Retroflex sounds in Taiwan (and/or Southern China)

[Olle Linge]

To make a long story short, I'd like to hear what people have to say about retroflex sounds in Taiwan and Southern China. I've always thought that they are significantly less prominent and sometimes completely absent in these variations of Standard Chinese compared with e.g. Beijing. I'm talking specifically about zh, ch, sh in case this is not obvious.

I pointed this out in a discussion on Reddit (see the discussion here (I'm Molndrake), but just read on because this is just background) and ended up in a long and not very constructive discussion with a native speaker who basically says that I'm wrong and that I should stop teaching Chinese because he or she thinks that my Chinese is not good enough if I make such a statement. I know I can hear and pronounce the difference between these sounds, that's not what I'm interested in here.

However, I don't discuss this because I want to win (you know what they say about winning on the Internet), but I do it because I'm genuinely curious. Am I wrong? Why? Why not? To make sure I wasn't hallucinating or something, I leafed through San Duanmu's The Phonology of Standard Chinese, which has a section about Taiwanese Standard Chinese (abbreviated TWSC below). This is from page 310:

Most TWSC speakers do not have the retroflex series or the front round vowel or glide.

This is followed by a table written in IPA that would take half an hour to type, so here is a photo:

So, from what I can read (and what i have heard during my stay in Taiwan), the retroflex zh, ch, sh are mostly not present in TWSC, or is at least reduced. I'm not saying this is true for all speakers at all times, but at least Mr. San here seems to think it's true in most cases.

What's going on here? I'm confused and a bit frustrated. I don't see where the problem is, even though perhaps it should be obvious. Can someone help?

[skylee]

There is a relevant thread - http://www.chinese-forums.com/index.php?/topic/36167-retroflex-pronounciation-in-the-south

[renzhe]

Be careful -- he says that most TWSC speakers don't use the retroflex. He didn't say that TWSC does not have retroflexes -- it certainly does, and high-level news anchors and actors use them. There is a difference between the standard and how people speak on the street.

Another issue is that the Taiwanese retroflex is often softer than the Mainland one, but is often present if you listen for it.

I can't quantify this -- a study with some hard numbers would be interesting. But my limited exposure to Taiwanese materials and some Taiwanese friends have taught me that there is quite a range in terms of what is spoken in Taiwan -- from perfect retroflexes to no retroflexes at all.

EDIT: I was going to link to that thread, but skylee beat me to it!

[Olle Linge]

Thank you for the link, Skylee.

Be careful -- he says that most TWSC speakers don't use the retroflex. He didn't say that TWSC does not have retroflexes -- it certainly does, and high-level news anchors and actors use them. There is a difference between the standard and how people speak on the street.

Another issue is that the Taiwanese retroflex is often softer than the Mainland one, but is often present if you listen for it.

Just to make sure, is this meant to be in contrast to what I said? It's basically what I think I said, even though I didn't bring up the difference in context, mostly because I didn't think about in when I posted.

I can't quantify this -- a study with some hard numbers would be interesting. But my limited exposure to Taiwanese materials and some Taiwanese friends have taught me that there is quite a range in terms of what is spoken in Taiwan -- from perfect retroflexes to no retroflexes at all.

Yeah, it would be very interesting. I'll have a look and see if I can find something. Regarding the range, I fully agree with you.

Update: Okay, he's not just guessing. He refers to this:

Duanmu, San; Wakefield, Gregory H.; Hsu, Yiping; Chritsina, G.; Qiu. Shanping. 1998. Taiwanese Putonghua Speech and Transcript Corpus. Linguistic Data Consortium.

However, I can't find this via the university library or freely available. Perhaps my search-fu is weak.

Update 2: This is what we want, but I doubt it's available online since it is a corpus of recorded speech and must be huge. Our best hope would probably de to search for papers or articles that refer to the corpus in some meaningful way. If someone happens to have access to LDC, that would of course make it easier.

[sleepy eyes]

It's been a long time since 1998. And we'd have to check that corpus, the title tells me nothing about its parameters. Taiwan is large and multidialectal.

All in all, renzhe's post was spot on.

[WestTexas]

Just my two cents, and I know this is a radical view, but sometimes I feel that they made a mistake when they created standard Chinese. It just seems to me like it would have been better for them to leave the s/sh distinction out of the standardization of the language. It would reflect the way many people naturally talk better, and it obviously isn't necessary for understanding the language, since Chinese people never seem to have trouble understanding each other when one or both of them doesn't make the distinction.

[rezaf]

since Chinese people never seem to have trouble understanding each other when one or both of them doesn't make the distinction

They do when they are not very familiar with the context. Ironically once a Taiwanese student told me he couldn't understand one of our professors well because he had a Zhejiangnese accent and couldn't make the distinction.

[anonymoose]

Just my two cents, and I know this is a radical view, but sometimes I feel that they made a mistake when they created standard Chinese. It just seems to me like it would have been better for them to leave the s/sh distinction out of the standardization of the language.

The standard language was based on the northern dialect, where most people do distinguish s/sh. (I know some people think otherwise.) Sure, many people from the south do not make the distinction, but for them, standard Mandarin is (or was at the time of standardisation) a second language. What you're essentially saying, is that sounds that second-language learners find difficult should have been left out.

[realmayo]

WestTexas: I think languages balance out to make sure that they're comprehensible. E.g. Southerners don't use the 儿 so they don't say 这儿, but they don't just say 这 either, they say 这里. When southerners speak their local dialects, I'd assume that absence of retroflex sh, zh, ch, does not mean there's any loss of comprehensibility, because there's no reason for a group of people to speak a language amongst themselves which relies for comprehension on a sound they can't produce.

But this is just a guess: I suppose you'd need to ask northerners about when they listen to someone speaking putonghua but who doesn't retroflex the zh, sh and ch: are there words which the northerner gets confused or mixed up while listening? And then go back to southern dialects (most usefully southern dialects close to the standard putonghua) and see if in those dialects these confused words are ever confused by speakers in those dialects.

For instance, 吃 without a retroflex in pronounced "ci", but in some parts of the south it is actually pronounced "qi", meaning there's no danger of non-retroflexers confusing 吃 with 次 or whatever.

[WestTexas]

In my experience northern Chinese speakers do not have any problems understanding people who do not make the s/sh distinction. Many people from Shenyang, for example, do not make the distinction, but they have no trouble going to Harbin or Changchun and being understood by everyone.

I do not have as much experience with a lack of c/ch distinction and z/zh distinction, as the people where I am, and in Shenyang where I lived last year, all seem to make that distinction.

I understand that the development of standard putonghua was based on northern Chinese, and that the distinction exists (for most people) there, but I guess I just feel that in practice the s/sh distinction is not nearly as universal as many would have you believe. In other words, I don't feel like the 'standard' for Chinese put out by the government and by textbooks, in this regard, accurately reflects the way people in much of China actually talk. If many people do not make the s/sh distinction, and it is not necessary to make this distinction to be understood, it seems counterproductive to include it in the standard language definition and the teaching literature. I don't see the purpose of such a distinction.

[xiaocai]

No the standard Chinese does not represent the natural speech of any Chinese person. All of us have to learn it in school. And to be specific standard Mandarin is based on Beijing accent and they do make the s/sh distinction there. Many of us do not make n/ng, n/l, r/l, ü/i and many other distinctions and theoretically and also based on much of my personal experience lacking of these distinctions did not cause too much misunderstanding as long as the topic of the conversation is clearly understood by both parties. Nevertheless the regulatory body chose to preserve them, because of many valid reasons. Just like you may argue that they can get rid of all the cases, moods, tenses, persons in many synthetic languages as they seem counterproductive and people speak languages such as Chinese which do not make those distinctions can also communicate just fine with each other. There are many things I don't see the purposes when I learn a new language, but I think it is one of the things making learning new languages fun. Don't you think so too?

[renzhe]

Just to make sure, is this meant to be in contrast to what I said?

I just felt that it wasn't clear from the material you linked, and might have contributed to the disagreement you had on reddit.

I understand that the development of standard putonghua was based on northern Chinese, and that the distinction exists (for most people) there

It's confusing to think about it as "distinction", they are completely separate phonemes. And they were always present in spoken Mandarin, where Mandarin is spoken as a native dialect. Putonghua was based on actual speech, it was not a completely arbitrary creation.

It sounds a bit like advocating the merger of "th" and "z" because Germans don't make the distinction. All native speakers of English have to relearn, because there are more Germans than English people.

[realmayo]

Some people advocated basing the standard language on something like Wuhan dialect, because although part of the same family as the language spoken further north it also has plenty of southern flavour to it too. In Wuhan words like 是 四 十 all are "si".

So I would say the comparison is more like advocating which way any putative standard British English should prounounce the "a" in "bath" or "last" etc.

[DespikableMi]

ʈʂ = ts, ʑ

ʈʂʰ = ts

ʂ = s

fəŋ = fʊŋ