It began life as an attempt to "reconstitute" my favourite of Tolkien's languages, Adûnaic (in the same vein as Vardelm's Quasi-Khuzdul, I guess). However, I rapidly realised that the idea was wrong-headed and futile, particularly as I kept messing about with it. Some elements of structure remain, as do several roots, but Qôni probably owes more now to Punic and Akkadian for inspiration.
History and situation
The heyday of spoken Qôni was some eleven centuries ago, in the cities of the Mafret peninsula, which at that time was part of the Tailancan Empire. While the current inhabitants of the area speak Çarronic varieties (an unrelated Kalpo-Lacaran language) in the north or Rahetian (a language collaterally related to Qôni) in the south, Qôni remains the liturgical language of the Dawwali denomination of Athaulism. As a living language, Qôni had ceased to be spoken in the cities of the Mafreti littoral by the eleventh century, although pockets of rural Qôni-speakers appear to have remained in the hill-country of modern Çarrón until only a few hundred years ago.
Qôni is itself a descendant of Achaunese (ʔaḳawnīyu in that language). Achaunese was the first language of high civilisation in western Adeia, and also the first in the region to acquire writing. The language and civilisation exerted a great deal of cultural prestige over the neighbouring Lacaran peoples, and the Tailancan alphabet ultimately (via Eteolacaran) finds its origins in the Achaunese syllabary. From the first arrival of the Lacaran people into western Adeia until the final collapse of Achaunese civilisation in the fifth century, the Tailancan language has borrowed numerous words from Achaunese: a number of which were exported back into Qôni after the peninsula's incorporation into the Tailancan Empire in the eighth century.
Phonology and transcription
Qôni is written in a variant of the Tailancan alphabet, augmented by the addition or modification of a few characters to indicate sounds not present in Classical Tailancan. Thus the Qôni alphabet (ramâl Qôni) bears roughly the same relationship to the Tailancan as Coptic does to Greek.
The consonant inventory attested in the manuscripts comprises eighteen phonemes:
Code: Select all
labial dental palatal velar glottal
nasal m n
voiceless stop t k
voiced stop b d g
ejective stop ṭ q
voiceless fric f s
voiced fric (v) z ḥ
ejective fric ṣ
continuant w y
lateral l
rhotic rThe vowel inventory was as follows, with a circumflex denoting length:
Code: Select all
front back
close i î u û
close-mid ê ô
mid (ǝ)
open-mid e o
open a âStress was generally on the final syllable if closed, the penultimate if not. However, stress was not always predictable. The native script does not mark irregular stress (although it can be contrastive), in this description it will be marked by an underline where relevant: Qôni /kʼoːˈni/, ruzed /ˈruzed/.
Root structure
Roots in Qôni essentially comprise either two or three consonants with a linked characteristic vowel, with those roots containing three consonants being the majority. Throughout this description, the consonants will be denoted by numbers, characteristic vowels by V and other vowels of insignificant quality by v. There are a few constraints on combinations of root consonants: 1 and 2 cannot be the same (2 and 3, however, can be), and aside from geminates, no more than one ejective can occur in a root. Similarly, characteristic vowels are limited to one of a i u. Thus √kokl and √ṭeqr would not be permissible roots, but √lukk and √ṭikr would. From these roots, by regular patterns of derivation we arrive at the surface forms of lexical words. For example, to the root √farz, which covers the semantic field of 'yellow, gold', we can apply the noun-deriving pattern 1V23 to get farz 'gold', the adjective-deriving pattern 1a2Vː3 to get farâz 'yellow', or the pattern i12V3ak (often associated with inanimate diminutives) to get ifrazak 'buttercup'.
The above bald statement, however, only obscures quite how fucked up sound-change has rendered these surface forms. The primary culprits are those with a medial or final weak consonant: w y m n ʔ - particularly galling is the fact that ʔ doesn't even have a surface instantiation in Qôni, having been lost in the transition from Achaunese.
The parent language of Qôni had a nice, regular, almost mechanical root-and-pattern structure; itself deriving from an even more regular agglutinative structure. Let us take for example, the regular masculine nouns karbu 'donkey', çawnu 'dog', jaynu 'land', banku 'man', raʔnu 'house', ramlu 'mark' and caklu 'shore'. In the plural, these were karabī, çawanī, jayanī, banakī, raʔanī, ramalī and cakalī. All following the same identical regular pattern.
(Why the change from 1V23u to 1a2V3ī? Underlyingly, the plural morpheme was -yi, which became -ī after a consonant by regular soundchange. In Achaunese's parent language, the forms would have been *karabu~karabyi, *çawanu~çawanyi etc. The loss of short intertonic vowels in open syllables rendered the singulars karbu, çawnu etc, but clusters of three consonants were not tolerated, preventing forms like **karbyi.)
Now, in Qôni these become karb~karabin, sôn~sawanin, zên~zayanin, bak~banakin, rân~rânin, rôl~ramalin and sakal~sakalin (don't worry about the sudden arrival of final -n in the plural forms, it doesn't pertain here) - all of which renders one original paradigm into seven different patterns. And this is before we even consider words in the same original paradigm with characteristic vowels other than a, let alone other paradigms (for the sake of argument: dûa~dâhin is a reasonable example of some of the complexities here).
To go through all the possible permutations in one go would be intemperate and, divorced from context, bewildering. The main processes to look out for then are listed below, with further complications discussed as they are encountered:
- Before a consonant or a word break, we have underlying aw uw iw becoming ô û î.
- In the same context, ay uy iy become ê û î.
- Underlying ʔ is lost both intervocalically and preconsonantally, triggering compensatory lengthening on the preceding vowel.
- Underlying geminate ʔʔ is realised as q (hence otherwise baffling verb forms as utûl 'he receives' vs. utaqol 'he has received').
- Following a consonant, underlying w y ʔ are lost, leaving compensatory lengthening on a following vowel.
- Preconsonantal m becomes w, and the resultant vowel plus w sequences are treated as above.
- Preconsonantal n fully assimilates to the following consonant.
- Stressed i u in an originally open syllable become e o. Thus dulg~dalogin 'black' from Achaunese dulgu~dalugī.
- An anaptyctic vowel (more commonly e, sometimes a) is inserted between two final consonants that would otherwise make up an impermissible final cluster (i.e. those that don't have r l ḥ as their first element).
Next up: nouns and (maybe) adjectives
