User:Soap/Birch

This is not the original Birch, but ratehr a languyage that ios spoekn in Mevumep. This thus derives ultimately from User:Soap/PC but there is a gap of probably 11,000 years.

Phoneme inventory

Rounded bilabials:    pʷ  mʷ  ɸ   w         /a i u ə/
Spread bilabials:     p   m                 /a     ə/
Linguolabials:        pʲ  mʲ                /a i u ə/
Dentals:              ṭ   ṇ                 /a   u ə/
Rounded alveolars:            sʷ      rʷ    /a i u ə/
Plain alveolars:      t   n   s   l   r     /a     ə/
Palatoalveolars:      č   ň   š   λ   ř     /a i u ə/
Palatals:                     ś   y         /a i u ə/
Palatovelars:         ć   ń                 /  i    /
Velars:               k   ŋ   h  (Ø)        /a     ə/
Labiovelars:          kʷ  ŋʷ                /a i u ə/

All four vowels could occur long, but unlike Play there were no superheavy syllables.

The codas were /-l -n -t/, but in some environments they combined with vocalic glides. The labialized forms were /-w -ŋʷ -kʷ/ (meaning thus that final /u/ and final /ul/ had the same raelization), and the palatalized forms were /-λ -ň -č/. However, there is a good chance that this coda λ might have shifted to /l/ already in the parent language, or perhaps sometimes to /l/ and sometimes to /j/, while /č/ could have shifted all the way to /s/.

There may have been retention of syllabic /l n/ instead of shifts to schwa.

There may have been a sequence /hi/ from a primordial /fi/ (seaparate from the labialized phi).

Possible shifts common to all branches

These would be kept separately for simplicity's sake, as many are only marginally phonemic.

Avoid relying on vowel elision, since this would make it more difficult to explain why the allowable codas are already reduxced to just /n t/ (and maybe /l/) in Proto-Birch.

1. The alveolar approximant r shifted to d in word-initial position and after /n/.
2. All coda l disappeared, becoming a vowel lengthener in most cases and a diphthong former in others. (Note that it disappears in the two main branches, behaving like a long vowel in both cases even though this behavior is different for each branch).

Commercial (Mercantile) branch

This moves towards a sesquisyllabic root strcture since the functional load of unstressed vowels was very low. It remains as a single language for a long period of time, so there are no languages that complete the shift only halfway. SImilar shifts in other branches may occur independently.

This becomes similar to contemporary Repilian and Moonshine in many ways, and forms the basis for a military cipher used by a few young Lilypads whose native language was Play and who had learned a small amount of Moonshine throuigh diplomacy.

1. If syllabic l n still existed, they shifted here to schwa.
2. The bilabial fricative ɸ shifted to f; it may have remained as a bilabial after /m/ long enough to reflex as /p/ after vowel nasalization, however.
3. Probably, all codas disappeared: -n caused long nasalized vowels, -l created long oral vowels, and -t created a checked tone comparable to Gold's, though this may have been rare and it may never have contrasted with a consonant sequence. Short nasalized vowels may have existed allophonically before nasals in open syllables, but were not yet phonemic.
4. The sequences ʲa ʷa shifted to ʲe ʷo. The coarticulations remained, and so the shift was not yet phonemic. NOTE: It may be best not to do this at all, since the /e o/ still have to delete in unstressed position, which means that they have to be assigned a "major" vowel to pattern with, and the most harmonious choice would be /a/. At the very most, this was simply an allophonic shift probably confined to stressed syllables.
5. The palatoalveolars č ň š λ ř shifted to c n s l z.
6. The labialized alveolars sʷ rʷ shifted to ṣ ṛ (pronunciation unclear; possibly velarized). If /ř/ becomes a fricative, /ṛ/ would simply shift to /r/.
7. Labialization disappeared before /u/; that is, ʷu became u. It probably disappeared before /i/ as well, but the resulting vowel would be [y], and so it is notationally convenient to continue to write it as /ʷi/. Since even /ʷu/ was [u] to begin with, this is more of a spelling change than a sound change.
8. Unstressed ə shifted to i. Some stressed /ə/ might also shift.
9. The unstressed vowel sequences ʲi i ʷi (including from earlier /ə/) shifted to y Ø w unconditionally. The coarticulations thus became full consonants. This allows for separate reflexes later on for pairs such as /kʷ wk/. These would soon cross syllable boundaries, phonemicizing the /e o/ of the earlier shift, but it happened conditionally and stepwise.
10. The sequences ʲə ʷə shifted to ʲi ʷu. THIS IS NOT BEING DONE. IT MAY JUST BE THAT ALL /ə/ > /i/. This time, the labialization before /u/ was audible. Some stressed ə became i here, such as after /t/.

    Remember that new labialization occurs later, so there's no need to make new /ʷu/ here.

11. The true palatals ć ń ś shifted to č ň š; of these, only /š/ occurred before all three vowels.
12. The only unstressed vowel, a, shifted now to ə, which probably no longer occurred in stressed position; it came to be seen as a consonant.
13. A final schwa aspirated a preceding stop. A following stop also became aspirated; this led to things like /tək/ > /tʰəkʰ/, and sometimes applied across word boundaries with constructions like /-t ək-/ > /-tʰ əkʰ/, where the schwa later disappeared.
14. Probably, a nasal touching a schwa in either direction becomes syllabic. This may apply to liquids too.
15. Then, ALMOSTG unconditionally, ʷi i > i e.
16. Likewise, almost unconditionally (but maybe with different exceptions), ʷa > o. Thus there were five vowels.
17. The palatalized labials pʲ mʲ became tʷ nʷ when bordering a /w/ in either direction.
18. Now the stranded extrasyllabic y w pre-onsets moved to the proper onset. That is, /wk/ > /κ/. ετ

Birght Colors branch

This branch evolves early on into something like a rich Old Japanese (similar, but more vowels, more conosnants, and more closed syllabes).

1. Stops become voiced after any coda /l n/.
2. Coda l became a vowel lengthener, except before another /l/ and possibly before all coronals. In the latter case, the coronals might become labialized to (partially) fill a gap. This means that voiced stops (aside from /d-/) now occur only after long vowels and nasals, but since there were preexisting long vowels before the shift, they are marginally phonemic now.
3. The sequence au shifted to ō, because coda -u was phonemically /lʷ/. Other shifts may foillow.
4. The vowel sequences wa wə shifted to ʷo o.
5. Then, ə shifted to e.
6. Then, wi wu delabialized to i u. There was no [y] because it was filling a gap of *[i].
7. FInally, ʷo o shifted to o ə. (This could also have happened in the beginning if the /ə/ > /e/ shift is worded differently.)
8. If syllabic l n still existed, they shift here to u ə.

Northern Branch P

This is the most conservative branch of the family. It fills a gap early on and thus becomes stable.

1. The labialized consonants kʷ ŋʷ shifted to pʷ mʷ before /a ə/ (the "shadow" shift), and otherwise to k ŋ (thus filling the gap). At the surface level, the latter shift was about IPA [y] shifting to [i] and IPA [u] staying as it was; that is, kʷu > ku was just a matter of analysis.
2. Meanwhile the labial fricative ɸ shifted to h before any /i u/.
3. The labial approximant w became b before /a ə/ and possibly in limited circumstances also before /i/ or /u/ (note that it is very common before these high vowels but was usually not audible, as it simply produced a rounded vowel; the exceptions would be in vowel sequences).

Branch Y

If this exists, this branch groups like rows of C and Cʷ together, seeing them as having different vowels, and it is thus somewhat like Old Japanese. The palatalized consonants may participate too but velars and palatals would not group together. However it may not be feasible.

Various pre-parent shifts

These are not inchronological order, and cover at least 11000 (perhaps as much as 17000) years.

Probably initial /p t k/ to /h/ at least conditionally, if assuming the feminine gender marker /h-/ is from original /t-/. The shift could be /p t k/ > /f s x/ > /h/, with each step conditional. Semantic shift might also be involved to get the final meaning to be feminine only. It may have only been the aspirates that shifted, especially if they became more common before the shift took place.

Loss of all voiced stops between vowels and perhaps also initially.