Dates of written Rgveda

Michael Witzel witzel at FAS.HARVARD.EDU
Mon Mar 13 17:44:02 UTC 2000

Dear Dr. Farmer,

I suppose we all get tired by now, but let me underline just one or two
points, as not to be misunderstood.

>I'll cut to the chase; you wrote:
>> I have no doubt that some 'progressive' Brahmins and others *may* have
>> tried to use written Veda texts during the time of the Smritis ... This
>>much is clear from the >>injunction NOT to use written texts.
>> But it did not become the main, major way of transmission until much later.

>Thanks -- that is the *only* admission that I was really looking
>or hoping for. I can live with this view. What you write here is,
>I think, a considerable move away from your original claim
>(possibly made polemically, and in haste) that the first written
>text of the Rgveda came no sooner than c. 1000 CE.

Not really. Please read this line together with the one where I stress,
that (your) case still is to be made by showing *clear miswritings* in
Vedic texts that can ONLY be due to the palaeographic development of the
Brahmi etc. scripts.  (not, to local pronunciation)

As long as that is not done, and if one cannot come up with clear
statements in the literature about Veda reciters with their manuscripts in
front of them, -- all remains (comparative) theory.

The history of Vedic quotations, too, indicates their ORAL origins.
Right from Yaska's Nirukta who slightly misquotes a Kathaka Samhita line,
to the RV commentary of SaayaNa (died 1387 CE, if I remember the date
correctly) the Veda quotations in commentaries show that their authors
relied on their -- not always perfect -- memory of texts that did NOT
BELONG to their own tradition (i.e.: a Rgvedin quoting a Samaveda Brahmana
text, etc.).
You always find the typical substitutions (words, collocations, etc.) which
we all make when reciting a poem/singing a song that we do not remember too
well anymore.   One does not do that so easily when one actually looks up
and  compares a written text (pace: the common copying mistakes we all
make; they are different in nature).
Just check out the list of such lapses of memory  by the great Sayana,
published already by Max Mueller in his Rgveda edition more than 100 years
ago (or is this just another one of his secret, destructive  schemes?). The
list can be extended. (I think I have some such data in my Caraka papers of

Also, please note that the available Veda MSS were habitually, until
(almost?) today,  corrected by reciters, from memory; they frequently also
put in the (red) accent marks. Much of this is even *visible* in extant MSS.

Modern habits -- Veda teachers with their printed texts/manuscripts, open,
in front of them (as I also have seen) -- does not tell us anything, after
some one thousand years of use of Veda MSS, about the practices of the
first millennium CE.

Second, I totally agree with S. Cousins in what he writes about early, even
Neolithic societies.
Neolithic Polynesia is a good test case. The main mythological texts,
transmitted orally by their "druids" (Kahuna, etc.) well into the 19th
century, often agree well across this vast region over some 2-3000 years,
-- more specifically in Tahiti, Hawaii, New  Zealand (in this case, origin
of settlement in the Society Islands,  roughly, after  the beginning of our
era). However, not literally as in India. And certainly not when it comes
to genealogies, which were -- as everywhere (compare the Indian Epics  &
the Vedas)  -- influenced by the interests of the local chieftains.  In
Polynesia,  large genealogical sections agree, but not all of them; and of
course, not, when it comes to local items.  No surprise to anyone who has
read Parry-Lord.


Michael Witzel
Department of Sanskrit & Indian Studies, Harvard University
2 Divinity Avenue, Cambridge MA 02138

ph. 617-496 2990 (also messages)
home page:

Elect. Journ. of Vedic Studies:

More information about the INDOLOGY mailing list