Old Nambāno

Old Nambāno, also known as Early Nambāno or Proto-Nambāno, is the ancestor of most language varieties within the Nambāno dialect cluster, believed to be the oldest Koyanic language spoken in Qnōp'.

Palatal obstruent
The palatal obstruent *j was likely a fricative [ʝ] or affricate [ɟʝ], though may have already been pronounced as an approximant [j] as it is most typically in Nambāno.

Vowels
Long vowels derive primarily from compensatory lengthening via two sources:


 * Loss of coda *h, itself a merger of older Proto-Namb *f and *x.
 * Loss of coda *r.

Five of the six vowels additionally had nasalised forms, *ã *ẽ *ĩ *õ *ũ. These resulted from loss of nasals at the end of words and before older fricatives and nasals, also lengthening the vowel, with the nasalised form of *ɨ merging with *ã. The bilabial nasal *m, however, was retained in these environments. Vowels before coda *n were likely also pronounced with nasalisation, but were not lengthened and simply represent allophones of the short vowels.

Only four sequences of two vowels are found in Old Nambāno, *ei, *ou, *io and *uo (or *ɨo). These may have acted as diphthongs, evidenced by the fact that they are usually reflected by single vowels in descendent languages.

A table of vowel reflexes in Nambāno languages is shown below:

Phonotactics
Old Nambāno had a maximal CCVC syllabic structure, with various constraints placed on the onset cluster and coda consonant.

Onset
Medial nasals in the onset could not have the same place of articulation as a preceding plosive, thus *bn- and *tŋʷ- are allowed onset clusters while *bm- and *tn- are forbidden.

The velar consonants *g *k *h *ŋ could occur preceding any vowel except *u, where they are neutralised into a labialised velar + *ɨ. These exact pronunciation of these syllables is not reconstructible, e.g. it is not known whether *gʷɨ was actually [gʷɨ] or [gu], but it is reconstructed as such due to the later nambāno form gua [gʷa] reflecting the regular *ɨ > /a/ shift seen elsewhere.

Coda
Obstruents could occur in coda position, though not at the end of words, but the labialised velars *gʷ *kʷ *hʷ *ŋʷ were not allowed in the coda.

Consonant clusters
Consonant clusters most typically arise at morpheme boundaries, since most underlying stems end in a consonant which resolves into one of the five rhymes when placed before an obstruent. Various morphophonological processes of assimilation limited the possible *-BB- clusters that were possible that could occur across syllable boundaries. These processes can be sorted into progressive and regressive assimilation:


 * Progressive assimilation
 * The velar fricative *h(ʷ) assimilates to any preceding consonant, e.g. *-th- *-rh- *-ŋh- > *-tt- *-rr- *-ŋŋ-.
 * *hʷ additionally causes a preceding bilabial consonant to shift to a labialised velar, thus *-mhʷ- *-bhʷ- *-phʷ- > *-ŋŋʷ- *-ggʷ- *-kkʷ-.
 * Regressive assimilation
 * Plosives assimilate to following nasals of the same place of articulation, e.g. *-pm- *-dn- *-kŋ- > *-mm- *-nn- *-ŋŋ-.
 * Labial stops *b *p *m assimilate to following velar stops, e.g. *-bg- *-pk- *-mgʷ- > *-gg- *-kk- *-ŋgʷ-.
 * Lingual nasals *n *ŋ assimilate to the place of articulation of any following consonant, e.g. *-nk- *-ŋn- *-ŋb- > *-ŋk- *-nn- *-mb-. Note that *m only assimilated to velar consonants as per the rule above, thus sequences like *-mt- and *-ml- were allowed.
 * Lingual plosives *d *t *g *k assimilate to the place of articulation and voicedness of any following plosives, *s and *j, e.g. *-dp- *-gs- *-tj- > *-pp- *-ss- *-jj-.

A full table of possible intrasyllabic clusters of two consonants subject to these rules is shown below:

Nouns
Each noun was either countable or uncountable. Countable nouns in the absolutive and dative cases distinguished singular, plural and numberless forms while uncountable nouns were always numberless. This separation ultimately derives from Proto-Koyanic's two animacy categories, though the countability of a noun was fairly unrelated to its animacy by the time of Old Nambāno.