`Conversational Sanskrit' vs `Real Sanskrit'

[Madhav Deshpande]

Robert Zydenbos recently wrote:

>Irrespective of our considerations on more abstract and debatable levels,
>what matters very much is the switch from "aha.m pa.thaami" to "mayaa
>pa.thitam", clearly visible in the surface grammar. Why should "aham" 
>suddenly become "mayaa"? Or, to take a parallel example from Urdu-Hindi: 
>"mai.m karataa huu.m", "mai.mne kiyaa hai", "mujh se karanaa hai". My
>point is that in Dravidian, the nominative case would be used to
>translate the agent in all three instances (hence I would actually
>prefer "agentive"  as a term for that case; but this does not matter
>here). This should be empirically absolutely clear, irrespective of
>whether we call something ergative, passive or anything else.  (If you
>want to call it ergative, that is fine with me, because it does not
>matter.) Hence "aha.m pa.thitavaan", etc.

	I think it would be useful to differentiate the preference for
different kinds of past participles in different varieties of Sanskrit.
In Hindi, we have (split) ergativity, in that the ergative appears in the
past tense, but not in the present tense.  Thus, generally aham pa.thaami
is fine, but instead of aham apa.tham, we are more likely to get mayaa
pa.thitam.  This is because in Hindi past tense, one has to say, without
any option, something like mai~ne pa.dhaa.  This is the kind of preference
for the -ta participle that is seen in the stories of
Vetaalapa~ncavi.m"sati etc. (See: Hans Hock's study: "P-Oriented
Constructions in Sanskrit", in South Asian Languages, Structure,
Convergence and Diglossia, ed. by Bh. Krishnamurti etal, Motilal
Banarsidass, 1986).
	Obviously, the preference for -tavat in the Bangalore Sanskrit of
Krishna Shastri is a distinct phenomenon.  Such a preference is not
manifest in the modern Sanskrit varieties coming from north India, or even
from Maharashtra.  The entire past tense systems of Hindi and Marathi are
derived from the derivatives of the -ta participle in Sanskrit, and there
is no trace of -tavat.  Where one can get the agentive use of -ta in
Sanskrit, such as aham gata.h (cf. Panini:  gatyarthaakarma"sli.sa-
"sii~gsthaasavasajanaruhajiiryatibhya" ca), we get agentive constructions
like mai~n gayaa (Hindi) and mii gelo (Marathi).  But for transitive verbs
like khaad, one cannot have an agentive -ta in Sanskrit, and hence one can
only say mayaa phalam khaaditam, the modern languages appropriately have
mai~ne phal khaayaa (Hindi) and mii phaL khaalle (Marathi).  These
languages do not have a construction corresponding aham phalam
khaaditavaan, because such a construction is dead in Prakrit a long time
ago.  Therefore, there is no preference for such a construction in
Sanskrit rooted in Hindi and Marathi. 
	In contrast, the preference for the -tavat construction in
southern Sanskrit (I do not know if this can be generalized) is a truly
distinctive phenomenon, and its underlying causes need to be distinguished
from ergativity in Hindi and Marathi.
	All the best,
				Madhav Deshpande