Achang ŋ-k-Alternation (?)

[LinguList]

There is one interesting pattern in Achang:

347196  laŋ³¹kɔʔ⁵⁵  stone   n   TBL 1824    Achang (Longchuan)  43

vs.

347197  luʔ⁵⁵koʔ⁵⁵  stone   n   TBL 1825    Achang (Xiandao)    43

Remember: Xiandao luʔ⁵⁵ vs. Achang laŋ³¹

Now compare this with our tree-example:

364051  saŋ³¹tseŋ⁵⁵ tree    n   TBL 1824    Achang (Longchuan)  372

vs.

364052  ʂɯk⁵⁵   tree    n   TBL 1825    Achang (Xiandao)    372

tseŋ⁵⁵ is classifier for "tree" in Achang (we need to include all classifiers in the dataset with more concepts, they are interesting and can give valuable hints). The glottal stop in Xiandao is strange, but the other varieties have k as final, maybe even error in the data. Anyway, given that ST languages often have a form like "laŋ" for stone, while we have the "lak"-like form in Burmish, with Achang as an exception, we might assume that you are originally right, and the saŋ for tree in Achang is also genuin. But it is too early to conclude this: we need to look deeper into this pattern.

[LinguList]

Further steps to investigate this obviously invole:

• looking up further sources on Achang and Xiandao
• looking for more potential examples

I am convinced that the "luk/ŋ"-part in "stone" originally means "stone/rock", as it recurs in the word for "rock" in Burmish languages. The etymology for Burmese remains unclear though, and it might be an innovation there, while Achang shows an old ST form, and the other Burmish languages show a Burmish-only innovation (which is also present in most cases of Achang).