Is "Sanskrit" Dravidian; and open-mindedness

Michael Witzel witzel at FAS.HARVARD.EDU
Sat Jun 26 21:29:25 UTC 1999

Just back from another  open-minded tour of Japan,  open-mindedly studying
not Indo-European,  but ancient Japanese myth and ritual, and open-mindedly
drinking, not IE *medhu,  but sacred Sake (miwa) offered by a Miko inside
the inner compound of one of the oldest Shinto shrines,
I was struck by the light-hearted naivite' of someone who constantly has
written against IE heritage in India and beyond:

At 19:31 -0700 6/10/99, Paul Kekai Manansala wrote:

>Yes, but Sanskrit is an Indian language.  Maybe it would be more
>relevant if -am was a common ending for IE nouns

And again, later:
>Are you really saying that Latin -um, Greek -on and Sanskrit -am are related?

And again:
>  Sanskrit is very similar in structure to Dravidian.

Sentences like these betray absolute innocence with regard to IE word
formation & grammar in general.
This has already been explained by other members . The upshot is that
Nothing  that is said about IE / non_IE  or Indo_Aryan/ non-IA in South
Asia can be taken seriously.

However, to put P.K.M. at ease, I will not only confess here an open-minded
"non-IE" marriage (with a Japanese), a "half-IE" son (Manabu), and our own
(linguistically very interesting)  German-Japanese- English pidgin/creole,
but rather to my recognition  of a quite substantial number of non-IE
substrate words in Vedic Sanskrit, stemming from Dravidian, Munda, and
OTHER, now lost languages. And different from what is usally supposed  as
far as the order, geographical distribution  and *immigration*  of Drav.,
Munda etc. into the subcontinent are concerned.  -- Details in EJVS in late
June (www address below).

>And many other words are of Austric or other etymology making
>up the vast majority of Sanskrit words.

The point of 'vast majority' has already been discussed by others. In Vedic
they range from c. 4% in the, after all, rather hieratic  RV to a larger
proportion in the Atharvaveda, etc.
The important point is not so much the exact percentage but the continuing
occurence of *new*
imports of such words into Vedic even after the Atharvaveda, indicative of
*when and where* the speakers of  Drav., Munda etc., were resident at the

The other importat point is that PKM is correct when he says that (Vedic)
Skt. is <<already>> an Indian language. It clearly has been influenced by
the local S. Asian substrate, WHILE Avestan,  ITS CLOSE RELATIVE, almost a
dialect of Vedic, has NOT.

This point (and local influence on IE/IA myth and ritual) is not always
understood properly, or rather neglected, by Indo-Europeanists.

PKM also is right when he says
> the Austric contribution as recognized by (most) modern scholars that
>appears to be small.  ...the >evidence suggests the contribution in all
>regards is quite great.

Kuiper (linguistics, myth) and Berger (myth) and now Sergent (myth) are
notable exceptions. More in EJVS.

and also :
> The Austric presence really extends beyond Eastern India

For example,  even today on the Tapti (Kurku, and their Nahali-speaking
neighbors), but already seen in the Rgvedic Panjab.

In short, the subject is in need of an open-minded, sine ira et studio
approach, --  not fixed attitudes such as  Aryan "invasion" or not,
Dravidian "autogenesis" in S.Asia,  Munda-minimizing   (not to speak of
other languages and their original speakers ).  More in next message.

Note that the late(?)  I.M.  Diakonoff has even sought to establish a
Munda- Sumerian family. (Mother Tongue III, 1997, 54- 62), with examples
from vocabulary and the complicated polysynthetic grammar.
(So much for the 'low level' culture of the Mundas and their relatives: Cf.
the Mon and Khmer and Sumerian?? civilizations, and -- how much do
Guatemalan jungle farmers look like bearers of the Maya civilization now?)

Michael Witzel                          Elect. Journ. of Vedic Studies
Harvard University        
my direct line (also for messages) :  617- 496 2990
home page:

More information about the INDOLOGY mailing list