Qa'oisi

Qa'oisi
Language family	Unknown
Era	today
Writing system	Unknown
Official status
Spoken in	Qoatzait
Speaker
Demonym	Unknown
Number of speakers	Unknown
Technical information
Usage	Unknown
Language code	qao

Qa'oisi (IPA: /qaʔoixi/, [k̠ɑ.ʔoɪ.ç̠ɪ]; English: Qoishi [kəˈwi.ʃi]) is reported to be a rare, not-well-documented language in the east of Ŋafrica, found around the jungle of Qoatzait. Not much is well documented on the language at this time, although its phonology is fairly well established due to its regularity, despite being somewhat unusual. It is not known to have any living relatives at this time.

Phonology

Qa'oisi is estimated to have a fairly small phonemic inventory, although its phonetic inventory is considerably larger.

Vowels

There are only 3 vowels in Qa'oisi: <Ii> /i/, <Aa> /a/, and <Oo> /o/. These vowels may appear next to each other or on their own, allowing for allophonic glides and diphthongs, although these are not considered unique and are in free variation with normal vowel hiatus. One analysis also preserves /ə/, which still appears phonetically between voiced stops, while becoming either an assimilating voiceless puff of frication or otherwise disappearing entirely in all other environments.

Consonants

By some counts, Qa'oisi may also have a very small inventory of consonants, with only 10 plosives, 2 fricatives, and 4 nasals. Glides are considered allophonic.

Base Consonants	Bilabial	Alveolar	Velar	Uvular	Glottal
Voiceless Plosive	Pp /p/	Tt /t/	Kk /k/	Qq /q/	Xx /ʔ/
Voiced Plosive	Bb /b/	Dd /d/	Gg /g/
Fricative		Zz /s/	Ss /x/
Nasal	Mm /m/	Nn /n/	Yy /ɲ/	Rr /ŋ/

This analysis is fairly simple, and matches the romanization (with the addition of ' to denote syllable boundaries), although it is not perfect. Assuming a CVC structure, it assumes /ə/ is still a fully-fledged vowel that disappears in most environments, while all consonants other than fricatives or a glottal stop may appear in coda position. This system is called the "ghost e" analysis, and can make words phonemically run a bit longer than other analyses, since a vowel and in some cases an additional missing unknown consonant need to be appended to stopped codas. Alternatively, it assumes a CVCC structure, with /ʔ/ being not only possible in coda position, but also after any other consonant (except /q/); this analysis is considered unrealistic and not true to the history of its phonetic realizations. This system is called the "ghost x" analysis.

Another, alternative analysis is that all consonants (except either /q/ or /ʔ/) have a duplicate, no-audible-release (or "stopped coda") version, and that these are phonemically distinct. (I.e. /p/ is distinct from /p̚/ in coda positions). The "dual coda" analysis suggests twice as many phonemes, which, along with somewhat more nuanced places of articulation, a linguist popularized with the below table.

Base Consonants	Bilabial	Dental	Palatal	Uvular	Glottal
Voiceless Plosive	/p/ /p̚/	/t/ /t̚/	/c/ /c̚/	/q/ (/q̚/)	/ʔ/
Voiced Plosive	/b/ /b̚/	/d/ /d̚/	/ɟ/ /ɟ̚/
Fricative		/s/ /s̚/	/ç/ /ç̚/
Nasal	/m/ /m̚/	/n/ /n̚/	/ɲ/ /ɲ̚/	/ɴ/ /ɴ̚/

The status of /q̚/ is disputed even among those who prefer the dual codas model, as it phonemically arises from /q/ much the same way other non-released stops are. However, it is not entirely clear if this is accurate, since it seems to be identical to /ʔ/, though there are some distinct impacts on following vowels rather than prior vowels. Since it is in complementary distribution with /ʔ/, they are sometimes instead analyzed one and the same, while others prefer a distinct /q̚/. This problem also persists in the simplified phonemic inventory, since the sound could still be represented better phonotactically with /q/ but phonetically with /ʔ/. Additionally, some linguists prefer to stick /ʔ/ next to /q/ and refer to the entire column as "gutteral".

It should be noted that the dental consonants may be dental-alveolar, the velar/palatal consonants seem to be more accurately prevelar, and the uvular consonant may be either postvelar or uvular - it's difficult to tell for sure which. This is why symbols have varied between linguists for palatal/velar and velar/uvular consonants.

Phonotactics

As stated previously, the basic syllable structure is (C)V(C), although it may be (C)V(C)(ʔ) under the ghost x analysis. Under both simple phoneme analyses, any consonant may begin a syllable, or be absent entirely. Most consonants can also optionally end a syllable; however, whereas the ghost e analysis argues that o /ʔ/, /s/, and /x/ cannot end a syllable, the ghost x analysis posits that all syllables may end a syllable, and all except /ʔ/, /s/, and /x/ may further be followed by /ʔ/. The dual coda analysis instead posits that all consonants may end syllable (except possible /ʔ/; see above paragraph), while only consonants with an audible release may begin one.

There is a phonemic distinction between two vowels separated by /ʔ/ and a plain vowel hiatus, the latter of which may optionally result in glides or diphthongs. However, /ʔ/ is only phonetically distinct intervocally, since utterance-initially, as well as after coda consonants, an epenthetic [ʔ] is inserted before a vowel. This may have once been a non-distinct realization as a boundary between morphemes, but later got engrained as a distinct phoneme only in certain environments. Consequently, some who use e analysis or see /q̚/ as distinct in dual coda analysis may argue /ʔ/ is only marginal phoneme, but it's widely agreed that its usage is common (and in some cases useful) enough for that not to be the case.

Additionally, while the language is phonemically fairly simple, if difficult to precisely qualify, it is also phonetically a bit more complicated. There is a distinction between /V.CV/ and /VC.V/, since /ʔ/ is epenthetic before vowels in these environments, although there is not a distinction between /VC.V/ and /VC.ʔV/. This may also cause voiceless plosives to become ejective in nature, although not necessarily. Additionally, there is the distinction between non-released consonants, which may be analyzed as either a distinction between /V.Cə.V/ and /VC.V/, /VC.(ʔ)V/ vs. /VCʔ.(ʔ)V/, or two different phonemes in /VC.V/

In addition to this distinction, there is also further distinctions between them and /VC.CV/, where the coda may be the same, different, or the non-audible version of the following consonant. Consonants which are fundamentally the same are both pronounced clearly when plosives, and cause gemination in fricatives and nasals. In cases where they are the same, but the coda consonant is inaudible, then this is reversed; plosives are geminate, while fricatives and nasals are "broken" with a glottal stop (or potentially a nonreleased stop of the same place of articulation).

Things are further complicated when the two consonants are different. While most of the time they are predictable, there are some nuances that must be given attention. Released codas are always fully pronounced, and never "melt" into the following consonant, while nonreleased codas may optionally lose their glottal-like quality when following another consonant. For instance, using dual coda analysis, /at.sa/ would be analyzed as [ʔat.sa], while /at̚.sa/ would be either [ʔat̚.sa] or, more frequently, [ʔa.t͡:sa].