Old Oxic
Anthologica Universe Atlas / Forums / Department of Creativity / Old Oxic

previous 1 2 next end
? dhok posts: 235
, Alkali Metal message
The lexicon is here.

Nearly a year ago, I started work on Lusitanic, a third branch of Italo-Celtic that was my first serious foray into Indo-European conlanging. It was imperfect, but I'm rather proud of it: it had a lexicon of about 400 words, a host of sample sentences and a fleshed-out syntax section. However, a year later, I'm feeling the IE bug biting again, and I'm under the impression I could do better these days: I've been doing more reading and can now read some Greek and Sanskrit in addition to Latin.

At first I had planned to create a Satem language with a plosive shift, separate reflexes of *a and *o, a shift of *s to /h/ and a firmly Greek/Sanskrit-style grammatical groundplan. I worked on sound changes for a couple hours until it became evident that I had independently reimagined Armenian, at which point I scrapped the project. This project, Old Oxic, is in the same area (the Amur Darya river valley) and time span (the late Hellenistic period until the 6th century) as that project would have been, but, like Tocharian, is a Centum language that has found its way east. It has ended up looking very much like a Centum version of Persian, mainly, I think, because of the vowels; but the vowel system is common enough.

Starting with late-stage Indo-European, here is a short sketch of sound changes with illustrative examples. (I don't feel like there's too much need to use early-stage PIE as my starting point, since Old Oxic is similar enough to Greek or Sanskrit that it doesn't really need it.)

0. Laryngeal vocalization.

As in Graeco-Phrygian and Armenian, word-initial h₁ h₂ h₃ -> e a o if /#_C.

1. Centumization: Ḱ -> K. Resulting Kw -> Kʷ.

2. First Palatalization: k g gʰ -> ts dz dz /_[iey], except s_.

3. Initial semivowel loss:
y -> Ø / #_F (mostly e)
w -> Ø / #_B (mostly o; stays before a)

4. Loss of labiovelars and *w in clusters:
o -> u / Cw_, Kʷ_
followed quickly by
w -> Ø / C_ (except #s_)
Kʷ -> K

5. Nasal loss (with lengthening) before aspirates and /s/:

VnCʰ -> VːCʰ
Vns -> Vːs

6. Grassmann's law:
Cʰ...Cʰ -> C...Cʰ

7. Plosive shift (note the differences from Grimm's law or the Armenian consonant shift!):

Word-initially:
p t k -> ɸ θ h / #_
bʰ dʰ gʰ -> pʰ tʰ kʰ -> p t k / #_
b d g: no change word-initially.

Medially:
b d g -> w ð ɣ / except when preceded by a non-syllabic nasal.
bʰ dʰ gʰ -> b d g
p t k : no change medially.

sw -> ɸ / #_ also happens at about this stage.

8. Various resonant changes:
ṃ́ ṇ́ -> ám án (when stressed)
ṃ ṇ -> a (when unstressed)
mb nd ndz ng -> m n z ŋ
mp nt nts nk -> b d dz g
ṛ ṛ: -> ar ār
ḷ ḷ: -> al āl

9. y -> w / #_o, -> Ø / #_u

10. ɸ -> w /_[ie]
ɸ -> h / otherwise

11. short *a *o merger:
ŏ -> ă (including in diphthongs)

12. Changes in clusters with /s/:
sp st sk -> ps ts ks / medially
sp st sk -> p t k / initially
remaining #sC -> #asC

13. i-devocalization:
ĭ -> y /C_#

14. Vocalization of labials and velars:
K -> y /_C
P -> w /_C

15. /l/ loss:
l -> w / V_V
l -> y / C_
l -> r / _C, #_

16. Second palatalization:
k g -> y / V_C
t d s ts dz n r -> ts dz ʃ tʃ dʒ ɲ rʲ/ _y
k g ŋ-> t d ɲ/ _y
w -> y / _y
Then:
y -> Ø /C_

17. tʃ dʒ -> ʃ ʒ

18. rʲ -> ʐ -> ʒ

19. ʒ -> y /#_
ʒ -> ʃ /V_V (likely, this was a merger of ʃ -> ʒ which later devoiced in all environments)

20. Vowel shift:

Firstly, in stressed or unstressed posttonic syllables:

ā -> ɔ
ē -> ɛ
ī -> e
ō -> ɔ
ū -> o

a -> a
e -> ɛ
i -> i
u -> u

ai -> ɛ
ei -> i

aw -> o
ew -> u

āj -> e
ēj -> i
ōj -> o

āw -> ɔ
ēw -> o
ōw -> u

In unstressed pretonic syllables, vowels simply lost their length distinction (as well as *ĕ -> ɛ), with the second element of diphthongs disappearing (except that ā ō -> ɔ, not a or o).

21.
ɔ -> o /if stressed
ɔ -> ɒ /otherwise
ɛ -> e /if stressed
ɛ -> æ /otherwise
a -> ɒ / if stressed
a -> æ / otherwise

22. ɣ, y -> Ø /V_V; either w or y appears in their place depending on whether the preceding vowel was front (i e æ) or back (u o ɒ).

23. θ -> s
ð -> z -> r

24. dz -> z
ts -> s /_#

25. m -> n /_#
ŋ -> g

26. Stress shifts to penult.

The end phonology should look something like this:

Plosives:  /p t k b d g/
Nasals: /m n ɲ/ <m n ñ>
Fricatives: /s z ʃ h/ <s z š h/
Approximants: /w r j/ < w r y>
Affricate: /ts/

Vowels: /i e æ u o ɒ/ <i e a u o ā>

I will soon be providing a link to a spreadsheet lexicon.
? Hallow XIII Primordial Crab
posts: 539
, 侯, Dalian, China
message
nice!

that said: why does the first palatalization only affect labiovelars? that seems a bit strange.
? thelettermu posts: 262
, Groovy Cat message
This change is not attested in any IE language, but I feel like it is plausible. Essentially, interconsonental laryngeals do not vocalize to i (as in Indo-Iranian) or to a (as everywhere else), but rather to their "underlying vowel" (e, o or a).

Am I mistaken or isn't that the case in Greek?

Edit:
2. First Palatalization: kʷ gʷ gʷʰ -> ts dz dz /_[iey], except s_.

that said: why does the first palatalization only affect labiovelars? that seems a bit strange.

Greek, again?
(kʷ → t /_V[+front])
? dhok posts: 235
, Alkali Metal message
Yeah, there's some Greek inspiration. I'm tempted to push the first palatalization onto the plain velars so as not to appear too derivative (except then it would be derivative of Phrygian- so who cares, really?). IIRC, though, Greek only differentiated the laryngeals when they were word-initial before a consonant.

I also don't like the vowel shift too much. I'm going to give it a rework this afternoon. I like the look of the language, though- it doesn't look much like Greek or Latin.

EDIT: Thinking about whether to palatalize mainly the plain or the labiovelars- if the former, then I'll just turn the labiovelars into velars under all circumstances; they don't appear as such right now. That'd give hatāras instead of tsatāras and waga instead of waza- slightly more recognizable, I suppose. Old Oxic is most closely related to Graeco-Phrygian, in any case, but these are separate developments. I'm still not sure and may flip a coin.

EDIT TO THE EDIT: The first palatalization now applies to the plain velars, with the labiovelars screwing with a few vowels and then merging to become plain velars. Also, I've come to peace with the vowel system- if Old Oxic is going to look like Persian, then so be it. I'm rather fond of Persian.
? dhok posts: 235
, Alkali Metal message
Moving onto the grammar. We'll take the nouns first. They've lost some cases from PIE (they are now reduced to a nominative, accusative, locative, and "oblique" derived from the genitive in the singular and dual and the dative/ablative in the plural), but have kept the dual and all three genders.

Here's the declension of the a-stem feminines. There are two subdeclensions of a-stems, those that were formerly accented on their stem (example ekā, ekās "mare"), and those that were formerly accented on their ending (apso, apsos "wasp"). We'll take the stem-accents first:

SingularDualPlural
Nominativeekāekaekās
Accusativeekānekaekās
Locativeekeekāwosekāsu
Obliqueekāsekāwosekābas


Fairly straightforward; note that the nominative and accusative are differentiated only in the singular, and that the locative and oblique merge in the dual (true for all genders). Here is an ending-accent:

SingularDualPlural
Nominativeapsoapseapsos
Accusativeapsonapseapsos
Locativeapsāasowosapsosu
Obliqueapsosasowosapsobas


Moving on to the o-stems. There are also two subdeclensions of o-stems, those which were formerly accented on their stem (example etas, etaša, "path"), and those which were formerly accented on their endings (example nisrās, nisrāša, "nest"). Here's etas:

SingularDualPlural
Nominativeetasetāetās
Accusativeetanetāetās
Locativeetietayosetasu
Obliqueetašaetayosetabas


Here's nisrās.

SingularDualPlural
Nominativenisrāsnisronisres
Accusativenisrānnisronisros
Locativenisrinisrāwosnisresu
Obliquenisrāšanisrāwosnisrābas


Note some differences in the endings, mainly changed vowels.

The o-stem neuters, to nobody's surprise, only differ from the masculines in their nominative and accusative. Here's a stem-accent, grewan, grewaša "reed, rush". Only the nom/acc needs to be presented.

SingularDualPlural
Nominative/Accusativegrewangrewagrewā


And here's an ending-accent, karān, karāša "greed".

SingularDualPlural
Nominative/Accusativekarānkarekaro



A subclass of o-stems, both masculine and neuter, have -u- in many forms. Here is an example (stem-accents and ending-accents are not differentiated in this class), ekus, ekuša, "horse":

SingularDualPlural
Nominativeekusekāekas
Accusativeekunekāekās
Locativeekiekuwāsekubas
Obliqueekušaekuwāsekesu


And here is a neuter, dārkun, darkuša "evening meal":

SingularDualPlural
Nominative/Accusativedārkundārkadārkā


I'll try and get some more nouns up soon, as well as pronouns.
? dhok posts: 235
, Alkali Metal message
Time to move on to the other noun declensions. We'll start with the u-stems. These may be either masculine or neuter, and do not have stem-accent and ending-accent subdeclensions; our model masculine (male model?) is kretus, kretus "step".

SingularDualPlural
Nominativekretuskretakretawas
Accusativekretunkretakretos
Locativekretuwikretuwoskretusu
Obliquekretuskretuwoskretubas


Our neuter is dāru, dārus "tree":

SingularDualPlural
Nominative/Accusativedārudāradārā


Here are the i-stems. These also don't have accent subclasses. Here's the feminine medis, medis, "mind".

SingularDualPlural
Nominativemedismedamedayas
Accusativemedinmedamedas
Locativemedimediyosmedisu
Obliquemedismediyosmedibas


...and a neuter, āsti, āstis "bone".

SingularDualPlural
Nominative/Accusativeāstiāstaāstā


Finally, we have the consonant stems, which are sort of a grab-bag declension where nothing else fits. Characteristic of the masculines/feminines is nes, netas, "night". The PIE thematic vowels here were a mix between e and o, with some i and ey imported from the i-stems. As far as stem mutation goes, these have all been treated as e or i (this causes palatalization of velars).

SingularDualPlural
Nominativenesnetanetas
Accusativenetanetanetas
Locativenetinetiyāsnetisu
Obliquenetasnetibānnetibas


And here's a neuter, nebas, nebasas "cloud":

SingularDualPlural
Nominative/Accusativenebasnebasanebasā


(You may be noticing a general theme with the neuters by now).

The consonant declension is the two-stem declension in Old Oxic, much like the grab-bag "third declension" of Latin and Greek, whether the stem must be derived from the genitive (or in this case, the oblique). It also includes the r-stems, such as hatar, hatras "father" and brātar, brātras "brother". R-stems have a nominative singular in -ar and a stem ending in -r elsewhere.

A final note on vocatives. Although Oxic had a vocative case in the singular, I have not shown it in the paradigms since it is fairly simple to form and would just add clutter. Here are the rules for vocatives:

- Masculine o-stems ending in -as have a vocative ending in -a.
- Masculine or feminine i-stems and u-stems ending in -is and -us have a vocative ending in -i or -u.
- R-stems replace the -ar ending of the nominative singular with -er
- Other nouns have no separate vocative.

And that's it: for non-neuter o-stems, i-stems and u-stems, just chop the final -u off, and for r-stems replace -ar with -er.

All in all, Oxic's noun system is not terribly painful. It has a number of sub-declensions, yes, but the accent-based ones are only found in the o-stems, and you can always decline a noun given its nominative, oblique and gender. I'mma call it a night.
? Nessari ?????? ?????? ????????
posts: 932
, Illúbequía, Seattle, Cascadia
message
I should do a once-over of Idlurian sometime. I managed to throw together a verb system that made some sort of sense in 3 hours like 8 years ago.
? KathTheDragon Beware the Dragon
posts: 92
, Baroness, United Kingdom
message
Chattaiumas' noun system is much more painful than yours
? Nessari ?????? ?????? ????????
posts: 932
, Illúbequía, Seattle, Cascadia
message
His point was how simple and unpainful it was…
? Jipí der saz ûf eime steine
posts: 291
, Transition Metal, Germany
message
A few centuries worth of sound changes more, and this could be a Baltic lookalike. I'm looking at you, Latvian.
? dhok posts: 235
, Alkali Metal message
This is a placeholder post for the verbs- right now I have redone a small vowel sound change with Consequences for the lexicon. Please hold.
? thelettermu posts: 262
, Groovy Cat message
This is a placeholder post for the verbs- right now I have redone a small vowel sound change with Consequences for the lexicon. Please hold.

This is a placeholder post for the comments regarding the verbs.
? Rhetorica Your Writing System Sucks
posts: 1292
, Kelatetía, Dis
message
This is a placeholder post for reminding you to make a frickin' wiki.
? dhok posts: 235
, Alkali Metal message
quoting Rhetorica:
This is a placeholder post for reminding you to make a frickin' wiki.

Yeah, I'm gonna do that.
? Morrígan Witch Queen of New York
posts: 303
, Marquise message
I like the loss of nasals with compensatory lengthening before aspirates, that's kind of a cool touch. As is the behavior of syllabic nasals with respect to stress.

Not much to say about the vowels; I'm never happy with my vowel systems or changes, they always feel wildly derivative, or boring. I need to get back to work on Kuma-Koban in the future; it's literally been like 3 years.
? Hallow XIII Primordial Crab
posts: 539
, 侯, Dalian, China
message
quoting Dhok:
The lexicon is (google doc).

well now that was blasphemous

Also: Morrígan! 欢迎光临!
? dhok posts: 235
, Alkali Metal message
quoting Morrígan:
I like the loss of nasals with compensatory lengthening before aspirates, that's kind of a cool touch. As is the behavior of syllabic nasals with respect to stress.

Not much to say about the vowels; I'm never happy with my vowel systems or changes, they always feel wildly derivative, or boring. I need to get back to work on Kuma-Koban in the future; it's literally been like 3 years.

Yeah, my first plan was to pull a Sanskrit, collapse everything but the high vowels, then shuffle things around from there. To the surprise of nobody, there were too many a's and ā's for my liking. (There are a lot of them here, too, but for different reasons). I am also thinking about replacing the laryngeal change (0) to the intial-laryngeals-vocalize-before-consonants change of Greek and Armenian, since Oxic is shaping up to be (most) closely related to Greek.

And welcome back! We've missed you.
? thelettermu posts: 262
, Groovy Cat message
Yeah, my first plan was to pull a Sanskrit, collapse everything but the high vowels,

Hasn't Sanskrit already (partly) done that?
? dhok posts: 235
, Alkali Metal message
quoting meficat:
Yeah, my first plan was to pull a Sanskrit, collapse everything but the high vowels,

Hasn't Sanskrit already (partly) done that?

Sanskrit merged PIE *e o ē ō, even in diphthings, to a ā. Most of its vowels are a or ā, which makes memorizing endings difficult (they all look the same).
? Morrígan Witch Queen of New York
posts: 303
, Marquise message
The o > a changes do look pretty good though, in terms of the results. Lots of as, sure, but it definitely looks good in the oblique plural -bas. Plus, it's like Slavic; not the most imitated of branches.
previous 1 2 next end