Help with a puzzle - asura and avuNar

[Sudalaimuthu Palaniappan]

The earliest Classical Tamil texts do not use the word 'asurar'. The word
used is  'avuNar' meaning 'asuras'. The etymology of this word is very
puzzling.  There does not seem to be any obvious Dravidian etymology for it.
 It is most probably of IA origin. (However, T. Burrows� derivation of avuNar
from asura is flawed.) After eliminating some other possible sources with the
help of George Thompson, I am narrowing the investigation to the Sanskrit
root 'vR'. Sanskrit 'R' can become 'u' in Tamil. For instance

Sanskrit amRta > Tamil amutu/amutam

So if we postulate a hypothetical 'avRNa' in Sanskrit, it can become Tamil
'avuNar'. 
However, I have not been able to verify any occurrence of 'avRna'. The
closest I have been able to find in connection with 'asura' is the following.

Wash E. Hale in his A'sura - in Early Vedic Religion p.104 says the
following.

"ye'na devA a'surANam
  o'jaMsy a'vRNIdhvam
 te'nA naH za'rma yacchata         AV 6.7.3

"O gods, grant us protection by means of that with which you repelled (or :
chose(?)) the strengths of the asuras."

There is a problem with this verse. The verb avRNIdhvam should be from the
root vR "to choose" rather than the root vR "to repel" since the latter makes
no other nA- formations. The AVP text (AVP 19.3.12) avoids this problem by
having yAni in place of ye'na and tebhin (for tebhis?) in place of te'na.
This allows one to translate: "O gods, grant us those protections, the
strengths of the asuras which you chose." But a'sura- still must not be used
here in its old meaning "lord" since the preceding verse in the text uses
a'sura- with a pejorative connotation."

Also, in vAjasaneyi samhita 1.13, we find, "yuSmA indroavRNIta vRtratUrye
yUyamindramavRNIdhvam vRtratUrye prokSitA stha". Regarding the words vRtra,
avRNIta, and avRNIdhvam,  Griffith says, "vRtra : the cloud-demon who
obstructs and withholds the rain. The root of the word is vR, to encompass,
also, to choose; hence the play on the words avRNIta, he elected, avRNIdhvam,
ye elected and vRtra. Cf. R.V. III.34.3, vRtram avRNot, he (Indra)
encompassed vRtra, the encompasser."

vAjasaneyi samhita 33.26 has "indro vRtramavRNocchardanIti pra
mAyinAmaminAdvarpaNIti". In explaining this, Griffith says, "..Encompassed
vRtra: vRtramavRnot: a play upon the words, both from vR, to encompass; vRtra
the drought demon  being the encompasser or obstructer of the seasonal
rains."

Thus even though avRNa as a word to indicate asura does not seem to be in
Vedic, it looks very probable that the root vR in the quintessential asura
vRtra seems to be the root the Tamil avuNar is based on. 

Does anybody know the occurrence of any equivalent of avRNa in later Sanskrit
or even any Pali/Prakrit texts, and if it does occur, what will be its
meaning -  chosen one or repelling one or covering one or encompassing one or
obstructing one?

If  Tamil avuNar is from the IA root vR, and if an equivalent form is not
found in Indo-Aryan, that raises a very interesting question. When and under
what conditions would the word have entered Tamil or Dravidian?

Comments from the members of the list will be appreciated. Thanks in advance.


Regards

S. Palaniappan