with more formal terminology.

  14. Syz says:

    @translit: That seems about right, roughly hypothesized like this: “a given dialect of any language will have about as much variation as another dialect of another language, but the actual features that vary (vowels, consonants, stress, etc) might be completely different.”

    Although, as a hypothesis, it might need some tightening up before we sic the PhD candidates on it.

    Congrats on the first non-spambot usage (the virgin voyage?) of “clitoris” on Sinoglot

Leave a Reply