Re: Re: [indic] Transliteration of Sinhala character

From: N. Ganesan (naa.ganesan@gmail.com)
Date: Wed Jul 05 2006 - 16:05:48 CDT

  • Next message: N. Ganesan: "Annotating U+0DA5 (Re: Transliteration of Sinhala character)"

    Richard Wordingham wrote:
    >The principle here is that these ligatures are
    >treated as having become new letters.

    U+0DA5 is *not* a new character in Sinhala,
    it is always a conjunct. This 0da5 "letter"
    is not listed as a Sinhala separate character
    in old or new books of Sinhala orthography.
    See Carl Faulmann's, Buch der Schrift, 1880 book
    or James Gair's (1996) Sinhala writing, in Daniels/Bright,
    World's writing systems (1996) or other books on Sinhalese
    script. Also, Sanskrit historically had less
    use amidst (Buddhist) Sinhalese folks.
    James Gair writes that Sinhala language/speech
    lacks aspirates (just as Tamil does, and unlike
    North Indian languages). Sinhala script is
    "an alphabet within an alphabet", the aspirates
    etc., are added in recent centuries to represent Sanskrit.

    Even the name tells that it's a conjunct: sanyooga = "saMyoga" (well-yoked).
    U+0DA5 - SINHALA LETTER TAALUJA SANYOOGA NAAKSIKYAYA - ඥ - Jnya
    So, to encode a conjunct <ja, virama, nya> as 0da5 was a mistake in Unicode.
    It is a normal practice that Unicode does not encode conjuncts,
    if so we'll have 1000s of codepoints for conjuncts for Indian letters.
    Even ksh, shrii are treated as conjuncts in Marathi, Kannada or Tamil.

    What to do now about U+0DA5? It will be good to follow what's
    done as annotation for "KANNADA LETTER FA"
    http://www.unicode.org/notes/tn27/

    http://www.unicode.org/errata/
    Please see for the glyph change for Zero
    in Malayalam, and letter RA for Tamil
    (glad to note that i also contributed to both
    changes).

    An annotation to write that 0da5 is a conjunct letter
    and is really representation of <ja, virama, nya> in
    Sinhala code chart can be done.

    >Another problem is that in virama-based scripts, decomposition would have to
    >be a two stage process, so you would have to introduce a new character just
    >to enable the decomposition!

    True. Atleast in Sinhala section, or introductory section of
    Brahmic scripts in the TUS, there can be a statement to
    help in transliteration of Sinhala half-nasals into other Indic scripts
     - to use candrabindu. If no distinction is needed, then use full nasals.

    N. Ganesan



    This archive was generated by hypermail 2.1.5 : Wed Jul 05 2006 - 16:44:01 CDT