On Thu, 14 Oct 2004 17:06:06 +0930, Adrian Morgan (aka Flesh-eating Dragon)
<[log in to unmask]> wrote:

>I would like to introduce a new consonant into Gzarondan phonology,
>namely a phonemic distinction between /x/ and /x_w/ (unlabialised
>versus labialised velar fricatives).
>
>I'm trying to decide what other changes to make to the phonology in
>order to make the new phoneme set look reasonably naturalistic. I
>would appreciate some advice.

I've forgotten who or what speaks Gzarondan. You said naturalistic, so I'm
assuming ordinary Earth-type humans.

>Intuitively, I feel that if there is a /x/ vs /x_w/ distinction then
>there ought to be a similar distinction for (at the very least) all
>fricatives and approximates at or behind the velar position. This
>would mean adding a /h/ vs /h_w/ distinction.

It seems to me that /k_w/ and /g_w/ are more likely with /x_w/ than /h_w/
is, but I haven't exactly studied a lot of phonologies -- there are
certainly much more unusual things in Natlangs.

>Alternatively, I could do away with /h/ (but how many natural
>languages don't have /h/ ?) and use the /x/ vs /x_w/ distinction in
>place of the /x/ vs /h/ distinction. This conserves the total number
>of phonemes.
>
>Building on the above approach, I could rewrite history to say that
>all unvoiced fricatives once distinguished labialised and
>non-labialised forms, but that /T_w/ later became /f/ and /s_w/ later
>became /S/, whereas /x_w/ remained as it was.

Your sound changes sound OK. The question I have is: why were only
voiceless fricatives labialized? You might want to widen your scope and
look at the whole recent *history*, including phonotactics and the vowel
system.

Hope this helps.

Jeff

>This sounds reasonable to me, but perhaps there's a better option.
>
>The current consonant phoneme inventory is listed below.
>
>Stops, Nasals & Trills:
>
>/p/ /b/ /m/
>/t/ /d/ /n/ /r/
>/k/ /g/
>
>Fricatives, Laterals & Approximants:
>
>/f/
>/T/ /D/
>/s/ /z/ /K/ /l/
>/S/
> /j/
>/x/
>/h/
>
>Adrian.