Sandhi in `real' Sanskrit vs `conversational' Sanskrit

[D.H. Killingley]

I've been following the dialogue with interest, and would like to 
comment on Vidyanath Rao's question about what is 'real' Sanskrit. 
The quotation marks on 'real' are apt, because reality in this context is 
relative. That's not to say it doesn't exist or can't be defined, but it 
has to be defined differently for different situations.
	For myself and many of the people I have taught Sanskrit to,
'real' Sanskrit is what we find in the published texts that we read. 
(These texts are mostly Vedic and Vedantic texts with their commentaries,
BhG with its commentaries, MBh.) While presenting students with specially
written beginners' Sanskrit, I have to warn them that when they come to
'real' Sanskrit they will find a much wider vocabulary, greater variety of
inflections, etc. If I were preparing them to converse in Sanskrit as
spoken by pandits, I would probably give them a different warning: that .r
is not distinguished from ri, that a vowel is repeated after visarga, and
that sandhi may be freer. If I were preparing them to converse in modern
spoken Sanskrit (as described by some participants in this exchange, or in
Hajime Nakamura's *A companion to contemporary Sanskrit*, Motilal
Banarsidass 1973), I would warn them to expect a _narrower_ range of
inflections, plural for dual, etc. But the 'reality' I am preparing them
for is the printed text. 

	That 'reality' is to some extent artificial, in that it has been
created by the people who took part in the growth of printed texts. For a
start, the kind of Sanskrit we read is in Devanagari, though I
occasionally read sources printed in Bengali or other script (and of
course quite often in Roman). Verses, which are run-on in manuscripts, are
set out in lines. And--to take up one of Vidyanath's points--sandhi is
regularised, 'correcting' the 'errors of sandhi' which occur in
manuscripts. (Here we come to the relative nature of 'reality': if the
'errors of sandhi' reflect the way a shishta scribe spoke the text, then
from the point of view of someone wanting to speak Sanskrit as pandits
spoke it, these 'errors' are 'real Sanskrit'.) Commas, quotation marks,
question marks, exclamation marks and dashes may be used. And sandhi
breaks are not only at sentence breaks, but sometimes between clauses in a
complex sentence, as a form of invisible punctuation. 

	This is not the only 'real' Sanskrit, but it is the first one that
many of us in the West encounter or want to encounter. It puts us in 
touch with a vast, varied and fascinating range of literature.

	Vidhyanath wrote: To return to my original question, who owns
Sanskrit? That is, who decides what is `real' Sanskrit? I'd say nobody
owns it, and everyone decides what, for themselves, is 'real' Sanskrit.
But as with any language, a merely private reality is a contradiction of
the communal and communicative nature of language itself. Also, while the
existence of varieties for different regions, registers, social situations
(including teaching and learning situations) and so on is essential to
language, one needs to be aware of which is which, and which belongs
where. Therefore: one shouldn't talk like a printed book (with long
sentences and no intonation) and expect to be understood; nor should one
use or teach a variety designed for beginners and try to pass it off as 
the language of a learned tradition.

	With best wishes to you all,

	Dermot Killingley

Dr Dermot Killingley
Dept of Religious Studies
University of Newcastle upon Tyne
Newcastle upon Tyne NE1 7RU
Phone 0191 222 6730    Fax 0191 222 5185