Re: +AFs-CONLANG+AF0- Vowel romanization

Joe:
> Herman Miller wrote:
> >And Rosta wrote:
> >>Your romanization options seem to go beyond what I would think of as a
> >>romanization, in that they use nonalphabetical characters. So I can't
> >>quite grasp the rationale or the problem: if you can go outside the
> >>roman alphabet, surely there is a great array of symbols available for
> >>use? If you have to stick with roman letters, then you'd have to fall
> >>back on diacritics and digraphs.
> >
> >The "open e" and "open o" are used in some African languages. The
> >letters with dots under them are used in Vietnamese (but not for the
> >same sounds -- the Vietnamese dot is a tone mark). The Cyrillic z for an
> >open-mid central vowel is an arbitrary substitution for a reversed open
> >e, but Zhuang uses it for a tone letter, which doesn't have anything to
> >do with its actual use in the Cyrillic alphabet. The main thing I want
> >to avoid are letters that don't have upper and lower case versions.
> >Limitations of font technology in Windows put restrictions on which
> >capital letters can have diacritics added to them. Some Greek letters
> >could potentially be of use, but the problem with Greek vowel letters is
> >that their upper case forms look like Latin letters. But I'd like to
> >stick with the Roman alphabet or extra letters that are traditionally
> >used with the Roman alphabet, rather than borrowing arbitrary characters
> >from other scripts. I've only considered the Cyrillic z because it looks
> >like the IPA character [3], and I haven't come up with a better
> >representation for that sound.
>
> Indeed.  And, if his argument is correct, English doesn't use the roman
> alphabet - w?  j? u?
w, j & u are letters of the roman alphabet, not the greek or cyrillic
or another.

As Herman implicitly says in his reply, the roman alphabet is a tradition,
a family of sets of letters, rather than a single set of letters. The
crux is whether such and such a character is 'traditionally' used with
the roman alphabet. For Livagian, I feel that the answer is Yes for
thorn, eth and yogh, but not for eng, so Livagian uses the first three
but n-diaeresis for eng; to use eng would, in the Livagian view, not
count as a romanization

If Herman's Tirelat notion of 'tradition' extends to the use of IPA
symbols in African orthographies, then he could simply extend that
tradition to use the symbols that best match the IPA symbol for
the vowel's value yet still have something that looks like an
upper case counterpart. Me, though, I'd go for something like
Remy's polygraph solution. (Livagian has two romanizations, one
that is restricted to ascii and uses digraphs, and one that isn't
restricted to ascii and uses thorn, eth, yogh and diacritics but
no digraphs. The ascii romanization is used for alphabetical
ordering in general, not only for typographically impoverished
media.)

--And.

Reply