On 19 Sep 2004 Andreas Johansson <andjo@FRE...> wrote:

> I believe I've indicated a lack of faith in the utility of such
> specifications at all.

  In my interpretation, this issue is about to help Rodlox to find
a correct phonetical representation for his/her specifications.
Unfortunately, we got involved with each other's argumentation

> The question, however, was whether the voiceless uvular fricative
> occurs in more well-known languages, which it certainly does.

  For me, this is not exactly the question.  The question is what
extent of the populace is aware of a given phonetical
characteristic.  In this respect, some features of well-known
languages can be less evident than other features of less-known

> Some books give [X] as the value of /x/ after back vowels in
> Modern Standard High German

  I do not bring into question of your information of standard High
German [X], but popular works as Duden's Aussprachewörterbuch do
not mention this feature.  My Ausprachewörterbuch (3rd edition,
1990) conveys |ach!| as [ax] and not [aX] despite of the fact that
it gives detailed allophone inventory for other phonemes, e.g.
enumerates four representations of syllable-initial /r/.

  Therefore a German example for [X] could be informative of
experts of High German phonology, but it could be misleading for
others, e.g. for those who are informed from Duden.

> I'll have it noted that X-SAMPA thinks that [J\] is a modified
> 'j'. The IPA symbol also seems to be based on 'j'.

  X-SAMPA thinks also that [J] is a modified 'j': in this respect
'j' means simply any palatal phoneme and the form of 'j' reflects
their frequency (j > J > j\ > J\).  But in Rodlox's notation this
feature is expressed already by modifying letter |y|.  Therefore
the first, base letter reflects rather other phonetical

  (As for IPA, voiceless pair of [J\] is a simple Latin letter [c]
and it has no tight connection with it phonetical value. Both this
assignment and [J\] seems to be rather odds and ends then a
systematic one.  Moreover, Unicode chart says about [J\]: "also
archaic phonetic for palatoalveolar (sic!) affricate 02A4 [dZ]",
i.e. this letter was invented originally for [dZ].)

> No, but orthographic devices that are not perfectly systematic are
> often approximately so. It would a priori be unsurprising if all
> all digraphs in -y denoted palatals.

  The only reason for |-y| = palatal assigment is Rodlox's notation
|jy| for [J\].  On the contrary you seemed to be scheptic about the
existence of "Eastern European" palatal fricatives.  My proposal
for Rodlox -- since he/she gave no exact phonetical assigments --
is that |-y| should mean prepalatal (i.e. alveolopalatal) locus,
that is |sy| = /s\/, |zy| = /z\/.  This makes possible also the
existence of true palatal "allophones" [C], [j\].  (Rodlox's
inventory lacks grapheme for [j], it would be a quite extraordinary
system which has phoneme /j\/ but /j/ is missing. This can be
another counter-argument against |zy| = /j\/.)

  For the reasons I outlined in my previous postings, and for the
symmetry of the orthography, I propose |jy| as /dz\/ (with a
possible allophonic representation [J\j\]).  This still allows
specification >>jy (like "DJoser" in Ancient Egyptian)<< and the
orthography remains systematic.  (The systematic solution for /J\/
would be rather |dy| even if |-y| would be treated as true palatal

> Rodlox, however, said it was like in EEan languages when
> describing the orthograpy, which certainly _suggests_ he believes
> some or other EEan language uses them in the same way.

  Rodlox acknowledged my preconception that this was only an
Eastern Europish impression: "well...I thought I had seen an
Eastern European name written like that...somewhere."

  My efforts aimed to establish a system that is really Eastern
European and could have graphemes like |sy|, |zy|, |jy| (if its
orthography would be re-designed).

  IMHO there are only two possible Eastern European prototypes:

1. Slavic languages having (nearly) full _palatalized_ series, like
e.g. Russian, Bulgarian. In these systems, |sy| would be [s'], |zy|
would be [z'] and |jy| would be [d'] (if we insist on plosive
representation, or [dZ'] if not).

2. On the contrary the above languages have near full palatalized
series.  The ones with defective series have either palatals or
prepalatals.  And there are no languages even of this type, where
palatal fricatives exist but palatal nasal is missing.  However,
alveolopalatal fricatives may occur in themselves, cf. Chinese.  To
return to Eastern Europe, we can find languages where palatal
sibillants are realized prepalatally, e.g. Polish.  But in Polish
common Western Slavic [J\] < [d'] is realized also as a prepalatal
affricate [dz\].  Moreover, in Polish (pre)palatals are conveyed
often (before vowels to be exact) as digraphs, however their second
element is |i|.  But if we would re-design the system with |y|
instead of |i|, we would get Rodlox-style |syano| [s\anO] 'hay',
|zyemya| [z\em'a] 'earth' (original spellings: |siano|, |ziemia|).
Therefore I think the best model in this field for Rodlox's system
is a variation of Polish where the palatals were lost and the
prepalatals retained (except unvoiced [ts\]).

> I originally asked specifically about _Latin_ orthographies, but
> good to know anyway.

  I do not want to contradict you but I do not remember explicit
narrowing for Latin script. After browsing postings I found still
general questions as e.g. 18/09: "Since you appear to be familiar
with EEan languages, do you know any that uses the digraphs 'sy'
and 'zy'?"

  I did not exlude Romanizations as for myself because Rodlox's
source for |sy| and |zy| could be a transcribed Cyrillic text.

  But to provide you an example from a Latin orthography, I present
Uzbek translation of Universal Declaration of Human Rights at
<>.  Here you can found a
word |xususyii|.