DesEsseintes wrote: ↑08 Sep 2018 14:51
Omzinesý wrote: ↑08 Sep 2018 12:08
DesEsseintes wrote: ↑04 Sep 2018 08:16
A possible gradation system for Limestone?
Code: Select all
p x w
ps xs y
pł xł r
px x(x) w
t s r/y
ts s(s) ry
tł ł r
tsx sx rw
k x w
ks xs y
kł xł r
kx x(x) w
What sound changes in what contexts cause them?
The general idea is to do something vaguely Finnish albeit not as thoroughgoing. Most stems would only show two grades, either 1 and 2, or 2 and 3, and I'd like to see these stem grades operating in valency marking on verbs for instance.
I have a competing idea of gemination occurring when a short vowel is shortened. Like this:
iit → it → itt
I haven’t decided whether to combine the two ideas. Do let me know if you have wild ideas!
Working out at least the original phonological conditions for the gradation is always helpful. It works also for deciding where and how to break the pattern by making the original conditioning unproductive. Especially if you are going to involve gradation in grammaticalised paradigmatic alternations, that's the way I'd proceed.
I still have a fair bit of ironing to do with Kišta, but the basic gradation pattern there follows roughly the Saamic style lengthening gradation. There's rhythmic lengthening of onset consonants in open syllables, governed by the stress pattern, followed by lenition of unlengthened onsets in unstressed syllables, yielding the strong and weak grades. Consonants or clusters falling at the start of syllables with primary or secondary stress are immune to both of these processes, forming a third neutral grade. Quite often the weak and neutral grades are identical.
Later agglutination processes can break the regular pattern. For example, the anaphoric SG3 human pronominal
-ti is a fairly recent development out of a cliticised deictic pronoun and doesn't appear in the expected strong grade. It simply attaches to the inflected 3rd person verb, leaving its original gradation pattern untouched. It does, however, get the regular weak grade form
-đin when combined with the pronominal plural
-n to create the corresponding anaphoric PL3 pronominal.
Code: Select all
nisi-k sleep-SG1 "I slept"
nisi-n sleep-SG2 "you slept"
nissi sleep.3 "(he/she/it/they ...) slept"
nissi-ti sleep-SG3.ANA "he/she slept" (instead of regular *nisi-tti)
nissi-đin sleep-PL3.ANA "they slept"
I'll probably also end up having a fair bit of irregular weak grades in Kišta for word forms that have just recently lost final consonants. If you wish, you could always describe these using the regular gradation rules by invoking a zero coda consonant, but I'm not sure if that's the description I want to go with.
I hope you'll find these musings at least somewhat helpful. The latest thing I found myself thinking concerning the Kišta gradation was what to do with the glides /ʋ j ɦ/ that don't exist as geminates when they fall in the strong grade. I may or may not go with adding an extra subsyllabic pulse to the preceding vowel. It's a neat idea but I'm not sure it plays well with the feel of the language.