Fojy

• Fojy * is an empire straddling the boundary of the tropics and subtropics, and consists of 7 states.

language

This is the source of the substratum of Dreamlandic, and of languages that were spoken as far north as 46N in territory that changed hands several times, ending up split between the Ghosts and Moonshine.

The early stop system was /p t k q pʷ tʷ kʷ qʷ/. However, the q's were never actually uvualar ... they varied depending on the vowel, unlike all other conosonants. They have variable reflexes, even within the same language, depending on whether they were analogized with one of the traditional stops or not. The Dreamlandic branch might shift all to /p pʷ/.

Most sound changes are on paper. Others are:

Shared changes

1. Before a high tone, the fricatives g h gʷ hʷ were lenited to Ø Ø w w. Before a low tone, these fricatives merged to q q qʷ qʷ. This /q/ was not a proper phoneme, but rather was determined by the surrounding vowels, and therefore could take on different values in different forms of the same word.
2. Before a low tone, the lateral approximant l shifted to r. This did not affect the dental form (if there was one).
3. Before a high tone, the voiceless stops pʷ p ṭ t tʷ k became the fricatives fʷ f þ s hʷ h.
4. Voiceless prenasalized sequences were reanalyzed as nasal + fricative. Voiced prenasalized sequences were reanalyzed as nasal + stop. Thus they were no longer phonemic. Nevertheless, they continued to occur in the syllable onset rather than splitting between the onset and the preceding coda.
5. The emphatic stops pʕʷ pʕ ṭʕ ḳ shifted to kpʷ kp kṭ kk "when the environment would support it".
6. The voiceless alveolar stop t affricated to c before the high vowels [i ə u].
7. Before the vowels a i (and probably /e/), the consonants p m f t n s all became palatalized to pʲ mʲ fʲ č ň š. The velars /k ŋ h/ became palatalized, but the spelling was retained. Likewise, palatalization of /l/ was not marked in spelling.
8. The rounded alveolars tʷ nʷ became kʷ ŋʷ.

   even though they had /q/ "on their side"? Possibly shift all nine, and let the daughter languages merge them in various ways.

9. The high central vowel ə shifted to i.

   Note that this shift opens up a three-way contrast of /t/~/č/~/k/ before [i], and the following shift adds [ə] and [a]. However, there was only a two-way contrast before [u]. This may be seen as related to the labialized phoneme /kʷ/ and might even trigger the loss of /u/ in some daughters.

10. The mid vowels e o shifted to ə a. This did not affect, and was not affected by, whether the previous consonant was labialized or not; the respective allophones shifted in tandem.
11. Tones were eliminated.
12. Syllabic consonants were resolved as sequences of /i/ + consonant, and assimilated to the place of articulation of a following consonant.

Thus the vowel inventory at this time was /a i u ə/, and the conosnants were:

Rounded bilabials:     pʷ  mʷ  fʷ  w
Spread bilabials:      p   m   f  (Ø)
Dentals:               ṭ   ṇ   þ   ḷ
Alveolars:             t   n   s   l   r
Postalveolars:         č   ň   š  
(Palatals:             ć   ń   ś        
Velars:                k   ŋ   h      
Labiovelars:           kʷ  ŋʷ  hʷ

unnamed branch

/t/ > /k/ unconditionally, followed by /ṭ ṇ/ > /t n/. However the original /n/ did not shift. This is similar to what happened in Andanese.

Dreamlandic branch

/k/ > /t/ unconditionally. This happened only after the wider change of /k t/ > /č/ when touching [i]. This /č/ may or may not have been a true phoneme, depending on whether there was ever [ya] or just [a].

In one language of this branch, labialization falls apart and produces a mix of labials and plain sounds of the original type. This is similar to Pabappa, but separated by thousands of years and thousands of miles. This language may also have a voicing distinction in stops, unlike most of the rest of the Fojyan languages.

In another branch, an Ogili-like setup appears, but with the extra vowels being /ʷa ʷi ʷu/ etc, and the true vowel inventory remaining small.

Primordial /f/ > /pʲ/, which fronts the "a" vowel to [æ]. <--- PROBABLY BAD (but even so, c.f. Marshallese)

Perhaps instead, /f/ > /p/ where preesxisting /p/ was palatalized, and survives intact elsewhere (later to /h/).

history

Aboriginal population density very low despite climate, hence Mumba settlers quickly become majority and are themselves seen as aboriginals by 1900 ad. They identify as Labans.

Set up "the Peace Zone", a supranational government of nations that would refuse to partake in war. However, unlike Paba and Moonshine, they maintained a strong military for defense.

tribes

Mumba are blonde people like those of Paba & Lenia. Sometimes they have considered themselves Lenians, other times aboriginals. The same racial divide was present in Laba.

climate

21st parallel to 31st parallel. Climate subtropical, not truly tropical, but winters are stable (no cold spells) ,so vegetation is primarily tropical.