Possessive suffixes and case endings

I realized last night that I never got around to figuring out how case endings and possessive suffixes interact, or even which order they go in. First of all, for clarity, the possessive formatives are:

1sg -ni
2sg -si
1pl -nne
2pl -sse
3prox *-o
3obv *-e

With nouns of the original four declension classes, these go as follows:

healmani, healmasi, healmanne, healmasse, healmō, healmē
laureni, lauresi, laurenne, lauresse, lauru, laurie
cuoroni, cuorosi, cuoronne, cuorosse, cuoruo, cuorō
talëni, talësi, talënne, talësse, talu, tali or taini, taisi, tainne, taisse

So far so good -- this is what we've had since 2005 or so. But what happens when the possessed noun is also, say, ergative?

kelve-s rāva-Ø cier-t-o-s
dog-ERG lion-ABS eat-PF-3-OBV
"The dog ate a lion"

Is fine; now what about "My dog..."? Should it be kelvesni or kelvenes or what?

kelve-ne-s rāva-Ø cier-t-o-s
dog-1SG-ERG lion-ABS eat-PF-3-OBV
"My dog ate a lion"

I think that's pretty much the way it has to go, because kelvesni is totally aesthetically unappealing, and the plural suffixes would be ridiculous: kelvesënne, etc. And I like the fact that the way I have it up there sort of resembles Turkish rather than Finnish for once.

The question, then, is what happens to the form of the possessives when case endings are added to them. -ni and -si seem pretty straightforward as -e stems; -nne and -sse must have a proto-coda for the vowel to have stayed mid, so maybe *-tnet and *-tset or something? But then, do these decline as -t stems, as per the preceding post? That would give us e.g. perhennennes, which I think is obviously inadmissible.

I think I'm going to call upon paradigmatic levelling to help me out here, and just call them stems as well, that happen to have an anomalous form in the absolutive singular for historical reasons. Let's go ahead and map out what we've got so far for all cases and numbers:

eāsi "house" with 1st- & 2nd-Person Possessive Suffixes

A eāseni, eāsesi, eāsenne, eāsesse
E eāsenes, eāsetes, eāsennes, eāsesses
I eāsenen, eāseten, eāsennen, eāsessen
P eāsenea, eāsetea, eāsennea, eāsessea
V eāsene, eāsete, eāsenne, eāsesse

eāsie "houses" with 1st- & 2nd-Person Possessive Suffixes

A eāsenie, eāsesie, eāsennie, eāsessie
E eāsenīs, eāsesīs, eāsennīs, eāsessīs
I eāsenīn, eāsesīn, eāsennīn, eāsessīn
P eāsenia, eāsesia, eāsennia, eāsessia
V eāsenē, eāsetē, eāsennē, eāsessē

Issue 1 is that I'm not at all sure I like the fact that singular and plural with possessive suffixes tends to be indicated solely by consonant length. I need to think about this; maybe I could have the plural fall into a different declension class or something? Anyway, the real problems are presented by the third-person suffixes, as we're bound to have a surfeit of vowels. I guess we should just take these one by one, tedious though it may be.

ceivasō & ceivasē "his bedroom"

A ceivasa-o > ceivasō; ceivasa + e > ceivasē
E ceivasa-o-nës > ceivasōs; ceivasa-e-nës > ceivasēs
I ceivasa-o-hën > ceivasōn; ceivasa-o-hën > ceivasēn
P ceivasa-o-da > ceivasōa > ceivasoā; ceivasa-e-da > ceivasēa > ceivaseā
V ceivasa-o-t > ceivasō; ceivasa-e-t > ceivasē

ceivasōi & ceivasēi "his bedrooms"

A ceivasa-o-i > ceivasōi; ceivasa-e-i > ceivasēi/ceivasaje
E ceivasa-o-i-nis > ceivasōis; ceivasa-e-i-nis > ceivasēis/ceivasāis
I ceivasa-o-i-hin > ceivasōin; ceivasa-e-i-hin > ceivasēin / ceivasāin
P ceivasa-o-i-da > ceivasōja; ceivasa-e-i-da > ceivasēja
V ceivasa-o-i-t > ceivasōi; ceivasa-e-i-t > ceivasēi

I need to leave the other three/five classes for later, as (1) I'm developing a pounding headache, and (2) as always I feel like these sound changes are more ad-hoc than really rigorously defined and it makes me feel like a hack. I really need to make that timeline and figure out exactly what happened when.

Hidden consonant stems

I was just realizing while lying in bed this morning that, since /t/ disappeared word-finally, there's bound to be a whole class of nouns that appear vowel-final (and might even end with /-e/ or /-o/ which otherwise don't exist in the absolutive because of the final raising rule) but decline differently in some cases than the genuine vowel stems. For example:

Stems in -t, Singular

A perhet-Ø > perhe
E perhet-nës > perhennes > perhes* (?)
G perhet-v > (perhetu)
I perhet-hën > perheten > perhen* (?)
S perhet-še > perheitsi > perheis
P perhet-da > perhetta
V perhet-t > perhe

* We really need to decide about that paradigmatic leveling issue ASAP.

Stems in -t, Plural

There's a big question here as to what happens to the /-t-/. Followed as it is by /i/, it ought to become an /s/ in every case; but it's also not inconceivable that it could have dropped out, leaving a bunch of vowels to resolve themselves into something pronounceable. Part of this is going to involve figuring out the chronology of the changes.

A perhet-i-Ø > perhesi > perhie (?)
E perhet-i-nis > perhesis > perhies (?)
G perhet-i-u > (perhesiu > perheju (?))
I perhet-i-hin > perhesin > perhien (?)
S perhet-i-še > perhesis > perhies/perheis (?)
P perhet-i-da > perhesia > perheja (?)
V perhet-i-t > perhesi > perhie (?)

I'm interested to see what happens when other vowels precede the /-t/. For example, from original rahat, velit, huonot, olut we'd have:

raha, rahas, rahan, rahais, rahatta, raha; rahai, rahais, rahain, rahaja, rahai
veli, velis, velin, velis, velitta, veli; velī, velīs, velīn, velija, velī
huono, huonos, huonon, huonois, huonotta, huono; huonoi, huonois, huonoin, huonoja, huonoi
olu, olus, olun, oluis, olutta, olu; olui, oluis, oluin, oluja, olui.

So, then, the cases in which we see a departure from vowel-final stems with the same ultimate vowel would be absolutive singular (-et and -ot stems), similative, partitive singular, ergative and instrumental plural (all except -ut stems), partitive plural (-et stems) and vocative plural (all except -it stems).

I have this feeling that maybe the -s- only rather recently dropped out, recently enough for those forms to be used in the modern language to evoke an old-fashioned or very formal feel.

Next question: What about verbs? Should they also exist with all possible vocalic stems? And what happens with /-t/ stems?

Declension of -i and -u stems

I've been realizing for some time that Seadi as originally designed has an artificially small number of declension classes. Back in 2005 I figured out consonant-stem nominals, and I knew that I was going to have to open up -i and -u eventually; apparently "eventually" meant about three years later, and here's my first draft as of this morning's BART ride.

Stems in -i, Singular

A turr-i-Ø > turri
E turr-i-nis > turris
G turr-i-v > (turriu)
I turr-i-hin > turrin
S turr-i-še > turrie
P turr-i-da > turria
V turr-i-t > turri

Stems in -i, Plural

A turr-i-i-Ø > turrī
E turr-i-i-nis > turrīs/turries?
G turr-i-i-v > (turriju)
I turr-i-i-hin > turrīn/turrien?
S turr-i-i-še > turrīs
P turr-i-i-da > turrija
V turr-i-i-t > turrī

Stems in -u, Singular

A karh-u-Ø > karhu
E karh-u-nus > karhus
G karh-u-v > (karhū)
I karh-u-hun > karhun
S karh-u-še > karhui
P karh-u-da > karhua
V karh-u-t > karhu

Stems in -u, Plural

A karh-u-i-Ø > karhui
E karh-u-i-nis > karhūs/karhīs
G karh-u-i-v > (karhuju)
I karh-u-i-hin > karhūn/karhīn
S karh-u-i-še > karhuis
P karh-u-i-da > karhuja
V karh-u-i-t > karhū

Now for the interesting part: I need to decide how paradigmatic leveling is going to affect not only these paradigms, but nominal inflection in general. Given that Seadi is inflectional and not agglutinative, there really ought to be a lot less consistency in the declensional endings.

What in god's name is going on with my phonology?

Can we finally motivate our diphthongs? We're doing short-vowel "breaking" à la Sámi:

a > ea as in ase > easi, kata > keata, but katu > katu, harma > harma
i > ie always, but serme > sermi

Something about blocking in polymoraic syllables?

o > uo as in coro > cuoru
o > oa as in koli > koali

We need to figure out the context rules, and also whether /i/ and /u/ change. i > ai, u > au, maybe?

So one isogloss thing between dialects could be whether/when these things happen. How about E/W/Capitol all do it (but ever so slightly differently), but N doesn't?

SO, how do we motivate long diphthongs? Do we have palatalized consonant phonemes in modern Seadi? We've been paying too much attention to orthography instead of phonology.

Is it possible that the short/long vowel distinction is being neutralized in southern dialects of Seadi in favor of quality differences? If /e/ never appears stressed in open syllables, but /ē/ does, it's a prime opportunity to lose the length feature since it's no longer contrastive.

But Jesus, this is going to mean a ****load of paradigmatic leveling with both verbs and nouns.

Declension of nominals in -ā

NOUNS IN -Ā
mā "land"

Singular

Abs: mā < *ma-a
Erg: mās < mānas < *ma-a-nës
Inst: mān < māhan < *ma-a-hën
Sim: māi or māisi < *ma-a-še
Part: māda < *ma-a-da
Voc: mā < *ma-a-t
3p: māu < *ma-a-o
4p: māi < *ma-a-e
(Gen: māu < *ma-a-v)

Plural

Abs: māi < *ma-a-i
Erg: mēs < mēnes < *ma-a-i-nës
Inst: mēn < mēhen < *ma-a-i-hën
(Sim: mās(i) < *ma-ā-še < *ma-a-i-še)
Part: māja < māida < *ma-a-i-da
Voc: māi < *ma-a-i-t, with paradigmatic leveling from *mā
3p: māju < *ma-a-i-o
4p: māji < *ma-a-i-e
(Gen: māju < *ma-a-i-v)

Rescued from the vaults: Complement Clauses

"I want you to love her."
Koanes vies ēlteta.


"I know that you love her."

you love her = vies ēlte
I know it = sanas teje

Sánas téie ó vei ēlteva
Sánas téie ēltevá vei

In conclusion, when a verb is nominalised, its agent receives the genitive.

general "I know that you love her" = Sanas teie o vei ēlteva
specific "I know that you love her" = Sanas teje o vei ēltēva

ēltesva > ēltehva > ēltēva
ēltekva > ēltekva > ēltehva

No, no, absolutive (patient) takes the genitive, and the verbal noun really shouldn't inflect for person and number—

that you love me = vies ēltevá sai, vies esái ēlteva
that I love you = sanas ēltevá vei, sanas evéi ēlteva


Infinitives:
-a- > -ān
-e- > -ein
-o- > -ūn

I think you want it:
Sanas think-as vies koanein