Re: Input method

--- In qalam@yahoogroups.com, Patrick Chew <patchew@...> wrote:

> At 06:17 PM 8/3/2004, "Richard Wordingham"
> <richard.wordingham@...> wrote:
> >"suzmccarth" <suzmccarth@...> wrote:
> > > How about o + k + u = k + ou ?
> > > This would then display properly as o + k + u. Is this

possible?

> >
> >I don't see this combination on the Tamil unicode page. Do you

mean

> >
> >visual <vowel sign e> + <k> + <aa> => stored <k> + <o>,

equivalent to

> >stored <k> + <vowel sign e> + <aa>?
>
> Do you mean in the last line "<k> + <vowel sign au>"?

No. It is TAMIL VOWEL SIGN O that decomposes to TAMIL VOWEL SIGN E
+ TAMIL VOWEL SIGN AA in Tamil.

> >In Thai, visual and stored <sara e> + <k> + <sara aa> is
> >pronounced /kau/ and seems to correspond historically. Other
> >possibly relevant groups are
>
> Unfortunately, this depends on which set of Indic scripts

you

> prefer to trace "pure" lineage to... northern or southern...
>
> ><sara e> + <k> + <mai han-akat> = /ke?/
> ><sara ae> + <k> + <mai han-akat> = /kE?/ (low-mid front vowel,

transcribed

> ><ae>)
>
> It's not actualy <mai han-akat>, but <sara a> that combines with
> <sara e> and <sara ae> to shorten them.

Drat! I thought I'd checked the chart. When I learnt the name, I
was told that the two vowels had the same name, so I find it
difficult to remember the Unicode distinction. Perhaps a menmonic
is to think that lone mark is more in the air ('AKAT').

> ><sara e> + <k> + <ii> + <y> = /kia/

...

> For the diphthongs, there is no large issue, excepting for sorting
> issues... the will all be sorted with <sara e>, as done

traditionally,

> rather than phonemically, which seems to be a large trend

nowadays...

I listed them to give an idea of possible issues if hitching a free
ride on Thai for Tamil stored in visual symbol order. On the second
point, written Thai is too ambiguous to sort phonemically.

> ><character o ang> is a glottal stop, but it also serves (mater

lectionis

> >in an abugida?) for the open-low back rounded vowel (/O/).

> urm.. wouldn't <character o ang> just be a conflation of two
> possible functions, which are often represented by different
> characters in other systems?

I understand that CHARACTER O ANG derives from the independent VOWEL
A; I don't know how it came to be used as a vowel symbol. I suppose
it may be compared to THAI CHARACTER RU, which serves as both a
dependent and an independent vowel. (It and THAI CHARACTER LU, not
used for Thai, are the only independent vowels in the Thai script.)
THAI CHARACTER WO WAEN also serves as both a consonant (/w/) and as
a vowel (/ua/ in closed syllables), but may be that's because the
sequence /awa/ is close enough to /ua/. The Thai representative of
Sanskrit /yavana/ 'Greek' is /yuan/ (one syllable), still spelt the
same: YO YAK, WO WAEN, NO NU. (What's the ISO 11940
transliteration? <y.ow.on>?) It was just that these _are_
effectively matres lectionis, which one shouldn't have in an abugida.

When Pali is written in Thai, it normally doesn't use SARA A or MAI
HAN-AKAT - it uses PHINTHU ('Thai virama'), so is an unambiguous
abugida. However, the only Pali text we have at home is written in
the Thai script without inherent vowels or PHINTHU - the inherent
vowel is written out using SARA A or MAI HAN-AKAT as appropriate!
By choosing the appropriate subset of the Thai script, the Pali is
actually written in what some would call an alphabet! (I feel it's
better described as a compulsorily pointed abjad, but that's a row
for another thread.)

Richard.