From: Kenneth Whistler (
Date: Tue Aug 22 2000 - 19:28:26 EDT

Chookij V. continued:

> ] >
> ] > Ex 3: Ka + Nukta + Virama + Ta + Nukta + Virama + Sa

I think the valid way to bracket this is:

[{( Ka + Nukta ) + Virama} + {( Ta + Nukta ) + Virama} + Sa]

Where each () pair indicates the nuktated consonantal value of a CV
(consonant + inherent vowel), and where each {} pair indicates a CVhalant,
namely a C (consonant with inherent vowel removed). And where the []
pair brackets the entire resultant syllable.

That results in the -CCCV- consonant structure implicit for the syllable

> ] >
> ] > This one, the Nukta applied to "Ta" with "Virama + Sa"
> ] > wouldn't be considered as part of syllable "Ka + Nukta + Virama + Ta".
> ] > Is this correct ?
> ]
> ] If you don't restrict the scope of the Nukta to the immediately
> ] preceding consonant, you have a scope ambiguity here -- which is
> ] not the intention.
> So, I think, for sure, "Nukta" must be applied to the immediately preceding
> consonant and now, the point is about how it's related to "syllable".

My suggestion would be as shown above with the bracketing.

> Would that make sense if, within syllable itself, the consonant cluster
> can have only one of "nukta applied immediately to preceding consonant"
> part of the cluster ?

I think this depends entirely on what non-native consonant combinations
(utilizing nuktated consonant representations) can occur in syllables, and
on how far Devangari is extended in transliteration of other languages.
I would certainly expect something like -aqfa- to be a possible consonant
combination for some Arabic or Persian-derived word, for example.

> I think this would require a little bit more of hindi linguistic for the
> decision.
> If yes, then, we can determine the boundary of "syllable" and if those nukta,
> even they are applied to immediately precedding consonant, can be determined
> whether they should be part of syllable or not. If not, then, it's up to
> the rendering system to render them seperately.

I think the syllable boundaries can be determined independent of whether
there are any nuktas in the picture. As suggested above, nuktas would
be the most tightly-bound elements in the structure. You should be able
to structurally substitute a ( Ka + Nukta ) for a ( Ka ) without in
any way impacting the syllable at the higher levels of structure.

The only real question here is what typographical practice in Devanagari
would be if two nuktated consonants came together in a halfform/fullform
juxtaposition or in a conjunct form involving two nuktated consonants
(do such conjunct forms even exist? -- I would doubt it).


> What do you think ?
> Chookij V.

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:13 EDT