Tyrbis ~ Bsarhus - a PIE lang

[Omzinesý]

Preliminary system of verbal affixes

tense(=augment) - root - aspect - mood - person agreement

Verbs, like nouns, have two conjugations: thematic and athematic ones.


The system of aspect is an innovation of Bsarhus. Old Indo-European perfective distinctions are lost.

There are two synthetic aspects
* momentous
* frequentative
and one analytic
* habitual

- The momentous is zero-market.
- The frequentative derives from PIE *lo particle, which marks past forms in Slavic languages. In Bsarhus the frequentative marker is /ol/ in the athematic conjugation in indicative 3rd and 2nd person forms. In all other contexts the frequentative marker is /l/.
- The habitual is formed with the auxiliary sy 'is' and present habitual participle -r. The -r is related to the actor noun -er in English. The participle agrees the number of the subject. See that 1-2 and 2-1 don't have plural subjects and the participle takes a singular form.

I haven't thought much of the aspect but I guess it will be a quite conservative: indicative, conjunctive, optative, imperative.

Person agreement
Person agreement is hierarchical. So 1st and 2nd persons are always marked, whether they were subjects, objects, or oblique arguments. 3rd person is zero marked. So the person agreement does not care the role of the argument. If there is both 1st and 2nd person arguments, plural 1st person forms are used.

sg1 athematic: -ëk, thematic -k (<- ego)
sg2: -te
3rd: zero
pl1: -um
pl2: -ti

Example: athematic verb bar 'to bear'
Momentous:
bar-ëk
bar-të
bar
bar-um
bar-ti

Frequentative:
bar-l-ëk
bar-ol-të
bar-ol
bar-l-um
bar-ol-ti

Habitual:
su bar-ë-r*
satë bar-ë-r
sy bar-ë-r
sym bar-r-ë-r 'I bear you/You bear me.'
sym bar-r-on 'We bear'
sati bar-r-on
sy bar-r-on

* The /ë/ is an epenthethic vowel between to /r/s.

Example: thematic verb këna 'to sing'
Momentous:
këna-k
këna-të
këna
kënom <këna-um
këna-ti

Frequentative:
këna-l-ëk
këna-l-të
këna-l
këna-l-um
këna-l-ti

Habitual:
su käna-r
satë käna-r
sy käna-r
sym käna-r-on
sati käna-r-on
sy käna-r-on

Meanings of the aspects:
Momentous:
- Something done once or described in the discourse as a single event.
- May be durative, perfective, or static aktionsart

Frequentative:
- Something done several times and seen as a cycle of events
- Something done to objects one after another, e.g. sing songs
- Something done by to agents separately, i.e. distributive action

Habitual:
- Something seen as an unchangeable state, habit, or a property of the subject


I would like to make an egophoric sensory evidence system for 1st person forms. They are though typologically very rare and don't appaer in PIE.

[Inkcube-Revolver]

The idea of using Grimm's Law for a non-Germanic lang is fascinating to me. I'm itching to screw around with that idea! Itching, I say! This conlang is also looking pretty neat, keep it up!

[Omzinesý]

The idea of using Grimm's Law for a non-Germanic lang is fascinating to me. I'm itching to screw around with that idea! Itching, I say! This conlang is also looking pretty neat, keep it up!

Yes, actually I just changed the vowel series and it happened to resemble Germanic :)

[Omzinesý]

Basics of the vowel changes

* a: -> o: -> o
*e -> a 
*e: -> a: -> a
*o -> ə
*u -> o -> ə
*o: -> u: -> u
*ue/*uo -> ɨ
*ie/*io -> ɨ
*ei/*oi -> e: -> e
*eu/*ou -> o: -> o
*i = i

[Omzinesý]

Hindustani forms unaccusative intransitivised verbs by ablaut. I guess it somehow derives from Sanskrit middle voice. Hindustani operates with Sanskrit participles though. In Greek middle voice seems not to cause ablaut.

By the message above, Bsarhus ablaut is:
(PIE zero) Zero. *e *o
(-) - a ë
(i-) i i y
(u-) ë y u
(-i) i e e
(-u) ë o o
(e) a i y
(o) ë u y
(a) a e o

The vowels changes aren't as systematic as those of Hindustani. I think somewhat leveling is needed.

[Omzinesý]

Athematic declension
I use quan 'woman' as the example word

Nominative: quans
Accusative: quan
Genitive/oblique: quani
DativeVocative: quane

Athematic singulars usually have a thematic plural with -i marker.
Nominative: quanis
Accusative: quani
Genitive/oblique: quanej
Vocative: quani

The same root in the thematic declension with a slight semantic difference
Nominative: quanos
Accusative: quano
Genitive/oblique: quany <o+i
Vocative: quano

Thematic singulars usually have an athematic plural with -n marker. The nominative does not have the -s marker.
Nominative: quanon
Accusative: quanon
Genitive/oblique: quanoni
vocative: quanone

I'll add one case more. It's a contraction of the locative and the instrumental.
It's also used with some postpositions to form more complex locational meanings.

The genitive also loses its use as the default oblique case. Having just one obligue case would be unnatural.

[Omzinesý]

Revising the declensions

The declensions actually are not direct descendants of the athematic and thematic declensions. Hereafter I'll call them just consonantal and vocalic declensions.

The consonantal declension is a descendant of PIE consonantal declensions and the -os declension, i.e. the thematic declension.
PIE *o -> Bsarhus ë (schwa)
The schwa has disappeared and everywhere but in the vocative case.

The vocalic declensions are descendants of PIE eh2 -> ā -> ō -> o, collective nouns ending in -a, and PIE stems ending in a semivowel, i.e. -is and -us conjugations. By sound laws -us should have merged with the consonantal conjugation but it has remained -us.

The situation somewhat resembles Catalan that has lost masculie -o but preserves feminine -a.

Bsarhus has 5 cases

Nominative: -s
Accusative: -Ø
Genitive:-i
Locative/instrumental: -ebh/-bh
Vocative:- ë/-Ø

[Omzinesý]

Consonantal:
Nominative: quans
Accusative: quan
Genitive: quani
Locative/instumental: quanebh
Vocative: quanë

Plural with i suffix:
Nominative: quanis
Accusative: quani
Genitive: quanje
Locative: quanibh
Vocative: quani

The same root with the ā ending:
Nominative: quanos
Accusative: quano
Genitive: quany <o+i
Locative/instumental: quanobh
Vocative: quano

Plural with n suffix.
Nominative: quanon
Accusative: quanon
Genitive: quanoni
Locative/instrumental: quanob <- quanomb <- quano-n-bh
vocative: quanonë

[Omzinesý]

Preliminary system of verbal affixes

tense(=augment) - root - aspect - mood - person agreement

Verbs, like nouns, have two conjugations: thematic and athematic ones.


The system of aspect is an innovation of Bsarhus. Old Indo-European perfective distinctions are lost.

There are two synthetic aspects
* momentous
* frequentative
and one analytic
* habitual

- The momentous is zero-market.
- The frequentative derives from PIE *lo particle, which marks past forms in Slavic languages. In Bsarhus the frequentative marker is /ol/ in the athematic conjugation in indicative 3rd and 2nd person forms. In all other contexts the frequentative marker is /l/.
- The habitual is formed with the auxiliary sy 'is' and present habitual participle -r. The -r is related to the actor noun -er in English. The participle agrees the number of the subject. See that 1-2 and 2-1 don't have plural subjects and the participle takes a singular form.

I haven't thought much of the aspect but I guess it will be a quite conservative: indicative, conjunctive, optative, imperative.

Person agreement
Person agreement is hierarchical. So 1st and 2nd persons are always marked, whether they were subjects, objects, or oblique arguments. 3rd person is zero marked. So the person agreement does not care the role of the argument. If there is both 1st and 2nd person arguments, plural 1st person forms are used.

sg1 athematic: -ëk, thematic -k (<- ego)
sg2: -te
3rd: zero
pl1: -um
pl2: -ti

Example: athematic verb bar 'to bear'
Momentous:
bar-ëk
bar-të
bar
bar-um
bar-ti

Frequentative:
bar-l-ëk
bar-ol-të
bar-ol
bar-l-um
bar-ol-ti

Habitual:
su bar-ë-r*
satë bar-ë-r
sy bar-ë-r
sym bar-r-ë-r 'I bear you/You bear me.'
sym bar-r-on 'We bear'
sati bar-r-on
sy bar-r-on

* The /ë/ is an epenthethic vowel between to /r/s.

Example: thematic verb këna 'to sing'
Momentous:
këna-k
këna-të
këna
kënom <këna-um
këna-ti

Frequentative:
këna-l-ëk
këna-l-të
këna-l
këna-l-um
këna-l-ti

Habitual:
su käna-r
satë käna-r
sy käna-r
sym käna-r-on
sati käna-r-on
sy käna-r-on

Meanings of the aspects:
Momentous:
- Something done once or described in the discourse as a single event.
- May be durative, perfective, or static aktionsart

Frequentative:
- Something done several times and seen as a cycle of events
- Something done to objects one after another, e.g. sing songs
- Something done by to agents separately, i.e. distributive action

Habitual:
- Something seen as an unchangeable state, habit, or a property of the subject


I would like to make an egophoric sensory evidence system for 1st person forms. They are though typologically very rare and don't appaer in PIE.

I'm going to make the verbs also agree gender of their (definite?) object, or maybe the subject as well at some circumstances.

Example:
Bar-u-k
bear-NEUTR-SG1
'I bear/have it.'

That is however difficult diachronically. I have two solutions:

1. A definite object pronoun is agglutinated to the end of the verb form.
A hunter killed that wolf. -> A hunter killed-that wolf.
- The basic word order is just SOV, so motivating the change would need some "drift" of grammaticalization.

2. The verb forms were (passive?) participles in the beginning.
A hunter killing-AGR wolf.
- The system of the verbal endings is however dramatically changed and the subject agreement suffixes are agglutinated pronouns.
- The orientation is a problem. To agree with the object, the participle should be passive.
- There should be some other traces of the participle endings than the frequentative -l.
- The augment used to appear with finite forms in PIE. I like the idea of having the augment and its causing changes of the first consonants of the stem. So how is the augment explained if the verb forms are participles originally?

Ideas? Which alternative would be better?

[Omzinesý]

I'll make some verbs form their perfect form bar 'have' and PIE *-to participle. The participle agrees the gender/number of a definite object (1). It doesn't agree if the object is indefinite (2). All of the verbs that form the periphrastic perfect are transitive or can be used transitively.

(1)
bar-m kyno-ts-a gan-t-i-Ø
have-SG1 wife-POSS-ACC hit-PTCP-FEM-ACC
'I have hit my wife.'
Lit. I have my wife hit.

(2)
bar-m kyn-Ø gan-t
have-SG1 woman-ACC hit-PTCP
'I have hit a woman.'

The gender number markers of the participle (and most adjectives) are:

MASC.SG -u
FEM.SG -i
MASC.PL -i
FEM.PL -on (after consonant), -un (after vowel)

To form a simple t-past (t standing for the participle suffix), not perfect,the auxiliary bar is omitted. Then personal clitics are added to the participle (3) and (4).

(3)
Kyno-ts-a gan-t-i-k
wife-aPOSS-ACC hit-PST-FEM-SG1
'I hit (PST) my wife.'

(4)
Kyn-Ø gan-t-ëk
woman-ACC hit-PST-SG1
'I hit a woman.'


I'm not sure how the simple t-past differs from the simple augmented past form (5).

(5)
Kyno-ts-a a-ghan-më.
wife-POSS-ACC PST-hit-SG1
'I hit my wife.'

I want Bsarhus not to have the IE triad of past tenses so that (5) was the imperfect, (3) and (4)the aorist, and (1) and (2) the perfect. Intransitive verbs don't however form t-past. The intransitive past (middle voice) would be like (6).

(6)
Me kyno-ts-Ø a-gn-a
My wife-POSS-NOM PAST-get_hit-TV
'My wife got hit.'

PS. The semantics of the clauses aren't misogyny. Bsarhus just has quite few words at the moment.

[Omzinesý]

The frequentative aspect still exists. It's formed by the PIE participal ending *-lo.

It also takes the gender/number suffixes agreeing with the object. I forms its past form with the augment. That's probably analogical to the momentous forms. Be cause it was a particple form itself in the beginning, it cannot form the t-past.

(7)
Kynotsa gan-l-i-k.
my.wife hit-FREQ-FEM-SG1
'I beat (PRS) my wife.' (i.e. repeatedly hit)

(8)
Kynotsa a-ghan-l-i-k.
my.wife PST-hit-FREQ-FEM-SG1
'I beat (PAST) my wife.' (i.e. repeatedly hit)

(9)
Kynotsa gan-l-i bar-m
My.wife hit-FREQ-FEM have-SG1
'I have to beaten my wife.'

[Omzinesý]

I think I'll drop labialized velars and threat them as velar + /u/ in sound changes.
So 'woman' is *kuin -> kɨn <kyn>

/ti/ -> /tsi/, /di/ -> /dsi/ [dzi]
Phonologically they are consonant clusters, but historically affricates.

When it comes to palatalizations and labializations, Bsarhus resembles Greek.

[Omzinesý]

Revised conjugations:

There are four productive conjugations. I name them mimicking the tradition of Latin.

I
The first conjugation derives from PIE -ah2 nouns.
Word final /ā/ shortens to /a/instead of rising to /ō/, which it does word-internally.

NOM -as
ACC -a
VOC -a
GEN -o <- ō <- ōd <- ād (PIE ablative)
LOC -e <ai

II
The second conjugation is a merger of PIE o and athematic conjugation.

NOM -s / -əs <ȯs> (-əs appears after consonant clusters)
ACC -Ø / -ə <ȯ> (-ə appears after consonant clusters)
VOC -ə <ȯ> (logical development from PIE *e would be Bsarhus -a but the vocative is apparent analogically ə, to keep it separate from a conjugation)
GEN -u <- ū <- ūd <- ōd <- od (PIE ablative of o-conjugation)
LOC -i <- i (PIE athematic ablative)

i-conjugation and u-conjugation are preliminary.

III
NOM -us
ACC -u
VOC -u
GEN -ɨ <y> <- ɨ: <- ɨ:d <- ūd <- ud
LOC -ɨ <y> <- ui

IV
NOM -is
ACC -i
VOC -i
GEN -e <- ei
LOC -ɨ <y> <- iu


Examples with my traditional words: I kynas 'wife, II kyns 'woman'

I
NOM kynas
ACC kyna
VOC kyna
GEN kyno
LOC kyne

II
NOM kyns
ACC kyn
VOC kynȯ
GEN kynu
LOC kyni

Most words in I are feminine.
Most words in II are masculine.
Most words in III are masculine.
Most words in IV are feminine.

[Omzinesý]

II conjugation has some stems ending in a resonant m, n, r, or l.

They lack the nominative -s marker, but the last vowel of the stem is lengthened.

NOM pȯtir <- *pte:r
ACC pȯtar <- *pter
VOC pȯtar <- *pter
GEN pȯtru <- *potr-
LOC pȯtri <- *potr-

The nouns originally belonging to the thematic o-conjugation do, however, have the -s marker. Many nouns, including kyns, have shifted to to the regula paradigm.

[Omzinesý]

On schwa

I have been writing the letter at least <ë>, <ȯ>, and <ə>. In this post I write it <ö> :)

In Bsarhus it is a vowel like the others. It can be stressed. It has reduction only in the word-final position.

It's the thematic vowel of II conjugation. It's though usually absent.
It however appears:
1. after voiced stops. There is no final devoicing in Bsarhus, the voiced stops just cannot appear word-finally. ö is added.
2. Bsarhus resenbles French in that schwa appears braking three-consonant clusters.

[Omzinesý]

Editing verbs

There are two parameters definining conjugations:
1. Thematic vs. Athematic

2. Strong vs. Weak

The historical imperfect(ive past) has got evidential senses. I'm not sure of its semantics yet. But the difference from the aorist, which takes care of basic affirmative sentnces, is not aspect anyway. I'll call the past tenses the simple past and the evidential past.

Weak verbs:
Present: the stem
Simple past: the stem + t (described more thoroughly earlier)
Evidential past: a (= augment) + the stem

Strong verbs
Present: the stem
Simple past: a + the aorist stem*
Evidential past: a + the present stem

* The aorist stem has avoided some lenitions of the stem.

Psa 'he writes' <- pika
Appika 'he wrote' <- a-pik-s-a
apsa 'he seems to have written(?)' <- a-pik-a

[Omzinesý]

Sound changes of consonants revisited

The difference between (1) and (2) has the same conditioning than Verner's law. (2) doesn's however merge with (4)
like in Germanic. (3), (4), and (5) equal Grimm's law of Germanic.

(1)
(When prededing syllable is stressed)
p = p
t = t
k = k
kʷ = kʷ => q

(2)
(When preceding syllable is not stressed)
p => f => h (=> ∅)
t => θ => s /V_
k => x => ʃ
kʷ => xʷ => ʃ

3.
b => p
d => t
(g => k)
(gʷ => kʷ =>q)

(4)
(In non-initial position)
bʰ => β
dʰ => ð
gʰ => ɣ
gʷʰ => ɣʷ => ʁ

(5)
(Ininitial position)
bʰ => β => b
dʰ => ð => d
gʰ => ɣ => g
gʷʰ => ɣʷ => ʁ => ɢ

[Omzinesý]

Vowels

PIE => Bsarkus

e => ə
o => a
a => a
ee => i:
oo => u:
a: => a
i => e
u => o
ie/ei => i
uo/ou => u
io/oi => ɨ
ue/eu => ɨ

Central vowels ɨ, ə, a develop long pairs in open syllables, which become phonemic after some changes of syllable structures.

Later
a: -> e: or o:

Some Vowel + nasal consonant sequences are nasalized creating ã:, ẽ:, õ:.

[Omzinesý]

Writing the vowels

i: <ij>, ɨ: <y>, u: <uw>
i <i>, ɨ <y>, u <u>
e: <ae>, ə: <ə>, o: <ao>
e <e>, ə <ə>, o <o>
(a: <a>)
a <a>

The length distinction in central vowels is still only partially phonemic, so it's not written.

[Omzinesý]

Bsargus has 7 vowels:

/i/, /ɨ/, /u/
/ʲe ~ je/, /wo/
/ɛ/, /ə/, /ɔ/
/ä/

In Cyrillic they are written:
<и>, <ы>, <у>
<(ь)e>, <во>
<э>, <ъ>, <о>
<а>

Code: Select all

Before Soviet era, <аи> or <ай> was used for /ɛ/ instead of <э>. That represents the older pronunciation. 
In Tyrbish, old /je/ has changed to /ʲe/ because of Russian influence. 
In Bsargus, old /je/ and /ʲe/ of Russian loan words are separate. 
In writing they are differentiated as <e> and <ьe> respectively. 
<e> is also used for /je/ word-initially and after a vowel. <e> is also used as the marker of Vocative in 3rd and 4th declensions. 

Romance writing of vowels is not as phonemic. /ɨ/ and /ə/ often alternate, /ɨ/ appearing in stressed syllables and /ə/ in unstressed syllables.
<i>, <y>, <u>
<e>, <wo>
<e>, <y>, <o>
<a>



Deriving the vowels from PIE

PIE => Bsargus
*i, *ey => /i/
*u, *ew => /u/
*e => /ɨ/ or /ə/
*o, *ā => /a/
*ow => /o/
*oy => /e/
*ē => je
*ō, => wo