Hallo conlangers!

On Tuesday 15 November 2011 10:29:38 R A Brown wrote:

> On 15/11/2011 05:48, George Corley wrote:
> 
> [snip]
> 
> > ............  Although, I think that aesthetics and
> > history can affect the former case -- particularly where
> > history justifies mirroring certain oddities in the
> > native script.
> 
> That's true - particularly if one is doing a
> _transliteration_ (i.e. grapheme by grapheme substitution)
> rather than a transcription.  E.g. Greek Θεός [4 graphemes,
> discounting the acute accent) is usually transcribed as
> _theos_ [5 graphemes] and _th_ may conveniently represent
> either the ancient [tʰ] or the modern [θ]. But a
> transliteration would demand something like θeós.  One
> modern system has _qeós_, apparently because of the supposed
> resemblance of upper case Q and Θ.
> 
> I'm afraid I don't find _q_ = [tʰ] and/or _q_ = [θ] either
> immediately accessible or aesthetic. But then _q_ has suffer
> almost as many strange uses as _c_ in different romanizations.

Using _q_ for [tʰ] or [θ] seems weird, but as you say, it has
been used for many different sounds.  The IPA convention of
using it for a voiceless uvular stop dates back, I think, to
Semitic transcription schemes where this usage is informed by
the fact that Latin _q_ evolved, via Greek koppa, from the
Phoenician letter qoph which was used for /q/.

My Latin transcription of Old Albic is not meant as a
transliteration (but the native script is pretty phonemic,
so that would not make much of a difference).

I use four digraphs: _ph_ for [ɸ], _th_ for [θ], _ch_ for [x]
and _ng_ for [ŋ].  I considered changing them into single
letters, perhaps using just the IPA symbols, but that would
mean changing all my files (as with _c_ for [k] and _v_ for [w]
which I sometimes considered changing into _k_ and _w_).

With the vowels, I use diacritics for length and intonation:
_a_ is short, _á_ is long with thrusting tone, _â_ is long with
slipping tone.  The front rounded vowels are spelled _ø_ and _y_.

I like the Old Albic transcription the way it is, and the
intrafictional justification for it is that Glastonian, the main
modern descendant language, uses these conventions as well;
extrafictionally, it is influenced by the spelling conventions
of Quenya, Sindarin, Irish and Welsh.

> [...]
> 
> > I would definitely write vowels.
> 
> Well, yes. If the Roman script is being used for a fictional
> natlang then it would IMO be perverse not to write vowels.
> Roman letter _transliterations_ of Semitic inscriptions do,
> of course, show only consonants; but, e.g. Maltese
> (basically Arabic with Italian borrowings) shows vowels in
> its Roman script.

Surely it does, and the usual transcription schemes for other
Semitic languages do so, too.  _Transliterations_ of Semitic
scripts or Egyptian hieroglyphs do not indicate vowels as they
are not indicated in the original scripts; but these systems
are used in citing the languages only where the vowels are
uncertain, as in Hieroglyphic Egyptian.  Transcribing a language
where the vowels are known without them would be bizarre.
 
> [...]
> 
> One thing that George's interesting article does not
> consider is the use of the Roman alphabet as a syllabary in
> the manner of Fuishiki Okamoto's "Babm' [bɔ'ɑ:bɔmu] or my
> own "Brx" ['puluçi] or the earlier system I have tried ever
> since I hit on the idea in the late 1950s      :)

Yes.

--
... brought to you by the Weeping Elf
http://www.joerg-rhiemeier.de/Conlang/index.html
"Bêsel asa Êm, a Êm atha cvanthal a cvanth atha Êmel." - SiM 1:1