User:Soap/PC

The PC languages were spoken by the aboriginals of the northeast of Rilola, and possibly also the northwest. They had been diverging for about 15000 years by the time of contact with Play speakers, but it is likely that one group overtook the rest at some point, perhaps multiple times, rather than having them diverge from each other for all 15000 years.

This may replace Haswaraba, despite the fact that cultures are in different places. See Outer Poswob languages for details, though I may have lost all records of the sound changes. It is possible that they are [1], and if so the phonology I was using for the parent language is very different from what it is now (it even had /f/).

The name PC describes its shifting the labiovelars to bilabials and the palatals to postalveolars, leaving just a single velar series. Note however that SE Laban, the branch that leads to MRCA, also does this.

Source phonology

Bilabials:      p   b   m   
Linguolabials:  þ   ð
Alveolars:      t   d   n   l   r   s
Postalveolars:          ň       ř
Palatals:       ć       ń   y
Velars:         k   ġ   ŋ  (Ø)      x
Labiovelars:    kʷ  ġʷ  ŋʷ (w)      xʷ
Glottals:                           h

The vowels were probably at least /a i u ə/ with two or more tones. There were no long vowels.

Primordial to Cave of the Swimmers

1. Syllable final -uŋ shifted to -um.
2. The schwa vowel ə shifted to ʲi when following a palatal, postalveolar, or linguolabial consonant.
3. The postalveolars ň ř shifted to n r.
4. The palatals ć ń shifted to š ň unconditionally.
5. Remaining pure i shifted to ʲi.
6. The schwa vowel ə disappeared in word-initial position.
7. The linguolabials þ ð shifted to p b.
8. The rounded back vowel u shifted to ə when NOT following a labialized consonant.
9. The labialized consonants kʷ ġʷ ŋʷ shifted to p b m; however xʷ hʷ shifted to f.
10. The glottal fricative h performed, shifting to š before a front vowel, to f before a back vowel, and disappearing to Ø otherwise.
11. The schwa vowel ə shifted to i, creating a three-vowel system.

SYLLABARY (Swimmer)

This format is used to highlight the CV gaps, which will lead to vowel collapse in the next stage.

 pa   pi   pu
 ba   bi   bu
 ma   mi   mu
 fa   fi   fu
 ta   ti
 da   di
 na   ni
 la   li
 ra   ri
 sa   si
      či
      ǯi
 ňa   ňi
      ři
 ša   ši
      ći
      ǵi
      ńi
      śi
 ya   yi  (yu)
 ka   ki
 ġa   ġi
 ŋa   ŋi
 xa   xi

Swimmer Stage II

1. The vowels i u shifted to ə before a nasal coda.
2. The sequences pu bu mu fu (when on a low tone) shifted to p b m f. If another consonant followed, they assimilated in place to that consonant, with f becoming aspiration. They may have nonetheless retained a labial element.
3. Any Cu/f/ also became aspiration. This happejned only for laibals.
4. The sequences ći ǵi ńi śi shifted to ć ǵ ń ś. It is likely that these too assimilated in POA but may have retained some palatal element.
5. The sequence ři shifted to r. It did not retain any palatal element.
6. The voiced aspirate stops bh dh ġh shifted to ph th kh. Others (ǯ etc) shifted likewise. NOTE: it is possible that these became voiceless fricatives instead, but if so, this shift should be further down, and the voiceless aspirates will still appear from their constitutents.
7. The velar fricative x shifted to h.
8. The consonants ǵ ǯ shifted to ž.
9. The consonants ś ć č (where they remained) all merged into š, but a phonemic geminate form resembling [č:] may have persisted, and the fricative [š:] would need a different analysis.
10. The voiced stops b d ġ (and any others) shifted to fricatives v ð g except when in clusters. A syllable coda, even at the end of a word, probably will behave as a cluster for this shift.
11. The palatal nasal ń shifted to ň.
12. All geminate stops became voiceless. Stop+nasal combinations like bm became plain voiced stops like b (this may even shift if the original is /pm/). The geminate /ǯ/ also existed from combinations like -giň-.
13. Then, ž shifted to z. (It is possible that /ð/ is really /z/, in which case this shift would not happen.)
14. Word-initial geminates simplified to singles. Many of these were labialized or palatalized, and probably none were plain.

It is possible that the plain voiced stops had become fricatives by this point and that clusters like /bb bm/ became the new voiced stops. However the lists currently do not show this option.

At this point the allowable codas were

p  b  m  f
   r  n     z
      ň  š  
      ŋ

But it is possible some shifts were missed above. For example, losing /ŋ/ would free mid vowels from needing a coda.

Consonant inventory (Swimmer II)

The full consonant inventory, assuming the voiced fricatives survived, and voiceless fricatives developed from /ʰ/ with these, would be

Labials:        pʰ  p   b   m   f           v
Alveolars:      tʰ  t   d   n   s   l   r   z
Postalveolars:          ǯ   ň   š           ž 
Palatals:           ć:         (ś:) y
Dorsals:        kʰ  k   ġ   ŋ   h           g

Voiceless geminates existed, but there were few if any voiceless aspirated geminates.

Swimmer Stage III

OPTION 1:

Since geminates were marked even between vowels, these now also simplified, leaving labialized and palatalized consonants in their wake. This implies the loss of codas generally, and that syllables will be front-loaded ("twa").

OPTION 2:

Geminates and other clusters can be kept, making the syllable CVC, but the three-way contrast between pal/plain/lab must reduce to two since otherwise there will be no plain geminates. This implies that syllables will be of the "tau" type.

These are not two languages, but two options for the same language. It is possible nonetheless that they will branch at near this point and end up taking these two paths but not at this precise point.

Stage III Option Ia

This assumes the mid vowel was /ə/.

1. Nasal+stop clusters become prenasals.
2. Any m+C cluster labialized the following consonant; any ń+C cluster palatalized the following consonant.
3. Any p b before a consonant labialized that consonant and then disappeared. There may not have been any /b/ in this position.
4. Additionally, p b m shifted to pw bw mw, though they likely shifted back under some conditions.
5. Any š z before a consonant palatalized that consonant and then disappeared. (This is a different pattern than the other branch.) This also canceled any labialization.
6. Any remaining non-nasal codas disappeared, lengthening the vowel if possible.
7. All nasal codas shifted to n.
8. The sequences wə wi shifted to o u.
9. The schwa vowel ə shifted to e. This creates /pe be me/ only when they had been palatalized.

Stage III Option Ib

Another way to do this is to retain the glides, and have an inventory like /yi i wu/ on the top row. This might eventually create a four-vowel system, as in Ogili, where two gaps in a five-vowel system closed up to make four full vowels.

This assumes the mid vowel was /ə/.

1. Nasal+stop clusters become prenasals.
2. Any m+C cluster labialized the following consonant; any ń+C cluster palatalized the following consonant.
3. Any p b before a consonant labialized that consonant and then disappeared. There may not have been any /b/ in this position. In this option, the plain labials did not become rounded.
4. Any š z before a consonant palatalized that consonant and then disappeared. (This is a different pattern than the other branch.) This also canceled any labialization.
5. The sequences wə wi shifted to wo wu. If this language follows the pattern of Ogili, this will later shift to /u wu/.
6. Word-final nasals and all other codas were deleted. Prenasals were retained.

The vowel system at this time was /ya a wa/ + /yə ə/ + /yi i/ + /wo wu/, the last of which might shift to /u wu/. It is however unlikely that this would shift further to /yu u/ as this would mean that /yu/ would be about as common as plain /u/, whereas for the front vowels the palatal forms would be much less common. It is possible that inherited /ň š/ would survive labialization.

Stage III Option II

This list assumes the mid vowel was /ə/.

1. Any ə i before a labial coda shifted to o u. This excludes the aspiration that came from earlier /f/.
2. Any z in the coda or adjacent to a voiceless consonant shifted to s. In fact, it probably all shifted either to s or to something else such as assimilation.
3. Any š in the coda shifted to s. THERE IS A CHANCE THIS COULD SURVIVE, but consider first whether it will affect preceding vowels or not. For example, it could actually labialize instead of palatalize.
4. Any ň in the coda shifted to n, if it was not already homorganic. This helps the mid vowel phonemicize.
5. The codas p b shifted to ʔ :, meaning a high tone and a long vowel. In Lava Bed languages, these are usually both considered tones. It is possible that the word-internal clusters continued to be written as such.
6. Any remaining coda f shifted to :ʔ, which is a long high tone, or a rising tone. But remember that most of this shifted to aspiration very early, and that this may have only existed in word-final position.
7. All other codas probably became toneless. (But remember that Leaper distinguished a reduced tone set on closed syllables.)
8. The schwa vowel moved to e.

The allowable codas were now /m n ŋ s r/.

The consonant inventory was

Labials:        pʰ  p   b   m
Alveolars:      tʰ  t   d   n   s   l   r    
Palataloids:                ň   š  (y) 
Dorsals:        kʰ  k   ġ   ŋ   h 

And the five vowels were /a e i o u/, but there were still major gaps for certain combinations of CV and VT (tone).

It is possible that the aspirates will shift to fricatives, while /h/ disappears, and then the new /x/ shifts to /h/.

Grammar

A closed class of adjectives for humans existed, where the stem of the adjective was surrounded by a circumfix marking gender and age.

THe gender prefixes t- l- probably trace all the way back to Primordial, since they are needed for this system, and it is unlikely that they had had a vowel padding. This means that vowel-initial adjectives existed in this closed class.

There may be a plural suffix -u.

For pronouns, see User:Soap/scratchpad#Gluons. There may have been a regular pronominal accusative suffix -l (despite it looking like a prefix) which passed into MRCA, and may or may not have been the sole marker in Primordial. Common nouns used -i by the time of MRCA but this may have been much more complicated and unpredictable in Prim.

The 1st and 2nd person identity/possessor suffixes are -m -t, but they may not have been suffixes originally. The corresponding agent prefixes appear to have been n- G-, where G is any consonant that can turn into MRCA /g~Ø/ and might most likely have been originally /h/.

Note that it's not clear the proto-language allowed final /-m -t/, so these might have been CV suffixes even if at a still earlier stage they really were /m t/.