Win Sinoglot's first-ever romanization prize!
What are we going to call it: the Sinoglot Sinitic Specialist Award?
The executive committee is still debating the exact value of the prize, but the contest is simple enough:
- Consider the following two lines of romanization* from a recording of a someone speaking a Sinitic language
- A,aqna.Geqkeq cio Cinsaenonminro,Geqthe raeyeugnin rokeq va?
- Naha la?Zyyau mentie ha la?Menmenkoe ya,Geqlaonshian raeyeugnin rokeq va?Dakae ze mmeqleq.
- On your honor, without peeking at the recording on Phonemica, be the first to name
- The Sinitic language being transcribed
- The type of romanization being used and a bit of its history
*Not guaranteed to be error-free, as Phonemica is crowd-sourced and editable by anyone!
I’m not allowed to guess, am I,
If we come to comments about comments about a Sinoglot post about a Phonemica recording, an outside observer might be forgiven for thinking we eat sleep & breathe Phonemica 😀
Going by the words “va” and “mentie”, I would guess it’s a Wu dialect. I don’t know what romanisation it is, but I would hazard a guess that it’s made by a Chinese person and not by a westerner. How close am I to the mark?
Certainly right on Wu dialect — I assume you’ve looked by now. But about the romanization I know nothing, embarrassingly. I was only hoping to get the answer from our always-erudite Sinoglot readers!
I’m guessing the language is a dialect of Wu:
The romanization shows a three-way contrast in “voicing” of initials (e.g., b-p-ph for voiced/weird-unvoiced-aspirated). Also, the presence of v indicates voicing.
The crazy double-m thing indicates there’s probably syllabic m’s.
Final stops exist, but only as glottal stops (= q).
Uses “va” as a question particle.
Has palatalized nasal initials (= gn; perhaps gnin = 人)
Geq seems plausible as the demonstrative 搿.
Just maybe the initial y’s and r’s represent ɦ
That said, I can’t figure out what dialect it is or the romanization scheme. Most romanization schemes for Wu look pretty similar, but this one differs in some details from all of them that I know (Long-short doesn’t use initial r’s, doesn’t have “ae”; this scheme doesn’t use q’s; this professor’s scheme has special notation for the voiced series rather than for the aspirated series…).
It also doesn’t help that I don’t actually speak any Wu dialect, so this is a highly academic exercise, and I might easily be wandering into incredibly-wrong territory.
Great stuff, Peter. Far more actual analysis and background research than I’d done!
For the record, I’m all in favor of guessing at languages one can’t speak…
Thanks for the encouragement.
I’d add one more question: How good of a system is it?
I can’t really say how good it is since I don’t know what it is, but unless there are sounds in obscure dialects that I’ve never dreamed of Chinese having, if the goal is to get English speakers to pronounce it correctly, it’s going to fail. (I can’t help but read all the q’s as uvular stops, for example, though I suppose that’s really the effect of other romanizations, not English itself.) On the other hand, if the goal is to find some way of consistently spelling the language in the roman alphabet, it might do just fine. Perhaps a bit like Korean romanization in that regard.
Fair enough. I for one would say Korean’s current system is quite acceptable.
As an armchair linguist, I am highly biased towards romanizations that remain unambiguous even after being typed on a standard US keyboard. Maybe like [a-zA-Z0-9,.? ]+ or something. Anyhow, many of the romanizations I’ve actually seen meet that criteria, but then all of them suffer from the problems: 1) Wu has too many damn vowels, and 2) 3-way voicing contrast is hard to represent in an “intuitive” way for English speakers.
It’s interesting how different people represent the voicing. Some are ph p b, others follow hanyu pinyin with p b bh where h represents voicing. I’m pretty sure a-zA-Z is guaranteed to lead to disappointment though, especially if you’re expecting any easy familiarity between European languages. Add tones and vowel quality (e.g. Vietnamese) and things fall apart quickly.
Speaking of too many vowels, take a look at Dônđäc/Jinhui Wu spoken outside of Shanghai. The claim is there are around 20 vowels (checked vs unchecked, nasalised vs not).
1. Another scheme I’ve seen is doubled-voiced letters for voiced, as in p-b-bb.
2. Yeah, the reason [A-Za-z] is nice isn’t because it’s readable or efficient… it’s because people inevitably ignore all the diacritics and special characters they can’t type, and you end up looking at something terribly ambiguous.
3. At least they fit into… uh… 11 series. I guess that’s still terrible.
This is what we used in Shanghainese class when I studied in Shanghai. I had no idea it was a formal type of romanization though.
Do you happen to recall what book you used?
Wow, so it’s in use in teaching. Good to know. I second Kellen and wonder — if there’s no book — if you have any other material on it you could send our direction.
Unfortunately, no. It was back in 2009, and I think the teacher made up all the materials herself; they were those low-quality mimeographs like you had in elementary school. At the end of the semester, I only took my textbooks from my real classes back, and Shanghainese was kind of just an after-school-extra thing rather than anything formal. I’m sorry to disappoint. But I am entirely sure this was it.
Really looking forward to your answer about what (if any) books it’s used with.
So to answer the question of what romanisation this is:
It’s a variaton on a system called 吳語拉丁式注音法. I say a variation because the transcript in question is for a non-urban Shanghainese dialect.
The top row is the transcript above; The bottom row is using the base system:
a aqna geqkeq cio cinsae nonminro geqthe raeyeugnin rokeq va
a aqna geqreq cio cinse nonminrau geqdaq reyoegnin raureq vaq
啊 阿奶 搿个叫 金山农民画 搿哒还有人画个伐