Chinese dictionary with IPA

[噴嚏疋]

IPA is very helpful when learning other European languages, because they have usually only a few extra symbols when compared to English and they're often covered at beginner's phonetics classes at Uni, such as /œ, ø, ɐ, ã, ɨ, y, x, ɲ, ç/.

Mandarin phonemes, on the other hand, have modified IPA symbols seen nowhere else, with tails, hooks and other fantastic chimeric tidbits. And different Sinologists use different sets of symbols, some are officially defunct according to the IPAssociation. They're more perplexing than Pinyin that it's much easier just learn the latter. Maybe to start, you can write in the IPAs yourself next to the Pinyin by using the conversion chart mentioned above. Very soon you'll find yourself not having to do that anymore.

[bancho]

I'm also looking for Chinese to Pinyin to IPA dictionary. I have read somewhere that the old Lonely Planet Mandarin phrasebooks did this before version 6 (which is currently available) when they reverted back to using Pinyin. Can anyone confirm this? I would be happy to pay quite a lot for one...

People have suggested that you can use a conversion chart to convert Pinyin to IPA. But that's against the point, really. Pinyin is not accurate enough to describe Mandarin pronunciation, just like our characters are not accurate enough to describe Spanish or French or English or Italian or German pronunciation. For example to pronounce French, you need to learn quite a few years of French to pronounce it correctly. And you learn by hearing the teacher say a word like "Bonjour" and reading the IPA pronunciation "bɔ̃ˈʒur" next to the written word "Bonjour". After a while you can read the IPA without the teacher and correlate it to written words.

That's what I need - a dictionary or phrasebook which has the IPA form written next to Pinyin. It needs to be written by a native speaker, since many syllables are written the same in Pinyin but are pronounced differently. And the whole dictionary should come from an established institute or at least a very knowledgeable person so that in reflects accurately the pronunciation of the language in a single geographic area, hopefully the Beijing area or maybe a little out of Beijing so that it's more common to what you hear on the streets (I mainly visit southern China...)

PS. I am a HSK level 2 certified Mandarin speaker, so I've got some experience, though there's still a long way to go.

[imron]

Pinyin is not accurate enough to describe Mandarin pronunciation

It seems to do a pretty good job to me.

[Phil]

It needs to be written by a native speaker' date=' since many syllables are written the same in Pinyin but are pronounced differently.[/quote']

I have to say, I agree with Bancho here. I was introduced to Chinese through the medium of Kan Qian's "Colloquial Chinese", which was the text used by my first tutor (and subsequently by the two that followed her), and to some of my peer group and I, there was a noticeable difference between the sound of the initial "r" in "ren" (人) and the initial "r" in "renshi" (認識) : the former was close to, though by no means identical to, a non-rhotic British "r" (untrilled), whilst the latter was better approximated by the "zh" sound of French "je" ("I"). Interestingly, this difference was less clear to our native-speaking Chinese tutors than it was to some (Western : British / Channel Islands) members of the class, suggesting that the two sounds are in fact members of the same continuum in spoken Chinese and never form a minimal pair. In fact, it was only when discussing this question with the son of one of my tutors that the very relevant fact emerged that the two "ren" syllables are transliterations of two quite different hanzi, and therefore (as he so correctly observed), there is no a priori reason why they should sound the same !

So, I echo the plea of the original writer in this thread, with a generalisation : if anyone is aware of a good Chinese-English dictionary that uses IPA rather than pinyin to represent the pronunciation of the Chinese, could they please respond with details. From my own perspective, this does not need to be online : a large-format, hard-bound, printed version would be absolutely ideal, although an online or softbound printed version would suffice (but not small format : at 64, my eyesight is no longer good enough for such things).

** Phil.

[Gharial]

Although not a dictionary, the following seems a pretty good survey, has what would appear to be pretty comprehensive conversion charts, and mentions several dictionaries in its bibliography:

http://www.sino-platonic.org/complete/spp052_chinese_ipa.pdf

[Phil]

Yes, I'm familiar with that chart, and forwarded a link to it some time ago to a fellow member of my class who was also interested in the inconsistent sound of initial "r". But despite the introduction, which casts serious doubt on the usefulness of Pinyin per se, it then (to my mind) falls into the very serious trap of accepting that there is a 1:1 mapping from Pinyin to IPA, which is an assertion that I find hard to accept (based on the evidence of my own ears, and of the ears of former colleagues who were/are professional linguists). What I have yet to locate, and what would be of considerably greater interest, would be a similar (but vastly larger) chart showing the 7000 most common Hanzi (the so-called "Tongyong Hanzi") and their IPA transcription, ideally for the Beijing fangyan of Putonghua, but at least for some well-defined topolect.

[roddy]

I had a quick look to see if I could spot anything on Amazon.cn - no joy. I can't see that you're going to find anything systematic - after all, you wouldn't want to suggest that China's home grown romanization scheme is inadequate, would you? You could try looking for materials either pre-pinyin, or from elsewhere in Greater China.

This looks interesting, but it's not what you want.

[Phil]

I think that "this" is actually the bound version of the Sino-Platonic paper to which Gharial kindly referred me in a previous message. And no, of course I wouldn't dare to suggest that the official Chinese romanization scheme could possibly be anything less than perfect (although some authorities have made reference to "The Great Chinese Romanisation Hoax", admittedly not with reference solely to Pinyin), but that still doesn't prevent me from searching for something even better !

[imron]

the former was close to, though by no means identical to, a non-rhotic British "r" (untrilled), whilst the latter was better approximated by the "zh" sound of French "je" ("I").

This is one of the main reasons the majority of English learners in China pronounce "usually" as "urually" - they are taught that the ʒ in usually is close to the pinyin 'r' and so that's what they use. Unfortunately, although these sounds might appear similar on the surface, or to the ear, they are pronounced in very different ways and it comes off giving people wacky pronunciation. Instead of IPA I would encourage you to find diagrams of tongue and mouth positions for pinyin and just go from there. For what it's worth, I don't think there's really much of a difference between the initial 'r' in 人 and 认识. Are you sure it's not the fact that the finals have different tones and this is affecting the sound?

[yonglin]

Something you may also want to keep in mind is that what noise is consider to be a "distinct" from some other noise depends on the language. From what I understand, initial "l" and "r" are not distinguished in Japanese (and some Japanese cannot even hear the difference, I've learnt). In Swedish, there are at least two different, interchangable ways of saying initial "sch" and "r", respectively (we can hear the difference, though). In Cantonese, most people merge initial n with l. I'm sure there are similar examples from every language. What variety of a sound is used might depend on the accent spoken, the individual speaker, the particular word (i.e., the combination with other noises), etc. However, when people "hear" different varieties, they interpret them as the same thing, because of the structure of the language. Learning what differences "matter" is just an intrinsic part of learning a language.

As for the "r" in "ren", I think there are all kinds of varieties from the "proper" r, to the French-like noise you are talking about, all the way to "y" mostly heard in southern accented Mandarin. When you get all the way to the "y", we usually say the accent is "non-standard". However, if you view this as a continuum, it's very tricky to determine where exactly it got "non-standard".

I would argue that pinyin can represent the pronunciation of mainland standard Chinese extremely accurately. This is because mainland standard Chinese is just a construct, which arguably is defined through a set of dictionaries using pinyin to indicate pronunciation. Indeed, pinyin was introduced for the very purpose of teaching people how to speak Chinese with "standard" pronunciation. If your tutor is a native speaker, pinyin can probably not exactly represent the way she speaks exactly, even if her accent is very "standard".

If you pronounce 人 and 认 the same (with different tones, obviously), I don't think anyone is ever going to call you out on that, even if it happens to be the case that they themselves happen to be pronouncing them slightly differently. As mentioned above, they might not even realize they pronounce them slightly differently.

[Gharial]

Another resource that might be of general interest is the 'PatPho for Chinese database' (which I stumbled upon from just now Googling 'a dictionary of mandarin arranged by ipa international phonetic alphabet transcription'; page 4 of the results had http://www.springerl...45/fulltext.pdf , which [cos the pg 7 Richmond link seems dead] led to me Googling 'xiaowei zhao psychology' to get http://sites.google....te/xiaoweizhao/ , and then clicking on the 'Computer Tools' sublink to get to the 'PatPho for Chinese', which is finally accessed by clicking on the 'To access our database, please click here' > ): http://www.personal....se_patphon.html . It seems to work fine with just entering Pinyin (in the box at 7) without any tone numbers.

Again, PatPho probably doesn't provide "inconsistent enough" (LOL! ) examples of e.g. initial "r" for your purposes, Phil (well, not unless you contrast the PatPho transcriptions with those in that Sino-Platonic pdf, which IIRC all generally differ from the PatPho as far as initial "r" is concerned), but it may be a bit more of a convenient way, for those wanting at least some form of IPA transcription in place of Pinyin, to look up/convert the syllables.

One thing that's always worth doing (obviously!) is browsing bookstores. For example, once in a store on the south side of Huai Hai Lu (about midway along IIRC, i.e. H.H. Zhong Lu) in Shanghai I came across a dictionary of English arranged primarily by entire strings of IPA, so it isn't inconceivable that there could be a dictionary of Mandarin (rather than always just regionalects) that incorporates IPA in a big way.

All that being said however (and turning the question here on its head), would learners of English be better served by dictionaries that used narrower (i.e. more detailed) rather than the usual broad (i.e. more general, "fuzzy") IPA transcriptions? I've never met a Chinese learner of English who felt that the broad transcriptions weren't adequate enough (though they did sometimes express frustration at English spelling/needing to sometimes guess the pronunciation of unknown words, but that's a different kettle of fish!), despite English being surely the trickier language at certainly the segmental level.

[Phil]

My thanks to all for their further comments. Just to pick up on a few points :

Instead of IPA I would encourage you to find diagrams of tongue and mouth positions for pinyin

Yes, I have tried such diagrams, but with little success : I have a nasty suspicion I am a "lazy" speaker of foreign languages, and if I can get a good enough approximation to a sound without having to make a conscious adjustment to my organs of speech, I tend to settle for that, allowing my ear to be the judge of how close I am getting to the target (that is, I prefer to aim at the target sound rather than at the the target tongue position, mouth shape, etc.)

Are you sure it's not the fact that the finals have different tones and this is affecting the sound?

Reasonably certain.

Something you may also want to keep in mind is that what noise is consider to be a "distinct" from some other noise depends on the language. From what I understand, initial "l" and "r" are not distinguished in Japanese (and some Japanese cannot even hear the difference, I've learnt).

Yes, this is a very important point. I long ago discovered that to a native Catalan speaker, "ben" and "ven" cannot be differentiated, but it was only while travelling in Japan with one of my Chinese teachers that I really began to understand the fundamental (and key) truth that some sounds in language A simply do not exist in language B. Japanese, as far as I can tell, does not have a sound that we in the West would think of as being associated with the letter "l", and so a Japanese person attempting to speak a foreign word which does contain an "l" will simply substitute the nearest sound from his/her native language, which in practice is the Japanese variant of "r".

However, when people "hear" different varieties, they interpret them as the same thing, because of the structure of the language. Learning what differences "matter" is just an intrinsic part of learning a language.

Yes, and if I understand the terms correctly, this is also the key difference between phonetics and phonemics.

I would argue that pinyin can represent the pronunciation of mainland standard Chinese extremely accurately. This is because mainland standard Chinese is just a construct, which arguably is defined through a set of dictionaries using pinyin to indicate pronunciation. Indeed, pinyin was introduced for the very purpose of teaching people how to speak Chinese with "standard" pronunciation. If your tutor is a native speaker, pinyin can probably not exactly represent the way she speaks exactly, even if her accent is very "standard".

I think you may well be correct, despite my earlier argument to the contrary. It may indeed be that,in the mindset of whatever was the Chinese equivalent to L'Académie française, the initial "r" in "ren" (人) and the initial "r" in "renshi" (認識) were to be pronounced identically, but of course the recordings to which I was exposed were made by real native speakers, each of whom would have been significantly influenced w.r.t. pronunciation by his/her normal accent. But there is another explanation, which may be worth considering :

If we accept that, to a native Mandarin speaker with no exposure to other languages, the "r" sound of <Br.E> "red" and the "zh" sound of <Fr.Fr> "je" cannot be differentiated, then the Chinese "Académie" would have been perfectly justified in using the same letter "r" to represent the initial of "ren" and the initial of "renshi", even if, to Western ears, these were consistently pronounced differently by the majority of native speakers. And thus it may well be that Pinyin is a perfectly good phonemic description of MSC, even if it is (arguably) very unsatisfactory as a phonetic description.

If you pronounce 人 and 认 the same (with different tones, obviously), I don't think anyone is ever going to call you out on that, even if it happens to be the case that they themselves happen to be pronouncing them slightly differently. As mentioned above, they might not even realize they pronounce them slightly differently.

Agreed. But that still doesn't stop me from wanting to try to sound closer to a native speaker than a slavish following of Pinyin will allow !

Another resource that might be of general interest is the 'PatPho for Chinese database'

Yes, agreed, a very useful resource : thank you !

One thing that's always worth doing (obviously!) is browsing bookstores. For example, once in a store on Huai Hai Lu in Shanghai I came across a dictionary of English arranged primarily by entire strings of IPA, so it isn't inconceivable that there could be a dictionary of Mandarin (rather than always just regionalects) that incorporates IPA in a big way.

Did that, on my first visit to Shanghai, in the company of two of my Chinese tutors (all three live in Shanghai); we failed to find anything, the IPA (when used at all) being used to describe the pronunciation of the English words rather than the Chinese (which makes perfect sense, for books produced and sold in China).

All that being said however (and turning the question here on its head), would learners of English be better served by dictionaries that used narrower (i.e. more detailed) rather than the usual broad (i.e. more general, "fuzzy") IPA transcriptions? I've never met a Chinese learner of English who felt that the broad transcriptions weren't adequate enough (though they did sometimes express frustration at English spelling/needing to sometimes guess the pronunciation of unknown words, but that's a different kettle of fish!), despite English being surely the trickier language at certainly the segmental level.

Well, I cannot answer from the perspective of a native speaker of Chinese seeking to learn English, but as a native speaker of British English seeking to learn to speak Scots Gaelic, I can assure you that the broad transcriptions that one finds in almost all English-Gaelic dictionaries and phrase books are totally inadequate for the task ! It really is a great shame that the IPA isn't taught in primary schools;if it were, we could, once and for all, lose those awful synthetic transliterations that phrase books tend to adopt, and represent the world's languages with the richness and accuracy that they properly deserve.

** Phil.

[yonglin]

Sorry, I'm not a linguist (as obvious from the above). However, it seems to me that these two objectives of yours

Yes, I have tried such diagrams, but with little success : I have a nasty suspicion I am a "lazy" speaker of foreign languages, and if I can get a good enough approximation to a sound without having to make a conscious adjustment to my organs of speech, I tend to settle for that, allowing my ear to be the judge of how close I am getting to the target (that is, I prefer to aim at the target sound rather than at the the target tongue position, mouth shape, etc.)

Agreed. But that still doesn't stop me from wanting to try to sound closer to a native speaker than a slavish following of Pinyin will allow !

are somewhat... inconsistent with each other? :unsure: As imron suggests, acquiring a native accent as a non-native Chinese speaker usually requires you to understand the correct positioning of the tongue in the mouth.

[Gharial]

One thing I wanted to add to the last paragraph of my previous reply (instead of or in addition to the vagueness of the bracketed "fuzzy"), but was struggling to find a word for (beyond the obvious 'allophone'), was I've now realized something like 'allophone-blind '.

That is, there may well be slight differences between how a phoneme is actually realized in one particular context/co-text/linguistic environment as opposed to another, but if the difference is of no real consequence (i.e. doesn't create a "minimal pair"), it tends to get "disregarded" by phoneticians in producing phonemically-serviceable (though perhaps not phonetically-exacting) transcription schemes for end users ("the average learner", lay person etc). (But hey, other posters have said similar).

That's not to say that the selection of one particular phoneme over another to be representative was ever straightforward, but once one has been selected it seems as though the phoneticians concerned (who I guess were as professional as the next) stick with it rather than introducing "unnecessary" or "complicating" distinctions into their scheme(s).

Regarding the tongue position diagrams, they may not work for everyone, but one would certainly need to be comfortable with a fair number of terms and descriptions instead (relating to place and manner of articulation) in order to make much sense of the full range of IPA symbols, so one could hardly be a lazy learner either way, Phil!

[Phil]

Sorry, I'm not a linguist (as obvious from the above). However, it seems to me that these two objectives of yours (lazy linguist not wanting to learn proper tongue position / mouth shape) (wanting to get closer to the sound of a native speaker than is possible by a tunnel-vision use of the pinyin as guide) are somewhat... inconsistent with each other? :unsure: As imron suggests, acquiring a native accent as a non-native Chinese speaker usually requires you to understand the correct positioning of the tongue in the mouth.

But does it ? (This is not meant to sound confrontational -- it is a genuine question). Surely no Chinese child is taught where to place his/her tongue/lips/whatever -- they hear a sound, repeat the sound, the parent/teacher corrects the sound, and in the end, the child ends up with a native speaker's accent without ever once having to worry about where his/her tongue/lips/whatever are. And I know from my own experience that my pronunciation of (say) "ni hao" is now very different to when I first started trying to speak Chinese, and that my mouth now naturally adopts a different shape to that which it would have if I were to say "knee how" in English. But my mouth has found that shape by itself, as does a Chinese child's, and it seems to me that this is the most "natural" way to become a better speaker, rather than the artificial way of forcing one's organs of speech into the shape/position/whatever that a skilled linguist might recommend.

** Phil.

[Gharial]

But my mouth has found that shape by itself, as does a Chinese child's, and it seems to me that this is the most "natural" way to become a better speaker, rather than the artificial way of forcing one's organs of speech into the shape/position/whatever that a skilled linguist might recommend.

But surely that's what narrower transcriptions in IPA (and even broad transcriptions in Pinyin) are ultimately? It's a bit harder now to know quite what you're after or suggesting, Phil. Learning simply by ear? If so, that would likely take too long (for learners to really notice and pick up on things) compared to more explicit and standard courses (in at least Pinyin).

[Phil]

But surely that's what narrower transcriptions in IPA (and even broad transcriptions in Pinyin) are ultimately ?

That is certainly one way of looking at IPA transcriptions, but it is not how I tend to think of them. Rather, I am interested in (for example) whether two different hanzi are normally pronounced with the same initial and/or final sound. Pinyin provides a very broad way of comparing their sounds; IPA is capable of far finer resolution. As I suggested in an earlier message, Pinyin is almost certainly fine for the phonemics of MSC, but if one is interested in learning and investigating how the language is actually used (i.e., spoken), then the narrower transcriptions possible with the IPA surely come into their own.

I return to my own original example : if, within a single set text, a single native speaker consistently pronounces the initial of "人" and the initial of "認" differently, then a Pinyin transliteration cannot represent this, but the IPA could. And if, in everyday Beijing usage, those two initials are typically pronounced differently, then even if this difference is not normally significant to a Beijingren, a foreigner wishing to acquire a good Beijing accent may well seek to emulate the difference. And he (or she) will also wish to know, when encountering further hanzi that are transliterated in Pinyin with an initial "r", whether that initial is usually pronounced like the initial of "人" or the initial of "認". For such people (of whom I am one, and my former Channel Islands colleague another), a dictionary which brought out those subtle differences in pronunciation would be a God-send, for neither of us really feel that a dictionary that uses a Pinyin translation can ever convey the full subtleties of pronunciation of what is undoubtedly a complex and challenging language, any more than it is possible to (say) adequately represent the equally complex sound system of Scots Gaelic using just the letters of the English alphabet.

[imron]

Surely no Chinese child is taught where to place his/her tongue/lips/whatever -- they hear a sound, repeat the sound, the parent/teacher corrects the sound, and in the end, the child ends up with a native speaker's accent without ever once having to worry about where his/her tongue/lips/whatever are.

It's the constant correction from the native speaker that gets the child placing their mouth tongue and lips in the correct position. Usually because when making the correction the parent makes exaggerated mouth movements until the child gets it close to correct. Children are great at mimicking this sort of thing.

This doesn't tend to happen with adult learners because a native speaker will just think "funny foreign accent", and won't spend 5-10 minutes talking to them in a baby voice.

[Gharial]

I've got the book but not the audio of the Kan Qian version of Colloquial Chinese.* I think the argument (from "the son of one of my tutors") that characters that look different should be pronounced differently (i.e. not be homophones) is spurious to say the least, Phil, but the main things I wanted to say are: 1) Imron's pointed out that the tones are different (though you say you are "reasonably certain" that this isn't a factor), to which I'd add that rén is obviously only one syllable, whereas rènshi is obviously two, which surely has something to do with it (that is, the speaker's subconscious anticipation of "what comes next" must be a factor), and 2) Can you really expect a dictionary to give much more than canonical readings for single items? (I mean, you don't find English dictionaries giving the pronunciation of 'hand' as potentially /hæm/ rather than /hænd/ just because the compound 'handbag' can [at least according to some CELTA trainers] sound in actual, rapid speech like /ˈhæmbæɡ/). It would certainly be quite an undertaking for a dictionary to enumerate all the phonetic variation there might be in the linked-up pronunciation of items as found in compounds and phrases rather than as individual items.

*(Just mentioning the following for the benefit of readers who might not be familiar with the CC coursebook): Taking a quick flick through this pretty much Pinyin-only course I see that there are only about 75 characters taught (that is, supplied with stroke-order diagrams) throughout it, plus a few dozen more (though without stroke orders) via a simple letter in the final lesson (and finally, in an appendix and simply for aiding visual recognition: some place names, sign vocab, supermarket section vocab, and currency characters, a few dozen extra items in all), and that the character is supplied for rén (人) in lesson 1, but none for rènshi (认识) in lesson 2.

[Phil]

I think the argument (from "the son of one of my tutors") that characters that look different should be pronounced differently (i.e. not be homophones) is spurious to say the least, Phil

To be fair to Rui, that is not what he said (and nor is it what I reported him as saying) : he said "there is no reason why they should be pronounced the same; they are different characters" , which is very different to "they should be pronounced differently : they are different characters".

but the main things I wanted to say are: 1) Imron's pointed out that the tones are different (though you say you are "reasonably certain" that this isn't a factor), to which I'd add that rén is obviously only one syllable, whereas rènshi is obviously two, which surely has something to do with it (that is, the speaker's subconscious anticipation of "what comes next" must be a factor)

Agreed. If I can work out how, I will post audio clips of the two sections so that others can decide for themselves as to whether or not there is a significant difference in pronunciation, and if so, what might be the underlying reason(s)/cause(s).

Can you really expect a dictionary to give much more than canonical readings for single items? (I mean, you don't find English dictionaries giving the pronunciation of 'hand' as potentially /hæm/ rather than /hænd/ just because the compound 'handbag' can [at least according to some CELTA trainers] sound in actual, rapid speech like /ˈhæmbæɡ/). It would certainly be quite an undertaking for a dictionary to enumerate all the phonetic variation there might be in the linked-up pronunciation of items as found in compounds and phrases rather than as individual items.

I agree. I think this really brings us right back to the crux of the question, which is "Is Pinyin an adequate system for the phonetic transcription of the typical speech of an educated Beijinren" ? If it is, then the debate could stop right here; if it is not, then I continue to believe that there is a significant lacuna in the existent documentation of the language which could best be filled by a Chinese (or Chinese-English) dictionary that uses IPA to represent the sound of the Chinese language.

At which point, I rest my case, because I feel that others must be getting bored by now with this discussion; I will post sound clips once I have finished editing them.

** Phil.