Pythagoras mentioned in Vedas?

Lars Martin Fosse lmfosse at ONLINE.NO
Tue Jun 15 19:36:22 UTC 1999

Bo Klintberg wrote:

> Dear Lars Martin Fosse:
> This is just a short letter to tell you that I am <truly> sorry if I have
> <personally> offended you. I thought that I just scrutinized some
> argumenents that potentially I thought were yours. Once again, I am truly
> sorry if I in any way disturbed or offended your person.
> That doesn't, however, mean that I like your way of doing statistics. I
> still don't. :-)

I disliked your argument about chimpanzees, which seemed to indicate that I had rather
primitive ideas about Indians. However, you are forgiven.

Now, my statistical views have to do with a certain kind of inferential logic that is used
extensively in the humanities, in politics and in law. I think a debate along such lines
would be interesting, and I invite you to such a debate. I'll promise to be serious, and
not make nasty remarks. My interest in this matter was stimulated by my reading of more
Indo-European homeland studies (both European and Indian) than any man should have to read
in his life. It struck me that much of the reasoning in cases where data are scanty is
based on analogy, or if you like, paradigmatic thinking. Because certain data (or
combinations of data) are associated with certain other things according to experience, the
data are interpreted along the same lines. In the homeland context, for instance, the
Indo-Aryan migration into India was a couple of generations ago seen as an invasion of
warlike tribes, overrunning and destroying the Harappan civilization. This was analogous to
for instance the situation in the late Roman empire. We know today that this interpretation
is wrong, and that there are other models or analogies available, so that the modalities of
the intrusion of Indo-Aryans into India can be interpreted in other ways. This leads me to
a more general consideration:

Assume that the phenomenon A always is accompanied by the phenomenon B. Then we have an
ironclad rule, and we know that whenever we find A, we have B as well (if B is not
physically present any more, we must assume that it once was). Then consider the following

A is accompanied by B in 50 % of the cases
                            by C in 25% of the cases
and                       by D in 25% of the cases.

Now if you find A (and only A), you have to make a choice between B, C and D. Since we have
statistics based on experience, the most likely candidate to accompany A is B, but both C
and D are possible. In other words, finding A, you make the inference that A is most likely
accompanied with B (even if you don't find B).

Now consider:

 A is accompanied by B, or C, or D. We have no statistics. We have in other words three
possible interpretations, but no way of deciding which one is the most probable. This is
the situation that obtains in a large number of cases, both in ancient history, in
philology and for that matter, in police work. And we need a kind of method for dealing
with such uncertainties. I hope you will see the connection to my reasoning around
Pythagoras. You do not have to like my statistics, but I would be interested to know how
you treat such conundrums methodically. Which is why I used the old trick question "Have
you stopped beating your wife?" (Yes or no!) in a slightly revamped version. If you are
ever dragged into court accused of wife-battering, you will soon find out that probability
thinking matters if physical proof or eyewitnesses are missing.

Best regards,

Lars Martin Fosse

More information about the INDOLOGY mailing list