No tone sandhi between semantically separated 3rd tones


When one zheng3ti3 (a semantic portion) ends with a 3rd tone and the following zheng3ti3 begins with a 3rd tone, many people have told me they don’t bian4diao4 (tone sandhi) because of the small separation between the two zheng3ti3. In English, we represent things like that with a comma, though they’re not so big on commas in Chinese. I’d be curious if people here would in fact bian4diao4 there.

More importantly, teacher insists that in this sentence:

Na4 jiu4 bu2 shi4 chuan3tong3 you3 guan3 de5 shi4.

the “tong3” in “chuan3tong3” doesn’t undergo tone sandhi because it’s in a separate zheng3ti3 from the following “you3.” Semantically, I don’t perceive it that way at all, and have no instinct to separate “chuan3tong” from “you3” for even a micro-instance. Within that attributive, “you3” is a verb whose subject is “chuan3tong3,” right? Even aside from any explication of the grammar, would you change “chuan3tong3” to “chuan3tong2 you3 guan1” in my example sentence?


I agree with this interpretation. I seem to remember your teacher being Chinese, but that’s a very “Taiwan gouyu” sentence.

I’m assuming the sentence is: 那就不是傳統有管的事。A more proper rendering would be: 那就不是傳統會管的事。

Without context, the meaning of the sentence is a little ambiguous. It could mean either “That’s not something that is traditionally controlled.” or “That’s not something that is controlled (dictated) by tradition.” In either case, the 傳統 would be one semantic unit, and the 有管 (is controlled) would be another.

Definitely not.


Right, teacher is mainland Chinese, but this discussion was only about the tone sandhi. It’s interesting that it seems like Taiwan word choice/order because we pronounce it as:


I’m much better with pinyin than hanzi, and unfortunately that deficiency has reared its ugly head here. As you now see, the “guan1” I mean is the “guan1” from “guan1xi5,” which is 关. My intended meaning of the sentence is “This matter has nothing to do with tradition.” So in that case, it seems to me that “chuan3tong3 you3 guan1 de,” as an attributive for “shi4,” must itself be considered an inseparable zheng3ti3, hence subject to tone sandhi. What do you think?


傳統有關的 can be further divided into 傳統 and 有關的 (related to), so there wouldn’t be sandhi between the two.


OK, but I just don’t understand why it must be subdivided that way because the attributive seems to me to be a conceptual unit, and I wouldn’t add a micro-pause between the “tong3” and the “you3.” So you’d add a micro-pause there, I take it, or else the consecutive third tones are rather uncomfortable.

What about:

Zher4 de chuan3tong3 hen3 duo1.

Does the “tong3” get changed to 2nd tone before the “hen3”?

Or, rather, I should ask: Must the “tong3” get changed to 2nd tone before the “hen3”? I could see doing it both ways if I were to set off the beginning for emphasis or illustration.



No. The sentence naturally divides into 這兒的傳統 and 很多. To carry over the sandhi from the first to the second would sound really weird.


When you say naturally divides, then, you mean that vocally you would always make at least a small, but still distinct separation there?


No, definitely not. It’s a natural division between components of the sentence, a subject and a complement. The sandhi doesn’t entirely depend on the tone of the following character in any string. It’s processed element by element.


I’d say it’s less of a micro-pause and more that the third tone before the break gets clipped a bit.


I get the idea–no tone change across zheng3ti3–but after all this time it’s going to be a hard habit to break. It’s disappointing that no teacher has ever pointed this out before, and there have been plenty of discussion of tone sandhi going way back.

What about sentences of the form, “wo3 hen3 …”. That “wo3” gets its tone changed, right?


Yes. Unless there’s sandhi on the “hen” caused by what follows it.

我很好。(no sandhi)


Ok, then. But as for the difference between “wo3” as a subject and “zher4 de chuan3tong3” as a subject, is it that we perceive “zher4 de chuan3tong3 hen3 duo1” as more of a topic-comment feel, but “wo3 hen3 e4” not so much?


What about:

gen1 ni3 you3 shen2me guan1xi?

Does that “ni3” change to 2nd tone?