A forum for all topics related to constructed languages
Post Reply
Posts: 1193
Joined: 30 Aug 2010 00:23


Post by Solarius » 02 Jan 2018 22:37

The Koteg Language (Koteg: <Ğghő ʔotegg> [ɣgo̰˥ ʔo˨.te˩g]) is my latest conlang project. It's intended to be an inversion of what I normally do in my conlangs: tonal, isolating, noun-incorporating, and loads of fricatives and front-rounded vowels. It's intended to be naturalistic but a bit weird as well, since I normally do languages which are highly naturalistic and easy to work with. It was inspired by Germanic languages and Oto-Manguean languages, as you can probably see in the phonology.

/i y e̞ ø̞ æ ɞ ɑ o ə ʉ~ɵ/

/p b t d g q ʔ ɸ f s z ç x ɣ χ ʁ m n ŋ j ɥ w/

There is a system of four register tones [˥ ˦ ˨ ˩]. Every syllable has an inherent tone.

Indepedently of the tone system is a system of phonation--modal voice (which often verges into breathy) and creaky voice, which is realized with a laryngeal component.

Phonotactics follow a (C)CV(C) structure. There's no adherence to the sonority hierarchy. The onsets *[st] and *[sn] are illicit; in areas where they would occur (e.g. compounds) they become [s.] Diphthongs and hiatus aren't allowed (may change this).

Allophony and Sociophonological Variation
I don't much like doing allophony (I'll probably come back to this later) but here are two major processes:
-The uvular stop /q/ is realized as [k] after front vowels and [q] after back/mid vowels. I analyze it here as /q/ because it is realized as such word initially and after consonants.
-The alveolar nasal stop /n/ is realized as [r] after consonants.

Koteg has several major lectal variants:
The standard language is the Northeastern dialect, described above. There are a few other major dialect groups:
The Western dialects vary phonologically in several important ways:
-Northeastern /q/ is Western /k/--it's [q] after back vowels, [k] elsewhere.
-/ŋ/ is merged to /n/. In some dialects the usual allophonic process of /Cn/-->[Cr] applies, in others not, which means that these dialects have a restricted phonemic /r/.
-In certain words, what is Northeastern word-initial /ʔ/ is Western /j/ e.g. NE [ʔys˩] is W [jys˩]. This reflects different onset insetions to words starting with a vowel.
-/χ ʁ/ are merged to [h].
-Western dialects retain a more conservative pronounciation of /ʉ~ɵ/ as /u~ʊ/.
-Western Koteg retains distinct phonemes /œ/ and /ɔ/, which are merged in NE Koteg as /ɞ/.
-Quite a few vowels are diphthongized: NE /e̞ ø̞ ɑ o/--> W /ie yœ uæ uɔ/
-All low tone vowels are creaky (this is common in colloquial NE Koteg too).

Southern dialects are little more divergent, retaining some conservatism as well as substrate influence.
-NE /ʔ/ is S /k/. The fronting of front-adjacent /q/ doesn't happen.
-As in Western /ŋ/ and /n/ have merged, but to [ŋ]. This merger goes further in codas; /m/, /n/, and /ŋ/ have all merged to [ŋ].
-/n/ isn't [r] after consonants.
-/d/ is [l] between vowels. Some dialects completely turn NE /d/ to /l/.
-Historic words starting with a vowel begin with a [ŋ] rather than a glottal stop, e.g. S /ŋy̰s˩/ for NE /ʔys˩/.
-Historic /st sn/ are retained.
-Significant vowel raising and offgliding has occured: NE /i y e̞ ʉ~ɵ o/--> /iç yɸ i uf u/
-Historic diphthongs /ai ay au/ are retained as /e ø̞ o/, where in the NE the offglide was lost. Compare NE /ga˩/ to S /gø̞˩/. The monophtongization happened after the raising.
-Schwa is /e/.
-As in Western /œ/ and /ɔ/ are unmerged.
-The tone system has contour tones: NE /˥ ˦ ˨ ˩/ vs. S /˥ ˥˧ ˩ ˨˩/.
-The registers are flipped--Northeastern and Western Koteg creaky voice is breathy voice in the south, and vice-versa. This because Old Koteg had a change of */hV/-->/ʔV/-->Creaky voice, but the first stage never happened and became a breathy voice--while other syllables became creaky in response.
-All low tone vowels are creaky here too.
Koteg is the language of a large cultural area, with influence widely dispersed as a prestige language/lingua franca. Outsiders tend to use much more conservative pronounciations, spelling pronounciations, and foreignisms (i.e. The city of <Syad Xnábbnos> "Fort Courage" is pronounced [swad˩ xna˥.bnost˩] by those who live there, while in Northeastern Standard Koteg it'd be [sɥɑ˩ xra˦.bros˨]. Compare Old Koteg [sɥad˩ xnag˩.bnost˩]). Many of these speakers speak these L2ish variants natively.
The variation between [ʉ] and [ɵ] isn't allophonic, but rather is sociolinguistic: women use [ʉ] and men use [ɵ]. I alternate between the two in transcription to avoid gender bias.

Morphophonological Processes
Pseudo-Rendaku: Voiceless stops and fricatives become voiced in compounds or with incorporated objects when followed by a voiced phoneme.
-/p t q ʔ ɸ f s ç x χ/-->[b d g g ɥ w s j ɣ ʁ]
/pF/ reduction: /p/ disappears before certain fricatives in onsets when preceded by another phoneme. This applies across word boundaries.
-/pɸ pf pç/--> /ɸ f ç/
Because this is so pervasive, onset clusters of the form /pɸ pf pç/ don't exist anywhere except utterance-initially.

My transcription of Koteg is intended to reflect the native syllabary, converted into Latin form.
The rough phoneme to letter correspondance is:
/i y e̞ ø̞ æ ɞ ɑ o ə ʉ~ɵ/<i ü e ö ä õ a o ë u>

/p b t d g q ʔ ɸ f s z ç x ɣ χ ʁ m n ŋ j ɥ w/<p b t d g q ʔ hw f s z ç x ğ x̆ ħ m n ŋ j y w>
This however ignores special uses of the letters and <h>, both of which are described below.

Tone is represented with a combination of acute accents and indicator consonants. Vowels with an acute accent are either high tone or mid-high tone, while vowels without an acute are either mid-low or low. Disambiguation is indicated by the helper consonants <b d g> which indicate that the preceding tone is lower. Thus, using <b> as an example:
CV́--> [CV˥]
CV--> [CV˨]

However, codas of any kind are permitted--so how is [CV˩] distinguished from [CVb˨] and [CVb˩]?
--A voiced stop coda with no influence on the preceding vowel (e.g. [CVb˨]) is written with a separating <h>--thus <núhg> [nʉg˥].
--A voiced stop coda in a syllable which requires a tone letter (e.g. [CVb˩]) is written by doubling the coda letter--thus <mdadd> [mdɑd˩]

These tone-indicating consonants are etymological, and often resurface in older compounds--thus <ged> [ge̞˩] "eye" becomes <gedŋöb> [ge̞˨.dŋø̞˩] "brown-eyed." It's worth noting that spelling regularization in the past 100 or so years has removed the etymological consonant from double codas--thus historic *<mdagd> "hair" became <mdadd>. These etymological consonants still do resurface in pronounciation (but not orthography!) though when in older compounds; thus <mdaddi> [mdɑ˨.gdi˨] "red-haired". These insertions are no longer productive.

Creaky voice is indicated with an <h> too, albeit placed before the vowel: <hwghü> [ɸgy̰˨] "come, venitive."

Pseudo-Rendaku is written phonetically. Thus <ʔös> [ʔø̞s˨] "child" + <net> [ne̞t˨] "hold" --> <ʔöznet> [ʔø̞˨.zre̞t˨] "baby-hold."

As for /pF/-reduction, the historic /p/ is unwritten, except when /p/ exists in etymologically transparent compounds/derivations or word initially:
-<pnufid> [prɵ˨.fi˩] "adze" (from historic */pnep.fid/)
-<näpçik> [næ˨.çik˨] "sock" (from <näp> "bed" + <çik> "foot")
-<pfṍ> [(p)fɞ˥] "andative" (even when not utterance initial).

However, one exception exists. The copula [fo̰˩] is written as <pfhod> utterance initially. However, due to its frequency it's abbreviated as <'fhod> elsewhere.

So yeah. I'll be talkin' about Tense and evidentiality next, or maybe Pronouns/Aspect. But first: for those who have read this far: There are two loanwords from real world languages in the examples above. What are they?
Check out Ussaria!

Posts: 1193
Joined: 30 Aug 2010 00:23

Re: Koteg

Post by Solarius » 06 Jan 2018 16:33

Tense and evidentiality are both marked periphrastically on verbs, in a way which is lightly intertwined.

An unmarked verb is understood to be direct and past.

hwõb "(Someone) prayed (and I saw it)"
pŋágg "(someone) died (and I heard it)"

To make the verb indirect, one adds the particle zlu after the verb.

hwõb zlu
"someone prayed (I was told)"

pŋágg zlu
"someone died (I figured)"

The present tense neuters distinctions of evidentiality. It's marked with the verb znghüdd "to lie [1]." It comes historically from a present progressive but it has extended semantically.

znghüdd hwõb
PRS pray
"is praying"

znghüdd pŋágg
PRS die

[1] In the sense of physically lying, rather than to tell a lie.
Check out Ussaria!

Post Reply