A bossible trend

For some reason I had the dictionary open the other day to 乒乓 (from the game: ping pong), that lovely pair of flip-flop characters that I’ve always pronounced as bīngbāng.

So what was my reaction when the ABC Dictionary told me the pronounciation is pīngpāng?

I’m ashamed to admit, that even though I had been quite sure of the pronunciation before that moment, I immediately assumed I’d just heard it wrong all these years, prostrating myself before the Lexicographer, as Ambrose Bierce defined him,

A pestilent fellow who, under the pretense of recording some particular stage in the development of a language, does what he can to arrest its growth, stiffen its flexibility and mechanize its methods. For your lexicographer, having written his dictionary, comes to be considered “as one having authority,” whereas his function is only to make a record, not to give a law. The natural servility of the human understanding having invested him with judicial power, surrenders its right of reason and submits itself to a chronicle as if it were a statue.

Natural servility indeed. But wait! A day later I was gobsmacked out of this ignominious condition when my mother-in-law started talking about bīngbāng qiú (乒乓球), quite clearly and not subject to any “correction” when I asked her to repeat it. So there was my bīngbāng — NOT to be included on my long list of misapprehended Mandarin vocabulary.

Does a third datapoint make a trend, though? Here are the other two:

  1. A couple of months back, there was the Bineapple survey. Although that went only 4 votes out of 62 for pōluó instead of the standard bōluó, that’s at least three folks other than me who vouch for the P as a native pronunciation.
  2. Then a while ago on Beijing Sounds a Beijing taxi driver executed a well-aspirated P in his pronunciation of 比 (bǐ)

Now this. Of course, it’s not as if all three examples show just one or the other, P->B or B->P. Also, (2) above may just be a production error (except that he sort of says it twice).

But maybe there’s something interesting in the whole flippertygibbetness of it. Anyone have other examples? Theories?

28 responses to “A bossible trend”

  1. Katie says:

    Huh. In my (very limited) experience, I’ve only heard pingpang. But oddly, my popup dictionary lists it as ping pang, but onomatopoetically, it’s bing (or ping) and bang. Maybe that’s a subtle indication that both pronunciations are out there? Maybe bingbang is a reading pronunciation that’s infiltrated the language? Yeah, that would be weird. How often are those words used onomatopoetically in writing? And how often do you read about ping pong? I’m theoried out now.

  2. Syz says:

    The more I ask folks about this the more they try to avoid me I think there’s something deeply interesting going on. My daughter actually started off with something bingbang-like, then when I asked her to say it again, said pingpang. So I asked her which it was, and she said it was pingpang, but bingbang if you say it fast. And now she says “the more I say it, the more confused I get.”

    This reminds me of another word: 糊涂, which is variously pronounced húdu or hútu. The latter is all I find in dictionaries, but I hear the former pretty often in Beijing. I’d always just attributed that to some intervocalic voicing rule or unstressed syllable rule that was too complicated for me to figure out. Maybe it is, or maybe it’s connected somehow?

  3. Daan says:

    When I was in Tai2wan1, I heard fluent speakers of Mandarin use Tai2pei3 every now and then when speaking Mandarin, where you really would have expected Tai2bei3. I’m not sure whether this was because of influence from other languages/dialects they spoke, or perhaps even the fact that Taipei is the commonly used romanised version of Tai2bei3, but it could also be related to this topic.

  4. Bruce says:

    A bit surprised to see hútu as the “standard” pronunciation. I’ve heard 糊涂 said many times, and I’ve asked about it. The only time I’ve heard it the second syllable pronounced as tu is when read very slowly (and thus not “naturally”); otherwise, it’s usually a “d” sound.

  5. Duncan says:

    Huh, well all this certainly intrigues (mystifies?) me as I’ve never heard either bing bang or hudu. Could it be a Beijing thing?

  6. Syz says:

    @Daan: Táipěi from a native Mandarin speaker is a riot. I don’t know what to make of that any more than the taxi driver saying pǐ for 比. Sounds like a possibly related phenomenon, anyway

    @Bruce: I still hear a predominance of hudu, but my 86-yr-old grandmother-in-law, for example, lifelong Beijinger, says hutu, so I’m not sure what to make of it all.

    @Duncan: maybe it’s Beijing. On the other hand, maybe no one hears it but me :-/. That was nearly the case with poluo for pineapple

  7. Zrv says:


    I think in Beijing Mandarin there is a tendency for intervocalic voicing of the initial of neutral-tone syllables in words like hútú, especially when they are high-frequency words. I’m referring to true voicing, i.e. in this case to [d], not pinyin d- [t-].

    The 乒乓 situation as you describe it is quite interesting. I’m pretty sure that pāngpāng is the original pronunciation — the aspirates seem to make more sense as onomatopoetics for the sound of the ball. But there are a number of possible factors that could be complicating things. (I’m making this all up, I have no data to rely on.) One is that the characters are deformed versions of 兵 bīng, which is pronounced with an unaspirated initial, so one could imagine a reading-influenced pronunciation, along the lines Katie suggested. A second factor is that 乒乓 almost always (perhaps always?) occurs in the compound 乒乓球, so that the second syllable tends to be short and in the neutral tone. The initial of that second syllable is after a voiced sound, and in such an environment we might expect it to de-aspirate (or even voice). The third factor is a presumed sense of the native speaker, whether from the onomatopoetics or the written characters, that the two syllables should have parallel pronunciations, i.e. either both starting with p- or both starting with b-. It’s not hard to imagine these three factors interacting in ways that lead to a high degree of variation in the speech community.

    Hypothetically: pingpangqiu comes out as pingbangqiu in rapid speech. When speakers slow down, it comes back out as pingpangqiu. But there may be other speakers for whom the pronunication pingbangqiu leads to assimilation, resulting in bingbangqiu. All three pronunciations could be floating around in the speech community, sounding perfectly natural until your persistent questioning forces a speaker to settle on a “correct” pronunciation.

  8. Syz says:

    @Zrv: I’m kind of dubious about the reading pronunciation theory since 乒乓 is such a commonly spoken word. But “intervocalic voicing followed by assimilation of first p” sounds nifty — very plausible. On the other hand, the mystery of Bs turning to Ps remains unsolved…

  9. Zrv says:

    @Syz: Frequently spoken, even highly colloquial words, are not immune to influence from reading pronunciations. Consider the English word “often”, which historically should have no “t” sound in it. Yet a supposedly correct reading pronunciation with [t] is now commonly heard. And “often” is often used!

  10. Syz says:

    @Zrv: good point, good point — my younger brother, no less (10 yrs younger) claims his native pronunciation of “often” includes the T. So that pronunciation might even be taking over!

  11. pot says:

    Another word along the lines of 糊涂 is 馄饨 (hún+tún), usually pronounced húndun. Incidentally, 涂 and 饨 have voiced initials in Middle Chinese and consequently in Wu.

    As for 菠萝, I’ve only heard bōluó. Pō is possibly a reading pronunciation (读半边): 波 (as in 波浪) is pronounced pō by a portion of older speakers.

  12. Katie says:

    @Zrv–that’s some nifty theorizing. I like it. And pot’s hun[d]un example fits right into the post-nasal unstressed syllable onset voicing theory. (Let’s violate all sorts of English voicing assimilation rules and call it PNUSOVT for short.) In fact, until I read his/her comment, I had no idea that this word was “supposed to” have a ‘t’ in it.

    Of course, that post also leads to a less interesting theory, which is that the historical voicing is just sticking around. Maybe not less interesting actually–we still have to come up with a reason why it sticks around in only certain words, and PNUSOVT might be a good place to start.

    This should be testable …

  13. Zrv says:

    @Katie: I like your idea of original voicing sticking around in the case of words like hun[d]un. The devoicing rule for Mandarin applies across the board to the reading pronunciation of individual characters, of course. And it seems to pretty consistently apply to most syllables of the spoken language. But it seems possible that in some bisyllabic colloquial words, where the morphemic identity of the second syllable isn’t clear (or doesn’t exist), and where that syllable may have been unstressed even very early (Song Dynasty?) in northern China, the devoicing rule wouldn’t apply. This may be a case where the modern pronunciation of a handful of words the colloquial language can reveal something about language history that the textual sources can’t ….

    Unfortunately, it may not be possible to decide between the no-devoicing and re-voicing (PNUSOVT or PVUSOVT) hypotheses for lack of historical evidence. In fact, I’d argue that we don’t really have any way of knowing what the second syllable of hun[d]un or hu[d]u ever sounded like prior to the modern era. An artificial reading pronunciation of the character(s) conventionally used to write the second syllable, which is what we get from medieval sources, isn’t telling us much about pronunciation in actual speech.

    To attempt more, I suppose the first step would be to collect all such examples.

    When is 乒乓 first attested? The 中文大詞典 or 汉语大词典 would be a good place to start dating these words. (I’m traveling now so don’t have access to any reference works.)

  14. pot says:

    @Zrv: I understand that character-based medieval rhyme books are not always very helpful on the pronunciation of polysyllabic words, especially when the characters involved have multiple readings, which is quite often. In the case of 餛飩, however, the second character was invented precisely to write the second syllable of this word, and in my opinion the editors of 廣韻 didn’t have much choice here but to record this syllable.

    It may also be possible to extract information from the choices of characters in the literature. For example, if we can establish that the contemporary word 跟頭 (gēntou) evolved from 觔斗 (jīndǒu) attested in 西遊記, this will be sort of an opposite of PNUSOVT.

  15. pot says:

    The logo of Sinoglot in Manchu writes gurun as kurun and gisun as kisun. This is rather incompatible with the expertise shown in Echoes of Manchu. Are these spellings chosen intentionally to go with this post?

  16. Syz says:

    @pot: great 馄饨 example. Like Katie, I had no idea it was “supposed to be” -tun, but I guess if I’d ever tried to type it and come up with 混沌 I would have realized my mistake 😀

    Pronouncing PNUSOVT sounds like it might get you into a fistfight in some slavic country, but I like it. And no doubt there will be revoicing in Beijing when we can finally muster sufficient evidence to decide PNUSOVT vs PVUSOVT.

    Oh and @pot: logos… Alas, as much as I’d like to claim that you’re right and that the g/k thing is an intentional easter egg, in the end it’s just a mistake that is not “incompatible with the expertise shown…” Our group effort in this case resulted in a bit of manchu-smatter. Paweł says he meant to fix it every time he caught sight of it, but he’s not the guy with the logo skills. Anyway, thanks and we’ll get it fixed.

  17. Syz says:

    Also: let’s add bīngqilín 冰淇淋 to the list. Second syllable pretty much voiced as far as I hear it.

  18. pot says:

    @Szy: That word is often pronounced bīngjīlíng and also written as 冰激凌. As in the case of 乒乓, it might be hard to tell which pronunciation is the original.

  19. Zrv says:

    I think in the case of 冰淇淋 ‘ice cream’ it’s pretty clear that the original pronunciation is with qi, not ji. Or to be more precise, that qi is the sound found in the Mandarin pronunciation of the characters first chosen to render a Cantonese compound of the native word for ‘ice’ plus a borrowing of English ‘cream’. The second syllable of the Chinese word would certainly have had an aspirated kh- initial to match the pronunciation of the first consonant of cream. (The third syllable ends in -m in Cantonese.)

    Forms like 冰激凌 can only be explained as changes that occurred in dialects where the connection to English cream was no longer apparent.

  20. Claw says:

    It’s interesting while the 淇淋 in 冰淇淋 is a Cantonese borrowing (pronounced /kʰei˩.lɐm˩/) of the English word cream, Cantonese speakers don’t generally refer to ice cream as 冰淇淋. It’s usually called 雪糕 (/syːt˧.kou˥/) instead. I wonder why that is.

    But anyway, yes, Zrv is right. The /kʰ-/ initial in the Cantonese word generally maps to the Pinyin q- initial in Mandarin (in words where the following vowel is i or ü).

  21. Katie says:

    And from a phonological perspective, forms like 冰激凌 are pretty strong evidence that PNUSOVT is (or was at some point since the word came into Mandarin) for real.

  22. Katie says:

    By which I mean–it is or was an active process, not just maintaining historical voicing but applying to new forms as well.

  23. pot says:

    I doubt that 淇淋 is a Cantonese borrowing. The vowels of 淋 and cream are too far apart. Also, there is already a Cantonese borrowing of the English word cream: 忌廉 gei6 lim1 /kei˨.lim˥/. Note the use of the unaspirated Cantonese initial g /k/ for the aspirated English consonant [kʰ]. This usage is quite common in Cantonese loanwords: 的士 dik1 si2 /tek˥.si˧˥/ from taxi, 布冧 bou3 lam1 /pou˧.lɐm˥/ from plum. So the mere fact that the first syllable of 激凌 has an unaspirated initial does not say much about its originality.

  24. Zrv says:

    @pot, Good points all. So we have to track down the southern dialect through which ice cream was borrowed into Chinese and based on which the characters 淇淋 were selected to write it. No doubt someone has already done this.

  25. Claw says:

    @Zrv/pot: Perhaps it’s from one of the Min dialects? According to the English-Amoy dictionary here, the ‘cream’ in ‘ice cream’ is 淇淋 and pronounced ki-lîm in POJ romanization. Working backwards to IPA, that should be /ki˥.lim˧˥/, the vowel value of which are closer to that of English.

  26. Zrv says:

    @Claw: that does look like a good candidate. The first syllable appears to have an unaspirated initial, which is interesting.

  27. yds says:

    About “冰淇淋”:i think that “汉语外来词” claimed it was absorbed through Shanghainese

  28. pott says:

    @yds: Cream -> /dʑiliɲ/? No comment.

Leave a Reply