Sarasvati (texts & arch.II)

Michael Witzel witzel at FAS.HARVARD.EDU
Wed May 27 12:16:01 UTC 1998

On Sun, 24 May 1998, Paul K. Manansala wrote:

> As I see it the RV is the main evidence for the early IE
> invasion/migration theory.  And the evidence is textual since we have
> no manuscripts from the Vedic period.  So every major proponent of
> the AIT has attempted some fairly liberal interpretations of the RV
> regarding geography and many other subjects.  Therefore, we cannot
> use the RV as proof in one direction but not the other.

As usual in this discussion the baby is thrown out with the bath water.
Of course, there are no RV time manuscripts. But, except for some minute,
well known details, the ORAL tradition has been fixed from the beginning.
No changes allowed. Thus, I call it a "tape recording" of the RV period.
It is as good as any well preserved inscription.

Of course, as Indologists have underlined for some 150 years, even such
texts are POLITICAL. If you can read between the lines, and that is of
course part of our training, you get a pretty good idea of what was going

The geographical data can be studied just like anything else in the text.
Some things are obvious, such as a RV area of the Greater Panjab and
surroundings. Others such as the samudra and the Rasaa need more study.

And we have to use the *RV* since we do not have anything else for the
whole period, say from 3000-1000 BC, except archeology, or, as we cannot
read it (Indus inscriptions). Or we can resign ourselves, and shut up.
Archaeology? - Archaeology alone did not tell us that the Mayas were not
as peaceful as they had been portrayed until we could read their

> My contention about the Sarasvati is that the AI theorists do seem to be
> "retrofitting" the data here and there.  The first and foremost thing
> is there is no hard archaeological evidence of a "Vedic" culture
> entering  India.
> We certainly cannot label PGW culture as Vedic anymore than IVC.

That last sentence alone shows that the data are not taken into account.
The data of the Indus civ. just do not fit the RV (we need a special
e-sattra on that), but the PGW has many correspondences, in spread,
material culture, etc., with the *POST*-RV texts.

Again the arch. evidence alone. Even tehre w have soem data as I have
already pointed out. Should I repeat myself again? -- What about language,
material and spiritual culture, customs, beliefs, rituals, religion etc as
found in RV?  These data have all the kinks with Iran and beyond, not with
the Indus civ.  etc.

> We have, at best, undoubted similarities between the medieval Vedic
> texts and the Avestan texts/inscriptions  dating from several
> centuries before this era.  From this we are attempting to
> extrapolate on events of anywhere from 1,500 BCE to 3,000+ BCE
> using mainly text interpretation..

medieval Vedic texts? The RV and te other Vedic texts were composed much
earlier: as even quotations in Panini and the Pali canon show... I give
Better to get the data right before drawing such hasty conclusions.

As for text interpretation: if we do or jb of understanding these texts
well, we can restore even damaged texts correctly, as indeed happend to me
long ago in my thesis, when I discovered the actual wording restored by
me, hidden under some broken off section of a birchbark manuscript that
was not visible in the film I used...

Every interpretation is of course an approximation, until you use time
travel and talk to the local people, lots of them... We have been at it
for some 150 years now... and know something about these texts, for
example all, those points I critized above. And more importantly, we know
quite well what we do *not* know, for example about the 'little tradition"
in the RV. (for that we read te Atharvaveda etc.)

There are no Avestan inscriptions. We would be so happy. If only the 3/4
of the lost Avesta corpus could be found in Balkh, allegedly written there
before "Alexander destroyed the texts"! Facts, again....

With our data coming straight from the Vedic period itself we do not
extrapolate backwards for 3000-odd years... we *listen* to what Visvamitra
or Vasistha composed. (Such tapes & records have been published).

> For example, our belief that the supposed "Vedic" peoples used horses
> is based entirely on the Vedic texts. Since we do know the dating of
> the entire corpus for sure,  how can we make any theories using these
> texts?  Is the use of the Vedas as historical sources even valid?

Using your wording, we "do know the dating for sure" - I suppose the
medieval MSS? That, we know for sure indeed. (otherwise the passage makes
not sense to me)  --- But we can also make very good educated guesses (and
do even better) about the historical situation of the time of the
composition. Of course, we can use poetic texts and mine them for some
historical information: Using the famous line from the Psalms where the
author (Salomon, I think) compares his girl friend's teeth to the parapets
of the walls of Jerusalem you can make some deductions about (a)
Jer.having a type of wall, and (b) about the ideal nature of teeth/ a
smile usual (or conversely the uneven nature of teeth) at the time...
Again, well known procedures.

By your standard we should give up the use of texts such as Homer's, the
Bible, Roman annals, old Chinese history, the Kojiki etc., since all of
them have "medieval MSS" (well, a few are a bit earlier). Tell that to the
Israelis,  the Chinese etc. !

The term "Vedic peoples" is not a good one. 'Tribes speaking Old
Indo-Aryan' or the 'Vedic variety of Old Indo-Aryan' would be correct.
"Vedic peoples"  would be those who use that language, and adhere to the
cultural norms spelled out in the texts. All well known (Kuiper,
Southworth etc.)_

Michael Witzel                       witzel at

More information about the INDOLOGY mailing list