`Akupa, was: 5 phonemes

No, I wasn't satisfied with Miituu. Danny was right saying that 'true'
geminates could be counted as phonemes. And the other complications
with syllabication can be suggestive of something like deep-level
glottal stop.

So I designed `Akupa <maapllpa> (not Mwakupu, as I erroneously stated
earlier). Historically, this is the same tribal name as Miituu
<mekprrp>. The phonetic correspondences between the two dialects are
quite straightforward ;)

Akupa (mor)phonology is much simpler since it doesn't allow for
'deep' closed syllables. Otherwise it resembles Miituu very much.

Is it flawless this time?


Basilius



1. Deep:

C: p t m
V: a l=

Allowed syllable structures: (C)V

No restrictions on syllable combinations within a word.


2. Surface realizations

C: /p t k s m n l w `/

V: /a i u/

Syllable structure: (C)(w)V

No vowel clusters are allowed within a word.

Consonant clusters: /sw kw/

Forbidden combinations: /wu/, /`i/

No other restrictions on sound combinations.


3. Deep to surface

a) Vowel contraction and emerging glides

Vowel sequences are processed from right to left. The subordination of
rules is as follows:

l= (+V) /l/

aal= /wi/
   otherwise al= /i/; otherwise l= /u/
   otherwise aa /wa/; otherwise a /a/

b) Reforming the clusters with /l/ and /w/

pl /k/
tl /s/
ml /n/
pw /kw/
tw /sw/
mw /`/; mwi /`u/

(_mw_ first became /Nw/ - analogous to _pw_ /kw/, then was simplified
to /N/, weakened to /`/; the idea of _pl_ /k/ was inspired by
Starostin's reconstructions for North Caucasian, showing that lateral
affricates often become velars)