[Oriya-group] Re: Oriya 'ya' and 'ba' and 'wa'
Rajat Kumar Mohanty
rkm at cse.iitb.ac.in
Thu Aug 31 23:37:55 IST 2006
Here is a few lines pertaining to ja-phala as mentioned in Dr. Nikhil's
mail and .pdf file
The point is whether we will design our software (1) keeping the current
practice of the mass in mind or (2)just because it is already approved by
some Govt. agencies/ISCII-1991 standard. First of all no native language
software should be designed considering the borrowed words seriously (I
don't know how many of you do agree with me).
(From the attached .pdf file in Dr. Nikhil's mail..."The current text
books, developed very professionally by competent people and supported
well by current grammatical practices, state the following:
A. Antastha ja (ya) and yya are two forms of the same letter. In earlier
times yya was not used at the beginning of a word, but is used now with
adapted words like Yunani and Europe.
In combination with other conjuncts it is treated as 'ja' phala.")
>From the above para, I wish to comment on the phrase "current grammatical
If the quoted lines are true among the Odia mass, what is the expected
situation if a software is given to a common man and he is asked to write
the Odia word 'kabya' (epic) for the first time. Without any special
instruction, he will definitely use y(0B2F)key, not yya(0B5F). Because at
the phychological level he reads it as ja-phala. That is the whole point.
Here comes a conceptual question: Do the software developers need to
consider the phychlogy of the human beings when they are developing a
machine for the human beings?
Now if we decide that " 'ja' phala to be formed with 'yya' (0B5F)", we are
definitely doing it only to maintain the ISCII-1991 standard. It is a fact
that ISCII-1991 is no longer in use now and is going to dead soon. It is
obvious that nobody will use ISCII-1991 once the Unicode based systems are
available for mass use. Then the question is whether the current
Unicode-based systems have the compatibility with the obsolete ISCII-1991
based systems or contents. Is there a grand necessity for that? If yes,
with a small engineering, that can be handled. The ISCII based contents
can be converted to the Unicode standard. And that will be done once by
experts. For example, the CILL corpus which is in ISCII will be converted
to Unicode only once. By just some replacement techniques, we can correct
the ISCII text and convert it to Unicode standard for the advanced NLP.
There after no body even look at the ISCII one.
Finally, I want to say that the mass may not get the opportunity to raise
the loud voice against 'ja' phala to be formed with 'yya' (0B5F) or some
other similar problem), because only a very small fraction of Oriya people
have access to the cutting edge technology. On the other side, only a
handful of people like us are playing with ICT and the future of this
language is solely dependent on this small community. But by doing this in
the name of maintaining the ISCII (once accepted inappropriate standard,
when there was no active Orissa-it or oriya-group yahoo groups or so), we
are forcing the mass to switch over to the IT practice from their usual
non-IT practice (i.e., from ja-phala to yya phala), isn't it?
Thanks for your patience for reading this boring topic.
> With reference to the current discussion on the Oriya conjucts with 'ya'
> and 'ba' and 'wa', I would like to suggest that the Indian Standards based
> on ISCII-1991 (and followed essentially by the Unicode org for modern,
> non-vedic text) should be adhered to.
> This would imply the following:
> 'ja' phala to be formed with 'yya' (0B5F),
> 'ba' phala to be formed with 'ba' (0B2C) and
> 0B71 to be taken as the glyph for letter 'wa'.
> I am attaching a detailed note (a bit long for the post) giving the
> background and rationale for my suggestion and some other supplementary
> material. Hope that this will be uploaded to the files section.
> Using non-standard rules will create more problems when other developers
> come up with their products. It is particularly important that big players
> like Microsoft (I do not know what role has OCAC played in the development
> of Kalinga font) and RedHat should stick to the standards.
> The Oriya rendering in Windows Vista with their Kalinga font has problems
> with 'ya/yya' and 'ba' letters and phalas as well as the letter 'wa'.
> Although the glyphs show up correctly in the charmap they do not appear so
> in Wordpad.
> The Utkal font (the most recent version bundled with Suse 10.1) works fine
> on Windows Vista, but the yya phala there forms with ja (ya). I have
> requested help for correcting this from several people and hope that it
> will be done soon.
More information about the Oriya-group