Re: Vedic Characters & Unicode

From: Antoine Leca (
Date: Thu Apr 13 2000 - 07:01:30 EDT


[ copies to both Indology and Unicode lists. ]

> Date: Tue, 11 Apr 2000 09:20:59 -0400
> From: John Robert Gardner <jrgardn@EMORY.EDU>
> Subject: Re: Vedic Characters & Unicode
> It is my understand that the vedic accent set, and quite a large one at
> that, is part of std. subsequent to ISCII-1988 (My copy is not here in the
> office), and this set is part of Annex G of the set.

Minor point: Is it ISCII-1988, or ISCII-1991 ?

> It's complete, but not implemented anywhere I know of. [...] It
> was IS 13194:1991 (which is out of synch with Unicode 1.0, though unicode
> is a superset of ISCII:1991). Alas, Annex G is expressly excepted from
> [Unicode].

Well, Unicode have four characters that are intended be used for Vedic:
  0953;DEVANAGARI GRAVE ACCENT;Mn;230;ON;;;;;N;;;;;
  0954;DEVANAGARI ACUTE ACCENT;Mn;230;ON;;;;;N;;;;;

Problems are, what are they supposed to mean, and how should they be used?

The problem is particularly acute (sorry) for the two "accents",
since I never saw accents used with Devanagari, but rather with Latin
transcriptions or translitterations. So what was the initial intent here?

Then come the "stress signs". ISCII-91, and most sources I had read
on this matter, distinguish three "signs": udatta, anudatta and svarita
(resp. high "tone", low or lacking "tone", and combined raise and
falling "tone"); it seems that tradition is well established for them
to be represented *in*transcription* as resp., acute accent, no accent,
and grave accent. On the other hand, use of these "stress signs" in
Devanagari does not appear to be consistent among the varying Veda
and vedic texts: for example, I understand that Rigveda (the most
important text) uses nothing for udatta, the sign represented as U+0952
for anudatta and the sign U+0951 for svarita. But others uses U+0951 for
udatta (Yajurveda's Maitrayaniya, according to ISCII-91), or U+0952 for
the only accent noted in Shatapatha (there is discussion on this
subject, for example see <URL:>).

ISCII-91 is mentionning (among others) kampa, and show for it two forms,
according to length: being Devanagari figure 1 (U+0966) or figure 3 (U+0968)
marked with both signs figured by U+0952 and U+0951 in Unicode.
I *guess* these should be encoded as resp. \u0966\u0953\u0952 and
\u0968\u0953\u0952, but I would like to have some confirmation from
the Unicode gurus.

Also, ISCII-91 mentions another special sound(s), described as half-visarga,
and known as jihvamuliya before gutturals (ka, kha), and upadhmaniya before
labials (pa, pha). ISCII-91 gives only one sign (looking like a x-cross),
but Pr. Stone gives detailled explanations on this matter and shows
two signs (<URL:>)
Does they have to be uniformly encoded in Unicode as visarga U+0903?

Thanks in advance for your answers,


Antoine Leca

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:01 EDT