Just when you thought the already esoteric nature of this blog couldn’t get any less universal, I now present an idea I had a while back but never got written down. It’s so ill-conceived and unlikely to be useful that I’ve mostly kept it to myself. However a recent conversation with Alan Lai on Twitter brought it back up, so now I’m finally putting pen to paper, or rather fingers to shiny backlit plastic.
Here it is:
Devanāgarī employed in writing Wu. Devanāgarī is the script used to write Sanskrit, Hindi and the subject of this post.
To take a sentence from Tatoeba as an example,
伊啥地方来个？¹
ɦi sa di fã lɛ ɦəˀ
Which in Devanāgarī could be
हि सा दि वं ले ह
Here the vowel issue isn’t really an issue. Long ā gets used to write the vowel in “sa” where short a is used (though not written in Devanāgarī) to denote the /ə/.
Another example:
吾有点小事体想请侬帮忙。²
ŋu ɦiɤ ti ɕiɔ zɿ tʰi ɕiã ʨʰiɲ noŋ pã mãङू हुी ति शुी स थि शिं थ्शिञ नोङ पं मं
There are pros and cons, though mostly cons. One pro is that in Devanāgarī, minimal pairs are actually minimal quartets. There are separate letters for [p] [pʰ] [b] and [bʰ] (प, फ, ब and भ repectively). Same goes for g/k and t/d, all but the aspirated voiced form are used in Wu. In the second example above, ति and थि are [ti] and [tʰi]. Then 地 [di] would be दि, as seen in the first example. In pinyin these three would be “di” and “ti” but then there’s no way to write [di] without resorting to weirdness. Vowels pose a problem in a couple areas, such as /ɔ/ which isn’t covered in Devanāgarī, but that could easily be remedied.
The other obvious drawback is that Devanāgarī isn’t used in China, where IPA and pinyin obviously are. So a modified pinyin is more intuitive, and IPA is just more known.
The other and in my mind more major drawback for me using it in my own notes is that I already know IPA, and often use IPA for Sanskrit in place of the standard Romanisation method. My notes from Sanskrit are a 50/50 mix where I grab from whichever system will make for the more legible version in my shameful handwriting. So ʈ loses out to ṭ but ʂ gets to stay in place of ṣ.
And so ends this thought experiment. Comments are welcome, but I’m not holding my breath.
edit: Note the same could be accomplished with the Gujarati script (ગુજરાતી લિપિ) and in fact could possibly be done better. In part this is because Gujarati includes /ɔ/ (written ઑ) as well as some variance to the vowels form Devanāgarī. Gujarati would also, in my opinion, be a bit quicker to write for being a bit more cursive and lacking the top line. Then again I’ve never really learned the Gujarati script and it’s certainly less common than Devanāgarī so who knows; I could be way off base here.
- – -
1. “Where are you from?”, from here.
2. “I have a few things I’d like your help with”, from here.
Kudos to you for completely bewildering me when I read the title of the post and wondered what in the world was going on. And you actually wrote it in Hindi. Anyway, I’d also experimented with writing Mandarin once or twice in Devnagari just for the novelty factor and because I was really bored. It does convey some of the consonants more “accurately” but vowels are sometimes a problem, especially ü. Amongst consonants, I found that I could distinguish sh and x in Devnagari but not zh and j, or ch and q. And I suppose one would just borrow a rolled r from the script and use it to denote the non-rolled variety. In any case, its use is probably limited to teaching Indian people how to pronounce Chinese.
I don’t think line vs. no line on top is a huge deal. When I write Hindi fast, I just omit the line anyway.
so now could you tell us how many languages and scripts you know?
@Karan ha. yeah I wrote it in Hindi (though keeping “话”), but mostly because I felt bad just transliterating “Shanghainese in Devanagari”, and I figured no one would notice. But I should have known you would.
ü (IPA /y/ eg. 云 [ɦyŋ]) is a problem, as is ö (IPA /ø/, e.g. 盘 [bø]), but the marks for vowels would be easier to deal with than making totally new consonants. Maybe ू could be -u while ु could be ü. At least अा could be -a while अ could be ə. At any rate, I never promised a 1:1 match. Meanwhile the appeal of Gujarati admittedly lies almost entirely in the ease of writing /ɔ/.
But then who knows. Maybe ‘phags pa would be a better fit. I’m petty in the dark when it comes to Indic scripts. Maybe you have a suggestion?
@diszho: I speak English well enough. I’ll leave it at that.
Shanghainese in Hindi would probably be शंघायी, which sounds almost the same as Shanghai itself, hehe.
Yeah, the thing, is, if you’re going to use Devnagari, which cannot represent the sounds of Shanghainese 1:1 on its own (apart from the consonants), then I don’t see any real advantage here. Why not just use some sort of 滬語拼音, just like Cantonese and 粵語拼音?
If you were to use an Indic script, I guess Devnagari would be the best choice, primarily because of popularity, but also probably because the Indic language that are richest in consonants are in the north, and most Indic scripts from the north have 1:1 (or very close) correspondences amongst themselves, so it does not really make a difference which one you use.
There have been multiple stabs taken at a 滬語拼音 but so far I’ve not been inspired to take up their cause. I agree that Devanagari is not really the answer.
Why hasn’t a Shanghaihua Pinyin system not been successful so far? I don’t see how it might be exceptionally challenging and it would be much easier to read than IPA.
I can only speculate. That said, I think the biggest issue is standardisation. We have 侬 or 你 depending on where you are. We have 啥 or 嗲 for “what” depending on where you are. You could do a pinyin for Shanghainese and just base it on the urban center, which has been done for a few books. However then the problem is that each book/writer is using their own pinyin.
There was some discussion here a while back on how in one text you find p, b and bh for [pʰ] [p] and [b], and how people thought that was a bit silly since h typically shows aspiration, not voicing. Another has ph, p and b for the same three sounds, but since b is [p] in Mandarin pinyin, that felt inconsistent.
Personally, I detest pinyin, though accept it in the absence of a better system. That is I’m not opposed to Romanisation, just to doing it rather shittily. I can type IPA about as fast as I can type pinyin, and IPA is standard. So for me there’s just little reason to bother with someone’s non-standard Shanghai pinyin. Though some readers here have said they support it, while others prefer IPA. For a while when I was doing weekly Twitter Shanghai hua ‘lessons’, I tried to do both IPA and pinyin. I’ve actually used a bit of a hybrid in my own notes sometimes, which may make a more useful post than “Devanāgarī for Wu”.
What about using Xu Bing’s New English? Granted, there would still be the same problem as a normal Roman-type alphabet, but I’m sure new symbols can be invented.
http://www.xubing.com/index.php/site/projects/year/1994/square_calligraphy_classroom
The benefit, of course, is that it looks Chinese.
I did forget about the tones, however. But I think with Shanghainese, you could mark the tones in a relatively simple way due to the tonal sandhi.
I’m not sure fake chinese characters would really be culturally appropriate. I like Xu Bing for his 天书 but not much else of his really strikes me. I’d support Hangul for Wu long before I’d consider Xu Bing’s work as a possible influence.
At any rate the need to create anything new is a major issue. The problem with using Zhuyin for Wu, for example, is that there are about 10 letters/characters/glyphs that were historically used but are not currently supported in Unicode. Though I tend to agree that the tone sandhi makes marking the tones not of primary importance. If you did mark them, by the time you put a sentence together, the markings would be of little use.
for the some glyphs which were not in unicode before, I think they’re now present since the last updates, after the problem is more about typing them, for example to type 𠲎 on linux I need to switch to the dedicated Wu IME , but which is far less convenient than mandarin IME as this Wu IME does not support “intelligent” typing
I think IME is the only reason I would like to see a “standardized” romanization for Wu
As antediluvian as I may sound, I still think the best bet for typing Shanghainese is probably with Chinese characters.
@Simon: 𠲎 is a good case of character issues, though the Zhuyin glyps are still absent. They were really only widely used by missionaries 90 years ago so I’m not expecting them to show up in Unicode any time soon. At any rate, intelligent typing would be rather difficult for “Wu” rather than “Shanghainese” for the difference in characters used.
@Karan: For communication between native speakers, I agree. For dictionaries and language learning books, it would probably be inadequate. I know I’ve passed over any book that relies solely on characters, buying only those with either pinyin or IPA. Especially when dealing with non-Shanghainese dialects.
Yeah, I was going to put in parentheses “unless you’re a foreigner trying to learn it”, but omitted it. A combination of some sort of pinyin along with characters is probably the best bet. Characters have the distinct advantage of allowing Mandarin cognates to be easily spotted.
I think that’s true for the most part but in some cases the cognate’s relation is lost when other characters are chosen for phonetic representation. As an example, and forgive the vulgarity, cena is Shanghainese for “fuck you”. It’s written 册那. It’s probably actually from 肏侬 or something similar but not you have two characters in an otherwise meaningless converstaion.
I don’t really know why this is the case. I think it’s probably because the 白话 drifted enough from the reading pronunciation and eventually other characters were reappropriated based on being similar in sound to the then-current 百姓 pronunciations. At any rate, it clouds some of the relationships between Mandarin, Wu, Min, Yue etc.
This has happened in Cantonese as well though to a lesser extent I think than Shanghainese. For example, the Cantonese word for “to come” is 嚟, which is etymologically the same as 来 but because the former is pronounced lèih (like in Shanghainese, in fact) and the latter lòih, people started using the character that matched the sound better. However, amongst Cantonese linguists, this is widely considered a “mistake”, and some people think that 嚟 should never have been used for 来. Another case is where 翻 is used instead of 返 for the meaning “to go back” because it matches the tone more accurately. However, I’m happy to see that more and more people are using the more etymologically correct 返, since tones are not that important to get right in these scenarios anyway as colloquial Cantonese has tonnes of tone changes.
I think the same should be done for Shanghainese as much as possible. People should try to use the correct cognate characters. If the pronunciations have drifted, people will just read the same character with a different pronunciation when reading 百话 as opposed to 书面语. I believe this distinction is already made in some dialects such as 闽南.
In particular for the 册那 case, you have to realize that even 操 in Mandarin is not the “correct” character for the word, so Shanghainese people have no reason to use a character that is neither etymologically correct not phonologically. As for 肏, my input method can’t even produce it, and no Chinese person I know knows that character, so I’m afraid it’s not the best example, since that character is mostly a curiosity for foreigners learning Chinese.
I’m not sure that this will happen with Wu. Those who are still using it to write have a certain sense of pride about the language, and so by using non-cognate characters it further distinguishes the speech as non-Guoyu. However I could see it happening if there were to be a Wu revival a generation from now when the native or potentially native speakers were much more entrenched with the characters as they appear in Mandarin, having lost track of whatever other weird one they were using.
The one that always bugged me was 就嗲 in 武进话 (dialect in the 常州/无锡 dialect group) which is just 做什么. 什么 is 嗲 which is fine because it’s so far off, but to write 做 as 就, when the sound isn’t that different from 做 the way they pronounce it, just seems a bit much. It’s said as something between “jiu dia” and “zou dia”.
At any rate, neither 册 nor 那 are anywhere close to the meaning of what’s meant to be conveyed. But yeah 操 isn’t much better. Just the first example of a major difference that came to mind.