Are native speakers aware of tone sandhi?

Just after I’d been thinking about the tone changes of yī (hanzi: 一) the other day, I happened to get momentarily puzzled over a word I say practically every day. Seeing this:


I asked my friend, a native speaker of Chinese:

“Is it dānyuán or dānyuán?”

“Huh? It’s dānyuán of course. Why would it be ?”

In this case, as I should have known, it’s pronounced because it’s functioning as a number, something like “Unit #1.”

But after further discussion as followup to her “Why would it be ?” comment, it became clear that she wasn’t aware that yī could change tones at all. It took only a few examples to demonstrate the phenomenon, but the phenomenon up until that point had been completely subconscious.

Maybe I shouldn’t be surprised. Native speakers of every language do similar things. Still, sandhi seem pretty salient to me… I’d like to know if this is an isolated case.

20 responses to “Are native speakers aware of tone sandhi?”

  1. I bet I could find a number of English speakers who aren’t, when first approached, really consciously aware that their concept of the letter pair “th” is actually two different sounds (that is, thin vs this). I use this specific sound as an example because I remember having heard people be initially surprised to find the sounds differentiated in Arabic, Albanian etc when first encountering those written languages. If you bring it to their attention, of course they’ll accept it as obvious. I’d reckon the sandhi thing is like this.

  2. Katie says:

    I think most phonology takes place at a subconscious level unless people have their attention called to it. I don’t think the average English speaker knows that the /t/ in ‘stop’ is different from the /t/ in ‘top’, or that the word ‘strength’ (almost?) always has a /k/ in it.

    [RA: “Strength” has (phonetic) [k], but not (phonemic) /k/, but your point is well-taken.]

    On the other hand, since standard Mandarin (including pronunciation, surely?) is taught in schools, it does surprise me a bit that people aren’t aware of it. As you pointed out in your post a few days ago, it even gets spelled sometimes.

  3. Randy Alexander says:

    I asked my kids’ new babysitter to read 七月 and 八月, and she read the numbers with a rising tone. I explained what I was doing, and she said people don’t pay attention to how they say things; they just say them.

  4. Phillip says:

    I wonder if the 3rd tone tone sandhi is more widely recognized as being a rule in the language by native speakers.

  5. Nicki says:

    I was using skritter during a class break a few days ago with one of my students looking over my shoulder. I went to mark the tones on 一定 (I think that was the one…) and marked it yídìng. She said I was wrong, then seemed really surprised when Skritter said I was right. I explained about the tone change, and after thinking it over and saying it under her breath a few times, she agreed. So I’d say no, they don’t seem to be aware of it until it is pointed out.

  6. F says:

    I’d have to agree that it’s mostly subconscious – recently had a conversation about the third tone sandhi, and the person tried to deny there was any change in tone for a while. He had no idea.

    I also have a question about that 3rd tone change – what happens when there’s more than three 3rd tones in a row? What if there’s 4 or even 5 3rd tones in a row? I remember one teacher saying something about some turning into a 4th tone, which sounded right at the time but has since been contradicted.

  7. Julen says:

    I like Kellen’s example with the th sounds, it is so true.

    I would add that students of foreign languages are generally more aware of this things than natives, because when you learn your own native language you don’t need to think of the rules.

    The “th” example is very obvious, but I am sure that even for us language conscious users, it would be quite easy to point out some quirk of our native language that we hadn’t thought of before, and that we wouldn’t believe until we stopped to pronounce it a few times…

    For example, I only realized the particularity of the Spanish “d” sound when I was in contact with non-native speakers and realized how different they pronounce that letter. Up to then I had always imagined that soft “d” of ours was just the normal way to say “d”.

  8. dlszho says:

    Yeah, native speakers are not aware of a lot of things–which is why they are native speakers. Imagine speaking English and wondering about which “th” needs to be used every time!

    I myself didn’t even realize Cantonese had tones until I was twelve, and didn’t know how many there were until I was sixteen or so. I just didn’t process the different tones in words as tones–just as completely different words.

  9. Katie says:

    @RA–Thanks for the correction– square brackets is what I meant :)

  10. Bruce says:

    It doesn’t surprise me in the slightest that most native Chinese speakers aren’t aware of tone sandhi. I’ve been speaking the language about 30 years, and there are many things that I do naturally that I have found I may need to explain “rationally” to foreign students of the language — which isn’t easy. To explain it, I first have to become aware of it “again,” as I must have had to do back in the 1970s!

    But more importantly, if you REALLY want to know how something is pronounced, how you ask it is vital. If you isolate a phrase or syllable, and ask someone how s/he pronounces it, the answer risks being untypical. “The” in “Give me the apple” may often be pronounced “thuh” in informal, rapid speech, but the odds are good if you point at the words “the apple” and ask a native speaker how to pronounce, you may get “thee apple.”

    Believe it or not, pronunciations can vary depending on who asks about it. Being a foreigner DOES sometimes alter the equation. An example of this is something that happened fairly frequently when I first took my snooty Beijing accent to Taipei.

    I remember asking a Taiwanese to pronounce the sentence I had written down: 你住几层楼?His answer, in pinyin: Ni zu ji cheng lou? I discovered gradually that my overly strong “correct” Beijing accent (couldn’t help it, all 3 of my teachers in the States were Beijingers!) made my Taiwanese interlocutors nervous, and their reaction often involved comical “hypercorrection,” in which “sh” “ch” and “zh” sounds appeared where a simple “s” or “c” or “z” was actually correct.

  11. Karan says:

    I explained the búshì tone change rule to a native once; definitely wasn’t surprised that she didn’t know about it. Heck, I found out that the words for ‘tomorrow’ and ‘yesterday’ are the same in Hindi only when I was 18 and came to the States. Somebody who was learning Hindi told me. I just never noticed… and every Indian person I tell this to is always shocked. 😛

  12. Daan says:

    I was using a phone that didn’t have a decent IME the other day and I needed to text a Taiwanese friend, so I just used Hànyǔ Pīnyīn instead, figuring she would still be able to understand with some effort. When I saw her that evening, she told me proudly she could read Pīnyīn very well (most Taiwanese people can’t) and that I’d made a mistake. She pointed out I should have written “Ke3yi3 ba” instead of “Ke2yi ba”.

    Of course, the rules of Hànyǔ Pīnyīn say tone sandhi should not be indicated, but that was not her point. She insisted the 可 in 可以 was pronounced as and the 以 as . When I then said ke3yi3 without tone sandhi, she shot me a strange look and said: no, we say it like this: ke2yi! When I pointed out there were no third tones in that entire word anywhere, she was confused to no end. Fortunately our burgers then arrived, and we had a nice dinner without any further discussion of tone sandhi issues more likely to confuse her than to contribute to the success of that particular night out 😉

  13. Helen says:

    No, that was not alone.

    I asked a native English speaker the other day about the use of MARRY, she did not know ‘the rule’ we foreigners have long been taught in school that MARRY SOMEONE should always be the case.

  14. Daan says:

    Sorry, Helen, but I’m not exactly sure what you mean. Do you mean the verb “marry” should always take a person as its object, i.e. the sentence “I married in 1982” would be wrong?

  15. arfo says:

    really nice article and discussion.
    I think the final verdict would be that it may be important to learn about the sandhi in the very beginning but as soon as you really start talking freely, you just have to ‘feel’ the language, rather than thinking about tone changes all the time.

    but for the stories: the funniest thing happened was with my first (student) chinese teacher. she prepared a lesson and taught me about the tone changes. after some weeks, I had a question about it and told her about the change in 不是。 she had completely forgotten about it and denied ever teaching it to me, she even said there was no change at all! :)

    //off topic: anyone knows something about the new HSK test? I gonna attend HSK L4 test in 2 weeks and don’t have a clue which characters are required..

  16. Syz says:

    @Karan: So you’re a native speaker of Hindi? Tomorrow vs yesterday is now way off topic but we can scoff at the weak will of the moderator…

    How does tomorrow/yesterday work then? By sheer coincidence I just came across another linguistic musing on tomorrow / yesterday similarities. Maybe you can go over there and comment.

  17. Kellen Parker says:

    Just had this today. Asked a question about sandhi to a Wu speaker and since the conversation was in English I had to define sandhi. Did so with a Mandarin example, which totally hijacked the conversation. My speaker never realised there was a system to all this sandhi stuff.

  18. Syz says:

    @Bruce: the hypercorrection story is lovely! I remember getting made fun of (gently) when I was talking to some non-northern Chinese in a California mall a while back. I got the feeling my erhuayin was seen as an affectation, but I honestly couldn’t get rid of it if I tried. I haven’t noticed hypercorrection yet, but I’ll keep a look out.

  19. Daan says:

    When speaking standard mainland Mandarin in Taiwan, that is to say with érhuàyīn and retroflexes, you get that feeling all the time. People will sometimes even tell comment on the fact you’re speaking “Běijīnghuàr”. Interesting enough, when you then tell them smilingly that huà ‘speech’ remains huà even in Běijīng, and that it’s only huà ‘painting’ that will be pronounced as huàr, they’re often surprised to learn there are actually rules governing érhuàyīn. Most native speakers of Taiwanese Mandarin I’ve met seem to think it’s a case of throwing in some R’s wherever you like.

  20. Syz says:

    @Daan: I guess the beijinghuar thing is sort of a hypercorrection too, come to think of it.

Leave a Reply