Kahichali 2.0.

A forum for all topics related to constructed languages
Post Reply
User avatar
Posts: 2598
Joined: 27 Aug 2010 07:17
Location: nowhere [naʊhɪɚ]

Kahichali 2.0.

Post by Omzinesý » 19 Feb 2019 15:52

Kahicali (the first version viewtopic.php?f=6&t=2690 ) is still my most ambitious project.
Its morphophonological processes are that complex that I myself get messed with them. The whole language suffered from what is called kitchensinkyness on this board. Anyways, it has good ideas.

Now that my abilities in linguistics - especially in describing things, and my English skills too - have got better, I'm going to revive Kahichali.
Some features will be simplified some maybe made more complicated.

Kàhičáli [kɑ:˩çi˧t͡ʃɑ:˥l̪i˧]
Last edited by Omzinesý on 19 Feb 2019 17:44, edited 1 time in total.

User avatar
Posts: 2598
Joined: 27 Aug 2010 07:17
Location: nowhere [naʊhɪɚ]

Re: Kahichali 2.0.

Post by Omzinesý » 19 Feb 2019 16:23


Phoneme inventories

Inventory of consonant phonemes:
Spirant affricates: t͡x, k͡x, k͡ʷʍ<th, kh, khw>
Sibilant affricates: t͡s, t͡ʃ, k͡ʃ <c, č, kš>
Plosives: p, t, k, kʷ, ʔ <p, t, k, kw, '>
Nasals: m, n <m, n>
Spirants: v, x, ʍ, h <v, h*, hw, h*>
Sibilants: s, ʃ <s, š>
Liquids: l, r <l, r>
Semivowels: j, w <j, w>

*/x/ and /h/ appear in same environments very rarely, so I think the homograph is not fatal.

Inventory of vowel phonemes:
High: y, ʲi, u <y*, i, u>
Mid-low: ʲɛ, ɛ <e*, e*>
Near-low: ɑ <a>

Palatalization before /ɛ/ is difficult for analyses because it is phonemic.
The non-palatalizing /ɛ/ is historically /ɑ/ after some fronting processes (comparable to what is called i-umlaut in Germanic).
The palatalizing /ʲɛ/ as an older phoneme.
I decided to leave that unwritten.

[y] can also appear both palatalizing and not palatalizing, but only the non-palatalizing /y/ is phomemic, a fronted pair of /u/. The palatalizing [ʲy] can be analyzed the result of phonetic process of contracting /i/ + /v/ or /ʲɛ/ + /v/ in unstressed syllables.
I leave that unwritten, too.


Stressing is clearly the most complex feature of Kahichali phonology.

Every odd or every even syllable in a Kahichali Word is stressed. The last syllable is, however, not stressed if the word does not have a contrastive focus stressing. Defining one of the stresses the main stress is not important. Maybe the first stress has the most strength.
The rhyme of Kahichali word foots can this be seen as trochee or iamb.

stressed - unstressed - ... - stressed - unstressed (-unstressed)
unstressed - stressed - ... - unstressed (-unstressed)

Every stressed syllable has either high tone (written with acute á) or low tone (written with gravis <à>. Unstressed syllables always have the mid-tone (no accent mark <a>).
All stressed syllables last two morae, i.e. their vowel either is longer, which is not a phonemic feature per se, or it has a coda consonant. Unstressed syllables last only one mora, i.e. do not have a long vowel or coda consonant.

This phoneme analyses is based on the premise that stressed syllables have the underlying representation of the syllables in question, and the full form of unstressed syllables is reduced. IMO, such premises always are descriptive stances not a truth.

Reducing the unstressed syllables

Most Kahichali morphemes have an inherent tone. Rhymes of stressing are governed be the leftmost morpheme. Kahichali is mostly a prefixing language. That is every root have a trochee stressing in some morphological environments and an iambic stressing some others. So adding a prefix makes (nearly) all stressed syllables rightwards from it unstressed and unstressed syllables rightwards from it stressed.

There are three processes in removing stress.
1) Removing possible coda consonant
2) Leniting the onset consonant
3) Removing tone

At least in principle, all of the changes are phonetic.
Most of them are also written.

1) Removing coda consonants
The following consonants can appear as codas of stressed syllables: /x/, /v/, /l/, /r/, /n/ - or rather a generic nasal assimilating in POA with the following consonant (anusvara) -, /ʔ/ - or rather generic stop assimilating in POA with the following consonant -.

When unstressed, /n/ nasalizes the preceding syllable and disappears.
Vn => Ṽ
High vowels tend to lower when nasalized, thus:
/yn/ => œ̃
/in/ => ʲɛ̃
/un/ => ɔ̃

In writing, the lowering is not marked. <n> is also preserved as a marker of vowel nasality.
When unstressed, /l/ or /r/ has a metathesis with the following consonant.
VlC => VCl

If the following consonant is, however, a nasal or a liquid, an new stressed schwa vowel (with the same value as the preceding vowel) appears after the /l/ or /l/.

If the following consonant is a semi-vowel, /w/ or /j/, the syllabic boundary only moves.
Vl.jV => V-ljV
When unstressed, the preceding vowel and /v/ assimilate.
A front vowel + /v/ creates /y/.
A back vowel + /v/ creates /u/.

Palatalization of the preceding consonant is preserved.
/ʲi/ or /ʲɛv/ => [ʲy]
When unstressed, /ʔ/ simply disappears in the coda.

When unstressed, /x/ develops a new stressed vowel after it, so that it is an onset of the new syllable.
Vʔ => V

2) Leniting the onset consonant
Onset lenition only concerns affricates. They become the corresponding fricatives.

t͡x, k͡x => [x] <h>
k͡ʷʍ => [ʍ] <hw>
t͡s => [s <s>
t͡ʃ, k͡ʃ => [ʃ] <š>

3) Removing tones

Like I said above, unstressed syllables do not have a contrastive (high or low) tone but a mid-tone.

Removing a high tone does not affect the consonants of the syllable (if Onset lenition in 2) is not seen as part of losing tone).
Removing a low tone makes the preceding consonant voiced. (I guess that can still be seen as allophony in an abstract level.)

So preceding a vowel that has lost its low tone when it's unstressed, the following changes of consonants do happen:
p, t, k, kʷ => b, d, g, gʷ <b, d, g, gw>
x, ʍ => ɣ, w̝ <q/j*, w>
s, ʃ => z, ʒ <z, ž>

*When palatalized.

Historically, it is apparent that tonogenesis happens the other way around, voiced vowels initiate a low tone.
Last edited by Omzinesý on 19 Feb 2019 18:11, edited 12 times in total.

User avatar
Posts: 1814
Joined: 11 Feb 2015 11:23

Re: Kahichali 2.0.

Post by gestaltist » 19 Feb 2019 17:02

I like the thought you put into your orthography. I feel most conlang orthographies are too exact and miss the underspecification of real languages.

Post Reply