Mr Cibu had challenged in his blog to find an example in Malayalam which brings meaning change with and without using ZWNJ.
An example was posted, no reply for that yet.
We all know, normally there wont be any meaning change, because there existed the type-writer script in Malayalam in which the reader has to interpret the meaning of the word according to the context. This script was only used by the official purposes and was rejected by the mainstream public. Any move to revive the script will bring strong resistance from the language lovers.
One reason that was told for the Atomic encoding was the meaning change theory.
Now I bring a similar example for non-joiners (This example is as acceptable as the vanyavanika,kanvalayam,manvikshobham examples),
sad_vaaram (means good week– with non-joiner) സദ്വാരം and sa_dvaaram (means with hole– without non joiner)സദ്വാരം . As there is no graphically independent Cil available for da, atomic encoding is not possible here, though da has a Cilling behaviour.
What is the solution here ?
A non-joining candrakkala ?
An atomic postbase va ?
Or is there a move to make the virama non-joining ?
These questions remain un-answered .
And a solution for this is not attempted.
The real solution for this and the chillus is the legalization of the current sequences. The joiners and non-joiners must become atleast second class citizens. And they must not be thrown out by any standard applications. There is no problem in IDN also with the existing sequences, as the joiners are mapped to null string. There wont be any compatibility issues and security threats.
But some persons told me that we have to use sat(സത്) instead of sad(സദ്).I can’t agree to it. We use sad_vaartha, sad_vichaaram, sambad_vyavasttha etc.
Even if that is the case(sat is to be used), I shall bring another example:
sat_varam (good blessing, nalla varam) and satvaram (speedy) the
സത്വരം , സത്വരം pair.
I think if we dig it more , we may find more pairs. Just we have to look into the postbase , prebase, belowbase conjuncts. If a non-expert like me can find these, experts can find more.
But there is no point in finding these pairs. If we preserve the (non)joiners, the issue never arise.
So one reason for atomic chillu encoding is void .
And if we look into all the stated available reasons for atomic chillus one by one, all of these are, according to me, simply excuses, not reasons.
[The real reason is “The Govt of India requested it. And the Govts of Kerala supported it”. Please tell us that openly. There was no need of these “mock” sessions.]
The basic argument that Cils are independent is void, because a glyph of Cil-ya is not in use. The very reference used to “prove” the independent nature of Cils, presents the Cilling property of YA using an example kaay_kaRikaL(കായ്കറികള്).
As there is no independent glyph for Cil-Y, atomic encoding is not possible
Let’s make the applications not remove/strip the joiners by suitable amendment in the standard. This is a solution and this is the solution.
If the vendors need their applications compliant to the standard, they will make necessary changes. There are open source solutions already working with the preserved joiners. These vendors also can follow the suit.
“Bad workers always blame their tools”