Sorry, yes. The version of the language represented by the script has
no phonemic voice distinctions, so [tajsO`r\] is phonemically
/dajzor/. Also, [a`] is an allophonic variant of /&/ before /r/,
ditto [O`] for /o/. And [aj] and [ej] are phonemically, or at least
underlyingly, or at the very least *written*, bisyllabic.

I dunno---I like the look and sound of it. It is ...

On 7/9/07, Dirk Elzinga <[log in to unmask]> wrote:
> On 7/9/07, Mark J. Reed <[log in to unmask]> wrote:
> > or something like that. I'm reëntering the conlang fray.
> >
> > Target: Okaikiar
> > Mission: Beautify.
> >
> > Okaikiar, as she is right now, I find aesthetically displeasing. It's all
> > in the eye and ear of the beholder, of course, but I just don't like the
> > phonology. As an example, here's the paternoster:
> >
> > Uza zi dinkmarkurd or 'uz& zi dink'mar\kUr\d Or\
> > Limzar don nudir 'limza\r don nu'dir\
> > Markiar don maid mar\'kiar\ don ma'id
> > M'ziar don ed m@'ziar don Ed
> > Maizim ram dinkmarkurd ma'jizim r&m diNk'mar\kur\d
> > Lokand m'lor zøn zyrkom zi rork 'lok&nd m@'lOr\ z2n 'zyr\kom zi rOr\k
> > Uz lymkiard zi dolk uz lym'kiar\d zi dolk
> > Ram zø lymkard zi dolkraz. r\&m z2 'lymkar\d zi 'dolkr\&z
> > Uz kondziark zim lolz zik uz kond'ziar\k zim lolz zik
> > L'ro m'kun zim lolz l@'ro m@'kun zim lolz
> > Ziin dan ømkraz markiard zi'in d&n '2mkr&z mar\'kiar\d
> > Uz k'ndard uz k@n'dar\d
> > Uz køkald uz 'k2kald
> > Ruzurd. 'r\uzur\d
> >
> > So I'm revisiting the phonology in an attempt to get a new one that is
> still
> > consistent with the established proper names and that still works with the
> > script. Suggestions welcome. Do you all find the above as
> > unsightly/harsh/hard on the ears as I do? If so, what do you think would
> > improve it?
>
> I dunno---I like the look and sound of it.
>
> > Known proper names: Dankar ['d&Nka`r\], Zan [z&n], Tysor ['tajsO`r\], Ral
> > [r&l], M'kei [m@'kej].
> > Vowels: & aj a`r\ ej o`r\
> > Consonants: d k l m n N r\ s t z
> >
> > Since we have both [d] and [t], chances are that since we have [k], we
> also
> > have [g]. The existence of [m] makes the existence of other bilabials
> > probable but not necessary. I find it quite probable that the [N] is an
> > allophone of /n/ occurring before velars.
>
> It is often the case that languages with a voicing distinction in the
> obstruents will still fail to have /g/ as a counterpart to /k/, so the
> lack of /g/ should be okay. By the same token, you can miss /p/ but
> still have /b/. (This has to do with the aerodynamics of voicing.)
>
> > The script as-is supports only 8 syllable onsets and 8 nuclei, but up to
> 64
> > codas, and it can be readily extended to support as many onsets as codas.
>
> It would seem then that the script doesn't capture all of the
> available distinctions. I count 9 consonants (excluding /N/ as
> allophonic) and 11 vowels (in the transcription of the Pater Noster).
> Is this right?
>
> Dirk
>


--
Mark J. Reed <[log in to unmask]>