Start with this phonology:
/m n ɲ/
/p pʰ b t tʰ d k kʰ g/
/s x h/
/w l j ɰ/
/u o a ɨ e i/
(s)(O)(R)V(N/(N)s/h/w(N)/l(N)/j(N)) (…or something along those lines)
Some morphological processes:
- Nasal assimilation to a following obstruent
- *h-deletion before an aspirate
Ignore this (or don't); it's scratch material/WIP stuff while I was trying to nail down how I wanted to proceed.
Spoiler:
/m n ȵ/ m n n(e)
/p b t d tɕ dʑ k g/ p b t d t(e) d(e) c g
/s ɕ h/ s s(e) h
/w l ȴ j/ u l l(e) i
/kw/ is written qu.
/o oː a aː e eː i iː/ o ō a ā e ē i ī
All alveolars are obligatorily palatalized before /e(ː)/ (they aren't before /i(ː)/ because *u came in
Sound changes:
Spoiler:
ui īs uiha onsa
ōn uisei uiha honā
uise uisei uā uōns
is uisas uisam uise
ōm uisem holēm uē
Spoiler:
Spoiler:
NOM -Ø -i/-j -pʰa -(n)tʰɨ
GEN -un/-wn -tʰi -kʰɨ -nɨh
ACC -kʰe -kʰi -kʰa -kʰons
DAT -s(ɨ) -satʰ -sɨm -swe
ABL -um/-wm -tʰem -lewm -pʰow
One thing I like is that some noun stems end up having strong and weak forms (for example, in the accusative) because the aspiration triggers the deaspiration of any previous aspirated stops.
One quirk of this language—and I seem to recall reading somewhere that one of Tolkien's languages did this—is that using the definite article with the partitive number gives the connotation of "many" (as opposed to just "some").
kwawn > kwa:n quān 'star' (nom. sg.)
kwawn-kHe > kwa:nxe > kwan:e quanne (acc. sg.)
Let's look at the word for "star", quān-.
quān
quānon
quanne
quanna
quānom
quānei
quanni
quanni
quānsas
quānsem
quamma
quanna
quanna
quānsam
quānlēm
quanni
quānā
quānons
quasse
quāme
Spoiler:
uē
dohon
uēse
uēsa
dohom
dohi
uesei
uesei
uēsas
uesem
doha
ueha
doha
uēsam
uesem
uesa
uēnā
dohons
uēse
dohe
Let's try something that will alternate the aspiration (*tʰom 'plateau' >):
som
smon
tosse
sons
smom
smi
tossei
tossei
tonsas
tonsem
toma
tona
tona
sonsam
somlēm
tossa
somnā
tonons
sosse
tome
Another one (*kʰu 'day, full day/night cycle' >):
sio
iōn
cise
ios
iōm
siō
cisei
cisei
cisas
cisem
ciha
ciha
ciha
uisam
uilēm