Pejo language

Pejo is a language spoken on the islands of Laba. It is descended from Tapilula.

Basic sound changes from Tapilula to proto-Pejo

• /ḳ/ > /ġ/ (a voiced velar stop).
• /tʷ dʷ nʷ/ > /kʷ ġʷ ŋʷ/. Thus all labialized consonants are velars.
• /f/ > /h/. Thus all fricatives (at this point) are velars as well.

0  l  j  h  ġ  k  ŋ  p  m  t
gʷ ŋʷ n  hʷ ġʷ kʷ g  b  f  d

• Probably /g gʷ/ > /0 w/, leaving the language with just /h hʷ/ for fricatives. Perhaps, though, before this shift happens, there is a vowel shift changing sequences like /àu/ into /ō/, like the languages of the mainland. Thus the /g/ > /0/ shift creates new sequences after a brief period in which the language had no vowel hiatus (except possibly at word boundaries). Also, /g/ after syllabic nasals creates a voiced stop instead of disappearing. Thus, sequences like /ṁb/ became more common. They had existed before, but only at the same rate as disharmonious sequences like /ṁt/, since the origin of the syllabic nasals had nothing to do with what consonants, if any, followed them.

0  l  j  h  ġ  k  ŋ  p  m  t
w  ŋʷ n  hʷ ġʷ kʷ    b  f  d

Thus Pejo is outside the Rilolan sprachbund where /g/ is common and /ġ/ is rare.

• Some of the new vowels created by the above shift might actually be rising diphthongs, which are not considered diphthongs in most Tapilula languages but rather sequences of /j/ or /w/ plus a vowel. Thus, Pejo gained new syllables that had a /j/ or /w/ between the initial consonant and the nucleus of the syllable. The primordial labiovelars merged with these, meaning that they can be analyzed as /kw/, /hw/, etc, and therefore the labiovelars disappeared from the phonology, although sequences of a velar consonant plus /w/ were still far more common than any of the other combinations.

The phonology at this stage would be for consonants /p b m w t d n l j k ġ ŋ h/. The dot over the /ġ/ is to conform with the spelling traditions of related languages, but is unnecessary since this language has no voiced velar fricative /g/. The vowels and the syllabic consonants remain the same as the parent language's, but there is probably at least one new tone, and possibly two. The maximal tone system would be /ă à ā á/, a common setup for the mainland branch of the family.

• Syllabic nasals remain. They cannot have an epenthetic vowel inserted, such as /ə/, because then there would be only one vowel in the language which could carry a syllable-final consonant.

Notes

While one might expect a pushchain sound shift of /b d ġ/ > /p t k/ > /f s x/, this never happened in any of the descendants because the parent language's voiceless stops were more common than the voiced stops, which would mean a language undergoing this shift unconditionally would then have more fricatives in its words than stops. However, this shift did not even occur conditionally, either. Instead, some languages changed the voiced stops /b d ġ/ into voiced fricatives /v z g/, which then in most branches became voiceless /f s x/.

Later sound changes

Here the language splits into daughters, but even so, there is one branch that is politically powerful and is the language called "Laban" by outsiders. (For comparison, imagine if Swahili or perhaps Arabic was called "African".) However, it will be referred to as Pejo here, the name of the tribe that spoke it.

• possibly /b d ġ/ > /v z g/ > /0 z 0/ > /0 s 0/. This creates an /s/, and gets rid of the "unwanted" consonants that would be rare for the geographical area Pejo is in. /v/ might cause labialization of the preceding vowel before it disappears, as happened in the parent language of Tapilula. Thus this is a repeating sound shift. If /v/ disappears, it probably takes /w/ with it.

• High tones cause following fricatives to become stops or affricates. /h g/ > /kʰ ġ/ or possible even /kʰ k/.

0  p  b  m  w  t  d  n  l  j  k  ġ  ŋ  h (f  s  x)
ʔ  pp pp mm pw tt tt nn tl ć  kk kk ŋŋ kʰ pf c  kʰ

The resolution of the voiced stops up above could be resolved by saying that /ʔ/ + voiced stop ---> /ʔ/ + voiceless stop. This would happen before the change of remaining voiced stops into voiced fricatives, thus removing the need for the sound change to happen twice.

But voiced stops must reappear somehow.