Re: Tamil Text Messaging in Mobile Phones

From: Michael Everson (
Date: Sun Jul 28 2002 - 13:24:31 EDT

At 16:22 +0000 2002-07-28, Sinnathurai Srivas wrote:

>1/ I'm not asking Tamil Script reform to be approved by Unicode.
>This will be done by Tami Nadu Government and other countries with
>Tamil speaking population. IFITT, Tamil_Reasearch and other relevant
>worldwide Institutions will make sure that this is done in an
>appropriate way.
>2/ I'm informing Tamil Script reform is imminent.

That is a great pity. There is nothing wrong with the Tamil writing
system, and nothing unnatural about reordrant vowel digraphs.
Reforming it in the way described will solve some very, very
short-term rendering problems on some computer systems, and in the
long run will make centuries of written literature unaccessible to
the population at large. Such a reform will doubtless outrage a great
many people as well.

If the Government of India is a member of the Unicode Consortium, it
would be very, very strange for them not to have communicated this
matter through normal channels.

>3/ Other systems with low resolution and limited bandwidth can
>immediately use reformed characters. For a Tamil spending half an
>hour will make sure that one can understand reformed Tamil. Hence,
>say instead of thousand of codes being transmitted for text
>messaging a picture of say "b" all you need is to send one code for
>the character "b". (consider the cost of bandwidth for sendin a
>picture and sending a code for a character.) It is not possible for
>all languages be rendered within the mobile phone itself, unless u
>store millions of MIBs within the mobile phones. Hence, Tamil will
>take the character transmission route and some languages may take
>picture transmission route.

Fix the technology. Don't change the orthography.

>6/ Luckily Unicode has ALREADY structured the coding in Linear Fashion.
>This means script reform for Tamil is ALREADY COMPLETED within Unicode.
>All we have to do is schedule the withering of various complex rendering.

It is agreed that the only difference is that the "reformed" Tamil
changes the shapes and positions of the glyphs of the vowels. I would
think that this would change reading speeds for fluent readers, as
word-shapes throughout the language would change. And why? Because
mobile phones (a new invention) aren't up to speed?

>7/ Original Tamil, even in BC3 was written in Linear fashion. The
>complex and fancy full (I call it uninteligent) changes to Tamil
>writing was introduced in the recent past. We are devicing a way to
>go back to the old scientific Tamil writing system, which is correct
>and sensible.

This sounds like a baseless and revisionist.

>8/ About inherent a and diacritics
>Inherent a and diacritics will need to be coded by Unicode. This
>will be requested officially at the appropriate time. The diacritics
>are for use in Pronounciation dictionary. (Example 18 consonants
>represents about 56 phonemes of consonants, similarly about 5 Vowels
>represent minimum of (not including long vowels) 10 phonemes, but in
>minute details about 25 Vowel phonemes.

I think this kind of reform is very ill-conceived, and does not
respect the natural development of Ashoka's script. Are you really
going to do away with the inherent vowel? Reprint every bit of
printed Tamil literature? Change not only computer systems but
typewiters and letterpress printing? Replace every street and shop
sign in Tamil Nadu and Tamil-speaking areas of Sri Lanka?

I would be VERY reluctant to add a whole lot of new characters to the
Unicode Standard and ISO/IEC 10646 based on this reform unless it
were very, very well documented and supported by the governments of
India, Sri Lanka, and Singapore.

Michael Everson *** Everson Typography ***

This archive was generated by hypermail 2.1.2 : Sun Jul 28 2002 - 11:42:54 EDT