An unnamed Uralic a-posteriari lang

[Omzinesý]

For some reason, I have put my old ideas to a Uralic lang, now. So the project repeats many ideas my old langs have but hopefully there is something new too.

The language is positioned somewhere in Finland, so I can easily borrow from Finnish and Swedish.
On the other hand the lang has some very atypical features for the area, namely verb-framed motion frames and quite productive aspect marking.

PU *kɒlɒ wetin ɒlɒnɒ ujɨ
This lang: kɒulɜ yjjɪs weidɛ̃ŋ alɜ̃ŋ

[Omzinesý]

On Vowels

I hereby presuppose Proto-Uralic (PU) had the following (stressed/ first syllable) consonants.

*y, *u, *ɨ, *u
*e, *o
*æ, *ɒ/ɑ

There long vowels *i: and *u: are supposed to be combinations of *i or *u and a glide.

Development of stressed vowels

1. *y -> i
2. *ɨ -> ɒ/ɑ(a common change in "West-Uralic)

3. Fronting: u -> y
Also diphthong /*au/ which appears in old loan words is fronted:
*au -> æy

4. Diphthongization in open syllables:
*e -> ei, o -> ou -> u
*æ -> æi, *ɒ/ɑ -> ɒu

Monophthongs are however restored in open syllables as well in roots where the syllable usually is closed.
For example verb menide 'to go' has a monophtong also in imperfective forms menis [me:nis] '(s)he goes about' because the root usually appears with perfective suffixes menti '(s)he went there' and the syllable is closed.

Diphthongs are written this way:
<é> ei, <ó> u*
<ai> æi, <ay> æy, á <ɒu>

Monophthongs are written this way:
<y> y/yi**, <i> i
<e> e, <o>, o
<ae> æ, <a> ɒ/ɑ>

*/u/ is historically a diphthong. It's written with <ó> in native words and in most Swedish loans. In Finnish loans it's usually written <u>.
** /y/ is pronounced [yi] in open syllables but it's just a phonetic realization.


Non-first /unstressed syllable vowels

PU had only two non-stressed vowels a high one *i and low one *a.
This lang has mostly preserved them. There is a vowel harmony so that, if the root has front vowels, the preceding vowels are front or [ɛ], and if the root has back vowels, the preceding vowels are mid [ɨ] or [ɜ]. They are always written <i> and <e>, respectively. The high vowel is mid-high ([ɪ] or [ɪ̈]) in closed syllables.

There is also secondary /u/ <u> that has developed from V + /v/. It only appears as Unergative intransitive derivational verb suffix and Instrumental case suffix. /u/ doesn't have vowel harmony.

Vowels before nasal /ŋ/ are nasalized and /ŋ/ is pronounced very weakly, so the language can be said to have nasal vowels too, but I analyse them as V + ŋ. Phonotaxis is more easily described that way.

[Omzinesý]

On Consonants

PU had (something like) these consonants:

*pp *tt *kk
*p *t *k
*m *n *ɲ, *ŋ
*t͡sʲ *t͡ʃ
*s *sʲ *ʃ
*r *l
*ʋ *ð *ðʲ *j

1. Voicing of single plosives in voice environments (inter-vocally or between a vowel and a sonorant):
*p *t *k -> b d g
2. Fortitation of *ð
*ð -> d

3. Merger of *s and *sʲ
*s *sʲ -> s
4. Voicing of sibilants (there are sporadic exceptions like 3rd person verb suffixes and some word initial sibilants)
s *ʃ -> z ʒ
5. Assibilation of *j (as an exception word-final *j of the plural oblique case and past tense)
*j -> ʒ
6. Lenitation of the affricates
*t͡sʲ *t͡ʃ -> s ʃ

7. Fortitation of *j word-finally
*j -> ç (-> ħ)
7. *ɲ and *ðʲ become /j/ word-initially
*ɲ *ðʲ -> /j/

8. Merger of *ɲ and *ŋ after vowels
*ɲ, *ŋ -> ŋ
9. Also *n and *m become ŋ before sonorants
*m *n -> ŋ

10. *v preceding a sonorant assimilates to it
*v + C[+son] -> CC[+son]

11. nasal/lateral + voiced plosive (created by 1.) lose the plosive when not followed by a vowel
mb nd ŋg -> m n ŋ
lb ld lg -> l l ʟ

12. Prestopped sonorants
In some stems there is consonants gradation where single sonorants of the stem are prestopped
l -> dl (ʟ -> gʟ) r -> dr
m -> bm n -> dn (ŋ -> gŋ)
j -> ɟj
Prestopped consonants often appear word-initially in loan words.


With those changes we end up to the following phoneme inventory:
p t k

Edit: b d g

m n ŋ
s ʃ
z ʒ
r l ʟ
ʋ j h(~ç ~ħ)
+ prestopped sonorants

Consonants can be geminated though it's rare of nasals and laterals.

Written
<
p t k

Edit: b d g

m n ň
s š
z ž
r l l'
v j h
>

[Omzinesý]

On Phonotaxis

Onset
a) Native word begin in a singe consonant or a vowel

b) Loanwords can begin with prestopped consonant, which is analysed as a single consonant. Phonetically its a sonorant preceded by a homorganic stop.

tran, trande 'beach', 'coast' <- Swe. stranda
trise 'crisis'
kl'ari 'clear' <- Swe. klar

c) A semivowel (ʋ or j) can appear between the onset consonant and vowel. They also only appear in loan words.
tvoksida 'to smell' <- Fin. tuoksua
njyte 'enjoy' <- Swe. njuta

Nucleus
Only monophthongs can appear in closed syllables.
Vowels in stressed open syllable are remarkably longer.

méne [meinɛ] 'I mean'
meni [me:ni] 'I go about'
mente [mentɛ] 'I went there'

The difference between open and closed syllables causes differences in conjugations and declensions as well.

Coda
More difficult to explain. Will be described later. Quite SAE.

[Omzinesý]

Morphology: nouns

Nouns have
two numbers: singular and plural
four cases: Nominative, Accusative, Genitive, and Instrumental.

Features distinguishing the declensions:
1) Is the preseding syllable stressed and open?
2) Is the stem consonant a sonorant?
3) is the an epenthetic vowel?

If the stressed syllable is open, i.e. there is only one consonant between the two syllables, the stem vowel is preserved in all cases.

Singular
NOM kále
ACC kádle
INSTR kálu
GEN kádleň

Plural
NOM kádlet
ACC kádleh
INSTR kédleh
GEN kálede

[to be continued]

[Frislander]

OK so I like the idea of a Uralic conlang. I do have a couple of questions though.

So you say you put it in Finland. However your language doesn't seem to be either Finnic or Saamic, which feels kinda strange to me, because it doesn't seem to line up with either of those groups in terms of its phonology or what we see so far of its grammar. I could kind of understand it if your proposing this as taking the place of much of Finnic's current range, but that's not what you're going for I think.

With those changes we end up to the following phoneme inventory:
p t k
m n ŋ
s ʃ
z ʒ
r l ʟ
ʋ j h(~ç ~ħ)
+ prestopped sonorants

Wait where did the voiced stops go? I didn't see you do anything which would cause their loss after you had generated them.

Finally I do like the case system, but I would like to ask where the instrumental comes from, and why the system is so reduced on its Uralic sisters (and therefore what replaced the lost cases).

[Omzinesý]

OK so I like the idea of a Uralic conlang. I do have a couple of questions though.

So you say you put it in Finland. However your language doesn't seem to be either Finnic or Saamic, which feels kinda strange to me, because it doesn't seem to line up with either of those groups in terms of its phonology or what we see so far of its grammar. I could kind of understand it if your proposing this as taking the place of much of Finnic's current range, but that's not what you're going for I think.

With those changes we end up to the following phoneme inventory:
p t k
m n ŋ
s ʃ
z ʒ
r l ʟ
ʋ j h(~ç ~ħ)
+ prestopped sonorants

Wait where did the voiced stops go? I didn't see you do anything which would cause their loss after you had generated them.

Finally I do like the case system, but I would like to ask where the instrumental comes from, and why the system is so reduced on its Uralic sisters (and therefore what replaced the lost cases).

I forgot the voiced stops. Now they are there.
I'm still considering if I allow /u/ in unstressed syllables. Instrumental is secondary anyway. The best etymology, I think, is V + the adjectival derivational suffix -va. The same suffix is also behind the accusative/anticausative derived verbs.

Finnish
liha 'flesh' -> lihava 'fat (A)'
väki 'strength, mana' -> väkevä 'strong'

Edit: Mordvic has -u as a prolative (via) marker. It also appears in Saami in some prolative adverb.
Erzya, Moksha: moda 'earth' -> modava Erzya modav, Moksha modu 'via earth'. It somehow derives from one of many Uralic lative (to) markers, the velar nasal.

This project resembles Scandinavic many ways, in syllable structure especially. I was thinking to locate it at Finnish west-coast where the main language is Swedish rather than Finnish.
The best known Uralic languages happen to have more cases than the worse known ones. So the reduction is not as heavy as it may seem. There are of course postpositions. Which is not typical for northern Europe, the transitive construction with a direct object is often used with motion verbs too. I find that areally problematic, but still like it. One option could be to locate language in Southern Europe and borrow the syllable structure from Italian instead in Scandinavic, but then there couldn't be loans from Finnish and Swedish.

Mon mente kóte.
I go.PST.SG1 house.ACC
'I entered the house.'

Mon laekte kóte.
I leave.PST.SG1 house.ACC
'I left the house.'

[Omzinesý]

Numbers

NOM - ACC
1 it [it] - itti <- ykti
2 kaet [kæt] ~ kat [kät] - kaette - katte <- kæktæ (æ and ɑ are merged in closed syllables, in many dialects)
3 kolim [ko:lin] - kolmi
4 néli [neili] - néli <- neljæ
5 kaide [kæidɛ] - kaite <- derived from kæti 'hand' (not fron witi)
6 kydi [kyidi] - kyti <- ku:ti
7 zeňs

Bigger numbers must be thought more about, because they are borrowed.

10 itluk
20 kaetluk ~ katluk
30 kolluk
40 nelluk
50 kaidluk
60 kydluk
70 seňsluk

[Frislander]

This project resembles Scandinavic many ways, in syllable structure especially. I was thinking to locate it at Finnish west-coast where the main language is Swedish rather than Finnish.

Yes, that does make sense.

The best known Uralic languages happen to have more cases than the worse known ones. So the reduction is not as heavy as it may seem.

Sure, I know Finnic and Hungarian are to some extent outliers on this, but even compared to the next nearest language group to yours (Saamic) it still seems a little small by comparison.

But anyway, as you were!

[Omzinesý]

Verbs

- Agreement with the subject
- Antipassive voice of transitive verbs
- Two tenses
- Two aspects (I still don't know how productive they are.)

Person
SG1 -[zero]
SG2 -d
SG/PL3 -s
PL1 -mi
PL2 -ddi
Impersonal -ssi

SG1 is thus just the bare stem. The Uralic -m has disappeared.
SG3 is the Uralic SG2 suffix -t, which has been sporadically voiced.
Third person -s derived from Uralic personal/demonstrative pronoun sV (Finnish se), which often appears as a verb suffix in Uralic langs.

PL1 is the normal Uralic suffix -mI/mA
PL2 is too sporadically voiced
The impersonal is apparently an analogical formation after -ddi and -s. Its older use is PL3 but it's mostly replaced by the singular form. So, -ssi has been left with general 'they' meaning.

tégi 'I do it'
tégid 'you do it'
tégis 'does/do it'
tégimi 'we do it'
tégiddi 'you all do it'
tégissi 'it is done'

Imperfertive past suffix in Uralic is -j. Word-final /j/ becomes /x/ in this language. The third person -s however assimilates with it creating /ʃ/.

tégih 'I did it'
tégihd 'you did it'
tégiš 'did it'
tégihmi 'we did it'
tégihdi 'you all did it'
tégišši 'it was done'

Some verbs do also have two perfective TAMs. They Are Past and Future. The past marker is -t, which is a Uralic past participle and Modern Hungarian forms its regular past form from it, and the future marker is -k, which appears as the imperative/subjunktive/present marker in many languages. Both of the still also appear as participles.

The perfective TAM markers are usually added to the consonant stem if it's phonologically possible. Because those suffixes cause closed syllables in the stem and thus omit diphthongs, many verbs that often appear in pperfective TAMs analogically acquire monophthongs also in imperfective TAMs. Verb "tegi" has actually been splitted: "tégi" meas 'I do it' and is always imperfective, while "tegi" means 'I make it (usually)' and "tegke" 'I will make it'.

tegke 'I will make it'
tegked 'You will make it'
tegki 'will make it'
tegkemi 'We will make it'
tegkeddi 'You all will make it'

tegte 'I made it'
tegted 'You made it'
tegti 'made it'
tegtemi 'We made it'
tegteddi 'You all made it'

The -k also appears as an imperative marker.
Tegik! 'Make it!'
Tegikki 'Make it, you all!' <- tegi-k-tti
Tegigni! 'Let's make it!' <- tegi-k-mi (gn is a prestopped velar nasal)

[Omzinesý]

I'm considering changing the changes of nasal + plosive clusters.

The new alternative:
(1) Homorganic clusters (= those with same POA) become voiced geminate stops.
m + p => mb > bb
n + t => nd => dd
ŋ + k => ŋg => gg

(I don't know if I should also create phoneme /ɟɟ/ from ɲ + t͡ʃ. (Hungarian gy)

(2) Non-homorganic clusters become homorganic clusters, having the POA of the stop.
ŋ + p => mb
ŋ + t => nd
m + t => nd
m + k => ŋg

The changes in (2) happen for lateral + plosive clusters regardless of POAs.


kunta => kyd(da)
ïmta => ande
vali-ma-k(i) => valimk(i) => valiŋ(gi) 'say-ing'

[Omzinesý]

[...]

Singular
NOM kále
ACC kádle
INSTR kálu
GEN kádleň

Plural
NOM kádlet
ACC kádleh
INSTR kédleh
GEN kálede

[to be continued]

I'll add a case meaning 'from' and is called Ablative. Its marker is -d and it derives boringly from PU ablative *-ta. Its plural is still a question though I think it will exist.

Singular
NOM kále
ACC kádle
INSTR kálu
GEN kádleň
ABL kádled


Ablative is used to express

1) source as a complement of a verb, while Accusative expresses a goal. This allows the motion verb constructions to be a bit less verb framed.

Pošk menes kóde.
boy.NOM goes house.ACC
'The boy is going to the house.'

Pošk menes kóded.
boy.NOM goes house.ABL
'The boy is going out from the house.'

Kóded menes pošk.
'A boy is going out from the house.'

2) source as an object of postpositions / specified by local adverbs.

Pošk menes kóded sis.
boy.NOM goes house.ABL inside
'The boy is going out from inside of the house.'

3) source of a state. Then it always appears before the verb.

Pošked tult kóži.
boy.ABL come.PST man.NOM
'The boy became a man.'

4) stimulus of mental states.

Borra tónd .
cry.SG1 you.ABL
' I'm crying for you.'

5) cause if the referent of the noun is abstract.

Typrused le taine.
stupidity.ABL be.SG1 here
' I'm here because of my stupidity'

6) the object of numbers.

kolim pošked
'three boys'

7) the partitive of uncountable words

Zéme védid.
drink.SG1 water.ABL
' I'm drinking water.'

Edit:
8) by-agent of non-finite constructions.

[kóžed taštilit] kiri
[man.ABL write-PTCP.ANTER].NOM book.NOM
'the book [written by the man]'

[Omzinesý]

Accusative and Genitive will be merged like in Finnish, Saami, and Mordvinic.

The new Genitive-Accusative will look like the genitive in my older posts, i.e. it has no suffix but usually differs from Nominative by its stem.

Semantically the merger actually explains the shift from a (non-German-)Germanic kind of a double object in ditransitive construction, like English I gave the man a book. PU Genitive apparently also had the Dative meaning, so the dative just is formally alike the accusative, though they historically are different cases.

The problem was infinite constructions. I wanted to preserve a separate genitive for they by-objects. But now the language has Ablative that can be used in that function. Using a source case in this function instead of the genitive is also quit a Germanic feature.

[kóžed taštilit] kiri
[man.ABL write-PTCP.ANTER].NOM book.NOM
'the book [written by the man]'

[kidri taštilit] kóže
[book.ACC write-PTCP.ANTER].NOM man.NOM
'the man [having written the book]'

[Omzinesý]

On Vowels

I hereby presuppose Proto-Uralic (PU) had the following (stressed/ first syllable) consonants.

*y, *u, *ɨ, *u
*e, *o
*æ, *ɒ/ɑ

There long vowels *i: and *u: are supposed to be combinations of *i or *u and a glide.

The long high vowels, however, behave like long vowels, so maybe PU should rather be reconstructed:

*i: *u:
*ʏ *ɪ *ɘ *ʊ
*e *o
*æ *ɑ~ɒ

Until now I've supposed *ɪ and *ʊ merger with *i: and *u: but they could also go downwards. More MOAs could make the language even more Germanic-like.

The vowel systen could for example be

y i u
e o
ɛ~æ ɔ
ä
+ diphthongs

[Omzinesý]

The anti-passive

This is a NOM-ACC language but it still has an anti-passive voice.
There are transitive verbs that presuppose a definite object. So if there is no explicit object, an anaphoric object "it/him/her/them" is understood, (1). So, if you want to kill the object, you have to use the anti-passive (2).

(1)
Mon zidme.
I.NOM love.SG1
'I love him/her/it/them.'

(2)
Mon zidme-n-e.
I.NOM love-ANTIP-SG1
'I love/am in love.'

Anti-passive marker is n. It derives from demonstrative *sen of which only the -n is left. The former meaning of the form is reflexive, so the demonstrative has earlier been a reflexive pronoun. It's added to the consonant stem if the stressed vowel is a monophthong and there aren't two different stem consonants (11).

Antipassive is also used if there is an object but it is indefinite. The object is then demoted to Instrumental case (3).

(3)
Mon zidme-n-e nežd-u.
I.NOM love-ANTIP-SG1 girl-INSTR
'I love a girl.'

(4)
Mon zidme nežd-i.
I.NOM love.SG1 girl-ACC
'I love the girl.'

Many motion verbs also presuppose an object expressing the goal, (5), (6), (7).

(5)
Mon men-ke
SG1.NOM go-FUT.SG1
'I will go there.'

(6)
Mon men-ke linn-e
SG1.NOM go-FUT.SG1 town-ACC
'I will go to the town.'

(7)
Mon men-ke-n-e linn-u
SG1.NOM go-FUT-ANTIP-SG1 town-INSTR
'I will go to a town.'

They can however have an argument in Ablative expressing source also. Then, no goal argument is presupposed. So (8) does not mean 'I will go from the town to there.'

(8)
Mon men-ke linn-ed
SG1.NOM go-FUT.SG1 town-ACC
'I will leave the town/go from the town.'

Some motion verbs, especially the imperfective ones do not presuppose a goal argument (9), though they can have it as an explicit NP (10) and even have the antipassive (11).

(9)
Mon žall'e.
'I am walking.'

(10)
Mon žall'e linn-e.
'I am walking towards the town.'

(11)
Mon žal'-n-e linn-u.
SG1.NOM walk-ANTIP-SG1 town-INSTR
'I am walking towards the town.'


--------------------------------------------------------------
Etymologies of the words used in examples:

lin(ne) 'town' Estinian: linn 'town', Finnish: linna 'castle', apparently borrowed
men(i) 'to go', usually appears in perfective TAMs while the stem has a closed syllable and diphthongization is blocked for all TAMs, PU *meni
nežd(e) 'girl', PU *nejti 'girl'
zidme- 'to love', derived from zidim 'heart', PU *sʲyðæmi 'heart'
žall'e 'to walk' , derived from žal'(ge) 'leg ~ food' , PU *jalka '

[Omzinesý]

Ditransitive clauses

PU Accusative *-m and PU Genitive-Dative *-n are merged. This language hasn't developed a new dative. So there is a zero-marked Genitive-Accusative-Dative (in the example just ACC) case.
Indefinite objects are however expressed by the Instrumental.

The distransitive construction much resembles that of Swedish and English. There ca be two objects. The order of the objects marks higher focusing/newer information so that the latter object is more focused. The recipient however demands the postposition par 'for/to' if it appears as the latter argument (2), like it does in Swedish and English.

(1)
Ante nežd-i bók-e
gave.SG1 girl-ACC book-ACC
'I gave the girl the book.'

(2)
Ante bók-e nežd-i par
gave.SG1 book-ACC girl-ACC/GEN for
'I gave the book to the/a girl.'

If the theme (what is given) is indefinite, it appears in the latter position in Instrumental (3). The primary object (first object) hardly ever is indefinite thus the verb does not take Anti-passive.
(3)
Ante nežd-i bók-u
gave.SG1 girl-ACC book-INSTR
'I gave the girl a book.'


There are no passives in this language, but either of the objects can be positioned in the beginning of the clause but then the recipient demands the postposition par and the theme must take Instrumantal case regardless of its definiteness, (4) and (5).

(4)
Nežd-i par (mon) ante bók-e
girl-ACC/GEN for (SG1.NOM) gave.SG1book-ACC
'The girl was given the book by me.'

(5)
Bók-u (mon) ante nežd-i.
book-INSTR (SG1.NOM) gave.SG1 girl-ACC/GEN for
'A/the book was given by me to the girl.

----------------------
The only interesting etymology may be:
par, PU para 'good'', the motivation behind the grammaticalization is 'for ones good'

[Frislander]

On Vowels

I hereby presuppose Proto-Uralic (PU) had the following (stressed/ first syllable) consonants.

*y, *u, *ɨ, *u
*e, *o
*æ, *ɒ/ɑ

There long vowels *i: and *u: are supposed to be combinations of *i or *u and a glide.

The long high vowels, however, behave like long vowels, so maybe PU should rather be reconstructed:

*i: *u:
*ʏ *ɪ *ɘ *ʊ
*e *o
*æ *ɑ~ɒ

Until now I've supposed *ɪ and *ʊ merger with *i: and *u: but they could also go downwards. More MOAs could make the language even more Germanic-like.

The vowel systen could for example be

y i u
e o
ɛ~æ ɔ
ä
+ diphthongs

Actually if you're gonna lower the *ɪ *ʊ vowels then it'd be nice to be consistent and have *ʏ lower to ø as well.

[Omzinesý]

On Vowels

I hereby presuppose Proto-Uralic (PU) had the following (stressed/ first syllable) consonants.

*y, *u, *ɨ, *u
*e, *o
*æ, *ɒ/ɑ

There long vowels *i: and *u: are supposed to be combinations of *i or *u and a glide.

The long high vowels, however, behave like long vowels, so maybe PU should rather be reconstructed:

*i: *u:
*ʏ *ɪ *ɘ *ʊ
*e *o
*æ *ɑ~ɒ

Until now I've supposed *ɪ and *ʊ merger with *i: and *u: but they could also go downwards. More MOAs could make the language even more Germanic-like.

The vowel systen could for example be

y i u
e o
ɛ~æ ɔ
ä
+ diphthongs

Actually if you're gonna lower the *ɪ *ʊ vowels then it'd be nice to be consistent and have *ʏ lower to ø as well.

But PU *ʏ merges with *ɪ anyway. Then of course it's loweved if *ɪ is.

[Omzinesý]

Participles and Action nominal

There are three participles that differ by relative tense. Thus they are Simultaneous (or present) -že, anterior (or perfect) -t, and posterior (or future ~ prospective) -k.
Their etymologies shouldn't be surprising. -že derives from PU participle *-ja, and -t and -k are the same participle/tense suffixes that are used in to form the perfective TAMs.

The participles aren't oriented, i.e. they do not code the role of their antecedent in the participal clause. "Tegit" can well mean either '(which is) done' or 'having done'.
All arguments that the participle has are positioned before it. The agent argument is usually coded by Ablative (1) while the patient argument is usually coded by Instrumental (2). If the agent is the same as that of the main clause (SS, same subject) , it is however expressed by the reflexive possessive suffix -n.

(1)
Kóže as lese nažed taždetit bóket.
kóze-t a-s lese naže-d tažde-t-it bóke-t
man-PL NEG-3 read.CONNEG woman-ABL write-PTCP.ANTERIOR-PL book-PL
’Men don’t read books written by women.’

(2)
Bókuh lesežet kóžet kevvembet-s bókuh lesežeh nažeh.
bók-u-h lese-že-t kóže-t kevve-mbi-t-s bók-u-h lese-že-h naže-h
book-INSTR-PL read-PTCP.SIM-PL man-PL few-COMP-PL-COP.3* book-INSTR-PL read-PTCP.SIM-PL.ABL woman-PL.ABL
’Men that read books are fewer than women that read books.’

(3)
Son as lese taždetnen bókine ytu.
son a-s lese tažde-t-ne-n bóki-ne ytu
SG3 NEG-3 read.CONNG write-PTCP.ANTERIOR-PL.ACC book-PL.ACC-POSS again
'(S)he doesn't read the books written by him/herself again.'

*Adjectives can be made "sami-verbs" by adding the third person agreement suffix -s (Present) or -š (Past).


The suffix of the action noun is -ng. Its etymology is apparently PU nominalizing derivational suffix *-n that appears e.g. in Hungarian and Permic infinitive -ni. But it usually appears after old PU lative *-k to form the infinitive -nk (historically 'to the Xing') it is velarized in all contexts. Another motivation for velarization is that Swedish has a similar suffix -ning (Englsih -ing).

[Examples may appear later]

[Omzinesý]

Plural

NOM kádlet
ACC-GEN kádlene ~ kádledne
ABL kádleh
INSTR kádluh

Plural nominative is also used for indefinite objects. -t is the old PU plural suffix
Plural Accusative-Genitive derives from demonstrative *nä- 'these' . It usually has a definite meaning. It may have the old plural suffix before it.
The -h in Ablative and Instrumental is the plural oblique suffix -j, which is a fricative word finally.
The /u/ in Instrumental is the same suffix as in the singular.