User:4pq1injbok/DLNAF/Diachrony

Sound changes

This is meant to be a chronologically correct list, at least in important detail. At present we focus on vowels. The intermediate displays may occasionally separate allophones, but aren't hypernarrow transcriptions.

• Nasalisation. (Shared with Pot.?) Coda [-n] loses closure, but retains its slot in the timing. Perhaps concomitantly, vowels before coda nasals are nasalised. The remaining coda nasal, [-m], no longer makes any place contrasts, so soon thereafter assimilates in place to following segments. It eventually relaxes to [-n] finally.

In the resulting system, /Ṽː Ṽn/ seem to be interpreted as two parallel sets of long nasal vowels or diphthongs with different degrees of nasality or closure.

[ã] quickly passes to [ɑ̃]. (This might be shared even more broadly.)

i ĩː ĩn	u ũː ũn
a	ɑ̃ː ɑ̃n

• Decay of weak consonants. (Part-shared with JS?) [w ð ɣ] are lost in an unstressed syllable. Like several of its congeners, pre-DNLAF segments words into trochaic feet from the right edge, with primary stress on the head of the rightmost trochee: there must have been secondary stress on the heads of the other trochees.

Vanishing [w] (but not surviving [w]) colours adjacent [a] to [ɑ], which is immediately phonemicised. That this originally subphonemic effect became fixed as relevant is probably due to the preexistence of [ɑ̃].

i ĩː ĩn	u ũː ũn
a	ɑ ɑ̃ː ɑ̃n

• Syneresis of sequences of identical V. Stress is not reassigned; syneretic vowels retain their (primary or secondary) stress.

i iː ĩː ĩn	u uː ũː ũn
a aː	ɑ ɑː ɑ̃ː ɑ̃n

• Raising of [a] to help maintain the [a] vs. [ɑ] contrast. (Shared with Tty.?)

i iː ĩː ĩn	u uː ũː ũn
ɛ ɛː	ɑ ɑː ɑ̃ː ɑ̃n

• Unstressed vowel weakening. (Several partial parallels.) Vowels in unstressed syllables shorten: the short vowels become extra-short and the long ones (unmarkedly) short. Nasalisation in unstressed syllables may also be lost already here.

In hiatus, the extra-short vowels become glides, except when a high vowel is adjacent to a non-high extra-short vowel; in this case the high vowel becomes the glide (unless it is long) and the non-high vowel is restored to (unmarkedly) short. (At some later point, [iw] also readjusted to [ju].)

ĭ i iː ĩː ĩn	ŭ u uː ũː ũn
ɛ̆ ɛ ɛː	ɑ ɑː ɑ̃ː ɑ̃n

• Palatalisation of consonants is prompted by centralisation of extra-short vowels displacing the [ĭ] vs. [ŭ] contrast. Succeeding high front vowels and glides trigger palatalisation; these are still exactly the continuants of Proto-Dumic *i.

i iː ĩː ĩn	ɨ˘	u uː ũː ũn
ɛ ɛː	ɜ̆	ɑ ɑː ɑ̃ː ɑ̃n

• High vowel lowering appears to have begun in the two classes of nasalised high vowel, where it was natural, but was extended by a fundamentally phonological analogy to all the long high vowels, resulting in the very odd situation that long but not short high vowels lowered.

i	ɨ˘	u
eː ẽː ẽn		oː õː õn
ɛ ɛː	ɜ̆	ɑ ɑː ɑ̃ː ɑ̃n

• Loss of [ɨ˘], and subsequent restitution of [ɜ̆] to [ɜ] > [ɛ].

The fall of [ɨ˘] created a broad class of coda consonants, at which point the interpretation of the /Ṽn/ elements as diphthongal gave way to the simpler interpretation that these nasals /n/ were coda consonants. Nasalisation before arose before the new coda nasals as well, rendering the nasalisation in the old /Ṽn/ elements non-phonemic.

i	u
e eː ẽː	o oː õː
ɛ ɛː	ɑ ɑː ɑ̃ː

• Loss of nasality, except allophonically, soon drew to completion.

i	u
e eː	o oː
ɛ ɛː	ɑ ɑː

• Merger of [e] into [ɛ]. The contrastive load of the distinction had been light, consonants before [e] but not before [ɛ] being palatalised.

i	u
	o oː
ɛ ɛː	ɑ ɑː

• Loss of length.

i	u
	o
ɛ	ɑ

• Restoration of the two-height system, into which [o] didn't fit; it broke into a diphthong [ɑw], with the [w] suppressed before a labial coda. Note that all instances of [o] had (primary or secondary) stress.

With the target of the /ɛ/ phoneme moving to a true low position, and [ɑ] taking on rounding to amplify the contrast with the new [a], we arrive at the synchronic vowel inventory.

i	u
a	ɒ