Round Robin Conlang/Observations

From FrathWiki
Jump to: navigation, search

Thought I'd look at what our phonology looks like so far. --Trɔpʏliʊmblah

Basic inventory

Consonants

It appears that we have four grades of consonants, plain, lenited, geminated and prenasalised, realisations of which may overlap. Roots appear to contain only plain consonants. --PeteBleackley 17:00, 13 January 2010 (UTC)

In fact, very few nonplain consonants are found in any underlying forms: there are only the /f/ of the first person marker and the /χ/ of the complicated past.

The status of /ts/ is unclear: no roots contain it, but it's not a non-plain grade of any plain consonant.


Plain consonants
Labial Coronal Palatal Velar Uvular Glottal
Stops /
affricates
Ejective tsʹ
Voiceless p t (ts) k q
Voiced b d ɡ
Nasals m n ŋ
Fricatives Voiceless s h
Voiced v
Liquid l
Presumed semivowel j

Vowels /i e ɛ a ɔ o u/; /oi au ai/ For purposes of vowel-harmonic suffixes, /a/ (phonetically open central [ä]?) counts as a back vowel.

Attested vowel clusters: /iɔ/

Attested vowel contractions: //ui// → /u/, //ɛi// → /e/, //ii// → /i/

Tone High and low. Low is unmarked. High tone remains in contractions, but in some proccesses of reduplication (not in verbal pluralization) the second of two is dissimilated to low.

Lenition

The following changes are attested:

Original p b t n g ŋ q
Lenited f v s ð̃ ɣ ɣ̃ χ

[ð̃ ɣ] have only been attested under spirant lenition thus far.

I would presume the process to apply regularly also to the "missing" buccal stops / nasals, ie. m, d, k → ṽ, z (ð? l??), x. /s, j/ appear to be unchanged under lenition as seen from gɔso, vijes. Whether the other consonants do anything remains to be seen.

Other alternations

Gemination appears to be regular for at least lenitable consonants, with bb, tt, nn, gg, ŋŋ attested. /tsʹ, s/ appears to resist gemination as seen from betsʼaq, ísasaq. /j/ becomes /ddz/ when geminated.

There is also prenasalization (apparently identical with gemination for nasals).

Syllable structure

Thus far (C)V(N)(C) seems sufficient (maximal example: boimb). Only clusters of two consonants have been observed medially, generally limited to geminates and nasal + consonant (but see hóvhov. Might coda /v/ be [w]?)