Re: A new member and a new conlang sketch

>> > [phoneme inventory]
>>
>> Doing nice so far. A somewhat European inventory maybe.
>
>How to un-Europeanize it? (I don't know if it matters enough, my goal has
>never been exoticness in itself.) I know close to nothing about the phoneme
>inventories of the worlds languages, what I know are mainly from here:
>
>http://www.eskimo.com/~ram/segmental_phonemes.png
Ah, that. It's a nice aid, but you might want to note that the analysis
1) covers "major" languages, not languages evenly from all over the world
2) omits some distinctions, as the footnote says
3) may have some questionable vowel aanalysis (eg. why the decided that
Finnish has /e 9 o/ is beyond me, common analyses speak of either hi-mid /e
2 o/ or lo-mid /E 9 O/ thruout) - but if you've settled on your vowels, this
shouldn't concern you.

If you want a wider look on the world's languages' phonologies, there's this
index I've been working on:
http://en.wikipedia.org/wiki/User:Tropylium/Shtuff
It's still very unfinished beyond Africa, but you should still be able to
get a picture...


>/v/ seems to be the most European phoneme in my inventory. But if you exclude
>all Indo-European languages in Europe, what will then be the cross-linguistic
>most average phoneme inventory?
I've seen /p t k s h m n r l w j/ + maybe /f tS/ quoted. If that's too few
obstruents for you, AIUI /b d S ?/ are some of the most common "additions",
and after those, /dZ g N/.


>> >/r/ has [4] as an allophone.
>> >/dZ/ has [Z] as an allophone.
>>
>> ...on what conditions?
>
>Well, that's mostly from the auxlangishness, to make it easier to people to
>pronounce it through allowing variation.
Yes, free variation works well for that purpose.


>But I'm considering to simplify the
>description and leave it out. Then you can ignore [4 Z A I].
You could just underspecify things in the initial description - eg. /r/
being simply a rhotic (not explicitely a trill or a flap), and expand on the
details in a later section...

BTW, any particular reason to have /i/ = [i ~ I], but no similar situation
with /u/?


>> As for the sibilants; I'm partial of using <x> for /S/, but other choices
>> might include the classic <sh>, as well as <s> + any diacritic whatsoever.
>> There's also <sy>. Similarly <ch> or <c> + diacritic or <cy> / <ty> / <ky>
>> (whichever you prefer) for /tS/. The affricates could also be done
>> compoundedly: <ts tx dj>, in this case probably furthermore using <j> for
>> [Z]. Alternately, the Pinyin-like single-char approach would be <c q j> (and
>> <x> for /S/).
>
>/S/ = <sh>, /tS/ = <c>, <ch> hmm..
>My tentative orthography in my notes are <' þ c q j y> for /? ts S tS dZ @/.
>Maybe I switch to <sh ch>, and keep the rest..
>/ Veoler
Thorn for /ts/... interesting. Feels kinda backwards to me, but I suppose
from a synchronic viewpoint it's fine. If you switched to <sh ch>, I'd also
switch to <c> for /ts/, but if you like thorn, why not...

John Vertical

Reply