Joshi and Roodbergen on Panini 1.4.49-51

Vidhyanath Rao vidynath at MATH.OHIO-STATE.EDU
Mon Jul 13 18:19:20 UTC 1998

This is tangentially related to the question of double accusatives.
Specifically, I am perplexed by how Joshi and Roodbergen deal with
Panini 1.4.49-51, and I will be thankful to anyone who can throw some
light on this.

According to them, ``tathaayokta.m caaniipsitam'' is about double
accusative and ``akathita.m ca'' is a deletion rule that makes sentences
like ``raamaaya dadaati'' correct. The argument for the latter is that
sampradaana assumes a karman which is missing in that sentence, and we
must use 1.4.51 to conclude that a karman present even if unexpressed.

But deletion of the subject is also found: For example MS 1.6.4:
``agni.m vai devaa vibhaaja.m naa"saknuvan ... tam a"svena
puurvavaahodvahan.'' Subject is typically deleted if it is an indefinite
person. We also see verbs deleted when understood from context:
``braahmanebhyo dadhi diiyataam. takra.m kau.n.dinyaaya.'' or an extreme
case (RV 10.168.4c) gho.saa id asya s.r.nvire, na ruupam''. I don't see
why deletion of the understood object needs special treatment, but not
of subject or verb.

I also fail to understand their claim that the object is always
`iipsitatama' of the agent. Surely we can talk about incidental or
accidental occurrences. Extant Sanskrit texts may not have in-your-face
examples like ``snaatiim pa"syan "svapurii.sam am.rdnaat'', but why
can't ``graama.m gacchan v.rk.smuulam upasarati'' refer to an
incidental happening, with the agent not caring if he nears the
tree-root or not?

The statement that secondary object is covered by `tathaa yukta.m
caaniipsitam' is equally puzzling: In ``bali.m vasudhaa.m yaacate'',
I can't see why earth is connected to begging the same way as
Bali is.

I don't understand their comments about causative of passive
(pp.144-147). According to Comrie (Ling. universals and lang. typology)
causative of passive is not common among languages with morphological
causatives. I don't see why we should assume that Panini's Bhaa.saa had
it, unless examples from literature support it.

More information about the INDOLOGY mailing list