G'amah phonology

OK, this message goes together with the nasal oddities
one. I'd like to present to you my latest attempt, a
language spoken by a people mostly consisting of orcs,
trolls, and men, in variable degrees of interbreeding.


As usual, constructive criticism is welcome. Destructive
criticism -- well, I'll do something with it too, probably.

The name of the language (provisionally) is _G'amah_,
where <g'> is implosive /g`/, <h> is pharyngeal /H/,
and the rest as in IPA. Stress, so far, in the last
syllable, but I'm toying with the idea of a pitch
accent system, without phonemical stress.

The phonology of G'amah is composed of:

Uvcd. stops:     p  t  ch  k
Eject. stops:    p' t' ch' k'
Vcd. stops:      b  d  j   g
Implos. stops:   b' d' j'  g'
Nasals:          m  n  nh  ng
Implos. nasals:  m' n' nh' ng'
Uvcd. frics.:       s  sh      h
Vcd. frics.:        z  zh
Flap/trill:         r
Laterals:           l  ll


Some informants have produced creaky voiced stops (with
creakiness spreading to following vowels) instead of
ejectives.

<h> is /H/ (uvcd pharyngeal fric), with allophones
    [x] after /i/
    [H] otherwise
<nh> is palatal(ized) /n^/ (Spanish <ñ>)
<ll> is [Z<lat>] (much like Welsh <ll>, but rather palatal)

<ch> is [tS], and <j> is [dZ]; both are included as stops
because they fill in the slot, functionally speaking. Note
<j'> is [J`] (?)(voiced palatal implosive *stop* -- not
affricate). Apparently the ancestor language had all palatal
stops, and they later dissimilated from alveolars.

/l/ gets velarized near velars and /u/; /r/ becomes
an uvular trill (or something very similar) in the
same position. In particular, /gr/ is practically [R:].

Vowels:

i     u
    o
  a

all as in IPA. /o/ sometimes alternates with [@_O].
I don't know if there was an /e/ at some point in time...

The syllable structure tends to be (S)(C)(L)V(C), with
(S) being any fricative, but mostly /s/, (L) being any of
/l r Z<lat>/, and V possibly a diphthong (not much info
on them so far). (S) is not terribly common. Open syllables
are quite common, but they tend to collapse in longer words.

There are some weirder patterns, too; I've heard a couple
of words beginning in <chh> /tSH/. Final clusters are
uncommon, mainly of the type stop + fricative.


Syllabic plain nasals are attested. Not having studied
prosody in detail, I cannot tell where the initial clusters
are in fact the onset of one syllable, or a syllabic sound
plus a simpler onset. Metrics would be useful, but these
people are not very interested in poetry or song. Anybody
can tell me something about mora-based langs? I think that
might be the path.

Assimilation is fairly common, the most usual changes being
voice agreement between stops (regressive, e. g. /kd/ > /gd/),
and POA for fricative (/SH/ > /S:/ > /S/). An interesting
kind of dissimilation is regularly found, too; when there are
two underlying glottalics of the same kind in adjacent syllables,
the second one becomes plain (for example, the reduplication of
<k'a> is <k'aka>, but <t'a> + <d'u> = <t'ad'u>, since <t'> is
ejective and <d'> is implosive -- though this has exceptions).


--Pablo Flores
  http://www.geocities.com/pablo-david/index.html
  http://www.geocities.com/pablo-david/draseleq.html