I've decided to actually properly work out a proto-lang for Sakata and Katamna (and any other relatives of those two I want to make) to descend from. Primarily because I think I'll do a better job of working out Katamna's morphology if I know what the morphology in the proto-lang looks like. It's named 'Proto-Kataic' after the word for 'people' in the languages of the family (e.g. Sakata
kát'á).
Here's the phonology:
Phonemes:
/ɓ b p tʼ d t tsʼ dz ts kʼ ɡ k/
/m n ŋ/
/z s/
/ɾ ɹ l/
/w j/
/iː uː eː oː aː/
/i u ə a/
Phonotactics:
The largest possible syllable is CCVC.
Onset clusters are limited to the following:
Obstruent + liquid
Obstruent + glide
Nasal + glide
Proto-Kataic is pretty permissive when it comes to medial clusters. The main restriction is that if both consonants in a cluster are obstruents, they must agree in voicing.
Glottalic consonants are not permitted in syllable codas, and only one is permitted per root.
The sequences əj and əw are not permitted in the surface form. They do not appear in roots, and if one occurs as the result of inflection, they simplify to [i] and [u] respectively.
Long vowels are only permitted in the first syllable of a root. If a lexical root is a single syllable, it must have a long vowel.
/l/ was realised as [ɫ] unless followed (or, in a coda, preceded) by a front vowel.
I'll write up the sound changes that result in Sakata and Katamna in a later post, but for now I've put a selection of cognates and some of the sound changes in the PDF below:
https://dl.dropboxusercontent.com/u/412 ... atamna.pdf