GROUPLANG: Phonetics

Carlos Thompson wrote:
>
>The phonetics, as propossed this far, is:
>
>basic sounds:
>vouels: /a/, /e/, /i/, /o/, /u/, /y/
>consonants:
>/p/ /b/ /t/ /d/ /k/ /g/ /q/ /q_voiced/
>/f/ /v/ /s/ /z/ /x/ /G/ /h/
>    /m/     /n/     /N/
>    /w/ /r/ /l/     /j/
>
Actually, we've discussed to drop /h/, mainly because it didn't
quite fit into the system. We could do with /x/, variants [x] and [h].

Dropping <h> would also allow us to use all the h-digraphs without
problems. Not only <kh>, <gh>, <qh> (which are invalid clusters, tho)
but also <nh> for /N/. If we keep /h/ = <h>, we'll have to write <ng>
or there'll be an ambiguity.

Note that <ng> for /N/ is not ambiguous; /ng/ could be written <ngg>,
because syllable-final nasals assimilate to the next place of
articulation.


>With the following orthography: all those sounds above, with lowercase
>letters in IPA will use the IPA value, but "x": a, b, d, e, f, g, i, j, k,
>l, m, n, o, p, q, r, s, t, u, v, w, y, z
>
>The other sounds are: /x/ as <kh> or <x>, /G/ as <gh>, /q_voiced/ as <qh> or
><qg>, /N/ as <nh> or <ng>, /h_voiced/
>
Let's decide about <h> and then I'll tell you :-)

>No values this far for <c> and <x>.
<c> could be /q_voiced/, but it's not very intuitive...
<x> could be /x/, but I think it's better to keep <kh> and <gh>
(they are more "mnemonic", and symmetrical).


[snip]
>Because of the VV posibility in a syllabe, in composing (or in adding
>affixes) there could rise V-V combinations.  The posibilities are:
>
>Avoid V-V combinations forcing C and F in every root.
>Asimilate V-V as a VV diphthong (even ae, ao, oe, etc.)
>Use something (both phonetically and orthographic to mark an hiatus, maybe a
>dieresis or an apostrophe)
>insert a sound, either a glide or a glotal stop.
>
>I vote for a glide in root combination and asimilation fo V-V into VV for
>affixes.
>
I vote for a glide in root combination AND for affixes.
I don't think we should force the roots to have any special form.


--Pablo Flores