|Reconstruction of||Vaniuan languages|
|Region||Vaniua, around the Melkanchuta Sea|
Proto-Vaniuan (PV) is the reconstructed hypothetical ancestor of the Vaniuan family of languages, one of the most widely spoken language families on Sahar owing to the extensive native range of languages in Miraria, and wider diaspora of speakers, particularly of Balak in former colonial nations such as Tuyo and Sonka.
PV is estimated to have been spoken as a single language from around 4000-2500 BCE, although estimates vary with some placing PV as far back as 6000 BCE, and others as recent as 1500 BCE. The location of the PV urheimat has long been proposed to lie on the shores of the Melkanchuta Sea but the exact location remains contested due to the reconstructed movements of daughter languages and the nature of loanwords in PV often providing seemingly conflicting evidence. The most widely accepted proposal places the original homeland of the Proto-Vaniuans around the northern and northwestern shores of the sea on the open grassland extending to the foothills of the Miralayas, given the locations of the oldest known daughter languages, and recovered material evidence from the northern coast of the Melkanchuta that is believed to have been associated with an early Vaniuan culture, if not the Proto-Vaniuans themselves.
As speakers of PV became isolated through migration, regional dialects of PV diverged and developed into independent daughter branches of the Vaniuan family. In the traditional analysis, nine such branches are recognised: Bodni-Cirnic, Central Vaniuan, Eastern Vaniuan, Lake Vaniuan, Northern Vaniuan (also called Orotic), Osheshdi, Ujaju, Western Vaniuan, and Ohanian - the oldest attested Vaniuan language. There are however disputes regarding the validity of some branches as some authors propose that similarities between certain daughter languages warrant them being grouped together in single branches, most notably with the hypothetical ‘Southern Vaniuan’ languages, whilst others consider branches like the extinct Lake Vaniuan languages to be polyphyletic groupings of only superficially similar languages rather than true genetic groupings. These debates continue but most authors accept the eight-branch analysis and the study of both modern and ancient Vaniuan languages has allowed PV to be reconstructed in detail through the comparative method. Today, the Vaniuan languages with the most native speakers are Balak, Vosan, and Koman.
Proto-Vaniuan phonology has been reconstructed in some detail through the comparative method and is traditionally seen as having the following phonemes, although competing theories do exist (see below).
The consonant system of PV was crosslinguistically fairly typical but with some unusual features. In the most common reconstruction, there was a two-way contrast in the plosives and sibilants between aspirated and unaspirated phonemes, with the exception of the palatal plosives, which lacked aspirated counterparts. There was also a pair of postalveolar plosives that are variously reconstructed as retroflex or retracted alveolar, as well as an unusual approximant inventory with a phonemic velar lateral approximant and a complete lack of phonemic semivowels.
The table gives the most common notation used in modern publications; suggested phonemic values are given below where the symbols differ from the IPA values.
*ṭʰ *ṭ: /ʈʰ ʈ/ but also reconstructed as /t͡ʂʰ t͡ʂ/, /t͡ʃʰ t͡ʃ/ or /t̠ʰ t̠/
*c *cʷ: /c cʷ/ but also reconstructed as /c͡ç c͡çʷ/
*ṣʰ *ṣ: /ʂʰ ʂ/ but also reconstructed as /ʃʰ ʃ/, /s̠ʰ s̠/ or /s̺ʰ s̺/
*x *xʷ: /x xʷ/ but also reconstructed as /χ χʷ/
*h: /h/ but also reconstructed as /ħ/
*‘: /ʔ/ but also reconstructed as /h/
A glottal stop *’ /ʔ/ is reconstructed by some authors as having occurred before vowels at the starts of words, and between vowels in hiatus, based largely on evidence from older stages of Ohanian which appear to have had a strict CV syllable structure with vowels separated by ⟨ʾ⟩. This consonant is also sometimes suggested to have been a fricative /h/ when *h is reconstructed as pharyngeal /ħ/, with the two phonemes subsequently often spelled *h₁ *h₂ respectively. Whether this glottal consonant was phonemic in PV however, or even simply allophonic, remains unclear as no other daughter languages reflect any evidence of it.
Plosive and Sibilant Series
The most widely accepted reconstruction of PV involves a series of two contrasting stops: voiceless unaspirated and voiceless aspirated: *t *tʰ. The aspiration distinction is not found in any living descendant language, but has been proposed for the extinct Ohanian language and Lake Vaniuan branch of languages based on evidence from recorded texts and loanwords both to and from other languages in the area. Unusually, the sibilants in PV patterned like the plosives, with the same unaspirated vs. aspirated distinction. Due to this and the reflexes of these consonants in some daughter languages, in particular Ohanian, which, being the oldest attested Vaniuan language, has traditionally been heavily relied upon for the reconstruction of PV, it has been suggested that the sibilants in PV were actually affricates /t͡sʰ t͡s/ and /t͡ʂʰ t͡ʂ/. Additionally, *c *cʷ, typically reconstructed as palatal plosives, may have been palatal fricatives /ʃ ʃʷ/, /ɕ ɕʷ/ or /ç çʷ/ according to proponents of this theory, providing a more balanced inventory than what is commonly reconstructed for PV. Evidence for this comes from the lenition of *c *cʷ to approximants in both the Bodni-Cirnic and Western Vaniuan language branches which is more likely to have happened to fricatives than plosives due to the sonority hierarchy. Additionally, written evidence of Ohanian dialects indicates a single reflex of both as a sibilant fricative ⟨s⟩ or ⟨š⟩ from an older *s which likely had a palatal articulation /ç/, suggesting a fricative origin.
However, most linguists who study the Vaniuan family reject these theories as conflicting diachronic evidence suggests that such interpretations would create too many irregularities in the sound change developments of descendant languages for them to be feasible. Notable evidence to the contrary comes from Proto-Eastern-Vaniuan (PEV) where *c *cʷ pattern like stops in the development of the language from PV, becoming voiced palatal occlusives *j *jw, whilst in Proto-Northern-Vaniuan (PNV), there appears to have been a merger of *c *cʷ with *k *kʷ, which is unlikely to have occurred were the proto consonants fricatives. As for the sibilants, although Ohanian reflects *sʰ *ṣʰ as aspirated affricates ⟨ř⟩ ⟨ĉ⟩ and PEV appears to show lenition to *s *š, there is little evidence in any of the other descendant languages to support this analysis.
Contrary to the reconstruction of the sibilants as affricates, PNV reflects *sʰ *ṣʰ as glottal consonants *h *hy through debuccalisation, which is an unlikely development for affricates, and Proto-Western Vaniuan (PWV) similarly reflects them as either null due to loss, or as plain sibilants. In addition, Ohanian reflects PV *s as ⟨h⟩, presumed to be a glottal fricative /h/, which, similarly to the PNV glottals, is unlikely to have developed from an affricate. It should also be noted that the very different reflexes of the aspirated sibilants from unaspirated sibilants in so many descendant languages provides evidence against the affricate analysis as it would be expected that affricates would pattern similarly to one another in their development. Aspirated fricatives are highly unstable and thus may be more likely to display a wider array of reflexes in descendant languages, as can be seen, lending stronger evidence to the fricative analysis. Additionally, the fricative reflexes of PV sibilants in descendants like PEV and PSV could simply reflect inherited fricatives rather than lenited forms. A transition from affricate to fricative prior to further changes is possible but considered less likely by most linguists who study Vaniuan languages as this assumes a greater span of time between the breakup of PV and the emergence of distinct daughter languages than is widely accepted, during which time other changes may be expected to have occurred that are not reflected.
PV has a series of postalveolar plosives and fricatives (see previous section for discussion on the exact qualities of *ṣʰ *ṣ) that appear to be in phonemic opposition to their dental equivalents. The most common reconstruction of these consonants is as retroflex /ʈʰ ʈ ʂʰ ʂ/ but there is some debate on the subject. *ṭʰ *ṭ are sometimes taken to be affricates /ʈ͡ʂʰ ʈ͡ʂ/ given the crosslinguistic tendency for retroflex stops to undergo affrication, although this theory faces criticism from the widespread reflexes of *ṭʰ *ṭ as stops in daughter languages where fricatives may be expected given the equally common tendency for affricates to undergo lenition.
However, not all authors even agree that a retroflex articulation is the true nature of PV’s postalveolars. Based largely on evidence from Ohanian and reconstructed Proto-Lake-Vaniuan (PLV), two of the most archaic descendant languages, as well as evidence from several old Southern Vaniuan languages such as Old Zinshan and Cirnese, it has been suggested that these consonants were in fact palato-alveolar /t͡ʃʰ t͡ʃ ʃʰ ʃ/. Both Ohanian and PLV display palatal reflexes of the postalveolars, and Old Zinshan reflects *ṭʰ as post-alveolar /t͡ʃ/. The only Vaniuan descendant branch with direct retroflex reflexes is Northern Vaniuan where *ṭʰ *ṭ show near-universal retention as either /ʈ ɖ/ or /ʈ͡ʂ ɖ͡ʐ/, whilst *ṣ is widely reflected as /ʂ/. Interestingly however, /t͡ʃ d͡ʒ ʃ/ are also attested as developments from PV *ṭʰ *ṭ *ṣ in some dialects, whilst *ṣʰ also often shows a palatal reflex /j/ from PNV *hy [hʲ~ç] which seemingly contradicts the retroflex analysis. It is possible that the quality of the sibilants was more open to free variation in PV, explaining the seemingly disparate reflexes in PNV. Some authors have put forward a theory of substratum influence on particular dialects but there is little evidence to support this theory beyond conjecture.
Another theory suggests that the postalveolars may have been retracted alveolar /t̠ʰ t̠ s̠ʰ s̠/ but there is no direct evidence that favours this interpretation over the others and given that a phonemic retracted alveolar contrast is crosslinguistically much rarer than a retroflex or palato-alveolar one, there is little support for this proposal. It has also been suggested that the distinction may have been lamino-dental /s̻ʰ s̻/ vs. apico-alveolar /s̺ʰ s̺/ (see next section) but this contrast is only widely attested for sibilants, not plosives, and faces similar criticisms to the retracted-alveolar hypothesis.
The only certain PV fricative phonemes are a set of (post)dorsal non-sibilants *x *xʷ *h, whose phonetic realisations are typically assumed to be equivalent to the common notation used: /x xʷ h/. It has been suggested however that *x *xʷ may have been uvular [χ χʷ] based on the reflexes of *x as a uvular stop *q in both PWV and PEV. Glottal *h has also been suggested to have actually been pharyngeal /ħ/, or, rarely, a glottal stop /ʔ/. This latter interpretation is however largely based on the phonology of Ohanian and is widely seen as outdated due to poor supporting evidence from the other Vaniuan descendant branches.
The placement of the PV sibilants in this category is still contested (see previous section) but is largely accepted as valid, with a two-way contrast akin to the plosives in *sʰ *s and *ṣʰ *ṣ. The exact phonetic realisations of *sʰ *s are assumed to be dental [s̪ʰ s̪] but are commonly represented with the broad phonemes /sʰ s/. The dental articulation is often assumed to better contrast with the postalveolar sibilants *ṣʰ *ṣ whose phonetic values vary by reconstruction from palato-alveolar [ʃʰ ʃ] to retracted alveolar [s̠ʰ s̠], although a retroflex articulation [ʂʰ ʂ] is the most common interpretation (see previous section). A competing theory proposes that the dental and postalveolar sibilant sets are instead lamino-dental /s̻ʰ s̻/ vs. apico-alveolar /s̺ʰ s̺/, although there is no direct evidence for this and, as the plosives pattern like the sibilants, it would be assumed that the stops would also show this distinction which is far less likely than the crosslinguistically more common dental vs. retroflex distinction.
Additionally, the consonants *c *cʷ, typically reconstructed as plosives or affricates, are sometimes suggested to be palatal fricatives /ʃ ʃʷ/, /ɕ ɕʷ/ or /ç çʷ/ (see section on plosives) but this interpretation isn’t widely accepted, although it is possible that they may have had fricative realisations in some dialects of PV given the fairly unstable natures of palatal obstruents, and reflexes as sibilants in some descendant languages.
PV had a fairly typical series of nasal consonants, with labial *m, dental *n, and velar *ŋ corresponding to the labial, dental and velar series of stops respectively. There were no corresponding nasals for the postalveolar or labiovelar series, which has led some authors to suggest that these consonant series may have evolved later, particularly those who support the Ryamaian macrofamily hypothesis that links Vaniuan to several small languages spoken in Northern and Central Miraria. Evidence for this however is scant and it is possible that these consonants once had corresponding nasals but their unstable natures led them to merge with other nasals, or that corresponding nasals simply never existed.
In addition to its unusual obstruents, PV had an unusual approximant series with two phonemic laterals: alveolar *l and velar *ł, but no semivowels. Some authors reconstruct allophonic semivowels [j w] when *i *u occured in hiatus with other vowels, but these interpretations are largely ignored in modern analyses due to conflicting diachronic evidence, especially from Archaic Ohanian which best preserves the reconstructed syllable structure of PV and shows no evidence of semivowel allophony, instead having hiatus between vowels.
The Vaniuan daughter languages display a number of similarities in the changes undergone by the PV consonants:
- Proto-Northern-Vaniuan and Proto-Bodnic-Cirnic both shifted the plosive series from an aspiration distinction to one of voicing: *pʰ *p > *p *b. The sibilant outcomes are slightly more varied as PNV shows debuccalisation of the aspirated sibilants and retention of the unaspirated ones: *sʰ *s *ṣʰ *ṣ > *h *s *hy *š, whilst PBC simply shifts to a distinction of voicing: *sʰ *s *ṣʰ *ṣ > *s *z *ṣ *ẓ.
- Proto-Eastern-Vaniuan and Proto-Western Vaniuan also display a transition from a phonemic distinction of aspiration to one of voice, but in their cases voicing was not a complete chain shift but instead part of a larger overall change in the qualities of obstruents that also involved lenition to fricatives.
- In PEV: *pʰ *tʰ *ṭʰ *kʰ *kʷʰ > *f *θ *th *x *kh, and *p *t *ṭ *k *kʷ > *b *d *t *k *k. Additionally, *p sporadically lenites to *v, and *kʷ lenites to *w when between vowels. *c *cʷ also undergo voicing to *j *jw.
- In PWV: *pʰ *tʰ *ṭʰ *kʰ *kʷʰ > *p *d *r *k *f, and *p *t *ṭ *k *kʷ > *b *t *d *g *g. Additionally, *pʰ *tʰ become *y *r respectively at the start of a word, and *tʰ lenites to *z between vowels. PWV *b *d further lenite to *w *y respectively and *pʰ *kʰ *kʷʰ all show voicing developments in certain vowel-determined environments, with *pʰ > *b before rounded vowels, and *kʰ *kʷʰ > *g when preceding a sequence of a vowel and alveolar lateral *l. Intervocalic lenition further changed the qualities of the stops in Late-PWV: *b > [β]. *c *cʷ show development to *y *c in Early-PWV, with further lenition of *c to either of *y or *s in Late-PWV.
- Ohanian and the Lake Vaniuan languages both show palatal reflexes of the PV postalveolars *ṭʰ *ṭ *ṣʰ *ṣ, and affricate reflexes of the PV aspirated sibilants *sʰ *ṣʰ.
- Ohanian also shows similarities to PSV in that both delabialised the PV labiovelar consonants *kʷʰ *kʷ *xʷ > *kʰ *k *x (further > *k *g *x in PSV). The change in PSV was however accompanied by a simultaneous shift of the PV velars *kʰ *k *x to palatals *c *j *ç, whereas in Ohanian the two sets simply merged. PWV also shows delabialisation of *kʷ > *g, but *kʷʰ *xʷ are reflected as labial *f *w respectively.
- A widespread shift in the qualities of the two approximants *l and *ł. Broadly speaking, there are 3 main outcomes for *ł: in Ohanian, PLV, PEV and PSV, *ł becomes a central velar approximant or fricative, in PNV and in some environments in PSV, *ł becomes a labiovelar approximant, and in PWV, *ł becomes an alveolar lateral. *l on the other hand is largely retained in descendant proto-languages, with the notable exception of PWV where the shift of PV *ł > *l is accompanied by PV *l > *r.
PV displayed a fairly typical seven-vowel system but with two levels of vowel harmony: a front-back distinction between *i *e vs. *u *o, and a height distinction between *i *u *e *o vs. *ɛ *ɔ *a, although both processes had become unproductive by the time PV split into its descendant languages. Harmony was progressive, only occurred in suffixes, and was determined by the last vowel in a word; roots themselves didn’t always obey harmony. Some prefixes also appear to have obeyed regressive harmony but this is only reflected in some daughter languages so it is unclear whether this is an archaic feature or an innovation of those languages.
There were no phonemic long vowels or diphthongs but sequences of vowels in hiatus could occur, which for sequences of the same vowel may have been phonetically realised as long vowels. It is possible that these were separated by glottal stops based on evidence from Ohanian, and that these stops were simply lost in the other descendant languages, but this remains unclear (see previous section on consonants).
Proto-Vaniuan had a (C)V(C) syllable structure, that is a vocalic nucleus with an optional consonant onset and/or coda. Some linguists believe however that PV had a CV(C) structure, that is a vocalic nucleus with an obligatory consonant onset and optional consonant coda. According to this theory, any vowel-initial syllables were actually preceded by a glottal consonant *’ or *h₁ (see section on consonants) but evidence for this is scarce and based largely on Ohanian and Proto-Northern-Vaniuan which may have innovated their glottalic onsets independently of one another, and separately from the other Vaniuan descendant branches, which appear to lack this feature.
All consonants could occur in onset position but *h could not occur in coda position.
No initial or final consonant clusters were allowed in PV so words could only begin or end with a maximum of one consonant. Within words however, clusters of two consonants were permitted including geminated consonants as in *całłao “eagle” and *ninnaŋ- “to be ill”. Whether aspirated consonants could be geminated however is unclear. There are no definitive reconstructions of aspirate geminates as the descendant languages variously reflect clusters, singleton consonants, or geminates. Texts indicate geminated aspirated plosives were present in Ohanian as in aĉĉim- “attractive” but the cognate term in Proto-Eastern-Vaniuan reflects a singleton plosive *athëm-. Unusually, nasals in clusters do not appear to have assimilated to following consonants given evidence from Ohanian that regularly displays *m as w and *n *ŋ as y when preceding a consonant, regardless of that consonant’s place of articulation, e.g: *cimṭɔŋulo “hero” and *manaŋseo “mother” > suĵug̃u(l) and mane₁šə from older siwĵug̃ul and manayše.
Within words and phrases, PV appears to have allowed long sequences of phonetic vowels without intervening consonants, pronounced instead with hiatus or possibly a glottal stop *’. It is possible that stress placement may have helped speakers track the timing when the vowels were identical, but as stress in PV is typically reconstructed as having been fixed on the initial syllable, this is unlikely. Examples of vowels in hiatus occur in words such as *iołɛło “amount; consideration; measure” and *pʰaniuŋo “family”. Sequences ending in *o were the most common as the surfix for the direct case singular number was *-o which was invariably suffixed after noun stems, regardless of whether they ended in a consonant or vowel, e.g: *haetʰeo “air; wind” and *kʰilcoo “beast; monster”. In many descendant languages, these sequences developed into diphthongs, such as in the PBC reflex of “air” *ḥayte with a diphthong *ay from the PV sequence *ae, or they became long vowels, such as in the Late-PNV reflex of “beast”
*kilġō with a long vowel *ō from the PV sequence *oo. In others, such as Ohanian, PEV, and sometimes PNV, these sequences were reduced to single vowels: *pʰaniuŋo > Ohanian panu(g̃) (from older paniʾug̃), PEV *fanuŋ, and Late-PNV *ponų̈.