They turned out to be (semi)vowels after all...
This is probably not worth it’s own post, but I feel the need to clarify the status of w y in Núta before going on to the next stage in this thread.
It is probably unnecessary to posit phonemic /w j/ in Núta as all surface glides can be explained according to the vowel sandhi rules outlined before. I’ll repost the table here for ease of reference:
Code: Select all
a e ı u ı2
a a e e ye ye
e ya e e yu wı
ı ya e ı yu wı
u wa we wı u ye
ı2 wa we yu u ı
{e ı} → y / V_V
{ı2 u} → w / V_V
As an example, the prefix nıw- has the allomorph nwı-. To understand why that is so, it is useful to understand that the underlying form of the prefix is |nıı2| from Proto-Núta *niu. Before a vowel, ı2 → w operates, while between two consonants, the sequence |ıı2| yields wı, as can be seen in the vowel sandhi table above.
Example:
nıı2 + ater → nıwá·ter
nıı2 + sehse → nwí·sėhse
We will soon see that analysing the glides as (semi)vowels will also be helpful when we tackle Vowel Dissimilation and Reduplication.
In this interpretation, the phoneme inventory of Núta is as follows:
/n/
/t k ʔ/
/s h/
/r/
/a e i i2 u/
Seven consonant phonemes and five vowels. That feels quite satisfying.