Dungan Phonology Questions

[DespikableMi]

Consonants

Cyrillic (IPA)

• р (r)
  
  • Is Dungan the only Central Plains Mandarin dialect with an alveolar trill?
    

  [*]ж (ʐ)

  • Depending on the speaker, the corresponding sound of this consonant can be pronounced either as "ɻ" or "ʐ" in Standard Chinese. However, I'm not sure this consonant be pronounced as " ɻ" in Dungan.
    

Vowels

Cyrillic (IPA)

• ан (æ̃)
  
• ян (iæ̃, jæ̃)
  
• уан (uæ̃)
  
• үан (yæ̃)
  
  • Has Dungan lost the syllable coda "n"?
    

Tones

Standard Gansu-Dungan doesn't distinguish tone 1 and tone 2 only in the final position of phonetic words

• Can anyone provide an example to better explain this? I'm confused.
  

[Michaelyus]

(Disclaimer: no knowledge of Dungan, nor even of any other Zhongyuan Mandarin variety)

I expect that the alveolar trill is only used in lexis of Russian and (Perso-?)Arabic derivation. I don't hear it much in the examples of Dungan that I can find online (which admittedly isn't the kind that would contain much Russian-derived vocabulary, although I expected to hear more Arabic loans). On the other hand, the retroflex approximant seems quite common, although it seems it is spelled (in Roman letters) with -r.

From my limited listening, the -n / -ng distinction does not still sound like it is maintained: the orthography also suggests that it is allophonic with the vowel (as it is in standard Beijing Mandarin too) except in -in, -en (-ƅn) or -yn where it seems the distinction has just disappeared. As to whether -n is more a nasal stop consonant or a nasalisation of the preceding vowel, I would agree that it is more commonly the latter. In isolation and in some environments (notably in front of n-) the stronger consonant version can still be heard.

百度 and Omniglot say 陕西 Dungan has four tones whilst standard 甘肃 Dungan has three. The statement is that the 阴平/阳平 distinction resurfaces in 甘肃 Dungan through different tone sandhi behaviour; this is corroborated within the Wikipedia article by the statement:

Like other Chinese languages, Dungan is tonal. There are two main dialects, one with 4 tones, and the other, considered standard, with 3 tones in the final position in phonetic words and 4 tones in the nonfinal position.

Still needs a citation though.

EDIT: Found something that backs this up: http://www.uusikaupunki.fi/~olsalmi/dungan/Dungan%20as%20Chinese%20Dialect.html. It's surprisingly detailed. It basically says 阴平 is low level (11) while 阳平 is low rising (while Wikipedia has mid rising; 13 ~ 24); the 阴平 merges into 阳平 in isolation, at the end of phonological words, and before another 阴平. (Hmmm... sounds like the behaviour of the low tone of another tonal language...)

That website linked to also states that ƶ- (ж-) is strongly fricated, implying more /ʐ/ than /ɻ/. I don't know how far back in the history of Chinese it was a medial (I thought in Middle Chinese at least it was thought of as an initial, and was probably a palatal or an alveolo-palatal nasal plosive). In any case, ƶ- is the initial of 人、热、饶 in the examples of the paper on aspects, on the same website, so I don't think the distinction between /ʐ/ and /ɻ/ can be too contrastive. Of course, not being very contrastive doesn't imply that the language does not favour one over the other, and I do agree that from my listening, the phoneme is strongly fricated.

I have to say that I am struck by how close to /β/ the equivalent to standard Mandarin /w/ is; the representation v- really was a good choice.