Re: new conlang - " *aiinodbus' "

On 22 Sep 2004 Rodlox <Rodlox@HOT...> wrote:

> phoneme pronounciation:
> gh
  Your original specification for |gh| was >>like "GHent" in
Holland<<. AFAIK it is phonetically /G/, i.e. voiced velar
fricative. However in this case, your phoneme inventory is slighly
unbalanced, since |kh| is defined as /X/, that is unvoiced _uvular_
fricative. If balancedness matters, you may consider all-uvular
solution (|gh| = /R/ and [kh| = /X/) or all-velar one (|gh| = /G/
and [kh| = /x/).

> o  (like "cOt" & "Octopus")
  Every time I make a statement on English phonology, I burn my
fingers  because of the confutations of various English 'lects :))
I do not know whether I dare to assign a phonetical value /Q/.

> u  (like "tUt & "bUnk") [U\] ?
  See previous comment. It is usually transcribed as /V/ but I was
told that it is rather a historical notation than phonetically
correct one.  I suppose it is near /6/.  ([U\] is much more close
and rounded.)

> I  [I] or [I\]
  Your orginal definition >>like "tIn" & "pIn"<< implies /I/. /I\/
has back characteristics in a binary classification. The symmetry
says that /I/ is a better choice because in this case you have two
front vowels |i| and |a| for two back ones |o| and |u|.

> ii  (like "sIght" and "strIve")
  /aI/, I presume.

> a  (like "bAt" & "cAn")
  Its usual transcription is /{/ in X-SAMPA or /&/ in CXS. But I
was told that some pronounce it a bit closer as it were /E/.

> d  [d]  (? [d'] or [d_<] ?)
  Implosive /d_</ would be an off-system sound (since |b| is not
/b_</). IMHO [d'] could be a possible (dialectal) allophone for
|dy| /dz\/. To reach a maximum phonematical contrast, I would
prefer /d/ (even in dental [d_d] variation).

> n  [n] or [N\]
  IMHO if a dento-alveolar stop /d/ exists, there should be also a
dento-alveolar nasal /n/ in the system.

> h  [h] (?) [h\] or [H\] ?)
  It could be a solution to enable allophonic representations for
/h/, e.g. [h\] intervocalically, [H\] at the end of the words (to
avoid omission) and
[h] otherwise.

Replies